kotovalexarian-likes-github/ruby--ruby
1
0
Fork 0
mirror of https://github.com/ruby/ruby.git synced 2022-11-09 12:17:21 -05:00
Code Releases Activity

* parse.y: patch, based on Nobu's, work to support \u escapes

Browse source 
also modifications for better coderange detection
* test/ruby/test_unicode_escapes.rb: test cases
* test/ruby/test_mixed_unicode_escapes.rb: mixed encoding test cases



git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@13836 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This commit is contained in:
davidflanagan 2007-11-07 22:55:27 +00:00
parent 89397d2f6e
commit baae1c4075
4 changed files with 605 additions and 84 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
ChangeLog
View file

@ -1,3 +1,9 @@
Thu Nov 8 07:54:22 UTC 2007 David Flanagan <davidflanagan@ruby-lang.org>
* parse.y: patch, based on Nobu's, work to support \u escapes
also modifications for better coderange detection
* test/ruby/test_unicode_escapes.rb: test cases
* test/ruby/test_mixed_unicode_escapes.rb: mixed encoding test cases
Thu Nov 8 07:14:37 UTC 2007 David Flanagan <davidflanagan@ruby-lang.org> Thu Nov 8 07:14:37 UTC 2007 David Flanagan <davidflanagan@ruby-lang.org>
* parse.y (rb_intern3): commented out broken code that prevented * parse.y (rb_intern3): commented out broken code that prevented
correct interning of multi-byte symbols. Without this patch correct interning of multi-byte symbols. Without this patch

418
parse.y
View file

@ -237,6 +237,7 @@ struct parser_params {
int has_shebang; int has_shebang;
int parser_ruby_sourceline; /* current line no. */ int parser_ruby_sourceline; /* current line no. */
rb_encoding *enc; rb_encoding *enc;
rb_encoding *utf8;
#ifndef RIPPER #ifndef RIPPER
/* Ruby core only */ /* Ruby core only */
@ -260,10 +261,12 @@ struct parser_params {
#endif #endif
}; };
#define UTF8_ENC() (parser->utf8 ? parser->utf8 : \
(parser->utf8 = rb_enc_find("utf-8")))
#define STR_NEW(p,n) rb_enc_str_new((p),(n),parser->enc) #define STR_NEW(p,n) rb_enc_str_new((p),(n),parser->enc)
#define STR_NEW0() rb_str_new(0,0) #define STR_NEW0() rb_str_new(0,0)
#define STR_NEW2(p) rb_enc_str_new((p),strlen(p),parser->enc) #define STR_NEW2(p) rb_enc_str_new((p),strlen(p),parser->enc)
#define STR_NEW3(p,n,m) parser_str_new((p),(n),STR_ENC(!ENC_SINGLE(m)),(m)) #define STR_NEW3(p,n,e,has8,hasmb) parser_str_new2((p),(n),(e),(has8),(hasmb))
#define STR_ENC(m) ((m)?parser->enc:rb_enc_from_index(0)) #define STR_ENC(m) ((m)?parser->enc:rb_enc_from_index(0))
#define ENC_SINGLE(cr) ((cr)==ENC_CODERANGE_SINGLE) #define ENC_SINGLE(cr) ((cr)==ENC_CODERANGE_SINGLE)
#define TOK_INTERN(mb) rb_intern3(tok(), toklen(), STR_ENC(mb)) #define TOK_INTERN(mb) rb_intern3(tok(), toklen(), STR_ENC(mb))
@ -4493,20 +4496,25 @@ none : /* none */
# define yylval (*((YYSTYPE*)(parser->parser_yylval))) # define yylval (*((YYSTYPE*)(parser->parser_yylval)))
static int parser_regx_options(struct parser_params*); static int parser_regx_options(struct parser_params*);
static int parser_tokadd_string(struct parser_params*,int,int,int,long*,int*); static int parser_tokadd_string(struct parser_params*,int,int,int,long*,int*,int*,rb_encoding**);
static void parser_tokaddmbc(struct parser_params *parser, int c, rb_encoding *enc);
static int parser_parse_string(struct parser_params*,NODE*); static int parser_parse_string(struct parser_params*,NODE*);
static int parser_here_document(struct parser_params*,NODE*); static int parser_here_document(struct parser_params*,NODE*);
# define nextc() parser_nextc(parser) # define nextc() parser_nextc(parser)
# define pushback(c) parser_pushback(parser, c) # define pushback(c) parser_pushback(parser, c)
# define newtok() parser_newtok(parser) # define newtok() parser_newtok(parser)
# define tokspace(n) parser_tokspace(parser, n) # define tokspace(n) parser_tokspace(parser, n)
# define tokadd(c) parser_tokadd(parser, c) # define tokadd(c) parser_tokadd(parser, c)
# define read_escape(m) parser_read_escape(parser, m) # define tok_hex(numlen) parser_tok_hex(parser, numlen)
# define tokadd_escape(t,m) parser_tokadd_escape(parser, t, m) # define tok_utf8(numlen,e) parser_tok_utf8(parser, numlen, e)
# define read_escape(flags,has8,hasmb,e) parser_read_escape(parser, flags, has8, hasmb, e)
# define tokadd_escape(t,has8,hasmb,e) parser_tokadd_escape(parser, t, has8,hasmb, e)
# define regx_options() parser_regx_options(parser) # define regx_options() parser_regx_options(parser)
# define tokadd_string(f,t,p,n,m) parser_tokadd_string(parser,f,t,p,n,m) # define tokadd_string(f,t,p,n,has8bit,hasmb,e) parser_tokadd_string(parser,f,t,p,n,has8bit,hasmb,e)
# define parse_string(n) parser_parse_string(parser,n) # define parse_string(n) parser_parse_string(parser,n)
# define tokaddmbc(c, enc) parser_tokaddmbc(parser, c, enc)
# define here_document(n) parser_here_document(parser,n) # define here_document(n) parser_here_document(parser,n)
# define heredoc_identifier() parser_heredoc_identifier(parser) # define heredoc_identifier() parser_heredoc_identifier(parser)
# define heredoc_restore(n) parser_heredoc_restore(parser,n) # define heredoc_restore(n) parser_heredoc_restore(parser,n)
@ -4829,6 +4837,15 @@ parser_str_new(const char *p, long n, rb_encoding *enc, int coderange)
return str; return str;
} }
static VALUE
parser_str_new2(const char *p, long n, rb_encoding *enc, int has8bit,int hasmb)
{
int coderange = ENC_CODERANGE_SINGLE;
if (hasmb) coderange = ENC_CODERANGE_MULTI;
