kotovalexarian-likes-github/jashkenas--coffeescript
1
0
Fork 0
mirror of https://github.com/jashkenas/coffeescript.git synced 2022-11-09 12:23:24 -05:00
Code Releases Activity
jashkenas--coffeescript/lib/coffeescript/parser.js

894 lines
108 KiB
JavaScript
Raw Normal View History

Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
/* parser generated by jison 0.4.17 */
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
/*
Returns a Parser object of the following structure:
Parser: {
yy: {}
}
Parser.prototype: {
yy: {},
trace: function(),
symbols_: {associative list: name ==> number},
terminals_: {associative list: number ==> name},
productions_: [...],
performAction: function anonymous(yytext, yyleng, yylineno, yy, yystate, $$, _$),
table: [...],
defaultActions: {...},
parseError: function(str, hash),
parse: function(input),
lexer: {
EOF: 1,
parseError: function(str, hash),
setInput: function(input),
input: function(),
unput: function(str),
more: function(),
less: function(n),
pastInput: function(),
upcomingInput: function(),
showPosition: function(),
test_match: function(regex_match_array, rule_index),
next: function(),
lex: function(),
begin: function(condition),
popState: function(),
_currentRules: function(),
topState: function(),
pushState: function(condition),
options: {
ranges: boolean (optional: true ==> token location info will include a .range[] member)
flex: boolean (optional: true ==> flex-like lexing behaviour where the rules are tested exhaustively to find the longest match)
backtrack_lexer: boolean (optional: true ==> lexer regexes are tested in order and for each matching regex the action code is invoked; the lexer terminates the scan when a token is returned by the action code)
},
performAction: function(yy, yy_, $avoiding_name_collisions, YY_START),
rules: [...],
conditions: {associative list: name ==> set},
}
}
token location info (@$, _$, etc.): {
first_line: n,
last_line: n,
first_column: n,
last_column: n,
range: [start_number, end_number] (where the numbers are indexes into the input string, regular zero-based)
}
the parseError function receives a 'hash' object with these members for lexer and parser errors: {
text: (matched text)
token: (the produced terminal token, if any)
line: (yylineno)
}
while parser (grammar) errors will also provide these members, i.e. parser errors deliver a superset of attributes: {
loc: (yylloc)
expected: (string describing the set of expected tokens)
recoverable: (boolean: TRUE when the parser has a error recovery rule available for this particular error)
}
*/
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
var parser = (function(){
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
var o=function(k,v,o,l){for(o=o||{},l=k.length;l--;o[k[l]]=v);return o},$V0=[1,22],$V1=[1,52],$V2=[1,86],$V3=[1,82],$V4=[1,87],$V5=[1,88],$V6=[1,84],$V7=[1,85],$V8=[1,60],$V9=[1,62],$Va=[1,63],$Vb=[1,64],$Vc=[1,65],$Vd=[1,66],$Ve=[1,53],$Vf=[1,40],$Vg=[1,54],$Vh=[1,34],$Vi=[1,71],$Vj=[1,72],$Vk=[1,33],$Vl=[1,81],$Vm=[1,50],$Vn=[1,55],$Vo=[1,56],$Vp=[1,69],$Vq=[1,70],$Vr=[1,68],$Vs=[1,45],$Vt=[1,51],$Vu=[1,67],$Vv=[1,76],$Vw=[1,77],$Vx=[1,78],$Vy=[1,79],$Vz=[1,49],$VA=[1,75],$VB=[1,36],$VC=[1,37],$VD=[1,38],$VE=[1,39],$VF=[1,41],$VG=[1,42],$VH=[1,89],$VI=[1,6,34,44,134],$VJ=[1,104],$VK=[1,92],$VL=[1,91],$VM=[1,90],$VN=[1,93],$VO=[1,94],$VP=[1,95],$VQ=[1,96],$VR=[1,97],$VS=[1,98],$VT=[1,99],$VU=[1,100],$VV=[1,101],$VW=[1,102],$VX=[1,103],$VY=[1,107],$VZ=[1,6,33,34,44,68,73,76,89,99,118,123,125,134,136,137,138,142,143,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$V_=[2,170],$V$=[1,113],$V01=[1,118],$V11=[1,114],$V21=[1,115],$V31=[1,116],$V41=[1,119],$V51=[1,112],$V61=[1,6,34,44,134,136,138,142,159],$V71=[1,6,33,34,42,43,44,68,73,76,87,88,89,90,91,92,95,99,116,117,118,123,125,134,136,137,138,142,143,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$V81=[2,98],$V91=[2,77],$Va1=[1,129],$Vb1=[1,134],$Vc1=[1,135],$Vd1=[1,137],$Ve1=[1,141],$Vf1=[1,139],$Vg1=[1,6,33,34,42,43,44,57,68,73,76,87,88,89,90,91,92,95,99,116,117,118,123,125,134,136,137,138,142,143,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$Vh1=[2,95],$Vi1=[1,6,34,44,68,73,76,89,99,118,123,125,134,136,137,138,142,143,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$Vj1=[2,29],$Vk1=[1,167],$Vl1=[2,65],$Vm1=[1,175],$Vn1=[1,187],$Vo1=[1,189],$Vp1=[1,184],$Vq1=[1,191],$Vr1=[1,6,33,34,42,43,44,57,68,73,76,87,88,89,90,91,92,95,99,101,116,117,118,123,125,134,136,137,138,142,143,159,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178],$Vs1=[2,117],$Vt1=[1,6,33,34,42,43,44,60,68,73,76,87,88,89,90,91,92,95,99,116,117,118,123,125,134,136,137,138,142,143,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$Vu1=[1,6,33,34,42,43,44,48,60,68,73,76,87,88,89,90,91,92,95,99,116,117,118,123,125,134,136,137,138,142,143,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$Vv1=[1,239],$Vw1=[42,43,117],$Vx1=[1,249],$Vy1=[1,248],$Vz1=[2,75],$VA1=[1,259],$VB1=[6,33,34,68,73],$VC1=[6,33,34,57,68,73,76],$VD1=[1,6,33,34,44,68,73,76,89,99,118,123,125,134,136,137,138,142,143,159,162,163,167,168,169,170,171,172,173,174,175,176,177],$VE1=[1,6,33,34,44,68,73,76,89,99,118,123,125,134,136,137,138,142,143,159,162,163,167,169,170,171,172,173,174,175,176,177],$VF1=[42,43,87,88,90,91,92,95,116,117],$VG1=[1,279],$VH1=[1,6,33,34,44,68,73,76,89,99,118,123,125,134,136,137,138,142,143,159],$VI1=[2,64],$VJ1=[1,291],$VK1=[1,293],$VL1=[1,298],$VM1=[1,300],$VN1=[2,191],$VO1=[1,6,33,34,42,43,44,57,68,73,76,87,88,89,90,91,92,95,99,116,117,118,123,125,134,136,137,138,142,143,149,150,151,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$VP1=[1,309],$VQ1=[6,33,34,73,118,123],$VR1=[1,6,33,34,42,43,44,57,60,68,73,76,87,88,89,90,91,92,95,99,101,116,117,118,123,125,134,136,137,138,142,143,149,150,151,159,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178],$VS1=[1,6,33,34,44,68,73,76,89,99,118,123,125,134,143,159],$VT1=[1,6,33,34,44,68,73,76,89,99,118,123,125,134,137,143,159],$VU1=[149,150,151],$VV1=[73,149,150,151],$VW1=[6,33,99],$VX1=[1,321],$VY1=[6,33,34,73,99],$VZ1=[6,33,34,60,73,99],$V_1=[6,33,34,57,60,73,99],$V$1=[1,6,33,34,44,68,73,76,89,99,118,123,125,134,136,137,138,142,143,159,162,163,169,170,171,172,173,174,175,176,177],$V02=[1,6,33,34,44,48,68,73,76,87,88,89,90,91,92,95,99,116,117,118,123,125,134,136,137,138,142,143,159,162,163,166,167,168,169,170,171,172,173,174,175,176,177],$V12=[14,30,36,40,42,43,46,47,50,51,52,53,54,55,63,64,65,66,70,71,86,89,97,100,102,110,120,121,122,128,132,133,136,138,140,142,152,158,160,161,162,163,164,165],$V22=[2,180],$V32=[6,33,34],$V42=[2,76],$V52=[1,336],$V62=[1,337],$V72=[1,6,33,34,44,68,73,76,89,99,118,123,125,130,131,134,136,137,138,142,143,154,156,159,162,163,16
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
var parser = {trace: function trace() { },
yy: {},
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
symbols_: {"error":2,"Root":3,"Body":4,"Line":5,"TERMINATOR":6,"Expression":7,"Statement":8,"FuncDirective":9,"YieldReturn":10,"AwaitReturn":11,"Return":12,"Comment":13,"STATEMENT":14,"Import":15,"Export":16,"Value":17,"Invocation":18,"Code":19,"Operation":20,"Assign":21,"If":22,"Try":23,"While":24,"For":25,"Switch":26,"Class":27,"Throw":28,"Yield":29,"YIELD":30,"FROM":31,"Block":32,"INDENT":33,"OUTDENT":34,"Identifier":35,"IDENTIFIER":36,"Property":37,"PROPERTY":38,"AlphaNumeric":39,"NUMBER":40,"String":41,"STRING":42,"STRING_START":43,"STRING_END":44,"Regex":45,"REGEX":46,"REGEX_START":47,"REGEX_END":48,"Literal":49,"JS":50,"UNDEFINED":51,"NULL":52,"BOOL":53,"INFINITY":54,"NAN":55,"Assignable":56,"=":57,"AssignObj":58,"ObjAssignable":59,":":60,"SimpleObjAssignable":61,"ThisProperty":62,"RETURN":63,"AWAIT":64,"HERECOMMENT":65,"PARAM_START":66,"ParamList":67,"PARAM_END":68,"FuncGlyph":69,"->":70,"=>":71,"OptComma":72,",":73,"Param":74,"ParamVar":75,"...":76,"Array":77,"Object":78,"Splat":79,"SimpleAssignable":80,"Accessor":81,"Parenthetical":82,"Range":83,"This":84,"Super":85,"SUPER":86,".":87,"INDEX_START":88,"INDEX_END":89,"?.":90,"::":91,"?::":92,"Index":93,"IndexValue":94,"INDEX_SOAK":95,"Slice":96,"{":97,"AssignList":98,"}":99,"CLASS":100,"EXTENDS":101,"IMPORT":102,"ImportDefaultSpecifier":103,"ImportNamespaceSpecifier":104,"ImportSpecifierList":105,"ImportSpecifier":106,"AS":107,"DEFAULT":108,"IMPORT_ALL":109,"EXPORT":110,"ExportSpecifierList":111,"EXPORT_ALL":112,"ExportSpecifier":113,"OptFuncExist":114,"Arguments":115,"FUNC_EXIST":116,"CALL_START":117,"CALL_END":118,"ArgList":119,"THIS":120,"@":121,"[":122,"]":123,"RangeDots":124,"..":125,"Arg":126,"SimpleArgs":127,"TRY":128,"Catch":129,"FINALLY":130,"CATCH":131,"THROW":132,"(":133,")":134,"WhileSource":135,"WHILE":136,"WHEN":137,"UNTIL":138,"Loop":139,"LOOP":140,"ForBody":141,"FOR":142,"BY":143,"ForStart":144,"ForSource":145,"ForVariables":146,"OWN":147,"ForValue":148,"FORIN":149,"FOROF":150,"FORFROM":151,"SWITCH":152,"Whens":153,"ELSE":154,"When":155,"LEADING_WHEN":156,"IfBlock":157,"IF":158,"POST_IF":159,"UNARY":160,"UNARY_MATH":161,"-":162,"+":163,"--":164,"++":165,"?":166,"MATH":167,"**":168,"SHIFT":169,"COMPARE":170,"&":171,"^":172,"|":173,"&&":174,"||":175,"BIN?":176,"RELATION":177,"COMPOUND_ASSIGN":178,"$accept":0,"$end":1},
