mirror of
https://github.com/ruby/ruby.git
synced 2022-11-09 12:17:21 -05:00
bfabc05a43
rb_ary_sort_bang() instead. [ruby-dev:24291] * enum.c (enum_sort_by): pedantic type check added. [ruby-dev:24291] * hash.c (rb_hash_foreach_iter): check iter_lev after each iteration. [ruby-dev:24289] * array.c (rb_ary_and): element size might change during comparison. [ruby-dev:24290] * array.c (rb_ary_or): ditto. [ruby-dev:24292] * array.c (rb_ary_equal): wrong fix. [ruby-dev:24286] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@6939 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2430 lines
54 KiB
C
2430 lines
54 KiB
C
/**********************************************************************
|
|
|
|
hash.c -
|
|
|
|
$Author$
|
|
$Date$
|
|
created at: Mon Nov 22 18:51:18 JST 1993
|
|
|
|
Copyright (C) 1993-2003 Yukihiro Matsumoto
|
|
Copyright (C) 2000 Network Applied Communication Laboratory, Inc.
|
|
Copyright (C) 2000 Information-technology Promotion Agency, Japan
|
|
|
|
**********************************************************************/
|
|
|
|
#include "ruby.h"
|
|
#include "st.h"
|
|
#include "util.h"
|
|
#include "rubysig.h"
|
|
|
|
#ifdef __APPLE__
|
|
#include <crt_externs.h>
|
|
#endif
|
|
|
|
#define HASH_DELETED FL_USER1
|
|
#define HASH_PROC_DEFAULT FL_USER2
|
|
|
|
static void
|
|
rb_hash_modify(hash)
|
|
VALUE hash;
|
|
{
|
|
if (!RHASH(hash)->tbl) rb_raise(rb_eTypeError, "uninitialized Hash");
|
|
if (OBJ_FROZEN(hash)) rb_error_frozen("hash");
|
|
if (!OBJ_TAINTED(hash) && rb_safe_level() >= 4)
|
|
rb_raise(rb_eSecurityError, "Insecure: can't modify hash");
|
|
}
|
|
|
|
VALUE
|
|
rb_hash_freeze(hash)
|
|
VALUE hash;
|
|
{
|
|
return rb_obj_freeze(hash);
|
|
}
|
|
|
|
VALUE rb_cHash;
|
|
|
|
static VALUE envtbl;
|
|
static ID id_hash, id_call, id_default;
|
|
|
|
static VALUE
|
|
eql(args)
|
|
VALUE *args;
|
|
{
|
|
return (VALUE)rb_eql(args[0], args[1]);
|
|
}
|
|
|
|
static int
|
|
rb_any_cmp(a, b)
|
|
VALUE a, b;
|
|
{
|
|
VALUE args[2];
|
|
|
|
if (a == b) return 0;
|
|
if (FIXNUM_P(a) && FIXNUM_P(b)) {
|
|
return a != b;
|
|
}
|
|
if (TYPE(a) == T_STRING && RBASIC(a)->klass == rb_cString &&
|
|
TYPE(b) == T_STRING && RBASIC(b)->klass == rb_cString) {
|
|
return rb_str_cmp(a, b);
|
|
}
|
|
if (a == Qundef || b == Qundef) return -1;
|
|
if (SYMBOL_P(a) && SYMBOL_P(b)) {
|
|
return a != b;
|
|
}
|
|
|
|
args[0] = a;
|
|
args[1] = b;
|
|
return !rb_with_disable_interrupt(eql, (VALUE)args);
|
|
}
|
|
|
|
VALUE
|
|
rb_hash(obj)
|
|
VALUE obj;
|
|
{
|
|
return rb_funcall(obj, id_hash, 0);
|
|
}
|
|
|
|
static int
|
|
rb_any_hash(a)
|
|
VALUE a;
|
|
{
|
|
VALUE hval;
|
|
|
|
switch (TYPE(a)) {
|
|
case T_FIXNUM:
|
|
case T_SYMBOL:
|
|
return (int)a;
|
|
break;
|
|
|
|
case T_STRING:
|
|
return rb_str_hash(a);
|
|
break;
|
|
|
|
default:
|
|
hval = rb_funcall(a, id_hash, 0);
|
|
if (!FIXNUM_P(hval)) {
|
|
hval = rb_funcall(hval, '%', 1, INT2FIX(536870923));
|
|
}
|
|
return (int)FIX2LONG(hval);
|
|
}
|
|
}
|
|
|
|
static struct st_hash_type objhash = {
|
|
rb_any_cmp,
|
|
rb_any_hash,
|
|
};
|
|
|
|
struct rb_hash_foreach_arg {
|
|
VALUE hash;
|
|
enum st_retval (*func)();
|
|
VALUE arg;
|
|
};
|
|
|
|
static int
|
|
rb_hash_foreach_iter(key, value, arg)
|
|
VALUE key, value;
|
|
struct rb_hash_foreach_arg *arg;
|
|
{
|
|
int status;
|
|
st_table *tbl = RHASH(arg->hash)->tbl;
|
|
struct st_table_entry **bins = tbl->bins;
|
|
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
status = (*arg->func)(key, value, arg->arg);
|
|
if (RHASH(arg->hash)->tbl != tbl ||
|
|
RHASH(arg->hash)->tbl->bins != bins) {
|
|
rb_raise(rb_eIndexError, "rehash occurred during iteration");
|
|
}
|
|
if (RHASH(arg->hash)->iter_lev == 0) {
|
|
rb_raise(rb_eArgError, "block re-entered");
|
|
}
|
|
return status;
|
|
}
|
|
|
|
static VALUE
|
|
rb_hash_foreach_call(arg)
|
|
struct rb_hash_foreach_arg *arg;
|
|
{
|
|
st_foreach(RHASH(arg->hash)->tbl, rb_hash_foreach_iter, (st_data_t)arg);
|
|
return Qnil;
|
|
}
|
|
|
|
static VALUE
|
|
rb_hash_foreach_ensure(hash)
|
|
VALUE hash;
|
|
{
|
|
RHASH(hash)->iter_lev--;
|
|
|
|
if (RHASH(hash)->iter_lev == 0) {
|
|
if (FL_TEST(hash, HASH_DELETED)) {
|
|
st_cleanup_safe(RHASH(hash)->tbl, Qundef);
|
|
FL_UNSET(hash, HASH_DELETED);
|
|
}
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
static int
|
|
rb_hash_foreach(hash, func, farg)
|
|
VALUE hash;
|
|
enum st_retval (*func)();
|
|
VALUE farg;
|
|
{
|
|
struct rb_hash_foreach_arg arg;
|
|
|
|
RHASH(hash)->iter_lev++;
|
|
arg.hash = hash;
|
|
arg.func = func;
|
|
arg.arg = farg;
|
|
return rb_ensure(rb_hash_foreach_call, (VALUE)&arg, rb_hash_foreach_ensure, hash);
|
|
}
|
|
|
|
static VALUE hash_alloc _((VALUE));
|
|
static VALUE
|
|
hash_alloc(klass)
|
|
VALUE klass;
|
|
{
|
|
NEWOBJ(hash, struct RHash);
|
|
OBJSETUP(hash, klass, T_HASH);
|
|
|
|
hash->ifnone = Qnil;
|
|
hash->tbl = st_init_table(&objhash);
|
|
|
|
return (VALUE)hash;
|
|
}
|
|
|
|
VALUE
|
|
rb_hash_new()
|
|
{
|
|
return hash_alloc(rb_cHash);
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* Hash.new => hash
|
|
* Hash.new(obj) => aHash
|
|
* Hash.new {|hash, key| block } => aHash
|
|
*
|
|
* Returns a new, empty hash. If this hash is subsequently accessed by
|
|
* a key that doesn't correspond to a hash entry, the value returned
|
|
* depends on the style of <code>new</code> used to create the hash. In
|
|
* the first form, the access returns <code>nil</code>. If
|
|
* <i>obj</i> is specified, this single object will be used for
|
|
* all <em>default values</em>. If a block is specified, it will be
|
|
* called with the hash object and the key, and should return the
|
|
* default value. It is the block's responsibility to store the value
|
|
* in the hash if required.
|
|
*
|
|
* h = Hash.new("Go Fish")
|
|
* h["a"] = 100
|
|
* h["b"] = 200
|
|
* h["a"] #=> 100
|
|
* h["c"] #=> "Go Fish"
|
|
* # The following alters the single default object
|
|
* h["c"].upcase! #=> "GO FISH"
|
|
* h["d"] #=> "GO FISH"
|
|
* h.keys #=> ["a", "b"]
|
|
*
|
|
* # While this creates a new default object each time
|
|
* h = Hash.new { |hash, key| hash[key] = "Go Fish: #{key}" }
|
|
* h["c"] #=> "Go Fish: c"
|
|
* h["c"].upcase! #=> "GO FISH: C"
|
|
* h["d"] #=> "Go Fish: d"
|
|
* h.keys #=> ["c", "d"]
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_initialize(argc, argv, hash)
|
|
int argc;
|
|
VALUE *argv;
|
|
VALUE hash;
|
|
{
|
|
VALUE ifnone;
|
|
|
|
rb_hash_modify(hash);
|
|
if (rb_block_given_p()) {
|
|
if (argc > 0) {
|
|
rb_raise(rb_eArgError, "wrong number of arguments");
|
|
}
|
|
RHASH(hash)->ifnone = rb_block_proc();
|
|
FL_SET(hash, HASH_PROC_DEFAULT);
|
|
}
|
|
else {
|
|
rb_scan_args(argc, argv, "01", &ifnone);
|
|
RHASH(hash)->ifnone = ifnone;
|
|
}
|
|
|
|
return hash;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* Hash[ [key =>|, value]* ] => hash
|
|
*
|
|
* Creates a new hash populated with the given objects. Equivalent to
|
|
* the literal <code>{ <i>key</i>, <i>value</i>, ... }</code>. Keys and
|
|
* values occur in pairs, so there must be an even number of arguments.