else if (has8bit) coderange = ENC_CODERANGE_UNKNOWN;
return parser_str_new(p, n, enc, coderange);
}
static inline int static inline int
parser_nextc(struct parser_params *parser) parser_nextc(struct parser_params *parser)
{ {
@ -4943,10 +4960,145 @@ parser_tokadd(struct parser_params *parser, int c)
} }
static int static int
parser_read_escape(struct parser_params *parser, int *mb) parser_tok_hex(struct parser_params *parser, int *numlen)
{ {
int c; int c;
c = scan_hex(lex_p, 2, numlen);
if (!*numlen) {
yyerror("invalid hex escape");
return 0;
}
lex_p += *numlen;
return c;
}
#if 0
static int
parser_tok_utf8(struct parser_params *parser, int *numlen, rb_encoding **encp)
{
int codepoint;
if (peek('{')) { /* handle \u{...} form */
nextc();
codepoint = scan_hex(lex_p, 6, numlen);
if (*numlen == 0) {
yyerror("invalid Unicode escape");
return 0;
}
if (codepoint > 0x10ffff) {
yyerror("illegal Unicode codepoint (too large)");
return 0;
}
lex_p += *numlen;
if (!peek('}')) {
yyerror("unterminated Unicode escape");
return 0;
}
nextc();
}
else { /* handle \uxxxx form */
codepoint = scan_hex(lex_p, 4, numlen);
if (*numlen < 4) {
yyerror("invalid Unicode escape");
return 0;
}
lex_p += 4;
}
if (codepoint >= 0x80) {
*encp = UTF8_ENC();
}
return codepoint;
}
#endif
static int
parser_tokadd_utf8(struct parser_params *parser, int *hasmb,
rb_encoding **encp, int string_literal, int symbol_literal)
{
/*
* If string_literal is true, then we allow multiple codepoints
* in \u{}, and add the codepoints to the current token.
* Otherwise we're parsing a character literal and return a single
* codepoint without adding it
*/
int codepoint;
int numlen;
if (peek('{')) { /* handle \u{...} form */
do {
nextc();
codepoint = scan_hex(lex_p, 6, &numlen);
if (numlen == 0) {
yyerror("invalid Unicode escape");
return 0;
}
if (codepoint > 0x10ffff) {
yyerror("illegal Unicode codepoint (too large)");
return 0;
}
lex_p += numlen;
if (codepoint >= 0x80) {
*hasmb = 1;
*encp = UTF8_ENC();
if (string_literal) tokaddmbc(codepoint, *encp);
}
else if (string_literal) {
if (codepoint == 0 && symbol_literal) {
yyerror("symbol cannot contain '\\u{0}'");
return 0;
}
tokadd(codepoint);
}
} while(string_literal && (peek(' ') || peek('\t')));
if (!peek('}')) {
yyerror("unterminated Unicode escape");
return 0;
}
nextc();
}
else { /* handle \uxxxx form */
codepoint = scan_hex(lex_p, 4, &numlen);
if (numlen < 4) {
yyerror("invalid Unicode escape");
return 0;
}
lex_p += 4;
if (codepoint >= 0x80) {
*hasmb = 1;
*encp = UTF8_ENC();
if (string_literal) tokaddmbc(codepoint, *encp);
}
else if (string_literal) {
if (codepoint == 0 && symbol_literal) {
yyerror("symbol cannot contain '\\u0000'");
return 0;
}
tokadd(codepoint);
}
}
return codepoint;
}
#define ESCAPE_CONTROL 1
#define ESCAPE_META 2
static int
parser_read_escape(struct parser_params *parser, int flags,
int *has8bit, int *hasmb, rb_encoding **encp)
{
int c;
int numlen;
switch (c = nextc()) { switch (c = nextc()) {
case '\\': /* Backslash */ case '\\': /* Backslash */
return c; return c;
@ -4974,6 +5126,7 @@ parser_read_escape(struct parser_params *parser, int *mb)
case '0': case '1': case '2': case '3': /* octal constant */ case '0': case '1': case '2': case '3': /* octal constant */
case '4': case '5': case '6': case '7': case '4': case '5': case '6': case '7':
if (flags & (ESCAPE_CONTROL|ESCAPE_META)) goto eof;
{ {
int numlen; int numlen;
@ -4981,21 +5134,19 @@ parser_read_escape(struct parser_params *parser, int *mb)
c = scan_oct(lex_p, 3, &numlen); c = scan_oct(lex_p, 3, &numlen);
lex_p += numlen; lex_p += numlen;
} }
if (mb && (c >= 0200)) *mb = ENC_CODERANGE_UNKNOWN; if (c >= 0200) *has8bit = 1;
return c; return c;
case 'x': /* hex constant */ case 'x': /* hex constant */
{ if (flags & (ESCAPE_CONTROL|ESCAPE_META)) goto eof;
int numlen; c = tok_hex(&numlen);
if (numlen == 0) return 0;
if (c >= 0x80) *has8bit = 1;
return c;
c = scan_hex(lex_p, 2, &numlen); case 'u': /* unicode constant: here only for char literal */
if (numlen == 0) { if (flags & (ESCAPE_CONTROL|ESCAPE_META)) goto eof;
yyerror("Invalid escape character syntax"); c = parser_tokadd_utf8(parser, hasmb, encp, 0, 0);
return 0;
}
lex_p += numlen;
}
if (mb && (c >= 0x80)) *mb = ENC_CODERANGE_UNKNOWN;
return c; return c;
case 'b': /* backspace */ case 'b': /* backspace */
@ -5005,30 +5156,32 @@ parser_read_escape(struct parser_params *parser, int *mb)
return ' '; return ' ';
case 'M': case 'M':
if (flags & ESCAPE_META) goto eof;
if ((c = nextc()) != '-') { if ((c = nextc()) != '-') {
yyerror("Invalid escape character syntax");
pushback(c); pushback(c);
return '\0'; goto eof;
} }
if ((c = nextc()) == '\\') { if ((c = nextc()) == '\\') {
if (mb) *mb = ENC_CODERANGE_UNKNOWN; *has8bit = 1;
return read_escape(0) | 0x80; int tmp;
return read_escape(flags|ESCAPE_META, &tmp, &tmp, encp) | 0x80;
} }
else if (c == -1) goto eof; else if (c == -1) goto eof;
else { else {
if (mb) *mb = ENC_CODERANGE_UNKNOWN; *has8bit = 1;
return ((c & 0xff) | 0x80); return ((c & 0xff) | 0x80);
} }
case 'C': case 'C':
if ((c = nextc()) != '-') { if ((c = nextc()) != '-') {
yyerror("Invalid escape character syntax");
pushback(c); pushback(c);
return '\0'; goto eof;
} }
case 'c': case 'c':
if (flags & ESCAPE_CONTROL) goto eof;
if ((c = nextc())== '\\') { if ((c = nextc())== '\\') {
c = read_escape(mb); int tmp;
c = read_escape(flags|ESCAPE_CONTROL, &tmp, &tmp, encp);
} }
else if (c == '?') else if (c == '?')