terminals_: {2:"error",6:"TERMINATOR",14:"STATEMENT",30:"YIELD",31:"FROM",33:"INDENT",34:"OUTDENT",36:"IDENTIFIER",38:"PROPERTY",40:"NUMBER",42:"STRING",43:"STRING_START",44:"STRING_END",46:"REGEX",47:"REGEX_START",48:"REGEX_END",50:"JS",51:"UNDEFINED",52:"NULL",53:"BOOL",54:"INFINITY",55:"NAN",57:"=",60:":",63:"RETURN",64:"AWAIT",65:"HERECOMMENT",66:"PARAM_START",68:"PARAM_END",70:"->",71:"=>",73:",",76:"...",86:"SUPER",87:".",88:"INDEX_START",89:"INDEX_END",90:"?.",91:"::",92:"?::",95:"INDEX_SOAK",97:"{",99:"}",100:"CLASS",101:"EXTENDS",102:"IMPORT",107:"AS",108:"DEFAULT",109:"IMPORT_ALL",110:"EXPORT",112:"EXPORT_ALL",116:"FUNC_EXIST",117:"CALL_START",118:"CALL_END",120:"THIS",121:"@",122:"[",123:"]",125:"..",128:"TRY",130:"FINALLY",131:"CATCH",132:"THROW",133:"(",134:")",136:"WHILE",137:"WHEN",138:"UNTIL",140:"LOOP",142:"FOR",143:"BY",147:"OWN",149:"FORIN",150:"FOROF",151:"FORFROM",152:"SWITCH",154:"ELSE",156:"LEADING_WHEN",158:"IF",159:"POST_IF",160:"UNARY",161:"UNARY_MATH",162:"-",163:"+",164:"--",165:"++",166:"?",167:"MATH",168:"**",169:"SHIFT",170:"COMPARE",171:"&",172:"^",173:"|",174:"&&",175:"||",176:"BIN?",177:"RELATION",178:"COMPOUND_ASSIGN"},
productions_: [0,[3,0],[3,1],[4,1],[4,3],[4,2],[5,1],[5,1],[5,1],[9,1],[9,1],[8,1],[8,1],[8,1],[8,1],[8,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[7,1],[29,1],[29,2],[29,3],[32,2],[32,3],[35,1],[37,1],[39,1],[39,1],[41,1],[41,3],[45,1],[45,3],[49,1],[49,1],[49,1],[49,1],[49,1],[49,1],[49,1],[49,1],[21,3],[21,4],[21,5],[58,1],[58,3],[58,5],[58,3],[58,5],[58,1],[61,1],[61,1],[61,1],[59,1],[59,1],[12,2],[12,1],[10,3],[10,2],[11,3],[11,2],[13,1],[19,5],[19,2],[69,1],[69,1],[72,0],[72,1],[67,0],[67,1],[67,3],[67,4],[67,6],[74,1],[74,2],[74,3],[74,1],[75,1],[75,1],[75,1],[75,1],[79,2],[80,1],[80,2],[80,2],[80,1],[56,1],[56,1],[56,1],[17,1],[17,1],[17,1],[17,1],[17,1],[17,1],[85,3],[85,4],[81,2],[81,2],[81,2],[81,2],[81,1],[81,1],[93,3],[93,2],[94,1],[94,1],[78,4],[98,0],[98,1],[98,3],[98,4],[98,6],[27,1],[27,2],[27,3],[27,4],[27,2],[27,3],[27,4],[27,5],[15,2],[15,4],[15,4],[15,5],[15,7],[15,6],[15,9],[105,1],[105,3],[105,4],[105,4],[105,6],[106,1],[106,3],[106,1],[106,3],[103,1],[104,3],[16,3],[16,5],[16,2],[16,4],[16,5],[16,6],[16,3],[16,4],[16,7],[111,1],[111,3],[111,4],[111,4],[111,6],[113,1],[113,3],[113,3],[113,1],[18,3],[18,3],[18,3],[18,3],[114,0],[114,1],[115,2],[115,4],[84,1],[84,1],[62,2],[77,2],[77,4],[124,1],[124,1],[83,5],[96,3],[96,2],[96,2],[96,1],[119,1],[119,3],[119,4],[119,4],[119,6],[126,1],[126,1],[126,1],[127,1],[127,3],[23,2],[23,3],[23,4],[23,5],[129,3],[129,3],[129,2],[28,2],[82,3],[82,5],[135,2],[135,4],[135,2],[135,4],[24,2],[24,2],[24,2],[24,1],[139,2],[139,2],[25,2],[25,2],[25,2],[141,2],[141,4],[141,2],[144,2],[144,3],[148,1],[148,1],[148,1],[148,1],[146,1],[146,3],[145,2],[145,2],[145,4],[145,4],[145,4],[145,6],[145,6],[145,2],[145,4],[26,5],[26,7],[26,4],[26,6],[153,1],[153,2],[155,3],[155,4],[157,3],[157,5],[22,1],[22,3],[22,3],[22,3],[20,2],[20,2],[20,2],[20,2],[20,2],[20,2],[20,2],[20,2],[20,2],[20,2],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,3],[20,5],[20,4],[20,3]],
Fix #3671: Allow step in optimized range comprehensions Allow the `by c` part in `for [a..b] by c then`. Continue disallowing a `when d` part, since it makes no sense having a guard that isn't given access to anything that changes on every iteration.
2015-01-09 20:31:56 -05:00
performAction: function anonymous(yytext, yyleng, yylineno, yy, yystate /* action[1] */, $$ /* vstack */, _$ /* lstack */) {
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
/* this == yyval */
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
var $0 = $$.length - 1;
switch (yystate) {
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
case 1:
return this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Block);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
case 2:
return this.$ = $$[$0];
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
case 3:
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(yy.Block.wrap([$$[$0]]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
case 4:
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])($$[$0-2].push($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
case 5:
this.$ = $$[$0-1];
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 6: case 7: case 8: case 9: case 10: case 11: case 12: case 14: case 15: case 16: case 17: case 18: case 19: case 20: case 21: case 22: case 23: case 24: case 25: case 26: case 27: case 28: case 37: case 42: case 44: case 58: case 59: case 60: case 61: case 62: case 63: case 75: case 76: case 86: case 87: case 88: case 89: case 94: case 95: case 98: case 102: case 103: case 111: case 191: case 192: case 194: case 224: case 225: case 243: case 249:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = $$[$0];
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 13:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.StatementLiteral($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 29:
Add support for standalone yield This breaks compatibility with -> yield for i in [1..3] i * 2 and -> yield i * 2 yield's behaviour now mirrors that of return in that it can be used stand alone as well as with expressions. Thus, it currently also inherits the above limitations.
2015-09-13 06:30:59 -04:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Op($$[$0], new yy.Value(new yy.Literal(''))));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 30: case 253: case 254: case 257:
Add support for standalone yield This breaks compatibility with -> yield for i in [1..3] i * 2 and -> yield i * 2 yield's behaviour now mirrors that of return in that it can be used stand alone as well as with expressions. Thus, it currently also inherits the above limitations.
2015-09-13 06:30:59 -04:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Op($$[$0-1], $$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 31:
Add support for standalone yield This breaks compatibility with -> yield for i in [1..3] i * 2 and -> yield i * 2 yield's behaviour now mirrors that of return in that it can be used stand alone as well as with expressions. Thus, it currently also inherits the above limitations.
2015-09-13 06:30:59 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Op($$[$0-2].concat($$[$0-1]), $$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 32:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Block);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 33: case 112:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])($$[$0-1]);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 34:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.IdentifierLiteral($$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 35:
Split out properties from identifiers - Split out a PROPERTY token from the IDENTIFIER token. - Split out Property from the Identifier in the grammar. - Split out PropertyLiteral from IdentifierLiteral.
