ruby-changes:2672
From: ko1@a...
Date: 10 Dec 2007 06:44:59 +0900
Subject: [ruby-changes:2672] akr - Ruby:r14163 (trunk): * re.c (rb_reg_names): new method Regexp#names.
akr 2007-12-10 06:44:19 +0900 (Mon, 10 Dec 2007)
New Revision: 14163
Modified files:
trunk/ChangeLog
trunk/lib/pp.rb
trunk/re.c
trunk/test/ruby/test_regexp.rb
Log:
* re.c (rb_reg_names): new method Regexp#names.
(rb_reg_named_captures): new method Regexp#named_captures
(match_regexp): new method MatchData#regexp.
(match_names): new method MatchData#names.
* lib/pp.rb (MatchData#pretty_print): show names of named captures.
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/pp.rb?r1=14163&r2=14162
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/ChangeLog?r1=14163&r2=14162
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/test/ruby/test_regexp.rb?r1=14163&r2=14162
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/re.c?r1=14163&r2=14162
Index: re.c
===================================================================
--- re.c (revision 14162)
+++ re.c (revision 14163)
@@ -523,7 +523,85 @@
return INT2NUM(options);
}
+static int
+reg_names_iter(const OnigUChar *name, const OnigUChar *name_end,
+ int back_num, int *back_refs, OnigRegex regex, void *arg)
+{
+ VALUE ary = (VALUE)arg;
+ rb_ary_push(ary, rb_str_new((const char *)name, name_end-name));
+ return 0;
+}
+/*
+ * call-seq:
+ * rxp.names => [name1, name2, ...]
+ *
+ * Returns a list of names of captures as an array of strings.
+ *
+ * /(?<foo>.)(?<bar>.)(?<baz>.)/.names
+ * #=> ["foo", "bar", "baz"]
+ *
+ * /(?<foo>.)(?<foo>.)/.names
+ * #=> ["foo"]
+ *
+ * /(.)(.)/.names'
+ * #=> []
+ */
+
+static VALUE
+rb_reg_names(VALUE re)
+{
+ VALUE ary = rb_ary_new();
+ onig_foreach_name(RREGEXP(re)->ptr, reg_names_iter, (void*)ary);
+ return ary;
+}
+
+static int
+reg_named_captures_iter(const OnigUChar *name, const OnigUChar *name_end,
+ int back_num, int *back_refs, OnigRegex regex, void *arg)
+{
+ VALUE hash = (VALUE)arg;
+ VALUE ary = rb_ary_new2(back_num);
+ int i;
+
+ for(i = 0; i < back_num; i++)
+ rb_ary_store(ary, i, INT2NUM(back_refs[i]));
+
+ rb_hash_aset(hash, rb_str_new((const char*)name, name_end-name),ary);
+
+ return 0;
+}
+
+/*
+ * call-seq:
+ * rxp.named_captures => hash
+ *
+ * Returns a hash representing information about named captures of <i>rxp</i>.
+ *
+ * A key of the hash is a name of the named captures.
+ * A value of the hash is an array which is list of indexes of corresponding
+ * named captures.
+ *
+ * /(?<foo>.)(?<bar>.)/.named_captures
+ * #=> {"foo"=>[1], "bar"=>[2]}
+ *
+ * /(?<foo>.)(?<foo>.)/.named_captures'
+ * #=> {"foo"=>[1, 2]}
+ *
+ * If there are no named captures, an empty hash is returned.
+ *
+ * /(.)(.)/.named_captures'
+ * #=> {}
+ */
+
+static VALUE
+rb_reg_named_captures(VALUE re)
+{
+ VALUE hash = rb_hash_new();
+ onig_foreach_name(RREGEXP(re)->ptr, reg_named_captures_iter, (void*)hash);
+ return hash;
+}
+
static Regexp*
make_regexp(const char *s, long len, rb_encoding *enc, int flags, onig_errmsg_buffer err)
{
@@ -604,6 +682,42 @@
/*
+ * call-seq:
+ * mtch.regexp => regexp
+ *
+ * Returns the regexp.
+ *
+ * m = /a.*b/.match("abc")
+ * m.regexp #=> /a.*b/
+ */
+
+static VALUE
+match_regexp(VALUE match)
+{
+ return RMATCH(match)->regexp;
+}
+
+/*
+ * call-seq:
+ * mtch.names => [name1, name2, ...]
+ *
+ * Returns a list of names of captures as an array of strings.
+ * It is same as mtch.regexp.names.