|
|
*
|
|
* Hash["a", 100, "b", 200] #=> {"a"=>100, "b"=>200}
|
|
* Hash["a" => 100, "b" => 200] #=> {"a"=>100, "b"=>200}
|
|
* { "a" => 100, "b" => 200 } #=> {"a"=>100, "b"=>200}
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_s_create(argc, argv, klass)
|
|
int argc;
|
|
VALUE *argv;
|
|
VALUE klass;
|
|
{
|
|
VALUE hash;
|
|
int i;
|
|
|
|
if (argc == 1 && TYPE(argv[0]) == T_HASH) {
|
|
hash = hash_alloc(klass);
|
|
|
|
RHASH(hash)->ifnone = Qnil;
|
|
RHASH(hash)->tbl = st_copy(RHASH(argv[0])->tbl);
|
|
|
|
return hash;
|
|
}
|
|
|
|
if (argc % 2 != 0) {
|
|
rb_raise(rb_eArgError, "odd number of arguments for Hash");
|
|
}
|
|
|
|
hash = hash_alloc(klass);
|
|
for (i=0; i<argc; i+=2) {
|
|
rb_hash_aset(hash, argv[i], argv[i + 1]);
|
|
}
|
|
|
|
return hash;
|
|
}
|
|
|
|
static VALUE
|
|
to_hash(hash)
|
|
VALUE hash;
|
|
{
|
|
return rb_convert_type(hash, T_HASH, "Hash", "to_hash");
|
|
}
|
|
|
|
static int
|
|
rb_hash_rehash_i(key, value, tbl)
|
|
VALUE key, value;
|
|
st_table *tbl;
|
|
{
|
|
if (key != Qundef) st_insert(tbl, key, value);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.rehash -> hsh
|
|
*
|
|
* Rebuilds the hash based on the current hash values for each key. If
|
|
* values of key objects have changed since they were inserted, this
|
|
* method will reindex <i>hsh</i>. If <code>Hash#rehash</code> is
|
|
* called while an iterator is traversing the hash, an
|
|
* <code>IndexError</code> will be raised in the iterator.
|
|
*
|
|
* a = [ "a", "b" ]
|
|
* c = [ "c", "d" ]
|
|
* h = { a => 100, c => 300 }
|
|
* h[a] #=> 100
|
|
* a[0] = "z"
|
|
* h[a] #=> nil
|
|
* h.rehash #=> {["z", "b"]=>100, ["c", "d"]=>300}
|
|
* h[a] #=> 100
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_rehash(hash)
|
|
VALUE hash;
|
|
{
|
|
st_table *tbl;
|
|
|
|
rb_hash_modify(hash);
|
|
tbl = st_init_table_with_size(&objhash, RHASH(hash)->tbl->num_entries);
|
|
st_foreach(RHASH(hash)->tbl, rb_hash_rehash_i, (st_data_t)tbl);
|
|
st_free_table(RHASH(hash)->tbl);
|
|
RHASH(hash)->tbl = tbl;
|
|
|
|
return hash;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh[key] => value
|
|
*
|
|
* Element Reference---Retrieves the <i>value</i> object corresponding
|
|
* to the <i>key</i> object. If not found, returns the a default value (see
|
|
* <code>Hash::new</code> for details).
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h["a"] #=> 100
|
|
* h["c"] #=> nil
|
|
*
|
|
*/
|
|
|
|
VALUE
|
|
rb_hash_aref(hash, key)
|
|
VALUE hash, key;
|
|
{
|
|
VALUE val;
|
|
|
|
if (!st_lookup(RHASH(hash)->tbl, key, &val)) {
|
|
return rb_funcall(hash, id_default, 1, key);
|
|
}
|
|
return val;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.fetch(key [, default] ) => obj
|
|
* hsh.fetch(key) {| key | block } => obj
|
|
*
|
|
* Returns a value from the hash for the given key. If the key can't be
|
|
* found, there are several options: With no other arguments, it will
|
|
* raise an <code>IndexError</code> exception; if <i>default</i> is
|
|
* given, then that will be returned; if the optional code block is
|
|
* specified, then that will be run and its result returned.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.fetch("a") #=> 100
|
|
* h.fetch("z", "go fish") #=> "go fish"
|
|
* h.fetch("z") { |el| "go fish, #{el}"} #=> "go fish, z"
|
|
*
|
|
* The following example shows that an exception is raised if the key
|
|
* is not found and a default value is not supplied.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.fetch("z")
|
|
*
|
|
* <em>produces:</em>
|
|
*
|
|
* prog.rb:2:in `fetch': key not found (IndexError)
|
|
* from prog.rb:2
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_fetch(argc, argv, hash)
|
|
int argc;
|
|
VALUE *argv;
|
|
VALUE hash;
|
|
{
|
|
VALUE key, if_none;
|
|
VALUE val;
|
|
long block_given;
|
|
|
|
rb_scan_args(argc, argv, "11", &key, &if_none);
|
|
|
|
block_given = rb_block_given_p();
|
|
if (block_given && argc == 2) {
|
|
rb_warn("block supersedes default value argument");
|
|
}
|
|
if (!st_lookup(RHASH(hash)->tbl, key, &val)) {
|
|
if (block_given) return rb_yield(key);
|
|
if (argc == 1) {
|
|
rb_raise(rb_eIndexError, "key not found");
|
|
}
|
|
return if_none;
|
|
}
|
|
return val;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.default(key=nil) => obj
|
|
*
|
|
* Returns the default value, the value that would be returned by
|
|
* <i>hsh</i>[<i>key</i>] if <i>key</i> did not exist in <i>hsh</i>.
|
|
* See also <code>Hash::new</code> and <code>Hash#default=</code>.
|
|
*
|
|
* h = Hash.new #=> {}
|
|
* h.default #=> nil
|
|
* h.default(2) #=> nil
|
|
*
|
|
* h = Hash.new("cat") #=> {}
|
|
* h.default #=> "cat"
|
|
* h.default(2) #=> "cat"
|
|
*
|
|
* h = Hash.new {|h,k| h[k] = k.to_i*10} #=> {}
|
|
* h.default #=> 0
|
|
* h.default(2) #=> 20
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_default(argc, argv, hash)
|
|
int argc;
|
|
VALUE *argv;
|
|
VALUE hash;
|
|
{
|
|
VALUE key;
|
|
|
|
rb_scan_args(argc, argv, "01", &key);
|
|
if (FL_TEST(hash, HASH_PROC_DEFAULT)) {
|
|
return rb_funcall(RHASH(hash)->ifnone, id_call, 2, hash, key);
|
|
}
|
|
return RHASH(hash)->ifnone;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.default = obj => hsh
|
|
*
|
|
* Sets the default value, the value returned for a key that does not
|
|
* exist in the hash. It is not possible to set the a default to a
|
|
* <code>Proc</code> that will be executed on each key lookup.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.default = "Go fish"
|
|
* h["a"] #=> 100
|
|
* h["z"] #=> "Go fish"
|
|
* # This doesn't do what you might hope...
|
|
* h.default = proc do |hash, key|
|
|
* hash[key] = key + key
|
|
* end
|
|
* h[2] #=> #<Proc:0x401b3948@-:6>
|
|
* h["cat"] #=> #<Proc:0x401b3948@-:6>
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_set_default(hash, ifnone)
|
|
VALUE hash, ifnone;
|
|
{
|
|
rb_hash_modify(hash);
|
|
RHASH(hash)->ifnone = ifnone;
|
|
FL_UNSET(hash, HASH_PROC_DEFAULT);
|
|
return ifnone;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.default_proc -> anObject
|
|
*
|
|
* If <code>Hash::new</code> was invoked with a block, return that
|
|
* block, otherwise return <code>nil</code>.