2016-03-05 11:41:15 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.PropertyName($$[$0]));
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 36:
Split out properties from identifiers - Split out a PROPERTY token from the IDENTIFIER token. - Split out Property from the Identifier in the grammar. - Split out PropertyLiteral from IdentifierLiteral.
2016-03-05 11:41:15 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.NumberLiteral($$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 38:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.StringLiteral($$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 39:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.StringWithInterpolations($$[$0-1]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 40:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.RegexLiteral($$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 41:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.RegexWithInterpolations($$[$0-1].args));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 43:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.PassthroughLiteral($$[$0]));
Fix #3597: Allow interpolations in object keys The following is now allowed: o = a: 1 b: 2 "#{'c'}": 3 "#{'d'}": 4 e: 5 "#{'f'}": 6 g: 7 It compiles to: o = ( obj = { a: 1, b: 2 }, obj["" + 'c'] = 3, obj["" + 'd'] = 4, obj.e = 5, obj["" + 'f'] = 6, obj.g = 7, obj ); - Closes #3039. Empty interpolations in object keys are now _supposed_ to be allowed. - Closes #1131. No need to improve error messages for attempted key interpolation anymore. - Implementing this required fixing the following bug: `("" + a): 1` used to error out on the colon, saying "unexpected colon". But really, it is the attempted object key that is unexpected. Now the error is on the opening parenthesis instead. - However, the above fix broke some error message tests for regexes. The easiest way to fix this was to make a seemingly unrelated change: The error messages for unexpected identifiers, numbers, strings and regexes now say for example 'unexpected string' instead of 'unexpected """some #{really long} string"""'. In other words, the tag _name_ is used instead of the tag _value_. This was way easier to implement, and is more helpful to the user. Using the tag value is good for operators, reserved words and the like, but not for tokens which can contain any text. For example, 'unexpected identifier' is better than 'unexpected expected' (if a variable called 'expected' was used erraneously). - While writing tests for the above point I found a few minor bugs with string locations which have been fixed.
2015-02-07 14:16:59 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 45:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.UndefinedLiteral);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 46:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.NullLiteral);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 47:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.BooleanLiteral($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 48:
Treat Infinity and NaN as reserved words Fixes #4218.
2016-03-05 15:32:20 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.InfinityLiteral($$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 49:
Treat Infinity and NaN as reserved words Fixes #4218.
2016-03-05 15:32:20 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.NaNLiteral);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 50:
Treat Infinity and NaN as reserved words Fixes #4218.
2016-03-05 15:32:20 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Assign($$[$0-2], $$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 51:
Treat Infinity and NaN as reserved words Fixes #4218.
2016-03-05 15:32:20 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Assign($$[$0-3], $$[$0]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 52:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Assign($$[$0-4], $$[$0-1]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 53: case 91: case 96: case 97: case 99: case 100: case 101: case 226: case 227:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Value($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 54:
Fix #3926: Disallow implicit objects as parameter destructuring
2015-08-28 17:11:47 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Assign(yy.addLocationDataFn(_$[$0-2])(new yy.Value($$[$0-2])), $$[$0], 'object', {
operatorToken: yy.addLocationDataFn(_$[$0-1])(new yy.Literal($$[$0-1]))
}));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 55:
Fix #3926: Disallow implicit objects as parameter destructuring
2015-08-28 17:11:47 -04:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Assign(yy.addLocationDataFn(_$[$0-4])(new yy.Value($$[$0-4])), $$[$0-1], 'object', {
operatorToken: yy.addLocationDataFn(_$[$0-3])(new yy.Literal($$[$0-3]))
}));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 56:
Implement ES2015-like destructuring defaults This let's you do things like: fullName = ({first = 'John', last = 'Doe'}) -> "#{first} #{last}" Note: CoffeeScrits treats `undefined` and `null` the same, and that's true in the case of destructuring defaults as well, as opposed to ES2015 which only uses the default value if the target is `undefined`. A similar ES2015 difference already exists for function parameter defaults. It is important for CoffeeScript to be consistent with itself. fullName2 = (first = 'John', last = 'Doe') -> "#{first} #{last}" assert fullName('Bob', null) is fullName2(first: 'Bob', last: null) Fixes #1558, #3288 and #4005.
2015-08-22 15:39:26 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Assign(yy.addLocationDataFn(_$[$0-2])(new yy.Value($$[$0-2])), $$[$0], null, {
operatorToken: yy.addLocationDataFn(_$[$0-1])(new yy.Literal($$[$0-1]))
}));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 57:
Implement ES2015-like destructuring defaults This let's you do things like: fullName = ({first = 'John', last = 'Doe'}) -> "#{first} #{last}" Note: CoffeeScrits treats `undefined` and `null` the same, and that's true in the case of destructuring defaults as well, as opposed to ES2015 which only uses the default value if the target is `undefined`. A similar ES2015 difference already exists for function parameter defaults. It is important for CoffeeScript to be consistent with itself. fullName2 = (first = 'John', last = 'Doe') -> "#{first} #{last}" assert fullName('Bob', null) is fullName2(first: 'Bob', last: null) Fixes #1558, #3288 and #4005.
2015-08-22 15:39:26 -04:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Assign(yy.addLocationDataFn(_$[$0-4])(new yy.Value($$[$0-4])), $$[$0-1], null, {
operatorToken: yy.addLocationDataFn(_$[$0-3])(new yy.Literal($$[$0-3]))
}));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 64:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Return($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 65:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Return);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 66:
Only allow `yield return` as a statement Fixes #4097. Also happens to fix #4096. I also took the liberty to simplify the error message for invalid use of `yield`.
2015-09-16 01:51:17 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.YieldReturn($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 67:
Only allow `yield return` as a statement Fixes #4097. Also happens to fix #4096. I also took the liberty to simplify the error message for invalid use of `yield`.
2015-09-16 01:51:17 -04:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.YieldReturn);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 68:
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.AwaitReturn($$[$0]));
break;
case 69:
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.AwaitReturn);
break;
case 70:
Only allow `yield return` as a statement Fixes #4097. Also happens to fix #4096. I also took the liberty to simplify the error message for invalid use of `yield`.
2015-09-16 01:51:17 -04:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Comment($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 71:
Only allow `yield return` as a statement Fixes #4097. Also happens to fix #4096. I also took the liberty to simplify the error message for invalid use of `yield`.
2015-09-16 01:51:17 -04:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Code($$[$0-3], $$[$0], $$[$0-1]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 72:
Only allow `yield return` as a statement Fixes #4097. Also happens to fix #4096. I also took the liberty to simplify the error message for invalid use of `yield`.
2015-09-16 01:51:17 -04:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Code([], $$[$0], $$[$0-1]));
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 73:
Only allow `yield return` as a statement Fixes #4097. Also happens to fix #4096. I also took the liberty to simplify the error message for invalid use of `yield`.
2015-09-16 01:51:17 -04:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])('func');
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 74:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])('boundfunc');
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 77: case 117:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])([]);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 78: case 118: case 137: case 157: case 186: case 228:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])([$$[$0]]);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 79: case 119: case 138: case 158: case 187:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])($$[$0-2].concat($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 80: case 120: case 139: case 159: case 188:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])($$[$0-3].concat($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 81: case 121: case 141: case 161: case 190:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-5], _$[$0])($$[$0-5].concat($$[$0-2]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 82:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Param($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 83:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Param($$[$0-1], null, true));
making debugger an expression.
2011-11-25 08:35:20 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 84:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Param($$[$0-2], $$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 85: case 193:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Expansion);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 90:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Splat($$[$0-1]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 92:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])($$[$0-1].add($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
Add Implicit Async Functions (#3757) * changed jison acceptable versions * added await support * wrong function bug fix * added tests for async/await * invalid to have await, yield(from) in same function * changed error handling and tests * bug fix * made error handling test more rigorous * consolidated harmony test files * added async constructor support and tests * removed .orig files * Fixed browser testing issue * Minor cleanup * Async test-suite and Cake support, simplified/removed funky tests * Skip async/await tests when not supported in runtime * cleanup * Replaced polyfill with native JS async/await * Oops * Make 'async' reserved word * Remove all async polyfills * fix merge conflict * make async testing opt-in * restore test, remove confusing polyfill language * Revert changes to test runners * Only run async tests where async/await is supported (Node 7+ with --harmony, for now) * remove 'async' from JS reserved words * The async tests should use their own special async-capable version of `global.test`, which is only loaded for the async tests and only loaded by async-capable environments * Reverting rename of `async`, it’s not a reserved word so there’s no longer a need for this change * async test refactoring and additions * oops * sync * better error reporting for `await` * more stuff geoffrey wants * fixed litcoffee tests * change test title
2016-11-02 11:51:26 -04:00
case 93:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Value($$[$0-1], [].concat($$[$0])));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 104:
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Super(yy.addLocationDataFn(_$[$0])(new yy.Access($$[$0]))));
break;
case 105:
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Super(yy.addLocationDataFn(_$[$0-1])(new yy.Index($$[$0-1]))));
break;
case 106:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Access($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 107:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Access($$[$0], 'soak'));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 108:
Split out properties from identifiers - Split out a PROPERTY token from the IDENTIFIER token. - Split out Property from the Identifier in the grammar. - Split out PropertyLiteral from IdentifierLiteral.
2016-03-05 11:41:15 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])([yy.addLocationDataFn(_$[$0-1])(new yy.Access(new yy.PropertyName('prototype'))), yy.addLocationDataFn(_$[$0])(new yy.Access($$[$0]))]);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 109:
Split out properties from identifiers - Split out a PROPERTY token from the IDENTIFIER token. - Split out Property from the Identifier in the grammar. - Split out PropertyLiteral from IdentifierLiteral.