return 0177; return 0177;
@ -5045,76 +5198,98 @@ parser_read_escape(struct parser_params *parser, int *mb)
} }
} }
#define tokcopy(n) memcpy(tokspace(n), lex_p - (n), (n))
static void
parser_tokaddmbc(struct parser_params *parser, int c, rb_encoding *enc)
{
int len = rb_enc_codelen(c, enc);
rb_enc_mbcput(c, tokspace(len), enc);
}
static int static int
parser_tokadd_escape(struct parser_params *parser, int term, int *mb) parser_tokadd_escape(struct parser_params *parser, int term,
int *has8bit, int *hasmb, rb_encoding **encp)
{ {
int c; int c;
int flags = 0;
first:
switch (c = nextc()) { switch (c = nextc()) {
case '\n': case '\n':
return 0; /* just ignore */ return 0; /* just ignore */
case '0': case '1': case '2': case '3': /* octal constant */ case '0': case '1': case '2': case '3': /* octal constant */
case '4': case '5': case '6': case '7': case '4': case '5': case '6': case '7':
if (flags & (ESCAPE_CONTROL|ESCAPE_META)) goto eof;
{ {
int numlen; int numlen;
int oct; int oct;
tokadd('\\'); oct = scan_oct(--lex_p, 3, &numlen);
pushback(c); if (numlen == 0) goto eof;
oct = scan_oct(lex_p, 3, &numlen); lex_p += numlen;
if (numlen == 0) { tokcopy(numlen + 1);
yyerror("Invalid escape character syntax"); if (oct >= 0200) *has8bit = 1;
return -1;
}
while (numlen--)
tokadd(nextc());
if (mb && (oct >= 0200)) *mb = ENC_CODERANGE_UNKNOWN;
} }
return 0; return 0;
case 'x': /* hex constant */ case 'x': /* hex constant */
if (flags & (ESCAPE_CONTROL|ESCAPE_META)) goto eof;
{ {
int numlen; int numlen;
int hex; int hex;
tokadd('\\'); hex = tok_hex(&numlen);
tokadd(c); if (numlen == 0) goto eof;
hex = scan_hex(lex_p, 2, &numlen); lex_p += numlen;
if (numlen == 0) { tokcopy(numlen + 2);
yyerror("Invalid escape character syntax"); if (hex >= 0x80) *has8bit = ENC_CODERANGE_UNKNOWN;
return -1;
}
while (numlen--)
tokadd(nextc());
if (mb && (hex >= 0x80)) *mb = ENC_CODERANGE_UNKNOWN;
} }
return 0; return 0;
case 'M': #if 0
if ((c = nextc()) != '-') { case 'u': /* Unicode constant */
yyerror("Invalid escape character syntax"); if (flags & (ESCAPE_CONTROL|ESCAPE_META)) goto eof;
pushback(c); {
return 0; int numlen;
int uc;
uc = tok_utf8(&numlen, encp);
if (numlen == 0) goto eof;
tokaddmbc(uc, *encp);
if (uc >= 0x80) *hasmb = 1;
} }
tokadd('\\'); tokadd('M'); tokadd('-'); return 0;
if (mb) *mb = ENC_CODERANGE_UNKNOWN; #endif
case 'M':
if (flags & ESCAPE_META) goto eof;
if ((c = nextc()) != '-') {
pushback(c);
goto eof;
}
tokcopy(3);
*has8bit = 1;
flags |= ESCAPE_META;
goto escaped; goto escaped;
case 'C': case 'C':
if (flags & ESCAPE_CONTROL) goto eof;
if ((c = nextc()) != '-') { if ((c = nextc()) != '-') {
yyerror("Invalid escape character syntax");
pushback(c); pushback(c);
return 0; goto eof;
} }
tokadd('\\'); tokadd('C'); tokadd('-'); tokcopy(3);
goto escaped; goto escaped;
case 'c': case 'c':
tokadd('\\'); tokadd('c'); if (flags & ESCAPE_CONTROL) goto eof;
tokcopy(2);
flags |= ESCAPE_CONTROL;
escaped: escaped:
if ((c = nextc()) == '\\') { if ((c = nextc()) == '\\') {
return tokadd_escape(term, mb); goto first;
} }
else if (c == -1) goto eof; else if (c == -1) goto eof;
tokadd(c); tokadd(c);
@ -5195,18 +5370,40 @@ static void
parser_tokadd_mbchar(struct parser_params *parser, int c) parser_tokadd_mbchar(struct parser_params *parser, int c)
{ {
int len = parser_mbclen(); int len = parser_mbclen();
do { tokadd(c);
tokadd(c); lex_p += --len;
} while (--len > 0 && (c = nextc()) != -1); if (len > 0) tokcopy(len);
} }
#define tokadd_mbchar(c) parser_tokadd_mbchar(parser, c) #define tokadd_mbchar(c) parser_tokadd_mbchar(parser, c)
static int static int
parser_tokadd_string(struct parser_params *parser, parser_tokadd_string(struct parser_params *parser,
int func, int term, int paren, long *nest, int *mb) int func, int term, int paren, long *nest,
int *has8bit, int *hasmb, rb_encoding **encp)
{ {
int c; int c;
int has_mb = 0;
rb_encoding *enc = *encp;
char *errbuf = 0;
static const char mixed_msg[] = "%s mixed within %s source";
#define mixed_error(enc1, enc2) if (!errbuf) { \
int len = sizeof(mixed_msg) - 4; \
len += strlen(rb_enc_name(enc1)); \
len += strlen(rb_enc_name(enc2)); \
errbuf = ALLOCA_N(char, len); \
snprintf(errbuf, len, mixed_msg, \
rb_enc_name(enc1), \
rb_enc_name(enc2)); \
yyerror(errbuf); \
}