|
|
*
|
|
* h = Hash.new {|h,k| h[k] = k*k } #=> {}
|
|
* p = h.default_proc #=> #<Proc:0x401b3d08@-:1>
|
|
* a = [] #=> []
|
|
* p.call(a, 2)
|
|
* a #=> [nil, nil, 4]
|
|
*/
|
|
|
|
|
|
static VALUE
|
|
rb_hash_default_proc(hash)
|
|
VALUE hash;
|
|
{
|
|
if (FL_TEST(hash, HASH_PROC_DEFAULT)) {
|
|
return RHASH(hash)->ifnone;
|
|
}
|
|
return Qnil;
|
|
}
|
|
|
|
static int
|
|
index_i(key, value, args)
|
|
VALUE key, value;
|
|
VALUE *args;
|
|
{
|
|
if (rb_equal(value, args[0])) {
|
|
args[1] = key;
|
|
return ST_STOP;
|
|
}
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.index(value) => key
|
|
*
|
|
* Returns the key for a given value. If not found, returns <code>nil</code>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.index(200) #=> "b"
|
|
* h.index(999) #=> nil
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_index(hash, value)
|
|
VALUE hash, value;
|
|
{
|
|
VALUE args[2];
|
|
|
|
args[0] = value;
|
|
args[1] = Qnil;
|
|
|
|
st_foreach(RHASH(hash)->tbl, index_i, (st_data_t)args);
|
|
|
|
return args[1];
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.delete(key) => value
|
|
* hsh.delete(key) {| key | block } => value
|
|
*
|
|
* Deletes and returns a key-value pair from <i>hsh</i> whose key is
|
|
* equal to <i>key</i>. If the key is not found, returns the
|
|
* <em>default value</em>. If the optional code block is given and the
|
|
* key is not found, pass in the key and return the result of
|
|
* <i>block</i>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.delete("a") #=> 100
|
|
* h.delete("z") #=> nil
|
|
* h.delete("z") { |el| "#{el} not found" } #=> "z not found"
|
|
*
|
|
*/
|
|
|
|
VALUE
|
|
rb_hash_delete(hash, key)
|
|
VALUE hash, key;
|
|
{
|
|
VALUE val;
|
|
|
|
rb_hash_modify(hash);
|
|
if (RHASH(hash)->iter_lev > 0) {
|
|
if (st_delete_safe(RHASH(hash)->tbl, (st_data_t*)&key, &val, Qundef)) {
|
|
FL_SET(hash, HASH_DELETED);
|
|
return val;
|
|
}
|
|
}
|
|
else if (st_delete(RHASH(hash)->tbl, (st_data_t*)&key, &val))
|
|
return val;
|
|
if (rb_block_given_p()) {
|
|
return rb_yield(key);
|
|
}
|
|
return Qnil;
|
|
}
|
|
|
|
struct shift_var {
|
|
int stop;
|
|
VALUE key;
|
|
VALUE val;
|
|
};
|
|
|
|
static int
|
|
shift_i(key, value, var)
|
|
VALUE key, value;
|
|
struct shift_var *var;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
if (var->stop) return ST_STOP;
|
|
var->stop = 1;
|
|
var->key = key;
|
|
var->val = value;
|
|
return ST_DELETE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.shift -> anArray or obj
|
|
*
|
|
* Removes a key-value pair from <i>hsh</i> and returns it as the
|
|
* two-item array <code>[</code> <i>key, value</i> <code>]</code>, or
|
|
* the hash's default value if the hash is empty.
|
|
*
|
|
* h = { 1 => "a", 2 => "b", 3 => "c" }
|
|
* h.shift #=> [1, "a"]
|
|
* h #=> {2=>"b", 3=>"c"}
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_shift(hash)
|
|
VALUE hash;
|
|
{
|
|
struct shift_var var;
|
|
|
|
rb_hash_modify(hash);
|
|
var.stop = 0;
|
|
st_foreach(RHASH(hash)->tbl, shift_i, (st_data_t)&var);
|
|
|
|
if (var.stop) {
|
|
return rb_assoc_new(var.key, var.val);
|
|
}
|
|
else if (FL_TEST(hash, HASH_PROC_DEFAULT)) {
|
|
return rb_funcall(RHASH(hash)->ifnone, id_call, 2, hash, Qnil);
|
|
}
|
|
else {
|
|
return RHASH(hash)->ifnone;
|
|
}
|
|
}
|
|
|
|
static enum st_retval
|
|
delete_if_i(key, value, hash)
|
|
VALUE key, value, hash;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
if (RTEST(rb_yield_values(2, key, value))) {
|
|
rb_hash_delete(hash, key);
|
|
}
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.delete_if {| key, value | block } -> hsh
|
|
*
|
|
* Deletes every key-value pair from <i>hsh</i> for which <i>block</i>
|
|
* evaluates to <code>true</code>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200, "c" => 300 }
|
|
* h.delete_if {|key, value| key >= "b" } #=> {"a"=>100}
|
|
*
|
|
*/
|
|
|
|
VALUE
|
|
rb_hash_delete_if(hash)
|
|
VALUE hash;
|
|
{
|
|
rb_hash_modify(hash);
|
|
rb_hash_foreach(hash, delete_if_i, hash);
|
|
return hash;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.reject! {| key, value | block } -> hsh or nil
|
|
*
|
|
* Equivalent to <code>Hash#delete_if</code>, but returns
|
|
* <code>nil</code> if no changes were made.
|
|
*/
|
|
|
|
VALUE
|
|
rb_hash_reject_bang(hash)
|
|
VALUE hash;
|
|
{
|
|
int n = RHASH(hash)->tbl->num_entries;
|
|
rb_hash_delete_if(hash);
|
|
if (n == RHASH(hash)->tbl->num_entries) return Qnil;
|
|
return hash;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.reject {| key, value | block } -> a_hash
|
|
*
|
|
* Same as <code>Hash#delete_if</code>, but works on (and returns) a
|
|
* copy of the <i>hsh</i>. Equivalent to
|
|
* <code><i>hsh</i>.dup.delete_if</code>.
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_reject(hash)
|
|
VALUE hash;
|
|
{
|
|
return rb_hash_delete_if(rb_obj_dup(hash));
|
|
}
|
|
|
|
static enum st_retval
|
|
select_i(key, value, result)
|
|
VALUE key, value, result;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
if (RTEST(rb_yield_values(2, key, value)))
|
|
rb_ary_push(result, rb_assoc_new(key, value));
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.values_at(key, ...) => array
|
|
*
|
|
* Return an array containing the values associated with the given keys.
|
|
* Also see <code>Hash.select</code>.
|
|
*
|
|
* h = { "cat" => "feline", "dog" => "canine", "cow" => "bovine" }
|
|
* h.values_at("cow", "cat") #=> ["bovine", "feline"]
|
|
*/
|
|
|
|
VALUE
|
|
rb_hash_values_at(argc, argv, hash)
|
|
int argc;
|
|
VALUE *argv;
|
|
VALUE hash;
|
|
{
|
|
VALUE result = rb_ary_new2(argc);
|
|
long i;
|
|
|
|
for (i=0; i<argc; i++) {
|
|
rb_ary_push(result, rb_hash_aref(hash, argv[i]));
|
|
}
|
|
return result;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.select {|key, value| block} => array
|
|
*
|
|
* Returns a new array consisting of <code>[key,value]</code>
|
|
* pairs for which the block returns true.
|
|
* Also see <code>Hash.values_at</code>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200, "c" => 300 }
|
|
* h.select {|k,v| k > "a"} #=> [["b", 200], ["c", 300]]
|
|
* h.select {|k,v| v < 200} #=> [["a", 100]]
|
|
*/
|
|
|
|
VALUE
|
|
rb_hash_select(hash)
|
|
VALUE hash;
|
|
{
|
|
VALUE result;
|
|
|
|
result = rb_ary_new();
|
|
rb_hash_foreach(hash, select_i, result);
|
|
return result;
|
|
}
|
|
|
|
static int
|
|
clear_i(key, value, dummy)
|
|
VALUE key, value, dummy;
|
|
{
|
|
return ST_DELETE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.clear -> hsh
|
|
*
|
|
* Removes all key-value pairs from <i>hsh</i>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 } #=> {"a"=>100, "b"=>200}
|
|
* h.clear #=> {}
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_clear(hash)
|
|
VALUE hash;
|
|
{
|
|
rb_hash_modify(hash);
|
|
st_foreach(RHASH(hash)->tbl, clear_i, 0);
|
|
|
|
return hash;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh[key] = value => value
|
|
* hsh.store(key, value) => value
|
|
*
|
|
* Element Assignment---Associates the value given by
|
|
* <i>value</i> with the key given by <i>key</i>.
|
|
* <i>key</i> should not have its value changed while it is in
|
|
* use as a key (a <code>String</code> passed as a key will be
|
|
* duplicated and frozen).
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h["a"] = 9
|
|
* h["c"] = 4
|
|
* h #=> {"a"=>9, "b"=>200, "c"=>4}
|
|
*
|
|
*/
|
|
|
|
VALUE
|
|
rb_hash_aset(hash, key, val)
|
|
VALUE hash, key, val;
|
|
{
|
|
rb_hash_modify(hash);
|
|
if (TYPE(key) != T_STRING || st_lookup(RHASH(hash)->tbl, key, 0)) {
|
|
st_insert(RHASH(hash)->tbl, key, val);
|
|
}
|
|
else {
|
|
st_add_direct(RHASH(hash)->tbl, rb_str_new4(key), val);
|
|
}
|
|
return val;
|
|
}
|
|
|
|
static int
|
|
replace_i(key, val, hash)
|
|
VALUE key, val, hash;
|
|
{
|
|
if (key != Qundef) {
|
|
rb_hash_aset(hash, key, val);
|
|
}
|
|
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.replace(other_hash) -> hsh
|
|
*
|
|
* Replaces the contents of <i>hsh</i> with the contents of
|
|
* <i>other_hash</i>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.replace({ "c" => 300, "d" => 400 }) #=> {"c"=>300, "d"=>400}
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_replace(hash, hash2)
|
|
VALUE hash, hash2;
|
|
{
|
|
hash2 = to_hash(hash2);
|
|
if (hash == hash2) return hash;
|
|
rb_hash_clear(hash);
|
|
st_foreach(RHASH(hash2)->tbl, replace_i, hash);
|
|
RHASH(hash)->ifnone = RHASH(hash2)->ifnone;
|
|
if (FL_TEST(hash2, HASH_PROC_DEFAULT)) {
|
|
FL_SET(hash, HASH_PROC_DEFAULT);
|
|
}
|
|
else {
|
|
FL_UNSET(hash, HASH_PROC_DEFAULT);
|
|
}
|
|
|
|
return hash;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.length => fixnum
|
|
* hsh.size => fixnum
|
|
*
|
|
* Returns the number of key-value pairs in the hash.
|
|
*
|
|
* h = { "d" => 100, "a" => 200, "v" => 300, "e" => 400 }
|
|
* h.length #=> 4
|
|
* h.delete("a") #=> 200
|
|
* h.length #=> 3
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_size(hash)
|
|
VALUE hash;
|
|
{
|
|
return INT2FIX(RHASH(hash)->tbl->num_entries);
|
|
}
|
|
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.empty? => true or false
|
|
*
|
|
* Returns <code>true</code> if <i>hsh</i> contains no key-value pairs.