2016-03-05 11:41:15 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])([yy.addLocationDataFn(_$[$0-1])(new yy.Access(new yy.PropertyName('prototype'), 'soak')), yy.addLocationDataFn(_$[$0])(new yy.Access($$[$0]))]);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 110:
Split out properties from identifiers - Split out a PROPERTY token from the IDENTIFIER token. - Split out Property from the Identifier in the grammar. - Split out PropertyLiteral from IdentifierLiteral.
2016-03-05 11:41:15 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Access(new yy.PropertyName('prototype')));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 113:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(yy.extend($$[$0], {
nodes: @soakNode -> @soak
2010-10-25 09:31:52 -04:00
soak: true
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}));
adding passed-through block comments back to the grammar/lexer/rewriter/nodes ... thanks, Trevor Burnham.
2010-07-01 21:26:33 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 114:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Index($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 115:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Slice($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 116:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Obj($$[$0-2], $$[$0-3].generated));
removing an unused rule in Root.
2010-11-13 18:13:09 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 122:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Class);
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 123:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Class(null, null, $$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 124:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Class(null, $$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 125:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Class(null, $$[$0-1], $$[$0]));
grammar: removed a remnant of Slice
2010-10-23 16:35:35 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 126:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Class($$[$0]));
Minor modifications to the grammar to allow a single trailing comma for function call arguments, array literal and object literals. Adjusted tests accordingly
2010-03-26 11:44:25 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 127:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Class($$[$0-1], null, $$[$0]));
removing the <- bind operator from CoffeeScript...
2010-06-12 11:00:56 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 128:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Class($$[$0-2], $$[$0]));
Minor modifications to the grammar to allow a single trailing comma for function call arguments, array literal and object literals. Adjusted tests accordingly
2010-03-26 11:44:25 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 129:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Class($$[$0-3], $$[$0-1], $$[$0]));
Minor modifications to the grammar to allow a single trailing comma for function call arguments, array literal and object literals. Adjusted tests accordingly
2010-03-26 11:44:25 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 130:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.ImportDeclaration(null, $$[$0]));
Minor modifications to the grammar to allow a single trailing comma for function call arguments, array literal and object literals. Adjusted tests accordingly
2010-03-26 11:44:25 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 131:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.ImportDeclaration(new yy.ImportClause($$[$0-2], null), $$[$0]));
first draft of range literals back...
2010-11-20 16:25:22 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 132:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.ImportDeclaration(new yy.ImportClause(null, $$[$0-2]), $$[$0]));
first draft of range literals back...
2010-11-20 16:25:22 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 133:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.ImportDeclaration(new yy.ImportClause(null, new yy.ImportSpecifierList([])), $$[$0]));
first draft of range literals back...
2010-11-20 16:25:22 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 134:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-6], _$[$0])(new yy.ImportDeclaration(new yy.ImportClause(null, new yy.ImportSpecifierList($$[$0-4])), $$[$0]));
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 135:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-5], _$[$0])(new yy.ImportDeclaration(new yy.ImportClause($$[$0-4], $$[$0-2]), $$[$0]));
first draft of range literals back...
2010-11-20 16:25:22 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 136:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-8], _$[$0])(new yy.ImportDeclaration(new yy.ImportClause($$[$0-7], new yy.ImportSpecifierList($$[$0-4])), $$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 140: case 160: case 173: case 189:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])($$[$0-2]);
Addressing some assignment issues.
2010-03-28 13:06:16 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 142:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.ImportSpecifier($$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 143:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.ImportSpecifier($$[$0-2], $$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 144:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.ImportSpecifier(new yy.Literal($$[$0])));
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 145:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.ImportSpecifier(new yy.Literal($$[$0-2]), $$[$0]));
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 146:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.ImportDefaultSpecifier($$[$0]));
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 147:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.ImportNamespaceSpecifier(new yy.Literal($$[$0-2]), $$[$0]));
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 148:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.ExportNamedDeclaration(new yy.ExportSpecifierList([])));
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 149:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.ExportNamedDeclaration(new yy.ExportSpecifierList($$[$0-2])));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 150:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.ExportNamedDeclaration($$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 151:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.ExportNamedDeclaration(new yy.Assign($$[$0-2], $$[$0], null, {
moduleDeclaration: 'export'
})));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 152:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.ExportNamedDeclaration(new yy.Assign($$[$0-3], $$[$0], null, {
moduleDeclaration: 'export'
})));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 153:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-5], _$[$0])(new yy.ExportNamedDeclaration(new yy.Assign($$[$0-4], $$[$0-1], null, {
moduleDeclaration: 'export'
})));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 154:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.ExportDefaultDeclaration($$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 155:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.ExportAllDeclaration(new yy.Literal($$[$0-2]), $$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 156:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-6], _$[$0])(new yy.ExportNamedDeclaration(new yy.ExportSpecifierList($$[$0-4]), $$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 162:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.ExportSpecifier($$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 163:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.ExportSpecifier($$[$0-2], $$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 164:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.ExportSpecifier($$[$0-2], new yy.Literal($$[$0])));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 165:
Allow imported module members to be named `default`
2016-12-04 21:44:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.ExportSpecifier(new yy.Literal($$[$0])));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 166:
CS1 tagged template literals (and CS2 interpolated strings as template literals) (#4352) * Add initial support for template literals with no interpolation * Change ‘unexpected string’ error message tests to use number not identifier prefix. Identifer prefixes are now valid as tagged template literals * Test tagged template literals for non-interpolated strings and tag function. * Tagged template literals work for pure Strings. Pull tagged template definition up to Invocation level in grammar, enabling chained invocation calls. We can view a tagged template is a special form of function call. * Readying for StringWithInterpolations work. * Tweaks. * Fix style * Pass StringWithInterpolations parameter straight into Call constructor. StringWithInterpolations will be output as template literal, so already in correct form for outputting tagged template literal. * Strip down compileNode for StringWithInterpolations * Done StringLiteral case for interpolated Strings * Remove need for TemplateLiteral * Simplify code. * Small code tidy * Interpolated strings now outputting as template literals. Still needs comprehensive testing. * Move error message tests into error_messages.coffee; remove test that is testing for a Node runtime error * Split up tests that were testing multiple things per test, so that each test tests only one thing * Edge cases: tagged template literals containing interpolated strings or even internal tagged template literals * Make more concise, more idiomatic style * Pull back extreme indentation * Restore and fix commented-out tests * Edge case: tagged template literal with empty string * Only use new ES2015 interpolated string syntax if we’re inside a tagged template literal; this keeps this PR safe to merge into CoffeeScript 1.x. Remove the code from this commit to make all interpolated strings use ES2015 syntax, for CoffeeScript 2. * Compiler now _doesn’t_ use template literals. * Expand tagged template literal tests * Move ‘Unexpected string’ error message tests into tagged template literal section. ‘Unexpected string’ is not reported in these test scenarios anymore. Instead, we error that the prefixing literal is not a function. * Don’t unwrap StringWithInterpolations. Saw bug with program consisting of “#{2}” not compiling with template literals. Root cause was that Block.compileNode was unwrapping interpolated string and so didn’t use compileNode logic at StringWithInterpolations level. * No need to bracket interpolated strings any more. When interpolated string looks like `hello ${2}`, no extract brackets are needed, as the `s mark the beginning and end. * Show html templating with tagged template literals * Multiline should match multiline * Comment out unnecessary `unwrap`, which is only needed for CoffeeScript 2 all-ES2015 syntax output
2016-11-18 13:25:03 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.TaggedTemplateCall($$[$0-2], $$[$0], $$[$0-1]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 167: case 168:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Call($$[$0-2], $$[$0], $$[$0-1]));
break;
Merge branch 'master' into 2 # Conflicts: # lib/coffee-script/lexer.js # lib/coffee-script/parser.js
2016-12-06 23:59:21 -05:00
case 169:
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.SuperCall(yy.addLocationDataFn(_$[$0-2])(new yy.Super), $$[$0], $$[$0-1]));
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
break;
Merge branch 'master' into 2 # Conflicts: # lib/coffee-script/lexer.js # lib/coffee-script/parser.js
2016-12-06 23:59:21 -05:00
case 170:
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(false);
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
break;
Merge branch 'master' into 2 # Conflicts: # lib/coffee-script/lexer.js # lib/coffee-script/parser.js
2016-12-06 23:59:21 -05:00
case 171:
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(true);
break;
case 172:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])([]);
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 174: case 175:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Value(new yy.ThisLiteral));
Addressing some assignment issues.
2010-03-28 13:06:16 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 176:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Value(yy.addLocationDataFn(_$[$0-1])(new yy.ThisLiteral), [yy.addLocationDataFn(_$[$0])(new yy.Access($$[$0]))], 'this'));
Restricted class and extends values to simple assignments.
2010-03-28 13:14:35 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 177:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Arr([]));
adding class methods to class definition syntax, using '@'
2010-03-29 21:43:12 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 178:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Arr($$[$0-2]));
first draft of CoffeeScript language extensions.
2010-03-09 23:04:16 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 179:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])('inclusive');
first draft of CoffeeScript language extensions.
2010-03-09 23:04:16 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 180:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])('exclusive');
allowing terminator/comment as a block, so that you can have comment-only blocks
2010-03-10 16:18:17 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 181:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Range($$[$0-3], $$[$0-1], $$[$0-2]));
Added the possibility to end an identifier with '::', refering directly to its prototype.
2010-03-16 17:39:36 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 182:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Range($$[$0-2], $$[$0], $$[$0-1]));
Added the '<-' currying operator, as heavily discussed in #251, along with tests.
2010-03-17 03:11:10 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 183:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Range($$[$0-1], null, $$[$0]));
Added the '<-' currying operator, as heavily discussed in #251, along with tests.