#define mixed_escape(beg, enc1, enc2) do { \
const char *pos = lex_p; \
lex_p = beg; \
mixed_error(enc1, enc2); \
lex_p = pos; \
} while (0)
while ((c = nextc()) != -1) { while ((c = nextc()) != -1) {
if (paren && c == paren) { if (paren && c == paren) {
@ -5227,6 +5424,7 @@ parser_tokadd_string(struct parser_params *parser,
} }
} }
else if (c == '\\') { else if (c == '\\') {
const char *beg = lex_p - 1;
c = nextc(); c = nextc();
switch (c) { switch (c) {
case '\n': case '\n':
@ -5239,17 +5437,43 @@ parser_tokadd_string(struct parser_params *parser,
if (func & STR_FUNC_ESCAPE) tokadd(c); if (func & STR_FUNC_ESCAPE) tokadd(c);
break; break;
case 'u':
if ((func & STR_FUNC_EXPAND) == 0) {
tokadd('\\');
break;
}
parser_tokadd_utf8(parser, hasmb, &enc, 1,
func & STR_FUNC_SYMBOL);
if (has_mb && enc != *encp) {
mixed_escape(beg, enc, *encp);
}
continue;
default: default:
if (func & STR_FUNC_REGEXP) { if (func & STR_FUNC_REGEXP) {
pushback(c); pushback(c);
if (tokadd_escape(term, mb) < 0) if ((c = tokadd_escape(term, has8bit, hasmb, &enc)) < 0)
return -1; return -1;
if (has_mb && enc != *encp) {
mixed_escape(beg, enc, *encp);
}
continue; continue;
} }
else if (func & STR_FUNC_EXPAND) { else if (func & STR_FUNC_EXPAND) {
int tmb = 0;
pushback(c); pushback(c);
if (func & STR_FUNC_ESCAPE) tokadd('\\'); if (func & STR_FUNC_ESCAPE) tokadd('\\');
c = read_escape(mb); c = read_escape(0, has8bit, &tmb, &enc);
if (tmb) {
*hasmb = tmb;
if (has_mb && enc != *encp) {
mixed_escape(beg, enc, *encp);
}
else {
tokaddmbc(c, enc);
}
continue;
}
} }
else if ((func & STR_FUNC_QWORDS) && ISSPACE(c)) { else if ((func & STR_FUNC_QWORDS) && ISSPACE(c)) {
/* ignore backslashed spaces in %w */ /* ignore backslashed spaces in %w */
@ -5260,8 +5484,13 @@ parser_tokadd_string(struct parser_params *parser,
} }
} }
else if (parser_ismbchar()) { else if (parser_ismbchar()) {
has_mb = 1;
if (enc != *encp) {
mixed_error(enc, *encp);
continue;
}
tokadd_mbchar(c); tokadd_mbchar(c);
if (mb) *mb = ENC_CODERANGE_MULTI; if (hasmb) *hasmb = 1;
continue; continue;
} }
else if ((func & STR_FUNC_QWORDS) && ISSPACE(c)) { else if ((func & STR_FUNC_QWORDS) && ISSPACE(c)) {
@ -5275,6 +5504,7 @@ parser_tokadd_string(struct parser_params *parser,
} }
tokadd(c); tokadd(c);
} }
*encp = enc;
return c; return c;
} }
@ -5287,7 +5517,8 @@ parser_parse_string(struct parser_params *parser, NODE *quote)
int func = quote->nd_func; int func = quote->nd_func;
int term = nd_term(quote); int term = nd_term(quote);
int paren = nd_paren(quote); int paren = nd_paren(quote);
int c, space = 0, mb = ENC_CODERANGE_SINGLE; int c, space = 0, has8bit=0, hasmb=0;
rb_encoding *enc = parser->enc;
if (func == -1) return tSTRING_END; if (func == -1) return tSTRING_END;
c = nextc(); c = nextc();
@ -5321,21 +5552,21 @@ parser_parse_string(struct parser_params *parser, NODE *quote)
tokadd('#'); tokadd('#');
} }
pushback(c); pushback(c);
if (tokadd_string(func, term, paren, &quote->nd_nest, &mb) == -1) { if (tokadd_string(func, term, paren, &quote->nd_nest,
&has8bit, &hasmb, &enc) == -1) {
ruby_sourceline = nd_line(quote);
if (func & STR_FUNC_REGEXP) { if (func & STR_FUNC_REGEXP) {
ruby_sourceline = nd_line(quote);
compile_error(PARSER_ARG "unterminated regexp meets end of file"); compile_error(PARSER_ARG "unterminated regexp meets end of file");
return tREGEXP_END; return tREGEXP_END;
} }
else { else {
ruby_sourceline = nd_line(quote);
compile_error(PARSER_ARG "unterminated string meets end of file"); compile_error(PARSER_ARG "unterminated string meets end of file");
return tSTRING_END; return tSTRING_END;
} }
} }
tokfix(); tokfix();
set_yylval_str(STR_NEW3(tok(), toklen(), mb)); set_yylval_str(STR_NEW3(tok(), toklen(), enc, has8bit, hasmb));
return tSTRING_CONTENT; return tSTRING_CONTENT;
} }
@ -5498,7 +5729,9 @@ parser_here_document(struct parser_params *parser, NODE *here)
} while (!whole_match_p(eos, len, indent)); } while (!whole_match_p(eos, len, indent));
} }
else { else {
int mb = ENC_CODERANGE_SINGLE, *mbp = &mb; /* int mb = ENC_CODERANGE_SINGLE, *mbp = &mb;*/
int has8bit=0, hasmb=0;
rb_encoding *enc = parser->enc;
newtok(); newtok();
if (c == '#') { if (c == '#') {
switch (c = nextc()) { switch (c = nextc()) {
@ -5513,16 +5746,17 @@ parser_here_document(struct parser_params *parser, NODE *here)
} }