|
|
*
|
|
* {}.empty? #=> true
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_empty_p(hash)
|
|
VALUE hash;
|
|
{
|
|
if (RHASH(hash)->tbl->num_entries == 0)
|
|
return Qtrue;
|
|
return Qfalse;
|
|
}
|
|
|
|
static enum st_retval
|
|
each_value_i(key, value)
|
|
VALUE key, value;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_yield(value);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.each_value {| value | block } -> hsh
|
|
*
|
|
* Calls <i>block</i> once for each key in <i>hsh</i>, passing the
|
|
* value as a parameter.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.each_value {|value| puts value }
|
|
*
|
|
* <em>produces:</em>
|
|
*
|
|
* 100
|
|
* 200
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_each_value(hash)
|
|
VALUE hash;
|
|
{
|
|
rb_hash_foreach(hash, each_value_i, 0);
|
|
return hash;
|
|
}
|
|
|
|
static enum st_retval
|
|
each_key_i(key, value)
|
|
VALUE key, value;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_yield(key);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.each_key {| key | block } -> hsh
|
|
*
|
|
* Calls <i>block</i> once for each key in <i>hsh</i>, passing the key
|
|
* as a parameter.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.each_key {|key| puts key }
|
|
*
|
|
* <em>produces:</em>
|
|
*
|
|
* a
|
|
* b
|
|
*/
|
|
static VALUE
|
|
rb_hash_each_key(hash)
|
|
VALUE hash;
|
|
{
|
|
rb_hash_foreach(hash, each_key_i, 0);
|
|
return hash;
|
|
}
|
|
|
|
static enum st_retval
|
|
each_pair_i(key, value)
|
|
VALUE key, value;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_yield_values(2, key, value);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.each_pair {| key_value_array | block } -> hsh
|
|
*
|
|
* Calls <i>block</i> once for each key in <i>hsh</i>, passing the key
|
|
* and value as parameters.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.each_pair {|key, value| puts "#{key} is #{value}" }
|
|
*
|
|
* <em>produces:</em>
|
|
*
|
|
* a is 100
|
|
* b is 200
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_each_pair(hash)
|
|
VALUE hash;
|
|
{
|
|
rb_hash_foreach(hash, each_pair_i, 0);
|
|
return hash;
|
|
}
|
|
|
|
static enum st_retval
|
|
each_i(key, value)
|
|
VALUE key, value;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_yield(rb_assoc_new(key, value));
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.each {| key, value | block } -> hsh
|
|
*
|
|
* Calls <i>block</i> once for each key in <i>hsh</i>, passing the key
|
|
* and value to the block as a two-element array. Because of the assignment
|
|
* semantics of block parameters, these elements will be split out if the
|
|
* block has two formal parameters. Also see <code>Hash.each_pair</code>, which
|
|
* will be marginally more efficient for blocks with two parameters.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.each {|key, value| puts "#{key} is #{value}" }
|
|
*
|
|
* <em>produces:</em>
|
|
*
|
|
* a is 100
|
|
* b is 200
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_each(hash)
|
|
VALUE hash;
|
|
{
|
|
rb_hash_foreach(hash, each_i, 0);
|
|
return hash;
|
|
}
|
|
|
|
static int
|
|
to_a_i(key, value, ary)
|
|
VALUE key, value, ary;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_ary_push(ary, rb_assoc_new(key, value));
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.to_a -> array
|
|
*
|
|
* Converts <i>hsh</i> to a nested array of <code>[</code> <i>key,
|
|
* value</i> <code>]</code> arrays.
|
|
*
|
|
* h = { "c" => 300, "a" => 100, "d" => 400, "c" => 300 }
|
|
* h.to_a #=> [["a", 100], ["c", 300], ["d", 400]]
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_to_a(hash)
|
|
VALUE hash;
|
|
{
|
|
VALUE ary;
|
|
|
|
ary = rb_ary_new();
|
|
st_foreach(RHASH(hash)->tbl, to_a_i, ary);
|
|
if (OBJ_TAINTED(hash)) OBJ_TAINT(ary);
|
|
|
|
return ary;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.sort => array
|
|
* hsh.sort {| a, b | block } => array
|
|
*
|
|
* Converts <i>hsh</i> to a nested array of <code>[</code> <i>key,
|
|
* value</i> <code>]</code> arrays and sorts it, using
|
|
* <code>Array#sort</code>.
|
|
*
|
|
* h = { "a" => 20, "b" => 30, "c" => 10 }
|
|
* h.sort #=> [["a", 20], ["b", 30], ["c", 10]]
|
|
* h.sort {|a,b| a[1]<=>b[1]} #=> [["c", 10], ["a", 20], ["b", 30]]
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_sort(hash)
|
|
VALUE hash;
|
|
{
|
|
VALUE entries = rb_hash_to_a(hash);
|
|
rb_ary_sort_bang(entries);
|
|
return entries;
|
|
}
|
|
|
|
static int
|
|
inspect_i(key, value, str)
|
|
VALUE key, value, str;
|
|
{
|
|
VALUE str2;
|
|
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
if (RSTRING(str)->len > 1) {
|
|
rb_str_cat2(str, ", ");
|
|
}
|
|
str2 = rb_inspect(key);
|
|
rb_str_buf_append(str, str2);
|
|
OBJ_INFECT(str, str2);
|
|
rb_str_buf_cat2(str, "=>");
|
|
str2 = rb_inspect(value);
|
|
rb_str_buf_append(str, str2);
|
|
OBJ_INFECT(str, str2);
|
|
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
static VALUE
|
|
inspect_hash(hash)
|
|
VALUE hash;
|
|
{
|
|
VALUE str;
|
|
|
|
str = rb_str_buf_new2("{");
|
|
st_foreach(RHASH(hash)->tbl, inspect_i, str);
|
|
rb_str_buf_cat2(str, "}");
|
|
OBJ_INFECT(str, hash);
|
|
|
|
return str;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.inspect => string
|
|
*
|
|
* Return the contents of this hash as a string.
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_inspect(hash)
|
|
VALUE hash;
|
|
{
|
|
if (RHASH(hash)->tbl == 0 || RHASH(hash)->tbl->num_entries == 0)
|
|
return rb_str_new2("{}");
|
|
if (rb_inspecting_p(hash)) return rb_str_new2("{...}");
|
|
return rb_protect_inspect(inspect_hash, hash, 0);
|
|
}
|
|
|
|
static VALUE
|
|
to_s_hash(hash)
|
|
VALUE hash;
|
|
{
|
|
return rb_ary_to_s(rb_hash_to_a(hash));
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.to_s => string
|
|
*
|
|
* Converts <i>hsh</i> to a string by converting the hash to an array
|
|
* of <code>[</code> <i>key, value</i> <code>]</code> pairs and then
|
|
* converting that array to a string using <code>Array#join</code> with
|
|
* the default separator.
|
|
*
|
|
* h = { "c" => 300, "a" => 100, "d" => 400, "c" => 300 }
|
|
* h.to_s #=> "a100c300d400"
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_to_s(hash)
|
|
VALUE hash;
|
|
{
|
|
if (rb_inspecting_p(hash)) return rb_str_new2("{...}");
|
|
return rb_protect_inspect(to_s_hash, hash, 0);
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.to_hash => hsh
|
|
*
|
|
* Returns <i>self</i>.
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_to_hash(hash)
|
|
VALUE hash;
|
|
{
|
|
return hash;
|
|
}
|
|
|
|
static int
|
|
keys_i(key, value, ary)
|
|
VALUE key, value, ary;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_ary_push(ary, key);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.keys => array
|
|
*
|
|
* Returns a new array populated with the keys from this hash. See also
|
|
* <code>Hash#values</code>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200, "c" => 300, "d" => 400 }
|
|
* h.keys #=> ["a", "b", "c", "d"]
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_keys(hash)
|
|
VALUE hash;
|
|
{
|
|
VALUE ary;
|
|
|
|
ary = rb_ary_new();
|
|
st_foreach(RHASH(hash)->tbl, keys_i, ary);
|
|
|
|
return ary;
|
|
}
|
|
|
|
static int
|
|
values_i(key, value, ary)
|
|
VALUE key, value, ary;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_ary_push(ary, value);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.values => array
|
|
*
|
|
* Returns a new array populated with the values from <i>hsh</i>. See
|
|
* also <code>Hash#keys</code>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200, "c" => 300 }
|
|
* h.values #=> [100, 200, 300]
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_values(hash)
|
|
VALUE hash;
|
|
{
|
|
VALUE ary;
|
|
|
|
ary = rb_ary_new();
|
|
st_foreach(RHASH(hash)->tbl, values_i, ary);
|
|
|
|
return ary;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.has_key?(key) => true or false
|
|
* hsh.include?(key) => true or false
|
|
* hsh.key?(key) => true or false
|
|
* hsh.member?(key) => true or false
|
|
*
|
|
* Returns <code>true</code> if the given key is present in <i>hsh</i>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.has_key?("a") #=> true
|
|
* h.has_key?("z") #=> false
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_has_key(hash, key)
|
|
VALUE hash;
|
|
VALUE key;
|
|
{
|
|
if (st_lookup(RHASH(hash)->tbl, key, 0)) {
|
|
return Qtrue;
|
|
}
|
|
return Qfalse;
|
|
}
|
|
|
|
static int
|
|
rb_hash_search_value(key, value, data)
|
|
VALUE key, value, *data;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
if (rb_equal(value, data[1])) {
|
|
data[0] = Qtrue;
|
|
return ST_STOP;
|
|
}
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.has_value?(value) => true or false
|
|
* hsh.value?(value) => true or false
|
|
*
|
|
* Returns <code>true</code> if the given value is present for some key
|
|
* in <i>hsh</i>.