2010-03-17 03:11:10 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 184:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Range(null, $$[$0], $$[$0-1]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 185:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])(new yy.Range(null, null, $$[$0]));
adding nice support for fancy implicit hashes
2010-07-25 02:52:02 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 195:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])([].concat($$[$0-2], $$[$0]));
adding nice support for fancy implicit hashes
2010-07-25 02:52:02 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 196:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Try($$[$0]));
Adding anonymous class support (gets a temporary variable for a name.) Issue #548
2010-07-27 22:05:55 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 197:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Try($$[$0-1], $$[$0][0], $$[$0][1]));
Disallowing Splats outside of ParamLists and ArgLists ... where they belong. This is in anticipation of the next commit...
2010-08-18 21:27:10 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 198:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Try($$[$0-2], null, null, $$[$0]));
Fixes #1838
2011-12-14 18:31:20 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 199:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Try($$[$0-3], $$[$0-2][0], $$[$0-2][1], $$[$0]));
Issue #621. Added the ability to leave the start and end index off of arrays. list[5..] is now valid CoffeeScript, slicing to the end of the array.
2010-08-18 21:51:44 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 200:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])([$$[$0-1], $$[$0]]);
Added expansion to destructuring
2014-01-24 11:00:34 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 201:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])([yy.addLocationDataFn(_$[$0-1])(new yy.Value($$[$0-1])), $$[$0]]);
Issue #621. Added the ability to leave the start and end index off of arrays. list[5..] is now valid CoffeeScript, slicing to the end of the array.
2010-08-18 21:51:44 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 202:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])([null, $$[$0]]);
Issue #621. Added the ability to leave the start and end index off of arrays. list[5..] is now valid CoffeeScript, slicing to the end of the array.
2010-08-18 21:51:44 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 203:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Throw($$[$0]));
Issue #621. Added the ability to leave the start and end index off of arrays. list[5..] is now valid CoffeeScript, slicing to the end of the array.
2010-08-18 21:51:44 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 204:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Parens($$[$0-1]));
first draft of range literals back...
2010-11-20 16:25:22 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 205:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Parens($$[$0-2]));
first draft of range literals back...
2010-11-20 16:25:22 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 206:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.While($$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 207:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.While($$[$0-2], {
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
guard: $$[$0]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 208:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.While($$[$0], {
adding until loops as the inverse of while loops
2010-04-28 22:08:00 -04:00
invert: true
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}));
adding until loops as the inverse of while loops
2010-04-28 22:08:00 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 209:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.While($$[$0-2], {
adding until loops as the inverse of while loops
2010-04-28 22:08:00 -04:00
invert: true,
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
guard: $$[$0]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}));
eliminating the IndentedAssignList nonterminal.
2010-04-25 21:22:29 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 210:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])($$[$0-1].addBody($$[$0]));
fixing static class properties that are implicit objects.
2010-09-18 22:25:45 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 211: case 212:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])($$[$0].addBody(yy.addLocationDataFn(_$[$0-1])(yy.Block.wrap([$$[$0-1]]))));
made `catch` optional
2010-10-04 16:53:32 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 213:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0], _$[$0])($$[$0]);
Fixes #1838
2011-12-14 18:31:20 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 214:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.While(yy.addLocationDataFn(_$[$0-1])(new yy.BooleanLiteral('true'))).addBody($$[$0]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 215:
Refactor `Literal` into several subtypes Previously, the parser created `Literal` nodes for many things. This resulted in information loss. Instead of being able to check the node type, we had to use regexes to tell the different types of `Literal`s apart. That was a bit like parsing literals twice: Once in the lexer, and once (or more) in the compiler. It also caused problems, such as `` `this` `` and `this` being indistinguishable (fixes #2009). Instead returning `new Literal` in the grammar, subtypes of it are now returned instead, such as `NumberLiteral`, `StringLiteral` and `IdentifierLiteral`. `new Literal` by itself is only used to represent code chunks that fit no category. (While mentioning `NumberLiteral`, there's also `InfinityLiteral` now, which is a subtype of `NumberLiteral`.) `StringWithInterpolations` has been added as a subtype of `Parens`, and `RegexWithInterpolations` as a subtype of `Call`. This makes it easier for other programs to make use of CoffeeScript's "AST" (nodes). For example, it is now possible to distinguish between `"a #{b} c"` and `"a " + b + " c"`. Fixes #4192. `SuperCall` has been added as a subtype of `Call`. Note, though, that some information is still lost, especially in the lexer. For example, there is no way to distinguish a heredoc from a regular string, or a heregex without interpolations from a regular regex. Binary and octal number literals are indistinguishable from hexadecimal literals. After the new subtypes were added, they were taken advantage of, removing most regexes in nodes.coffee. `SIMPLENUM` (which matches non-hex integers) had to be kept, though, because such numbers need special handling in JavaScript (for example in `1..toString()`). An especially nice hack to get rid of was using `new String()` for the token value for reserved identifiers (to be able to set a property on them which could survive through the parser). Now it's a good old regular string. In range literals, slices, splices and for loop steps when number literals are involved, CoffeeScript can do some optimizations, such as precomputing the value of, say, `5 - 3` (outputting `2` instead of `5 - 3` literally). As a side bonus, this now also works with hexadecimal number literals, such as `0x02`. Finally, this also improves the output of `coffee --nodes`: # Before: $ bin/coffee -ne 'while true "#{a}" break' Block While Value Bool Block Value Parens Block Op + Value """" Value Parens Block Value "a" "break" # After: $ bin/coffee -ne 'while true "#{a}" break' Block While Value BooleanLiteral: true Block Value StringWithInterpolations Block Op + Value StringLiteral: "" Value Parens Block Value IdentifierLiteral: a StatementLiteral: break
2016-01-31 14:24:31 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.While(yy.addLocationDataFn(_$[$0-1])(new yy.BooleanLiteral('true'))).addBody(yy.addLocationDataFn(_$[$0])(yy.Block.wrap([$$[$0]]))));
Allowing both end-points of slices to be implicit.
2011-12-20 19:21:26 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 216: case 217:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.For($$[$0-1], $$[$0]));
Closes #971. Existensial slices now supported.
2011-01-22 01:56:14 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 218:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.For($$[$0], $$[$0-1]));
Fixes #2508 -- existential access of the prototype.
2013-03-04 04:07:47 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 219:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])({
Fixes #2508 -- existential access of the prototype.
2013-03-04 04:07:47 -05:00
source: yy.addLocationDataFn(_$[$0])(new yy.Value($$[$0]))
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
allowing Klass::['dynamic-property'] syntax. Issue #392
2010-05-31 22:32:43 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 220:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])({
Fix #3671: Allow step in optimized range comprehensions Allow the `by c` part in `for [a..b] by c then`. Continue disallowing a `when d` part, since it makes no sense having a guard that isn't given access to anything that changes on every iteration.
2015-01-09 20:31:56 -05:00
source: yy.addLocationDataFn(_$[$0-2])(new yy.Value($$[$0-2])),
step: $$[$0]
});
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 221:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])((function () {
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
$$[$0].own = $$[$0-1].own;
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
$$[$0].ownTag = $$[$0-1].ownTag;
Removing 'do', in favor of a trailing ->
2010-12-21 18:54:36 -05:00
$$[$0].name = $$[$0-1][0];
$$[$0].index = $$[$0-1][1];
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
return $$[$0];
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}()));
adding slightly optimized syntax for range comprehensions that could care less about the index variable. Issue #547
2010-07-27 22:38:38 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 222:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])($$[$0]);
adding slightly optimized syntax for range comprehensions that could care less about the index variable. Issue #547
2010-07-27 22:38:38 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 223:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])((function () {
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
$$[$0].own = true;
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
$$[$0].ownTag = yy.addLocationDataFn(_$[$0-1])(new yy.Literal($$[$0-1]));
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
return $$[$0];
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}()));
fixing static class properties that are implicit objects.
2010-09-18 22:25:45 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 229:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])([$$[$0-2], $$[$0]]);
Fixes #2580 -- Try/Catch is allowed to destructure the error object.
2012-10-23 17:57:22 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 230:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])({
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
source: $$[$0]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 231:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])({
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
source: $$[$0],
first draft of range literals back...
2010-11-20 16:25:22 -05:00
object: true
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
first draft of range literals back...
2010-11-20 16:25:22 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 232:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])({
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
source: $$[$0-2],
guard: $$[$0]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 233:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])({
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
source: $$[$0-2],
guard: $$[$0],
first draft of range literals back...
2010-11-20 16:25:22 -05:00
object: true
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
Removed unprocessed values on a for loop from the grammar.
2010-03-28 14:32:01 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 234:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])({
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
source: $$[$0-2],
step: $$[$0]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
Removed unprocessed values on a for loop from the grammar.
2010-03-28 14:32:01 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 235:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-5], _$[$0])({
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
source: $$[$0-4],
guard: $$[$0-2],
step: $$[$0]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 236:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-5], _$[$0])({
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
source: $$[$0-4],
step: $$[$0-2],
guard: $$[$0]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
});
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 237:
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])({
source: $$[$0],
from: true
});
Allowing for negative indices in slice literals.
2010-03-21 10:59:41 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 238:
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])({
source: $$[$0-2],
guard: $$[$0],
from: true
});
Allowing for negative indices in slice literals.
2010-03-21 10:59:41 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 239:
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Switch($$[$0-3], $$[$0-1]));
enabling condition-less switches.
2010-05-10 22:57:51 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 240:
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
this.$ = yy.addLocationDataFn(_$[$0-6], _$[$0])(new yy.Switch($$[$0-5], $$[$0-3], $$[$0-1]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 241:
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Switch(null, $$[$0-1]));
enabling condition-less switches.