do { do {
pushback(c); pushback(c);
if ((c = tokadd_string(func, '\n', 0, NULL, mbp)) == -1) goto error; if ((c = tokadd_string(func, '\n', 0, NULL,
&has8bit, &hasmb, &enc)) == -1) goto error;
if (c != '\n') { if (c != '\n') {
set_yylval_str(STR_NEW3(tok(), toklen(), mb)); set_yylval_str(STR_NEW3(tok(), toklen(), enc, has8bit,hasmb));
return tSTRING_CONTENT; return tSTRING_CONTENT;
} }
tokadd(nextc()); tokadd(nextc());
if (mbp && mb == ENC_CODERANGE_UNKNOWN) mbp = 0; /* if (mbp && mb == ENC_CODERANGE_UNKNOWN) mbp = 0;*/
if ((c = nextc()) == -1) goto error; if ((c = nextc()) == -1) goto error;
} while (!whole_match_p(eos, len, indent)); } while (!whole_match_p(eos, len, indent));
str = STR_NEW3(tok(), toklen(), mb); str = STR_NEW3(tok(), toklen(), enc, has8bit,hasmb);
} }
heredoc_restore(lex_strterm); heredoc_restore(lex_strterm);
lex_strterm = NEW_STRTERM(-1, 0, 0); lex_strterm = NEW_STRTERM(-1, 0, 0);
@ -5782,6 +6016,8 @@ parser_yylex(struct parser_params *parser)
int space_seen = 0; int space_seen = 0;
int cmd_state; int cmd_state;
enum lex_state_e last_state; enum lex_state_e last_state;
rb_encoding *enc;
int has8bit = 0, hasmb = 0;
int mb; int mb;
#ifdef RIPPER #ifdef RIPPER
int fallthru = Qfalse; int fallthru = Qfalse;
@ -6123,25 +6359,28 @@ parser_yylex(struct parser_params *parser)
return '?'; return '?';
} }
newtok(); newtok();
enc = parser->enc;
if (parser_ismbchar()) { if (parser_ismbchar()) {
mb = ENC_CODERANGE_MULTI; hasmb = 1;
tokadd_mbchar(c); tokadd_mbchar(c);
} }
else if ((rb_enc_isalnum(c, parser->enc) || c == '_') && else if ((rb_enc_isalnum(c, parser->enc) || c == '_') &&
lex_p < lex_pend && is_identchar(lex_p, lex_pend, parser->enc)) { lex_p < lex_pend && is_identchar(lex_p, lex_pend, parser->enc)) {
goto ternary; goto ternary;
} }
else if (c == '\\' && (c = read_escape(0)) >= 0x80) { else if (c == '\\' && (c = read_escape(0, &has8bit, &hasmb, &enc)) >= 0x80) {
rb_encoding *enc = parser->enc; if (hasmb) {
mb = ENC_CODERANGE_UNKNOWN; tokaddmbc(c, enc);
rb_enc_mbcput(c, tokspace(rb_enc_codelen(c, enc)), enc); }
else {
tokadd(c);
}
} }
else { else {
mb = ENC_CODERANGE_SINGLE;
tokadd(c); tokadd(c);
} }
tokfix(); tokfix();
set_yylval_str(STR_NEW3(tok(), toklen(), mb)); set_yylval_str(STR_NEW3(tok(), toklen(), enc, has8bit, hasmb));
lex_state = EXPR_ENDARG; lex_state = EXPR_ENDARG;
return tCHAR; return tCHAR;
@ -7211,6 +7450,17 @@ list_concat_gen(struct parser_params *parser, NODE *head, NODE *tail)
return head; return head;
} }
static void
literal_concat0(struct parser_params *parser, VALUE head, VALUE tail)
{
if (!rb_enc_compatible(head, tail)) {
compile_error(PARSER_ARG "string literal encodings differ (%s / %s)",
rb_enc_name(rb_enc_get(head)),
rb_enc_name(rb_enc_get(tail)));
}
rb_str_buf_append(head, tail);
}
/* concat two string literals */ /* concat two string literals */
static NODE * static NODE *
literal_concat_gen(struct parser_params *parser, NODE *head, NODE *tail) literal_concat_gen(struct parser_params *parser, NODE *head, NODE *tail)
@ -7228,7 +7478,7 @@ literal_concat_gen(struct parser_params *parser, NODE *head, NODE *tail)
switch (nd_type(tail)) { switch (nd_type(tail)) {
case NODE_STR: case NODE_STR:
if (htype == NODE_STR) { if (htype == NODE_STR) {
rb_str_concat(head->nd_lit, tail->nd_lit); literal_concat0(parser, head->nd_lit, tail->nd_lit);
rb_gc_force_recycle((VALUE)tail); rb_gc_force_recycle((VALUE)tail);
} }
else { else {
@ -7238,7 +7488,7 @@ literal_concat_gen(struct parser_params *parser, NODE *head, NODE *tail)
case NODE_DSTR: case NODE_DSTR:
if (htype == NODE_STR) { if (htype == NODE_STR) {
rb_str_concat(head->nd_lit, tail->nd_lit); literal_concat0(parser, head->nd_lit, tail->nd_lit);
tail->nd_lit = head->nd_lit; tail->nd_lit = head->nd_lit;
rb_gc_force_recycle((VALUE)head); rb_gc_force_recycle((VALUE)head);
head = tail; head = tail;

25
test/ruby/test_mixed_unicode_escapes.rb Normal file
View file

@ -0,0 +1,25 @@
# -*- coding: sjis -*-
# This test is in a differnt file than TestUnicodeEscapes
# So that we can have a different coding comment above
require 'test/unit'
class TestMixedUnicodeEscape < Test::Unit::TestCase
def test_basic
# Unicode escapes do work in an sjis encoded file, but only
# if they don't contain other multi-byte chars
assert_equal("A", "\u0041")