|
|
*
|
|
* h = { "a" => 100, "b" => 200 }
|
|
* h.has_value?(100) #=> true
|
|
* h.has_value?(999) #=> false
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_has_value(hash, val)
|
|
VALUE hash;
|
|
VALUE val;
|
|
{
|
|
VALUE data[2];
|
|
|
|
data[0] = Qfalse;
|
|
data[1] = val;
|
|
st_foreach(RHASH(hash)->tbl, rb_hash_search_value, (st_data_t)data);
|
|
return data[0];
|
|
}
|
|
|
|
struct equal_data {
|
|
int result;
|
|
st_table *tbl;
|
|
};
|
|
|
|
static int
|
|
equal_i(key, val1, data)
|
|
VALUE key, val1;
|
|
struct equal_data *data;
|
|
{
|
|
VALUE val2;
|
|
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
if (!st_lookup(data->tbl, key, &val2)) {
|
|
data->result = Qfalse;
|
|
return ST_STOP;
|
|
}
|
|
if (!rb_equal(val1, val2)) {
|
|
data->result = Qfalse;
|
|
return ST_STOP;
|
|
}
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
static VALUE
|
|
hash_equal(hash1, hash2, eql)
|
|
VALUE hash1, hash2;
|
|
int eql; /* compare default value if true */
|
|
{
|
|
struct equal_data data;
|
|
|
|
if (hash1 == hash2) return Qtrue;
|
|
if (TYPE(hash2) != T_HASH) {
|
|
if (!rb_respond_to(hash2, rb_intern("to_hash"))) {
|
|
return Qfalse;
|
|
}
|
|
return rb_equal(hash2, hash1);
|
|
}
|
|
if (RHASH(hash1)->tbl->num_entries != RHASH(hash2)->tbl->num_entries)
|
|
return Qfalse;
|
|
if (eql) {
|
|
if (!(rb_equal(RHASH(hash1)->ifnone, RHASH(hash2)->ifnone) &&
|
|
FL_TEST(hash1, HASH_PROC_DEFAULT) == FL_TEST(hash2, HASH_PROC_DEFAULT)))
|
|
return Qfalse;
|
|
}
|
|
|
|
data.tbl = RHASH(hash2)->tbl;
|
|
data.result = Qtrue;
|
|
st_foreach(RHASH(hash1)->tbl, equal_i, (st_data_t)&data);
|
|
|
|
return data.result;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh == other_hash => true or false
|
|
*
|
|
* Equality---Two hashes are equal if they each contain the same number
|
|
* of keys and if each key-value pair is equal to (according to
|
|
* <code>Object#==</code>) the corresponding elements in the other
|
|
* hash.
|
|
*
|
|
* h1 = { "a" => 1, "c" => 2 }
|
|
* h2 = { 7 => 35, "c" => 2, "a" => 1 }
|
|
* h3 = { "a" => 1, "c" => 2, 7 => 35 }
|
|
* h4 = { "a" => 1, "d" => 2, "f" => 35 }
|
|
* h1 == h2 #=> false
|
|
* h2 == h3 #=> true
|
|
* h3 == h4 #=> false
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_equal(hash1, hash2)
|
|
VALUE hash1, hash2;
|
|
{
|
|
return hash_equal(hash1, hash2, Qfalse);
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.eql?(other_hash) => true or false
|
|
*
|
|
* Returns true if two hashes are equal, i.e they have same key-value set,
|
|
* and same default values.
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_eql(hash1, hash2)
|
|
VALUE hash1, hash2;
|
|
{
|
|
return hash_equal(hash1, hash2, Qtrue);
|
|
}
|
|
|
|
static int
|
|
rb_hash_invert_i(key, value, hash)
|
|
VALUE key, value;
|
|
VALUE hash;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_hash_aset(hash, value, key);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.invert -> aHash
|
|
*
|
|
* Returns a new hash created by using <i>hsh</i>'s values as keys, and
|
|
* the keys as values.
|
|
*
|
|
* h = { "n" => 100, "m" => 100, "y" => 300, "d" => 200, "a" => 0 }
|
|
* h.invert #=> {0=>"a", 100=>"n", 200=>"d", 300=>"y"}
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_invert(hash)
|
|
VALUE hash;
|
|
{
|
|
VALUE h = rb_hash_new();
|
|
|
|
st_foreach(RHASH(hash)->tbl, rb_hash_invert_i, h);
|
|
return h;
|
|
}
|
|
|
|
static int
|
|
rb_hash_update_i(key, value, hash)
|
|
VALUE key, value;
|
|
VALUE hash;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
rb_hash_aset(hash, key, value);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
static int
|
|
rb_hash_update_block_i(key, value, hash)
|
|
VALUE key, value;
|
|
VALUE hash;
|
|
{
|
|
if (key == Qundef) return ST_CONTINUE;
|
|
if (rb_hash_has_key(hash, key)) {
|
|
value = rb_yield_values(3, key, rb_hash_aref(hash, key), value);
|
|
}
|
|
rb_hash_aset(hash, key, value);
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.merge!(other_hash) => hsh
|
|
* hsh.update(other_hash) => hsh
|
|
*
|
|
* Adds the contents of <i>other_hash</i> to <i>hsh</i>, overwriting
|
|
* entries with duplicate keys with those from <i>other_hash</i>.
|
|
*
|
|
* h1 = { "a" => 100, "b" => 200 }
|
|
* h2 = { "b" => 254, "c" => 300 }
|
|
* h1.merge!(h2) #=> {"a"=>100, "b"=>254, "c"=>300}
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_update(hash1, hash2)
|
|
VALUE hash1, hash2;
|
|
{
|
|
hash2 = to_hash(hash2);
|
|
if (rb_block_given_p()) {
|
|
st_foreach(RHASH(hash2)->tbl, rb_hash_update_block_i, hash1);
|
|
}
|
|
else {
|
|
st_foreach(RHASH(hash2)->tbl, rb_hash_update_i, hash1);
|
|
}
|
|
return hash1;
|
|
}
|
|
|
|
/*
|
|
* call-seq:
|
|
* hsh.merge(other_hash) -> a_hash
|
|
*
|
|
* Returns a new hash containing the contents of <i>other_hash</i> and
|
|
* the contents of <i>hsh</i>, overwriting entries in <i>hsh</i> with
|
|
* duplicate keys with those from <i>other_hash</i>.
|
|
*
|
|
* h1 = { "a" => 100, "b" => 200 }
|
|
* h2 = { "b" => 254, "c" => 300 }
|
|
* h1.merge(h2) #=> {"a"=>100, "b"=>254, "c"=>300}
|
|
* h1 #=> {"a"=>100, "b"=>200}
|
|
*
|
|
*/
|
|
|
|
static VALUE
|
|
rb_hash_merge(hash1, hash2)
|
|
VALUE hash1, hash2;
|
|
{
|
|
return rb_hash_update(rb_obj_dup(hash1), hash2);
|
|
}
|
|
|
|
static int path_tainted = -1;
|
|
|
|
static char **origenviron;
|
|
#ifdef _WIN32
|
|
#define GET_ENVIRON(e) (e = rb_w32_get_environ())
|
|
#define FREE_ENVIRON(e) rb_w32_free_environ(e)
|
|
static char **my_environ;
|
|
#undef environ
|
|
#define environ my_environ
|
|
#elif defined(__APPLE__)
|
|
#undef environ
|
|
#define environ (*_NSGetEnviron())
|
|
#define GET_ENVIRON(e) (e)
|
|
#define FREE_ENVIRON(e)
|
|
#else
|
|
extern char **environ;
|
|
#define GET_ENVIRON(e) (e)
|
|
#define FREE_ENVIRON(e)
|
|
#endif
|
|
|
|
static VALUE
|
|
env_str_new(ptr, len)
|
|
const char *ptr;
|
|
long len;
|
|
{
|
|
VALUE str = rb_tainted_str_new(ptr, len);
|
|
|
|
rb_obj_freeze(str);
|
|
return str;
|
|
}
|
|
|
|
static VALUE
|
|
env_str_new2(ptr)
|
|
const char *ptr;
|
|
{
|
|
if (!ptr) return Qnil;
|
|
return env_str_new(ptr, strlen(ptr));
|
|
}
|
|
|
|
static VALUE
|
|
env_delete(obj, name)
|
|
VALUE obj, name;
|
|
{
|
|
char *nam, *val;
|
|
|
|
rb_secure(4);
|
|
SafeStringValue(name);
|
|
nam = RSTRING(name)->ptr;
|
|
if (strlen(nam) != RSTRING(name)->len) {
|
|
rb_raise(rb_eArgError, "bad environment variable name");
|
|
}
|
|
val = getenv(nam);
|
|
if (val) {
|
|
VALUE value = env_str_new2(val);
|
|
|
|
ruby_setenv(nam, 0);
|
|
#ifdef ENV_IGNORECASE
|
|
if (strcasecmp(nam, PATH_ENV) == 0)
|
|
#else
|
|
if (strcmp(nam, PATH_ENV) == 0)
|
|
#endif
|
|
{
|
|
path_tainted = 0;
|
|
}
|
|
return value;
|
|
}
|
|
return Qnil;
|
|
}
|
|
|
|
static VALUE
|
|
env_delete_m(obj, name)
|
|
VALUE obj, name;
|
|
{
|
|
VALUE val;
|
|
|
|
val = env_delete(obj, name);
|
|
if (NIL_P(val) && rb_block_given_p()) rb_yield(name);
|
|
return val;
|
|
}
|
|
|
|
static VALUE
|
|
rb_f_getenv(obj, name)
|
|
VALUE obj, name;
|
|
{
|
|
char *nam, *env;
|
|
|
|
StringValue(name);
|
|
nam = RSTRING(name)->ptr;
|
|