2010-05-10 22:57:51 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 242:
Add a `for .. from ..` loop for generators, see #4306, #3832 (#4355) * Added support for for-from loop, see #3832 * for-from: remove extra newline and add support for ranges * for-from: tidy up the lexer * for-from: add support for patterns * for-from: fix bad alignment * for-from: add two more tests * for-from: fix test "for-from loops over generators" See explanation here: https://github.com/jashkenas/coffeescript/pull/4306#issuecomment-257066877 * for-from: delete leftover console.log * Refactor the big `if` block in the lexer to be as minimal a change from `master` as we can get away with * Cleanup to make more idiomatic, remove trailing whitespace, minor performance improvements * for-from: move code from one file to another * for-from: clean up whitespace * for-from: lexer bikeshedding * Move "own is not supported in for-from loops" test into error_messages.coffee; improve error message so that "own" is underlined * Revert unnecessary changes, to minimize the lines of code modified by this PR
2016-11-08 02:40:01 -05:00
this.$ = yy.addLocationDataFn(_$[$0-5], _$[$0])(new yy.Switch(null, $$[$0-3], $$[$0-1]));
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 244:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])($$[$0-1].concat($$[$0]));
Allowing for negative indices in slice literals.
2010-03-21 10:59:41 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 245:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])([[$$[$0-1], $$[$0]]]);
Allowing for negative indices in slice literals.
2010-03-21 10:59:41 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 246:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])([[$$[$0-2], $$[$0-1]]]);
Added Statement to the grammar.
2010-04-05 10:26:23 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 247:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.If($$[$0-1], $$[$0], {
Issue #958. Removing UNLESS tokens to make them just inverted IFs.
2010-12-20 22:50:49 -05:00
type: $$[$0-2]
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 248:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])($$[$0-4].addElse(yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.If($$[$0-1], $$[$0], {
Issue #958. Removing UNLESS tokens to make them just inverted IFs.
2010-12-20 22:50:49 -05:00
type: $$[$0-2]
Speed up `updateLocationDataIfMissing`. * Avoid excessive search for missing `locationData` * Fix `locationData` for `ELSE IF`.
2013-06-20 20:47:29 -04:00
}))));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 250:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])($$[$0-2].addElse($$[$0]));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 251: case 252:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.If($$[$0], yy.addLocationDataFn(_$[$0-2])(yy.Block.wrap([$$[$0-2]])), {
Issue #958. Removing UNLESS tokens to make them just inverted IFs.
2010-12-20 22:50:49 -05:00
type: $$[$0-1],
adding complete documentation for the grammar
2010-03-07 11:41:56 -05:00
statement: true
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 255:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Op('-', $$[$0]));
removing an unused rule in Root.
2010-11-13 18:13:09 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 256:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Op('+', $$[$0]));
Allowing for negative indices in slice literals.
2010-03-21 10:59:41 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 258:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Op('--', $$[$0]));
being stricter about body-less scoped loops.
2010-12-21 20:28:48 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 259:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Op('++', $$[$0]));
Fixes #1838
2011-12-14 18:31:20 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 260:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Op('--', $$[$0-1], null, true));
Allowing both end-points of slices to be implicit.
2011-12-20 19:21:26 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 261:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Op('++', $$[$0-1], null, true));
Fixes #2274 -- allow @variables as loop variables
2012-04-24 12:21:47 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 262:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-1], _$[$0])(new yy.Existence($$[$0-1]));
lexer/parser: split out `null` and `undefined` from `BOOL` token also made explicit AST nodes `Bool`, `Null`, and `Undefined`
2012-05-08 16:11:23 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 263:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Op('+', $$[$0-2], $$[$0]));
removed support for '->*" and '=>*'
2014-09-06 06:55:27 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 264:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Op('-', $$[$0-2], $$[$0]));
lexer/parser: split out `null` and `undefined` from `BOOL` token also made explicit AST nodes `Bool`, `Null`, and `Undefined`
2012-05-08 16:11:23 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 265: case 266: case 267: case 268: case 269: case 270: case 271: case 272: case 273: case 274:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Op($$[$0-1], $$[$0-2], $$[$0]));
Fixes #2580 -- Try/Catch is allowed to destructure the error object.
2012-10-23 17:57:22 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 275:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])((function () {
rebuilt parser with Jison 0.2.0
2010-12-19 20:04:49 -05:00
if ($$[$0-1].charAt(0) === '!') {
Revert #2599
2013-02-28 08:37:30 -05:00
return new yy.Op($$[$0-1].slice(1), $$[$0-2], $$[$0]).invert();
combination of satyr's three comment patches ... comments are now statements, not expressions...
2010-11-08 23:39:21 -05:00
} else {
Revert #2599
2013-02-28 08:37:30 -05:00
return new yy.Op($$[$0-1], $$[$0-2], $$[$0]);
combination of satyr's three comment patches ... comments are now statements, not expressions...
2010-11-08 23:39:21 -05:00
}
Fix merge problem, and rebuild parser.js
2013-02-21 20:17:06 -05:00
}()));
allowing empty parens.
2010-08-17 21:31:32 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 276:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Assign($$[$0-2], $$[$0], $$[$0-1]));
coco 98271e6: made 'extends' chainable
2010-11-09 00:26:31 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 277:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-4], _$[$0])(new yy.Assign($$[$0-4], $$[$0-1], $$[$0-3]));
Disallowing Splats outside of ParamLists and ArgLists ... where they belong. This is in anticipation of the next commit...
2010-08-18 21:27:10 -04:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 278:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-3], _$[$0])(new yy.Assign($$[$0-3], $$[$0], $$[$0-2]));
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
break;
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
case 279:
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
this.$ = yy.addLocationDataFn(_$[$0-2], _$[$0])(new yy.Extends($$[$0-2], $$[$0]));
Issue #633. 'not instanceof'
2010-08-21 08:56:25 -04:00
break;
Allowing for negative indices in slice literals.
2010-03-21 10:59:41 -04:00
}
},
[CS2] Compile all super calls to ES2015 super (#4424) * Compile all super calls to ES2015 super This breaks using `super` in non-methods, meaning several tests are failing. Self-compilation still works. * Use bound functions for IIFEs containing `super` `super` can only be called directly in a method, or in an arrow function. * Fix handling of `class @A extends A` This behaviour worked 'for free' when the parent reference was being cached by the executable class body wrapper. There now needs to be special handling in place to check if the parent name matches the class name, and if so to cache the parent reference. * Fix tests broken by compiling ES2015 `super` * Disallow bare super This removes syntax support for 'bare' super calls, e.g.: class B extends A constructor: -> super `super` must now always be followed with arguments like a regular function call. This also removes the capability of implicitly forwarding arguments. The above can be equivalently be written as: class B extends A constructor: -> super arguments... * Support super with accessors `super` with following accessor(s) is now compiled to ES2015 equivalents. In particular, expressions such as `super.name`, `super[name]`, and also `super.name.prop` are all now valid, and can be used as expected as calls (i.e. `super.name()`) or in expressions (i.e. `if super.name? ...`). `super` without accessors is compiled to a constructor super call in a constructor, and otherwise, as before, to a super call to the method of the same name, i.e. speak: -> super() ...is equivalent to speak: -> super.speak() A neat side-effect of the changes is that existential calls now work properly with super, meaning `super?()` will only call if the super property exists (and is a function). This is not valid for super in constructors. * Prevent calling `super` methods with `new` This fixes a bug in the previous super handling whereby using the `new` operator with a `super` call would silently drop the `new`. This is now an explicit compiler error, as it is invalid JS at runtime. * Clean up some old super handling code This was mostly code for tracking the source classes and variables for methods, which were needed to build the old lookups on `__super__`. * Add TODO to improve bare super parse error * Add some TODOs to improve some of the class tests
2017-02-04 15:03:17 -05:00