# 8-bit character escapes are okay.
assert_equal("B\xFF", "\u0042\xFF")
# sjis mb chars mixed with Unicode shound not work
assert_raise(SyntaxError) { eval %q("é<>\u1234")}
assert_raise(SyntaxError) { eval %q("\u{1234}é<>")}
# String interpolation turns into an expression and we get
# a different kind of error, but we still can't mix these
assert_raise(ArgumentError) { eval %q("\u{1234}#{nil}é<>")}
assert_raise(ArgumentError) { eval %q("é<>#{nil}\u1234")}
end
end

240
test/ruby/test_unicode_escape.rb Normal file
View file

@ -0,0 +1,240 @@
# -*- coding: utf-8 -*-
require 'test/unit'
class TestUnicodeEscape < Test::Unit::TestCase
def test_basic
assert_equal('Matz - 松本行弘',
"Matz - \u677E\u672C\u884C\u5F18")
assert_equal('Matz - まつもと ゆきひろ',
"Matz - \u307E\u3064\u3082\u3068 \u3086\u304D\u3072\u308D")
assert_equal('Matz - まつもと ゆきひろ',
"Matz - \u{307E}\u{3064}\u{3082}\u{3068} \u{3086}\u{304D}\u{3072}\u{308D}")
assert_equal('Matz - まつもと ゆきひろ',
"Matz - \u{307E 3064 3082 3068 20 3086 304D 3072 308D}")
assert_equal("Aoyama Gakuin University - \xE9\x9D\x92\xE5\xB1\xB1\xE5\xAD\xA6\xE9\x99\xA2\xE5\xA4\xA7\xE5\xAD\xA6",
"Aoyama Gakuin University - \u9752\u5C71\u5B66\u9662\u5927\u5B66")
assert_equal('Aoyama Gakuin University - 青山学院大学',
"Aoyama Gakuin University - \u9752\u5C71\u5B66\u9662\u5927\u5B66")
assert_equal('青山学院大学', "\u9752\u5C71\u5B66\u9662\u5927\u5B66")
assert_equal("Martin D\xC3\xBCrst", "Martin D\u00FCrst")
assert_equal('Martin Dürst', "Martin D\u00FCrst")
assert_equal('ü', "\u00FC")
assert_equal("Martin D\xC3\xBCrst", "Martin D\u{FC}rst")
assert_equal('Martin Dürst', "Martin D\u{FC}rst")
assert_equal('ü', "\u{FC}")
assert_equal('ü', %Q|\u{FC}|)
assert_equal('ü', %W{\u{FC}}[0])
# \u escapes in here documents
assert_equal('Matz - まつもと ゆきひろ', <<EOS.chop)
Matz - \u307E\u3064\u3082\u3068 \u3086\u304D\u3072\u308D
EOS
assert_equal('Matz - まつもと ゆきひろ', <<"EOS".chop)
Matz - \u{307E 3064 3082 3068} \u{3086 304D 3072 308D}
EOS
assert_not_equal('Matz - まつもと ゆきひろ', <<'EOS'.chop)
Matz - \u{307E 3064 3082 3068} \u{3086 304D 3072 308D}
EOS
# single-quoted things don't expand \u
assert_not_equal('ü', '\u{FC}')
assert_not_equal('ü', %q|\u{FC}|)
assert_not_equal('ü', %w{\u{FC}}[0])
assert_equal('\u00fc', "\\" + "u00fc")
# \u in %x strings
assert_equal(`echo "\u0041"`.chop, "A")
assert_equal(%x{echo "\u0041"}.chop, "A")
assert_equal(`echo "\u{FC}"`.force_encoding("utf-8"), "ü\n")
# \u in quoted symbols
assert_equal(:A, :"\u0041")
assert_equal(:a, :"\u0061")
assert_equal(:ま, :ま)
assert_equal(:ü, :ü)
assert_equal(:"\u{41}", :"\u0041")
assert_equal(:ü, :"\u{fc}")
# the NUL character is not allowed in symbols
assert_raise(SyntaxError) { eval %q(:"\u{0}")}
assert_raise(SyntaxError) { eval %q(:"\u0000")}
assert_raise(SyntaxError) { eval %q(:"\u{fc 0 0041}")}
assert_raise(SyntaxError) { eval %q(:"\x00")}
assert_raise(SyntaxError) { eval %q(:"\0")}
end
def test_regexp
# Compare regexps to regexps
assert_equal(/Yukihiro Matsumoto - 松本行弘/,
/Yukihiro Matsumoto - \u677E\u672C\u884C\u5F18/)
assert_equal(/Yukihiro Matsumoto - 松本行弘/,
/Yukihiro Matsumoto - \u{677E 672C 884C 5F18}/)
assert_equal(/Matz - まつもと ゆきひろ/,
/Matz - \u307E\u3064\u3082\u3068 \u3086\u304D\u3072\u308D/)
assert_equal(/Aoyama Gakuin University - 青山学院大学/,
/Aoyama Gakuin University - \u9752\u5C71\u5B66\u9662\u5927\u5B66/)
assert_equal(/青山学院大学/, /\u9752\u5C71\u5B66\u9662\u5927\u5B66/)
assert_equal(/Martin Dürst/, /Martin D\u00FCrst/)
assert_equal(/ü/, /\u00FC/)
assert_equal(/Martin Dürst/, /Martin D\u{FC}rst/)
assert_equal(/ü/, /\u{FC}/)
assert_equal(/ü/, %r{\u{FC}})
assert_equal(/ü/i, %r{\u00FC}i)