if (strlen(nam) != RSTRING(name)->len) {
|
|
rb_raise(rb_eArgError, "bad environment variable name");
|
|
}
|
|
env = getenv(nam);
|
|
if (env) {
|
|
#ifdef ENV_IGNORECASE
|
|
if (strcasecmp(nam, PATH_ENV) == 0 && !rb_env_path_tainted())
|
|
#else
|
|
if (strcmp(nam, PATH_ENV) == 0 && !rb_env_path_tainted())
|
|
#endif
|
|
{
|
|
VALUE str = rb_str_new2(env);
|
|
|
|
rb_obj_freeze(str);
|
|
return str;
|
|
}
|
|
return env_str_new2(env);
|
|
}
|
|
return Qnil;
|
|
}
|
|
|
|
static VALUE
|
|
env_fetch(argc, argv)
|
|
int argc;
|
|
VALUE *argv;
|
|
{
|
|
VALUE key, if_none;
|
|
long block_given;
|
|
char *nam, *env;
|
|
|
|
rb_scan_args(argc, argv, "11", &key, &if_none);
|
|
block_given = rb_block_given_p();
|
|
if (block_given && argc == 2) {
|
|
rb_warn("block supersedes default value argument");
|
|
}
|
|
StringValue(key);
|
|
nam = RSTRING(key)->ptr;
|
|
if (strlen(nam) != RSTRING(key)->len) {
|
|
rb_raise(rb_eArgError, "bad environment variable name");
|
|
}
|
|
env = getenv(nam);
|
|
if (!env) {
|
|
if (block_given) return rb_yield(key);
|
|
if (argc == 1) {
|
|
rb_raise(rb_eIndexError, "key not found");
|
|
}
|
|
return if_none;
|
|
}
|
|
#ifdef ENV_IGNORECASE
|
|
if (strcasecmp(nam, PATH_ENV) == 0 && !rb_env_path_tainted())
|
|
#else
|
|
if (strcmp(nam, PATH_ENV) == 0 && !rb_env_path_tainted())
|
|
#endif
|
|
return rb_str_new2(env);
|
|
return env_str_new2(env);
|
|
}
|
|
|
|
static void
|
|
path_tainted_p(path)
|
|
char *path;
|
|
{
|
|
path_tainted = rb_path_check(path)?0:1;
|
|
}
|
|
|
|
int
|
|
rb_env_path_tainted()
|
|
{
|
|
if (path_tainted < 0) {
|
|
path_tainted_p(getenv(PATH_ENV));
|
|
}
|
|
return path_tainted;
|
|
}
|
|
|
|
static int
|
|
envix(nam)
|
|
const char *nam;
|
|
{
|
|
register int i, len = strlen(nam);
|
|
char **env;
|
|
|
|
env = GET_ENVIRON(environ);
|
|
for (i = 0; env[i]; i++) {
|
|
if (
|
|
#ifdef ENV_IGNORECASE
|
|
strncasecmp(env[i],nam,len) == 0
|
|
#else
|
|
memcmp(env[i],nam,len) == 0
|
|
#endif
|
|
&& env[i][len] == '=')
|
|
break; /* memcmp must come first to avoid */
|
|
} /* potential SEGV's */
|
|
FREE_ENVIRON(environ);
|
|
return i;
|
|
}
|
|
|
|
void
|
|
ruby_setenv(name, value)
|
|
const char *name;
|
|
const char *value;
|
|
{
|
|
#if defined(_WIN32)
|
|
/* The sane way to deal with the environment.
|
|
* Has these advantages over putenv() & co.:
|
|
* * enables us to store a truly empty value in the
|
|
* environment (like in UNIX).
|
|
* * we don't have to deal with RTL globals, bugs and leaks.
|
|
* * Much faster.
|
|
* Why you may want to enable USE_WIN32_RTL_ENV:
|
|
* * environ[] and RTL functions will not reflect changes,
|
|
* which might be an issue if extensions want to access
|
|
* the env. via RTL. This cuts both ways, since RTL will
|
|
* not see changes made by extensions that call the Win32
|
|
* functions directly, either.
|
|
* GSAR 97-06-07
|
|
*
|
|
* REMARK: USE_WIN32_RTL_ENV is already obsoleted since we don't use
|
|
* RTL's environ global variable directly yet.
|
|
*/
|
|
SetEnvironmentVariable(name,value);
|
|
#elif defined __CYGWIN__
|
|
#undef setenv
|
|
#undef unsetenv
|
|
if (value)
|
|
setenv(name,value,1);
|
|
else
|
|
unsetenv(name);
|
|
#else /* WIN32 */
|
|
|
|
int i=envix(name); /* where does it go? */
|
|
|
|
if (environ == origenviron) { /* need we copy environment? */
|
|
int j;
|
|
int max;
|
|
char **tmpenv;
|
|
|
|
for (max = i; environ[max]; max++) ;
|
|
tmpenv = ALLOC_N(char*, max+2);
|
|
for (j=0; j<max; j++) /* copy environment */
|
|
tmpenv[j] = strdup(environ[j]);
|
|
tmpenv[max] = 0;
|
|
environ = tmpenv; /* tell exec where it is now */
|
|
}
|
|
if (!value) {
|
|
if (environ != origenviron) {
|
|
char **envp = origenviron;
|
|
while (*envp && *envp != environ[i]) envp++;
|
|
if (!*envp)
|
|
free(environ[i]);
|
|
}
|
|
while (environ[i]) {
|
|
environ[i] = environ[i+1];
|
|
i++;
|
|
}
|
|
return;
|
|
}
|
|
if (!environ[i]) { /* does not exist yet */
|
|
REALLOC_N(environ, char*, i+2); /* just expand it a bit */
|
|
environ[i+1] = 0; /* make sure it's null terminated */
|
|
}
|
|
else {
|
|
if (environ[i] != origenviron[i])
|
|
free(environ[i]);
|
|
}
|
|
environ[i] = ALLOC_N(char, strlen(name) + strlen(value) + 2);
|
|
#ifndef MSDOS
|
|
sprintf(environ[i],"%s=%s",name,value); /* all that work just for this */
|
|
#else
|
|
/* MS-DOS requires environment variable names to be in uppercase */
|
|
/* [Tom Dinger, 27 August 1990: Well, it doesn't _require_ it, but
|
|
* some utilities and applications may break because they only look
|
|
* for upper case strings. (Fixed strupr() bug here.)]
|
|
*/
|
|
strcpy(environ[i],name); strupr(environ[i]);
|
|
sprintf(environ[i] + strlen(name),"=%s", value);
|
|
#endif /* MSDOS */
|
|
|
|
#endif /* WIN32 */
|
|
}
|
|
|
|
void
|
|
ruby_unsetenv(name)
|
|
const char *name;
|
|
{
|
|
ruby_setenv(name, 0);
|
|
}
|
|
|
|
static VALUE
|
|
env_aset(obj, nm, val)
|
|
VALUE obj, nm, val;
|
|
{
|
|
char *name, *value;
|
|
|
|
if (rb_safe_level() >= 4) {
|
|
rb_raise(rb_eSecurityError, "cannot change environment variable");
|
|
}
|
|
|
|
if (NIL_P(val)) {
|
|
env_delete(obj, nm);
|
|
return Qnil;
|
|
}
|
|
|
|
StringValue(nm);
|
|
StringValue(val);
|
|
name = RSTRING(nm)->ptr;
|
|
value = RSTRING(val)->ptr;
|
|
if (strlen(name) != RSTRING(nm)->len)
|
|
rb_raise(rb_eArgError, "bad environment variable name");
|
|
if (strlen(value) != RSTRING(val)->len)
|
|
rb_raise(rb_eArgError, "bad environment variable value");
|
|
|
|
ruby_setenv(name, value);
|
|
#ifdef ENV_IGNORECASE
|
|
if (strcasecmp(name, PATH_ENV) == 0) {
|
|
#else
|
|
if (strcmp(name, PATH_ENV) == 0) {
|
|
#endif
|
|
if (OBJ_TAINTED(val)) {
|
|
/* already tainted, no check */
|
|
path_tainted = 1;
|
|
return val;
|
|
}
|
|
else {
|
|
path_tainted_p(value);
|
|
}
|
|
}
|
|
return val;
|
|
}
|
|
|
|
static VALUE
|
|
env_keys()
|
|
{
|
|
char **env;
|
|
VALUE ary = rb_ary_new();
|
|
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s) {
|
|
rb_ary_push(ary, env_str_new(*env, s-*env));
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return ary;
|
|
}
|
|
|
|
static VALUE
|
|
env_each_key(ehash)
|
|
VALUE ehash;
|
|
{
|
|
VALUE keys = env_keys();
|
|
long i;
|
|
|
|
for (i=0; i<RARRAY(keys)->len; i++) {
|
|
rb_yield(RARRAY(keys)->ptr[i]);
|
|
}
|
|
return ehash;
|
|
}
|
|
|
|
static VALUE
|
|
env_values()
|
|
{
|
|
char **env;
|
|
VALUE ary = rb_ary_new();
|
|
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s) {
|
|
rb_ary_push(ary, env_str_new2(s+1));
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return ary;
|
|
}
|
|
|
|
static VALUE
|
|
env_each_value(ehash)
|
|
VALUE ehash;
|
|
{
|
|
VALUE values = env_values();
|
|
long i;
|
|
|
|
for (i=0; i<RARRAY(values)->len; i++) {
|
|
rb_yield(RARRAY(values)->ptr[i]);
|
|
}
|
|
return ehash;
|
|
}
|
|
|
|
static VALUE
|
|
env_each_i(ehash, values)
|
|
VALUE ehash;
|
|
int values;
|
|
{
|
|
char **env;
|
|
VALUE ary = rb_ary_new();
|
|
long i;
|
|
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s) {
|
|
rb_ary_push(ary, env_str_new(*env, s-*env));
|
|
rb_ary_push(ary, env_str_new2(s+1));
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