table: [{1:[2,1],3:1,4:2,5:3,7:4,8:5,9:6,10:25,11:26,12:20,13:21,14:$V0,15:23,16:24,17:7,18:8,19:9,20:10,21:11,22:12,23:13,24:14,25:15,26:16,27:17,28:18,29:19,30:$V1,35:73,36:$V2,39:59,40:$V3,41:83,42:$V4,43:$V5,45:61,46:$V6,47:$V7,49:28,50:$V8,51:$V9,52:$Va,53:$Vb,54:$Vc,55:$Vd,56:27,62:74,63:$Ve,64:$Vf,65:$Vg,66:$Vh,69:35,70:$Vi,71:$Vj,77:57,78:58,80:43,82:29,83:30,84:31,85:32,86:$Vk,97:$Vl,100:$Vm,102:$Vn,110:$Vo,120:$Vp,121:$Vq,122:$Vr,128:$Vs,132:$Vt,133:$Vu,135:46,136:$Vv,138:$Vw,139:47,140:$Vx,141:48,142:$Vy,144:80,152:$Vz,157:44,158:$VA,160:$VB,161:$VC,162:$VD,163:$VE,164:$VF,165:$VG},{1:[3]},{1:[2,2],6:$VH},o($VI,[2,3]),o($VI,[2,6],{144:80,135:105,141:106,136:$Vv,138:$Vw,142:$Vy,159:$VJ,162:$VK,163:$VL,166:$VM,167:$VN,168:$VO,169:$VP,170:$VQ,171:$VR,172:$VS,173:$VT,174:$VU,175:$VV,176:$VW,177:$VX}),o($VI,[2,7],{144:80,135:108,141:109,136:$Vv,138:$Vw,142:$Vy,159:$VY}),o($VI,[2,8]),o($VZ,[2,16],{114:110,81:111,93:117,42:$V_,43:$V_,117:$V_,87:$V$,88:$V01,90:$V11,91:$V21,92:$V31,95:$V41,116:$V51}),o($VZ,[2,17],{93:117,114:120,81:121,87:$V$,88:$V01,90:$V11,91:$V21,92:$V31,95:$V41,116:$V51,117:$V_}),o($VZ,[2,18]),o($VZ,[2,19]),o($VZ,[2,20]),o($VZ,[2,21]),o($VZ,[2,22]),o($VZ,[2,23]),o($VZ,[2,24]),o($VZ,[2,25]),o($VZ,[2,26]),o($VZ,[2,27]),o($VZ,[2,28]),o($V61,[2,11]),o($V61,[2,12]),o($V61,[2,13]),o($V61,[2,14]),o($V61,[2,15]),o($VI,[2,9]),o($VI,[2,10]),o($V71,$V81,{57:[1,122]}),o($V71,[2,99]),o($V71,[2,100]),o($V71,[2,101]),o($V71,[2,102]),o($V71,[2,103]),{87:[1,124],88:[1,125],114:123,116:$V51,117:$V_},o([6,33,68,73],$V91,{67:126,74:127,75:128,35:130,62:131,77:132,78:133,36:$V2,76:$Va1,97:$Vl,121:$Vb1,122:$Vc1}),{32:136,33:$Vd1},{7:138,8:140,12:20,13:21,14:$V0,15:23,16:24,17:7,18:8,19:9,20:10,21:11,22:12,23:13,24:14,25:15,26:16,27:17,28:18,29:19,30:$Ve1,35:73,36:$V2,39:59,40:$V3,41:83,42:$V4,43:$V5,45:61,46:$V6,47:$V7,49:28,50:$V8,51:$V9,52:$Va,53:$Vb,54:$Vc,55:$Vd,56:27,62:74,63:$Ve,64:$Vf1,65:$Vg,66:$Vh,69:35,70:$Vi,71:$Vj,77:57,78:58,80:43,82:29,83:30,84:31,85:32,86:$Vk,97:$Vl,100:$Vm,102:$Vn,110:$Vo,120:$Vp,121:$Vq,122:$Vr,128:$Vs,132:$Vt,133:$Vu,135:46,136:$Vv,138:$Vw,139:47,140:$Vx,141:48,142:$Vy,144:80,152:$Vz,157:44,158:$VA,160:$VB,161:$VC,162:$VD,163:$VE,164:$VF,165:$VG},{7:142,8:140,12:20,13:21,14:$V0,15:23,16:24,17:7,18:8,19:9,20:10,21:11,22:12,23:13,24:14,25:15,26:16,27:17,28:18,29:19,30:$Ve1,35:73,36:$V2,39:59,40:$V3,41:83,42:$V4,43:$V5,45:61,46:$V6,47:$V7,49:28,50:$V8,51:$V9,52:$Va,53:$Vb,54:$Vc,55:$Vd,56:27,62:74,63:$Ve,64:$Vf1,65:$Vg,66:$Vh,69:35,70:$Vi,71:$Vj,77:57,78:58,80:43,82:29,83:30,84:31,85:32,86:$Vk,97:$Vl,100:$Vm,102:$Vn,110:$Vo,120:$Vp,121:$Vq,122:$Vr,128:$Vs,132:$Vt,133:$Vu,135:46,136:$Vv,138:$Vw,139:47,140:$Vx,141:48,142:$Vy,144:80,152:$Vz,157:44,158:$VA,160:$VB,161:$VC,162:$VD,163:$VE,164:$VF,165:$VG},{7:143,8:140,12:20,13:21,14:$V0,15:23,16:24,17:7,18:8,19:9,20:10,21:11,22:12,23:13,24:14,25:15,26:16,27:17,28:18,29:19,30:$Ve1,35:73,36:$V2,39:59,40:$V3,41:83,42:$V4,43:$V5,45:61,46:$V6,47:$V7,49:28,50:$V8,51:$V9,52:$Va,53:$Vb,54:$Vc,55:$Vd,56:27,62:74,63:$Ve,64:$Vf1,65:$Vg,66:$Vh,69:35,70:$Vi,71:$Vj,77:57,78:58,80:43,82:29,83:30,84:31,85:32,86:$Vk,97:$Vl,100:$Vm,102:$Vn,110:$Vo,120:$Vp,121:$Vq,122:$Vr,128:$Vs,132:$Vt,133:$Vu,135:46,136:$Vv,138:$Vw,139:47,140:$Vx,141:48,142:$Vy,144:80,152:$Vz,157:44,158:$VA,160:$VB,161:$VC,162:$VD,163:$VE,164:$VF,165:$VG},{7:144,8:140,12:20,13:21,14:$V0,15:23,16:24,17:7,18:8,19:9,20:10,21:11,22:12,23:13,24:14,25:15,26:16,27:17,28:18,29:19,30:$Ve1,35:73,36:$V2,39:59,40:$V3,41:83,42:$V4,43:$V5,45:61,46:$V6,47:$V7,49:28,50:$V8,51:$V9,52:$Va,53:$Vb,54:$Vc,55:$Vd,56:27,62:74,63:$Ve,64:$Vf1,65:$Vg,66:$Vh,69:35,70:$Vi,71:$Vj,77:57,78:58,80:43,82:29,83:30,84:31,85:32,86:$Vk,97:$Vl,100:$Vm,102:$Vn,110:$Vo,120:$Vp,121:$Vq,122:$Vr,128:$Vs,132:$Vt,133:$Vu,135:46,136:$Vv,138:$Vw,139:47,140:$Vx,141:48,142:$Vy,144:80,152:$Vz,157:44,158:$VA,160:$VB,161:$VC,162:$VD,163:$VE,164:$VF,165:$VG},{7:145,8:140,12:20,13:21,14:$V0,15:23,16:24,17:7,18:8,19:9,20:10,21:11,22:12,23:13,24:14,25:15,26:16,27:17,28:18,29:19,30:$Ve1,35:73,36:$V2,39:59,40:$V3,41:83,42:$V4,43:$V5,45:61,46:$V6,47
defaultActions: {71:[2,73],72:[2,74],246:[2,115],367:[2,147]},
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
parseError: function parseError(str, hash) {
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
if (hash.recoverable) {
this.trace(str);
} else {
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
function _parseError (msg, hash) {
this.message = msg;
this.hash = hash;
}
_parseError.prototype = Error;
throw new _parseError(str, hash);
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
}
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
},
parse: function parse(input) {
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
var self = this, stack = [0], tstack = [], vstack = [null], lstack = [], table = this.table, yytext = '', yylineno = 0, yyleng = 0, recovering = 0, TERROR = 2, EOF = 1;
'->*' and '=>*' now produce generators
2013-12-19 21:08:25 -05:00
var args = lstack.slice.call(arguments, 1);
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
var lexer = Object.create(this.lexer);
var sharedState = { yy: {} };
for (var k in this.yy) {
if (Object.prototype.hasOwnProperty.call(this.yy, k)) {
sharedState.yy[k] = this.yy[k];
}
}
lexer.setInput(input, sharedState.yy);
sharedState.yy.lexer = lexer;
sharedState.yy.parser = this;
if (typeof lexer.yylloc == 'undefined') {
lexer.yylloc = {};
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
}
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
var yyloc = lexer.yylloc;
revised patch for #1234; consolidated Value#push and Value#concat into Value#add; removed unnecessary INDEX_PROTO
2011-08-12 13:38:34 -04:00
lstack.push(yyloc);
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
var ranges = lexer.options && lexer.options.ranges;
if (typeof sharedState.yy.parseError === 'function') {
this.parseError = sharedState.yy.parseError;
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
} else {
this.parseError = Object.getPrototypeOf(this).parseError;
}
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
function popStack(n) {
stack.length = stack.length - 2 * n;
new version of Jison vendored ... rebuild the grammar, tests pass.
2010-04-21 22:52:20 -04:00
vstack.length = vstack.length - n;
revised patch for #1234; consolidated Value#push and Value#concat into Value#add; removed unnecessary INDEX_PROTO
2011-08-12 13:38:34 -04:00
lstack.length = lstack.length - n;
new version of Jison vendored ... rebuild the grammar, tests pass.
2010-04-21 22:52:20 -04:00
}
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
_token_stack:
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
var lex = function () {
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
var token;
token = lexer.lex() || EOF;
if (typeof token !== 'number') {
token = self.symbols_[token] || token;
}
return token;
Support import and export of ES2015 modules (#4300) This pull request adds support for ES2015 modules, by recognizing `import` and `export` statements. The following syntaxes are supported, based on the MDN [import](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import) and [export](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/export) pages: ```js import "module-name" import defaultMember from "module-name" import * as name from "module-name" import { } from "module-name" import { member } from "module-name" import { member as alias } from "module-name" import { member1, member2 as alias2, … } from "module-name" import defaultMember, * as name from "module-name" import defaultMember, { … } from "module-name" export default expression export class name export { } export { name } export { name as exportedName } export { name as default } export { name1, name2 as exportedName2, name3 as default, … } export * from "module-name" export { … } from "module-name" ``` As a subsitute for ECMAScript’s `export var name = …` and `export function name {}`, CoffeeScript also supports: ```js export name = … ``` CoffeeScript also supports optional commas within `{ … }`. This PR converts the supported `import` and `export` statements into ES2015 `import` and `export` statements; it **does not resolve the modules**. So any CoffeeScript with `import` or `export` statements will be output as ES2015, and will need to be transpiled by another tool such as Babel before it can be used in a browser. We will need to add a warning to the documentation explaining this. This should be fully backwards-compatible, as `import` and `export` were previously reserved keywords. No flags are used. There are extensive tests included, though because no current JavaScript runtime supports `import` or `export`, the tests compare strings of what the compiled CoffeeScript output is against what the expected ES2015 should be. I also conducted two more elaborate tests: * I forked the [ember-piqu](https://github.com/pauc/piqu-ember) project, which was an Ember CLI app that used ember-cli-coffeescript and [ember-cli-coffees6](https://github.com/alexspeller/ember-cli-coffees6) (which adds “support” for `import`/`export` by wrapping such statements in backticks before passing the result to the CoffeeScript compiler). I removed `ember-cli-coffees6` and replaced the CoffeeScript compiler used in the build chain with this code, and the app built without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-piqu) * I also forked the [CoffeeScript version of Meteor’s Todos example app](https://github.com/meteor/todos/tree/coffeescript), and replaced all of its `require` statements with the `import` and `export` statements from the original ES2015 version of the app on its `master` branch. I then updated the `coffeescript` Meteor package in the app to use this new code, and again the app builds without errors. [Demo here.](https://github.com/GeoffreyBooth/coffeescript-modules-test-meteor-todos) The discussion history for this work started [here](https://github.com/jashkenas/coffeescript/pull/4160) and continued [here](https://github.com/GeoffreyBooth/coffeescript/pull/2). @lydell provided guidance, and @JimPanic and @rattrayalex contributed essential code.