# match strings to regexps
assert_equal("Yukihiro Matsumoto - 松本行弘" =~ /Yukihiro Matsumoto - \u677E\u672C\u884C\u5F18/, 0)
assert_equal("Yukihiro Matsumoto - \u677E\u672C\u884C\u5F18" =~ /Yukihiro Matsumoto - \u677E\u672C\u884C/, 0)
assert_equal("Yukihiro Matsumoto - 松本行弘" =~ /Yukihiro Matsumoto - \u{677E 672C 884C 5F18}/, 0)
assert_equal(%Q{Yukihiro Matsumoto - \u{677E 672C 884C 5F18}} =~ /Yukihiro Matsumoto - \u{677E 672C 884C 5F18}/, 0)
assert_equal("Matz - まつもと ゆきひろ" =~ /Matz - \u307E\u3064\u3082\u3068 \u3086\u304D\u3072\u308D/, 0)
assert_equal("Aoyama Gakuin University - 青山学院大学" =~ /Aoyama Gakuin University - \u9752\u5C71\u5B66\u9662\u5927\u5B66/, 0)
assert_equal("青山学院大学" =~ /\u9752\u5C71\u5B66\u9662\u5927\u5B66/, 0)
assert_equal("Martin Dürst" =~ /Martin D\u00FCrst/, 0)
assert_equal("ü" =~ /\u00FC/, 0)
assert_equal("Martin Dürst" =~ /Martin D\u{FC}rst/, 0)
assert_equal("ü" =~ %r{\u{FC}}, 0)
assert_equal("ü" =~ %r{\u00FC}i, 0)
# Flip order of the two operands
assert_equal(/Martin D\u00FCrst/ =~ "Martin Dürst", 0)
assert_equal(/\u00FC/ =~ "testü", 4)
assert_equal(/Martin D\u{FC}rst/ =~ "fooMartin Dürstbar", 3)
assert_equal(%r{\u{FC}} =~ "fooübar", 3)
# Put \u in strings, literal character in regexp
assert_equal("Martin D\u00FCrst" =~ /Martin Dürst/, 0)
assert_equal("test\u00FC" =~ /ü/, 4)
assert_equal("fooMartin D\u{FC}rstbar" =~ /Martin Dürst/, 3)
assert_equal(%Q{foo\u{FC}bar} =~ %r<ü>, 3)
end
def test_syntax_variants
# all hex digits
assert_equal("\xC4\xA3\xE4\x95\xA7\xE8\xA6\xAB\xEC\xB7\xAF", "\u0123\u4567\u89AB\uCDEF")
assert_equal("\xC4\xA3\xE4\x95\xA7\xE8\xA6\xAB\xEC\xB7\xAF", "\u0123\u4567\u89AB\uCDEF")
assert_equal("\xC4\xA3\xE4\x95\xA7\xE8\xA6\xAB\xEC\xB7\xAF", "\u0123\u4567\u89ab\ucdef")
assert_equal("\xC4\xA3\xE4\x95\xA7\xE8\xA6\xAB\xEC\xB7\xAF", "\u0123\u4567\u89ab\ucdef")
assert_equal("\xC4\xA3\xE4\x95\xA7\xE8\xA6\xAB\xEC\xB7\xAF", "\u0123\u4567\u89aB\uCdEf")
assert_equal("\xC4\xA3\xE4\x95\xA7\xE8\xA6\xAB\xEC\xB7\xAF", "\u0123\u4567\u89aB\ucDEF")
end
def test_fulton
# examples from Hal Fulton's book (second edition), chapter 4
# precomposed e'pe'e
assert_equal('épée', "\u00E9\u0070\u00E9\u0065")
assert_equal('épée', "\u00E9p\u00E9e")
assert_equal("\xC3\xA9\x70\xC3\xA9\x65", "\u00E9\u0070\u00E9\u0065")
assert_equal("\xC3\xA9\x70\xC3\xA9\x65", "\u00E9p\u00E9e")
# decomposed e'pe'e
assert_equal('épée', "\u0065\u0301\u0070\u0065\u0301\u0065")
assert_equal('épée', "e\u0301pe\u0301e")
assert_equal("\x65\xCC\x81\x70\x65\xCC\x81\x65", "\u0065\u0301\u0070\u0065\u0301\u0065")
assert_equal("\x65\xCC\x81\x70\x65\xCC\x81\x65", "e\u0301pe\u0301e")
# combinations of NFC/D, NFKC/D
assert_equal('öffnen', "\u00F6\u0066\u0066\u006E\u0065\u006E")
assert_equal("\xC3\xB6ffnen", "\u00F6\u0066\u0066\u006E\u0065\u006E")
assert_equal('öffnen', "\u00F6ffnen")
assert_equal("\xC3\xB6ffnen", "\u00F6ffnen")
assert_equal('öffnen', "\u006F\u0308\u0066\u0066\u006E\u0065\u006E")
assert_equal("\x6F\xCC\x88ffnen", "\u006F\u0308\u0066\u0066\u006E\u0065\u006E")
assert_equal('öffnen', "o\u0308ffnen")
assert_equal("\x6F\xCC\x88ffnen", "o\u0308ffnen")
assert_equal('öffnen', "\u00F6\uFB00\u006E\u0065\u006E")
assert_equal("\xC3\xB6\xEF\xAC\x80nen", "\u00F6\uFB00\u006E\u0065\u006E")
assert_equal('öffnen', "\u00F6\uFB00nen")
assert_equal("\xC3\xB6\xEF\xAC\x80nen", "\u00F6\uFB00nen")
assert_equal('öffnen', "\u006F\u0308\uFB00\u006E\u0065\u006E")
assert_equal("\x6F\xCC\x88\xEF\xAC\x80nen", "\u006F\u0308\uFB00\u006E\u0065\u006E")
assert_equal('öffnen', "o\u0308\uFB00nen")
assert_equal("\x6F\xCC\x88\xEF\xAC\x80nen", "o\u0308\uFB00nen")
# German sharp s (sz)
assert_equal('Straße', "\u0053\u0074\u0072\u0061\u00DF\u0065")