|
|
for (i=0; i<RARRAY(ary)->len; i+=2) {
|
|
if (values) {
|
|
rb_yield_values(2, RARRAY(ary)->ptr[i], RARRAY(ary)->ptr[i+1]);
|
|
}
|
|
else {
|
|
rb_yield(rb_assoc_new(RARRAY(ary)->ptr[i], RARRAY(ary)->ptr[i+1]));
|
|
}
|
|
}
|
|
return ehash;
|
|
}
|
|
|
|
static VALUE
|
|
env_each(ehash)
|
|
VALUE ehash;
|
|
{
|
|
return env_each_i(ehash, Qfalse);
|
|
}
|
|
|
|
static VALUE
|
|
env_each_pair(ehash)
|
|
VALUE ehash;
|
|
{
|
|
return env_each_i(ehash, Qtrue);
|
|
}
|
|
|
|
static VALUE
|
|
env_reject_bang()
|
|
{
|
|
volatile VALUE keys;
|
|
long i;
|
|
int del = 0;
|
|
|
|
rb_secure(4);
|
|
keys = env_keys();
|
|
|
|
for (i=0; i<RARRAY(keys)->len; i++) {
|
|
VALUE val = rb_f_getenv(Qnil, RARRAY(keys)->ptr[i]);
|
|
if (!NIL_P(val)) {
|
|
if (RTEST(rb_yield_values(2, RARRAY(keys)->ptr[i], val))) {
|
|
FL_UNSET(RARRAY(keys)->ptr[i], FL_TAINT);
|
|
env_delete(Qnil, RARRAY(keys)->ptr[i]);
|
|
del++;
|
|
}
|
|
}
|
|
}
|
|
if (del == 0) return Qnil;
|
|
return envtbl;
|
|
}
|
|
|
|
static VALUE
|
|
env_delete_if()
|
|
{
|
|
env_reject_bang();
|
|
return envtbl;
|
|
}
|
|
|
|
static VALUE
|
|
env_values_at(argc, argv)
|
|
int argc;
|
|
VALUE *argv;
|
|
{
|
|
VALUE result = rb_ary_new();
|
|
long i;
|
|
|
|
for (i=0; i<argc; i++) {
|
|
rb_ary_push(result, rb_f_getenv(Qnil, argv[i]));
|
|
}
|
|
return result;
|
|
}
|
|
|
|
static VALUE
|
|
env_select()
|
|
{
|
|
VALUE result;
|
|
char **env;
|
|
|
|
result = rb_ary_new();
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s) {
|
|
VALUE k = env_str_new(*env, s-*env);
|
|
VALUE v = env_str_new2(s+1);
|
|
if (RTEST(rb_yield_values(2, k, v))) {
|
|
rb_ary_push(result, rb_assoc_new(k, v));
|
|
}
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
|
|
return result;
|
|
}
|
|
|
|
static VALUE
|
|
env_clear()
|
|
{
|
|
volatile VALUE keys;
|
|
long i;
|
|
|
|
rb_secure(4);
|
|
keys = env_keys();
|
|
|
|
for (i=0; i<RARRAY(keys)->len; i++) {
|
|
VALUE val = rb_f_getenv(Qnil, RARRAY(keys)->ptr[i]);
|
|
if (!NIL_P(val)) {
|
|
env_delete(Qnil, RARRAY(keys)->ptr[i]);
|
|
}
|
|
}
|
|
return envtbl;
|
|
}
|
|
|
|
static VALUE
|
|
env_to_s()
|
|
{
|
|
return rb_str_new2("ENV");
|
|
}
|
|
|
|
static VALUE
|
|
env_inspect()
|
|
{
|
|
char **env;
|
|
VALUE str = rb_str_buf_new2("{");
|
|
VALUE i;
|
|
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
|
|
if (env != environ) {
|
|
rb_str_buf_cat2(str, ", ");
|
|
}
|
|
if (s) {
|
|
rb_str_buf_cat2(str, "\"");
|
|
rb_str_buf_cat(str, *env, s-*env);
|
|
rb_str_buf_cat2(str, "\"=>");
|
|
i = rb_inspect(rb_str_new2(s+1));
|
|
rb_str_buf_append(str, i);
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
rb_str_buf_cat2(str, "}");
|
|
OBJ_TAINT(str);
|
|
|
|
return str;
|
|
}
|
|
|
|
static VALUE
|
|
env_to_a()
|
|
{
|
|
char **env;
|
|
VALUE ary = rb_ary_new();
|
|
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s) {
|
|
rb_ary_push(ary, rb_assoc_new(env_str_new(*env, s-*env),
|
|
env_str_new2(s+1)));
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return ary;
|
|
}
|
|
|
|
static VALUE
|
|
env_none()
|
|
{
|
|
return Qnil;
|
|
}
|
|
|
|
static VALUE
|
|
env_size()
|
|
{
|
|
int i;
|
|
char **env;
|
|
|
|
env = GET_ENVIRON(environ);
|
|
for(i=0; env[i]; i++)
|
|
;
|
|
FREE_ENVIRON(environ);
|
|
return INT2FIX(i);
|
|
}
|
|
|
|
static VALUE
|
|
env_empty_p()
|
|
{
|
|
char **env;
|
|
|
|
env = GET_ENVIRON(environ);
|
|
if (env[0] == 0) {
|
|
FREE_ENVIRON(environ);
|
|
return Qtrue;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return Qfalse;
|
|
}
|
|
|
|
static VALUE
|
|
env_has_key(env, key)
|
|
VALUE env, key;
|
|
{
|
|
char *s;
|
|
|
|
s = StringValuePtr(key);
|
|
if (strlen(s) != RSTRING(key)->len)
|
|
rb_raise(rb_eArgError, "bad environment variable name");
|
|
if (getenv(s)) return Qtrue;
|
|
return Qfalse;
|
|
}
|
|
|
|
static VALUE
|
|
env_has_value(dmy, value)
|
|
VALUE dmy, value;
|
|
{
|
|
char **env;
|
|
|
|
if (TYPE(value) != T_STRING) return Qfalse;
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s++) {
|
|
long len = strlen(s);
|
|
if (RSTRING(value)->len == len && strncmp(s, RSTRING(value)->ptr, len) == 0) {
|
|
FREE_ENVIRON(environ);
|
|
return Qtrue;
|
|
}
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return Qfalse;
|
|
}
|
|
|
|
static VALUE
|
|
env_index(dmy, value)
|
|
VALUE dmy, value;
|
|
{
|
|
char **env;
|
|
VALUE str;
|
|
|
|
StringValue(value);
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s++) {
|
|
long len = strlen(s);
|
|
if (RSTRING(value)->len == len && strncmp(s, RSTRING(value)->ptr, len) == 0) {
|
|
str = env_str_new(*env, s-*env-1);
|
|
FREE_ENVIRON(environ);
|
|
return str;
|
|
}
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return Qnil;
|
|
}
|
|
|
|
static VALUE
|
|
env_to_hash()
|
|
{
|
|
char **env;
|
|
VALUE hash = rb_hash_new();
|
|
|
|
env = GET_ENVIRON(environ);
|
|
while (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s) {
|
|
rb_hash_aset(hash, env_str_new(*env, s-*env),
|
|
env_str_new2(s+1));
|
|
}
|
|
env++;
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return hash;
|
|
}
|
|
|
|
static VALUE
|
|
env_reject()
|
|
{
|
|
return rb_hash_delete_if(env_to_hash());
|
|
}
|
|
|
|
static VALUE
|
|
env_shift()
|
|
{
|
|
char **env;
|
|
|
|
env = GET_ENVIRON(environ);
|
|
if (*env) {
|
|
char *s = strchr(*env, '=');
|
|
if (s) {
|
|
VALUE key = env_str_new(*env, s-*env);
|
|
VALUE val = env_str_new2(getenv(RSTRING(key)->ptr));
|
|
env_delete(Qnil, key);
|
|
return rb_assoc_new(key, val);
|
|
}
|
|
}
|
|
FREE_ENVIRON(environ);
|
|
return Qnil;
|
|
}
|
|
|
|
static VALUE
|
|
env_invert()
|
|
{
|
|
return rb_hash_invert(env_to_hash());
|
|
}
|
|
|
|
static int
|
|
env_replace_i(key, val, keys)
|
|
VALUE key, val, keys;
|
|
{
|
|
if (key != Qundef) {
|
|
env_aset(Qnil, key, val);
|
|
if (rb_ary_includes(keys, key)) {
|
|
rb_ary_delete(keys, key);
|
|
}
|
|
}
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
static VALUE
|
|
env_replace(env, hash)
|
|
VALUE env, hash;
|
|
{
|
|
volatile VALUE keys = env_keys();
|
|
long i;
|
|
|
|
if (env == hash) return env;
|
|
hash = to_hash(hash);
|
|
st_foreach(RHASH(hash)->tbl, env_replace_i, keys);
|
|
|
|
for (i=0; i<RARRAY(keys)->len; i++) {
|
|
env_delete(env, RARRAY(keys)->ptr[i]);
|
|
}
|
|
return env;
|
|
}
|
|
|
|
static int
|
|
env_update_i(key, val)
|
|
VALUE key, val;
|
|
{
|
|
if (key != Qundef) {
|
|
if (rb_block_given_p()) {
|
|
val = rb_yield_values(3, key, rb_f_getenv(Qnil, key), val);
|
|
}
|
|
env_aset(Qnil, key, val);
|
|
}
|
|
return ST_CONTINUE;
|
|
}
|
|
|
|
static VALUE
|
|
env_update(env, hash)
|
|
VALUE env, hash;
|
|
{
|
|
if (env == hash) return env;
|
|
hash = to_hash(hash);
|
|
st_foreach(RHASH(hash)->tbl, env_update_i, 0);
|
|
return env;
|
|
}
|
|
|
|
/*
|
|
* A <code>Hash</code> is a collection of key-value pairs. It is
|
|
* similar to an <code>Array</code>, except that indexing is done via
|
|
* arbitrary keys of any object type, not an integer index. The order
|
|
* in which you traverse a hash by either key or value may seem
|
|
* arbitrary, and will generally not be in the insertion order.