2016-09-14 14:46:05 -04:00
};
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
var symbol, preErrorSymbol, state, action, a, r, yyval = {}, p, len, newState, expected;
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
while (true) {
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
state = stack[stack.length - 1];
ahh, finally. switched up our Jison configuration to provide better syntax errors. Now you get the unexpected token, instead of the expected one.
2010-06-01 20:21:12 -04:00
if (this.defaultActions[state]) {
action = this.defaultActions[state];
} else {
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
if (symbol === null || typeof symbol == 'undefined') {
ahh, finally. switched up our Jison configuration to provide better syntax errors. Now you get the unexpected token, instead of the expected one.
2010-06-01 20:21:12 -04:00
symbol = lex();
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
}
ahh, finally. switched up our Jison configuration to provide better syntax errors. Now you get the unexpected token, instead of the expected one.
2010-06-01 20:21:12 -04:00
action = table[state] && table[state][symbol];
}
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
if (typeof action === 'undefined' || !action.length || !action[0]) {
var errStr = '';
new version of Jison vendored ... rebuild the grammar, tests pass.
2010-04-21 22:52:20 -04:00
expected = [];
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
for (p in table[state]) {
if (this.terminals_[p] && p > TERROR) {
expected.push('\'' + this.terminals_[p] + '\'');
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
}
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
}
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
if (lexer.showPosition) {
errStr = 'Parse error on line ' + (yylineno + 1) + ':\n' + lexer.showPosition() + '\nExpecting ' + expected.join(', ') + ', got \'' + (this.terminals_[symbol] || symbol) + '\'';
Removing the last Jison hack. (Jison 0.1.25)
2010-10-28 01:39:35 -04:00
} else {
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
errStr = 'Parse error on line ' + (yylineno + 1) + ': Unexpected ' + (symbol == EOF ? 'end of input' : '\'' + (this.terminals_[symbol] || symbol) + '\'');
new version of Jison vendored ... rebuild the grammar, tests pass.
2010-04-21 22:52:20 -04:00
}
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
this.parseError(errStr, {
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
text: lexer.match,
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
token: this.terminals_[symbol] || symbol,
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
line: lexer.yylineno,
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
loc: yyloc,
expected: expected
});
new version of Jison vendored ... rebuild the grammar, tests pass.
2010-04-21 22:52:20 -04:00
}
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
if (action[0] instanceof Array && action.length > 1) {
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
throw new Error('Parse Error: multiple actions possible at state: ' + state + ', token: ' + symbol);
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
}
switch (action[0]) {
case 1:
stack.push(symbol);
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
vstack.push(lexer.yytext);
lstack.push(lexer.yylloc);
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
stack.push(action[1]);
symbol = null;
if (!preErrorSymbol) {
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
yyleng = lexer.yyleng;
yytext = lexer.yytext;
yylineno = lexer.yylineno;
yyloc = lexer.yylloc;
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
if (recovering > 0) {
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
recovering--;
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
}
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
} else {
symbol = preErrorSymbol;
preErrorSymbol = null;
updating Jison
2010-02-13 18:19:59 -05:00
}
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
break;
case 2:
len = this.productions_[action[1]][1];
yyval.$ = vstack[vstack.length - len];
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
yyval._$ = {
first_line: lstack[lstack.length - (len || 1)].first_line,
last_line: lstack[lstack.length - 1].last_line,
first_column: lstack[lstack.length - (len || 1)].first_column,
last_column: lstack[lstack.length - 1].last_column
};
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
if (ranges) {
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
yyval._$.range = [
lstack[lstack.length - (len || 1)].range[0],
lstack[lstack.length - 1].range[1]
];
Allowing both end-points of slices to be implicit.
2011-12-20 19:21:26 -05:00
}
'->*' and '=>*' now produce generators
2013-12-19 21:08:25 -05:00
r = this.performAction.apply(yyval, [
yytext,
yyleng,
yylineno,
Make CoffeeScript work with jison 0.4.14+ Since zaach/jison commit 3548861b, `parser.lexer` is never modified anymore (a copy of it is made, and that copy is modified instead). CoffeeScript itself modifies `parser.lexer` and then accesses those modifications in the custom `parser.yy.parseError` function, but that of course does not work anymore. This commit puts the data that `parser.yy.parseError` needs directly on the `parser` so that it is not lost. Supersedes #3603. Fixes #3608 and zaach/jison#243.
2015-01-15 11:47:07 -05:00
sharedState.yy,
'->*' and '=>*' now produce generators
2013-12-19 21:08:25 -05:00
action[1],
vstack,
lstack
].concat(args));
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
if (typeof r !== 'undefined') {
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
return r;
}
if (len) {
stack = stack.slice(0, -1 * len * 2);
vstack = vstack.slice(0, -1 * len);
lstack = lstack.slice(0, -1 * len);
}
stack.push(this.productions_[action[1]][0]);
vstack.push(yyval.$);
lstack.push(yyval._$);
newState = table[stack[stack.length - 2]][stack[stack.length - 1]];
stack.push(newState);
break;
case 3:
return true;
Fixes #2274 -- allow @variables as loop variables
2012-04-24 12:21:47 -04:00
}
the parser seems too big to compile
2010-01-30 23:17:36 -05:00
}
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
return true;
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
}};
'->*' and '=>*' now produce generators
2013-12-19 21:08:25 -05:00
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
function Parser () {
this.yy = {};
}
Parser.prototype = parser;parser.Parser = Parser;
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
return new Parser;
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
})();
Better handling of initial indent at file start. * Detect initial indentation before the first token and enforce it. * Don't add `INDENT` token (or the matching `OUTDENT, TERMINATOR`).
2013-06-13 18:21:47 -04:00
revised patch for #1234; consolidated Value#push and Value#concat into Value#add; removed unnecessary INDEX_PROTO
2011-08-12 13:38:34 -04:00
if (typeof require !== 'undefined' && typeof exports !== 'undefined') {
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
exports.parser = parser;
Parse compound assignment followed by a terminator. Closes #2532.
2012-11-27 20:11:01 -05:00
exports.Parser = parser.Parser;
Revert #2599
2013-02-28 08:37:30 -05:00
exports.parse = function () { return parser.parse.apply(parser, arguments); };
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
exports.main = function commonjsMain(args) {
Revert #2599
2013-02-28 08:37:30 -05:00
if (!args[1]) {
console.log('Usage: '+args[0]+' FILE');
process.exit(1);
removing our vendor'd copy of Jison ... you should install it from NPM now.
2010-06-28 00:26:45 -04:00
}
Merge branch 'master' of github.com:jashkenas/coffeescript into 2 # Conflicts: # lib/coffee-script/nodes.js # test/error_messages.coffee
2017-01-22 19:32:38 -05:00
var source = '';
Fix stack trace (#4428) * Revert aee27fbff03870c5479c6c33e6b1f1a32219420c * Patch Jison’s output so that it requires `fs` only if we’re truly in a CommonJS/Node environment, not a browser environment that may happen to have globals named `require` and `exports` (as would be the case if require.js is being used). Fixes #4391. * Temporary fix for exceptions getting thrown when trying to generate a stack trace for a file that has been deleted since compilation; fixes #3890, but not well. A better solution would not try to recompile the file when trying to retrieve its stack trace. * Save the test REPL history in the system temp folder, not in the CoffeeScript project folder * Rewrite `getSourceMap` to never read a file from disk, and therefore not throw IO-related exceptions; source maps are either retrieved from memory, or the related source code is retrieved from memory to generate a new source map. Fixes #3890 the proper way. * Add test to verify that stack traces reference the correct line number. Closes #4418. * Get the parser working in the browser compiler again; rather than detecting a CommonJS environment generally, just check for `fs` before trying to use it * Follow Node’s standard of 4-space indentation of stack trace data * Better .gitignore * Fix caching of compiled code and source maps; add more tests to verify correct line numbers in stack traces * Better fallback value for the parser source * Fix the stack traces and tests when running in a browser * Update the browser compiler so that @murrayju doesn’t have any extra work to do to test this branch
2017-01-22 16:20:18 -05:00
var fs = require('fs');
if (typeof fs !== 'undefined' && fs !== null)
source = fs.readFileSync(require('path').normalize(args[1]), "utf8");
removing our vendor'd copy of Jison ... you should install it from NPM now.
2010-06-28 00:26:45 -04:00
return exports.parser.parse(source);
Revert #2599
2013-02-28 08:37:30 -05:00
};
nagated relational operators are now grouped into NOT_RELATED, fixing #720
2010-10-05 11:43:44 -04:00
if (typeof module !== 'undefined' && require.main === module) {
Revert #2599
2013-02-28 08:37:30 -05:00
exports.main(process.argv.slice(1));
It's beginning to come alive.
2010-02-10 23:24:05 -05:00
}
Fixes #2359 -- tweak grammar to use new name
2013-02-13 22:34:32 -05:00
}
Copy permalink