assert_equal("\x53\x74\x72\x61\xC3\x9F\x65", "\u0053\u0074\u0072\u0061\u00DF\u0065")
assert_equal('Straße', "Stra\u00DFe")
assert_equal("\x53\x74\x72\x61\xC3\x9F\x65", "Stra\u00DFe")
assert_equal('Straße', "\u{53}\u{74}\u{72}\u{61}\u{DF}\u{65}")
assert_equal("\x53\x74\x72\x61\xC3\x9F\x65", "\u{53}\u{74}\u{72}\u{61}\u{DF}\u{65}")
assert_equal("\x53\x74\x72\x61\xC3\x9F\x65", "\u{53 74 72 61 DF 65}")
assert_equal('Straße', "Stra\u{DF}e")
assert_equal("\x53\x74\x72\x61\xC3\x9F\x65", "Stra\u{DF}e")
end
def test_edge_cases
# start and end of each outer plane
assert_equal("\xF4\x8F\xBF\xBF", "\u{10FFFF}")
assert_equal("\xF4\x80\x80\x80", "\u{100000}")
assert_equal("\xF3\xBF\xBF\xBF", "\u{FFFFF}")
assert_equal("\xF3\xB0\x80\x80", "\u{F0000}")
assert_equal("\xF3\xAF\xBF\xBF", "\u{EFFFF}")
assert_equal("\xF3\xA0\x80\x80", "\u{E0000}")
assert_equal("\xF3\x9F\xBF\xBF", "\u{DFFFF}")
assert_equal("\xF3\x90\x80\x80", "\u{D0000}")
assert_equal("\xF3\x8F\xBF\xBF", "\u{CFFFF}")
assert_equal("\xF3\x80\x80\x80", "\u{C0000}")
assert_equal("\xF2\xBF\xBF\xBF", "\u{BFFFF}")
assert_equal("\xF2\xB0\x80\x80", "\u{B0000}")
assert_equal("\xF2\xAF\xBF\xBF", "\u{AFFFF}")
assert_equal("\xF2\xA0\x80\x80", "\u{A0000}")
assert_equal("\xF2\x9F\xBF\xBF", "\u{9FFFF}")
assert_equal("\xF2\x90\x80\x80", "\u{90000}")
assert_equal("\xF2\x8F\xBF\xBF", "\u{8FFFF}")
assert_equal("\xF2\x80\x80\x80", "\u{80000}")
assert_equal("\xF1\xBF\xBF\xBF", "\u{7FFFF}")
assert_equal("\xF1\xB0\x80\x80", "\u{70000}")
assert_equal("\xF1\xAF\xBF\xBF", "\u{6FFFF}")
assert_equal("\xF1\xA0\x80\x80", "\u{60000}")
assert_equal("\xF1\x9F\xBF\xBF", "\u{5FFFF}")
assert_equal("\xF1\x90\x80\x80", "\u{50000}")
assert_equal("\xF1\x8F\xBF\xBF", "\u{4FFFF}")
assert_equal("\xF1\x80\x80\x80", "\u{40000}")
assert_equal("\xF0\xBF\xBF\xBF", "\u{3FFFF}")
assert_equal("\xF0\xB0\x80\x80", "\u{30000}")
assert_equal("\xF0\xAF\xBF\xBF", "\u{2FFFF}")
assert_equal("\xF0\xA0\x80\x80", "\u{20000}")
assert_equal("\xF0\x9F\xBF\xBF", "\u{1FFFF}")
assert_equal("\xF0\x90\x80\x80", "\u{10000}")
# BMP
assert_equal("\xEF\xBF\xBF", "\uFFFF")
assert_equal("\xEE\x80\x80", "\uE000")
assert_equal("\xED\x9F\xBF", "\uD7FF")
assert_equal("\xE0\xA0\x80", "\u0800")
assert_equal("\xDF\xBF", "\u07FF")
assert_equal("\xC2\x80", "\u0080")
assert_equal("\x7F", "\u007F")
assert_equal("\x00", "\u0000")
end
def test_chars
assert_equal(?\u0041, ?A)
assert_equal(?\u{79}, ?\x79)
assert_equal(?\u{0}, ?\000)
assert_equal(?\u0000, ?\000)
end
# Tests to make sure that disallowed cases fail
def test_fail
assert_raise(SyntaxError) { eval %q("\uabc") } # too short
assert_raise(SyntaxError) { eval %q("\uab") } # too short
assert_raise(SyntaxError) { eval %q("\ua") } # too short
assert_raise(SyntaxError) { eval %q("\u") } # too short
assert_raise(SyntaxError) { eval %q("\u{110000}") } # too high
assert_raise(SyntaxError) { eval %q("\u{abcdeff}") } # too long
assert_raise(SyntaxError) { eval %q("\ughij") } # bad hex digits
assert_raise(SyntaxError) { eval %q("\u{ghij}") } # bad hex digits
assert_raise(SyntaxError) { eval %q("\u{123 456 }")} # extra space
assert_raise(SyntaxError) { eval %q("\u{ 123 456}")} # extra space
assert_raise(SyntaxError) { eval %q("\u{123 456}")} # extra space
# The utf-8 encoding object currently does not object to codepoints
# in the surrogate blocks, so these do not raise an error.
# assert_raise(SyntaxError) { "\uD800" } # surrogate block
# assert_raise(SyntaxError) { "\uDCBA" } # surrogate block
# assert_raise(SyntaxError) { "\uDFFF" } # surrogate block
# assert_raise(SyntaxError) { "\uD847\uDD9A" } # surrogate pair
end
end