|
|
*
|
|
* Hashes have a <em>default value</em> that is returned when accessing
|
|
* keys that do not exist in the hash. By default, that value is
|
|
* <code>nil</code>.
|
|
*
|
|
*/
|
|
|
|
void
|
|
Init_Hash()
|
|
{
|
|
id_hash = rb_intern("hash");
|
|
id_call = rb_intern("call");
|
|
id_default = rb_intern("default");
|
|
|
|
rb_cHash = rb_define_class("Hash", rb_cObject);
|
|
|
|
rb_include_module(rb_cHash, rb_mEnumerable);
|
|
|
|
rb_define_alloc_func(rb_cHash, hash_alloc);
|
|
rb_define_singleton_method(rb_cHash, "[]", rb_hash_s_create, -1);
|
|
rb_define_method(rb_cHash,"initialize", rb_hash_initialize, -1);
|
|
rb_define_method(rb_cHash,"initialize_copy", rb_hash_replace, 1);
|
|
rb_define_method(rb_cHash,"rehash", rb_hash_rehash, 0);
|
|
|
|
rb_define_method(rb_cHash,"to_hash", rb_hash_to_hash, 0);
|
|
rb_define_method(rb_cHash,"to_a", rb_hash_to_a, 0);
|
|
rb_define_method(rb_cHash,"to_s", rb_hash_to_s, 0);
|
|
rb_define_method(rb_cHash,"inspect", rb_hash_inspect, 0);
|
|
|
|
rb_define_method(rb_cHash,"==", rb_hash_equal, 1);
|
|
rb_define_method(rb_cHash,"eql?", rb_hash_eql, 1);
|
|
rb_define_method(rb_cHash,"[]", rb_hash_aref, 1);
|
|
rb_define_method(rb_cHash,"fetch", rb_hash_fetch, -1);
|
|
rb_define_method(rb_cHash,"[]=", rb_hash_aset, 2);
|
|
rb_define_method(rb_cHash,"store", rb_hash_aset, 2);
|
|
rb_define_method(rb_cHash,"default", rb_hash_default, -1);
|
|
rb_define_method(rb_cHash,"default=", rb_hash_set_default, 1);
|
|
rb_define_method(rb_cHash,"default_proc", rb_hash_default_proc, 0);
|
|
rb_define_method(rb_cHash,"index", rb_hash_index, 1);
|
|
rb_define_method(rb_cHash,"size", rb_hash_size, 0);
|
|
rb_define_method(rb_cHash,"length", rb_hash_size, 0);
|
|
rb_define_method(rb_cHash,"empty?", rb_hash_empty_p, 0);
|
|
|
|
rb_define_method(rb_cHash,"each", rb_hash_each, 0);
|
|
rb_define_method(rb_cHash,"each_value", rb_hash_each_value, 0);
|
|
rb_define_method(rb_cHash,"each_key", rb_hash_each_key, 0);
|
|
rb_define_method(rb_cHash,"each_pair", rb_hash_each_pair, 0);
|
|
rb_define_method(rb_cHash,"sort", rb_hash_sort, 0);
|
|
|
|
rb_define_method(rb_cHash,"keys", rb_hash_keys, 0);
|
|
rb_define_method(rb_cHash,"values", rb_hash_values, 0);
|
|
rb_define_method(rb_cHash,"values_at", rb_hash_values_at, -1);
|
|
|
|
rb_define_method(rb_cHash,"shift", rb_hash_shift, 0);
|
|
rb_define_method(rb_cHash,"delete", rb_hash_delete, 1);
|
|
rb_define_method(rb_cHash,"delete_if", rb_hash_delete_if, 0);
|
|
rb_define_method(rb_cHash,"select", rb_hash_select, 0);
|
|
rb_define_method(rb_cHash,"reject", rb_hash_reject, 0);
|
|
rb_define_method(rb_cHash,"reject!", rb_hash_reject_bang, 0);
|
|
rb_define_method(rb_cHash,"clear", rb_hash_clear, 0);
|
|
rb_define_method(rb_cHash,"invert", rb_hash_invert, 0);
|
|
rb_define_method(rb_cHash,"update", rb_hash_update, 1);
|
|
rb_define_method(rb_cHash,"replace", rb_hash_replace, 1);
|
|
rb_define_method(rb_cHash,"merge!", rb_hash_update, 1);
|
|
rb_define_method(rb_cHash,"merge", rb_hash_merge, 1);
|
|
|
|
rb_define_method(rb_cHash,"include?", rb_hash_has_key, 1);
|
|
rb_define_method(rb_cHash,"member?", rb_hash_has_key, 1);
|
|
rb_define_method(rb_cHash,"has_key?", rb_hash_has_key, 1);
|
|
rb_define_method(rb_cHash,"has_value?", rb_hash_has_value, 1);
|
|
rb_define_method(rb_cHash,"key?", rb_hash_has_key, 1);
|
|
rb_define_method(rb_cHash,"value?", rb_hash_has_value, 1);
|
|
|
|
#ifndef __MACOS__ /* environment variables nothing on MacOS. */
|
|
origenviron = environ;
|
|
envtbl = rb_obj_alloc(rb_cObject);
|
|
rb_extend_object(envtbl, rb_mEnumerable);
|
|
|
|
rb_define_singleton_method(envtbl,"[]", rb_f_getenv, 1);
|
|
rb_define_singleton_method(envtbl,"fetch", env_fetch, -1);
|
|
rb_define_singleton_method(envtbl,"[]=", env_aset, 2);
|
|
rb_define_singleton_method(envtbl,"store", env_aset, 2);
|
|
rb_define_singleton_method(envtbl,"each", env_each, 0);
|
|
rb_define_singleton_method(envtbl,"each_pair", env_each_pair, 0);
|
|
rb_define_singleton_method(envtbl,"each_key", env_each_key, 0);
|
|
rb_define_singleton_method(envtbl,"each_value", env_each_value, 0);
|
|
rb_define_singleton_method(envtbl,"delete", env_delete_m, 1);
|
|
rb_define_singleton_method(envtbl,"delete_if", env_delete_if, 0);
|
|
rb_define_singleton_method(envtbl,"clear", env_clear, 0);
|
|
rb_define_singleton_method(envtbl,"reject", env_reject, 0);
|
|
rb_define_singleton_method(envtbl,"reject!", env_reject_bang, 0);
|
|
rb_define_singleton_method(envtbl,"select", env_select, 0);
|
|
rb_define_singleton_method(envtbl,"shift", env_shift, 0);
|
|
rb_define_singleton_method(envtbl,"invert", env_invert, 0);
|
|
rb_define_singleton_method(envtbl,"replace", env_replace, 1);
|
|
rb_define_singleton_method(envtbl,"update", env_update, 1);
|
|
rb_define_singleton_method(envtbl,"inspect", env_inspect, 0);
|
|
rb_define_singleton_method(envtbl,"rehash", env_none, 0);
|
|
rb_define_singleton_method(envtbl,"to_a", env_to_a, 0);
|
|
rb_define_singleton_method(envtbl,"to_s", env_to_s, 0);
|
|
rb_define_singleton_method(envtbl,"index", env_index, 1);
|
|
rb_define_singleton_method(envtbl,"size", env_size, 0);
|
|
rb_define_singleton_method(envtbl,"length", env_size, 0);
|
|
rb_define_singleton_method(envtbl,"empty?", env_empty_p, 0);
|
|
rb_define_singleton_method(envtbl,"keys", env_keys, 0);
|
|
rb_define_singleton_method(envtbl,"values", env_values, 0);
|
|
rb_define_singleton_method(envtbl,"values_at", env_values_at, -1);
|
|
rb_define_singleton_method(envtbl,"include?", env_has_key, 1);
|
|
rb_define_singleton_method(envtbl,"member?", env_has_key, 1);
|
|
rb_define_singleton_method(envtbl,"has_key?", env_has_key, 1);
|
|
rb_define_singleton_method(envtbl,"has_value?", env_has_value, 1);
|
|
rb_define_singleton_method(envtbl,"key?", env_has_key, 1);
|
|
rb_define_singleton_method(envtbl,"value?", env_has_value, 1);
|
|
rb_define_singleton_method(envtbl,"to_hash", env_to_hash, 0);
|
|
|
|
rb_define_global_const("ENV", envtbl);
|
|
#else /* __MACOS__ */
|
|
envtbl = rb_hash_s_new(0, NULL, rb_cHash);
|
|
rb_define_global_const("ENV", envtbl);
|
|
#endif /* ifndef __MACOS__ environment variables nothing on MacOS. */
|
|
}
|