+ *
+ * /(?<foo>.)(?<bar>.)(?<baz>.)/.match("hoge").names
+ * #=> ["foo", "bar", "baz"]
+ *
+ * m = /(?<x>.)(?<y>.)?/.match("a") #=> #<MatchData "a" x:"a" y:nil>
+ * m.names #=> ["x", "y"]
+ */
+
+static VALUE
+match_names(VALUE match)
+{
+ return rb_reg_names(RMATCH(match)->regexp);
+}
+
+/*
* call-seq:
* mtch.length => integer
* mtch.size => integer
@@ -1362,11 +1476,12 @@
int i;
int num_regs = RMATCH(match)->regs->num_regs;
struct backref_name_tag *names;
+ VALUE regexp = RMATCH(match)->regexp;
names = ALLOCA_N(struct backref_name_tag, num_regs);
MEMZERO(names, struct backref_name_tag, num_regs);
- onig_foreach_name(RREGEXP(RMATCH(match)->regexp)->ptr,
+ onig_foreach_name(RREGEXP(regexp)->ptr,
match_inspect_name_iter, names);
str = rb_str_buf_new2("#<");
@@ -2818,6 +2933,8 @@
rb_define_method(rb_cRegexp, "options", rb_reg_options_m, 0);
rb_define_method(rb_cRegexp, "encoding", rb_obj_encoding, 0); /* in encoding.c */
rb_define_method(rb_cRegexp, "fixed_encoding?", rb_reg_fixed_encoding_p, 0);
+ rb_define_method(rb_cRegexp, "names", rb_reg_names, 0);
+ rb_define_method(rb_cRegexp, "named_captures", rb_reg_named_captures, 0);
rb_define_const(rb_cRegexp, "IGNORECASE", INT2FIX(ONIG_OPTION_IGNORECASE));
rb_define_const(rb_cRegexp, "EXTENDED", INT2FIX(ONIG_OPTION_EXTEND));
@@ -2830,6 +2947,8 @@
rb_undef_method(CLASS_OF(rb_cMatch), "new");
rb_define_method(rb_cMatch, "initialize_copy", match_init_copy, 1);
+ rb_define_method(rb_cMatch, "regexp", match_regexp, 0);
+ rb_define_method(rb_cMatch, "names", match_names, 0);
rb_define_method(rb_cMatch, "size", match_size, 0);
rb_define_method(rb_cMatch, "length", match_size, 0);
rb_define_method(rb_cMatch, "offset", match_offset, 1);
Index: ChangeLog
===================================================================
--- ChangeLog (revision 14162)
+++ ChangeLog (revision 14163)
@@ -1,3 +1,12 @@
+Mon Dec 10 06:41:00 2007 Tanaka Akira <akr@f...>
+
+ * re.c (rb_reg_names): new method Regexp#names.
+ (rb_reg_named_captures): new method Regexp#named_captures
+ (match_regexp): new method MatchData#regexp.
+ (match_names): new method MatchData#names.
+
+ * lib/pp.rb (MatchData#pretty_print): show names of named captures.
+
Mon Dec 10 01:35:06 2007 Yukihiro Matsumoto <matz@r...>
* parse.y (expr): redefinable not (!) operator.
Index: lib/pp.rb
===================================================================
--- lib/pp.rb (revision 14162)
+++ lib/pp.rb (revision 14163)
@@ -473,10 +473,24 @@
class MatchData
def pretty_print(q)
+ nc = []
+ self.regexp.named_captures.each {|name, indexes|
+ indexes.each {|i| nc[i] = name }
+ }
q.object_group(self) {
q.breakable
- q.seplist(1..self.size, lambda { q.breakable }) {|i|
- q.pp self[i-1]
+ q.seplist(0...self.size, lambda { q.breakable }) {|i|
+ if i == 0
+ q.pp self[i]
+ else
+ if nc[i]
+ q.text nc[i]
+ else
+ q.pp i
+ end
+ q.text ':'
+ q.pp self[i]
+ end
}
}
end
Index: test/ruby/test_regexp.rb
===================================================================
--- test/ruby/test_regexp.rb (revision 14162)
+++ test/ruby/test_regexp.rb (revision 14163)
@@ -2,7 +2,9 @@
class TestRegexp < Test::Unit::TestCase
def test_ruby_dev_24643
- assert_nothing_raised("[ruby-dev:24643]") { /(?:(?:[a]*[a])?b)*a*$/ =~ "aabaaca" }
+ assert_nothing_raised("[ruby-dev:24643]") {
+ /(?:(?:[a]*[a])?b)*a*$/ =~ "aabaaca"
+ }
end
def test_ruby_talk_116455
@@ -58,7 +60,8 @@
assert_equal(8, m.end(:foo))
assert_equal([5,8], m.offset(:foo))
- assert_equal("aaa [amp] yyy", "aaa & yyy".sub(/&(?<foo>.*?);/, '[\k<foo>]'))
+ assert_equal("aaa [amp] yyy",
+ "aaa & yyy".sub(/&(?<foo>.*?);/, '[\k<foo>]'))
assert_equal('#<MatchData "& y" foo:"amp">',
/&(?<foo>.*?); (y)/.match("aaa & yyy").inspect)
@@ -72,9 +75,29 @@
/(?<id>[A-Za-z_]+)/ =~ "!abc"
assert_equal("abc", Regexp.last_match(:id))
- /a/ =~ "b"
+ /a/ =~ "b" # doesn't match.
assert_equal(nil, Regexp.last_match)
assert_equal(nil, Regexp.last_match(1))
assert_equal(nil, Regexp.last_match(:foo))
+
+ assert_equal(["foo", "bar"], /(?<foo>.)(?<bar>.)/.names)
+ assert_equal(["foo"], /(?<foo>.)(?<foo>.)/.names)
+ assert_equal([], /(.)(.)/.names)
+
+ assert_equal(["foo", "bar"], /(?<foo>.)(?<bar>.)/.match("ab").names)
+ assert_equal(["foo"], /(?<foo>.)(?<foo>.)/.match("ab").names)
+ assert_equal([], /(.)(.)/.match("ab").names)
+
+ assert_equal({"foo"=>[1], "bar"=>[2]},
+ /(?<foo>.)(?<bar>.)/.named_captures)
+ assert_equal({"foo"=>[1, 2]},
+ /(?<foo>.)(?<foo>.)/.named_captures)
+ assert_equal({}, /(.)(.)/.named_captures)
end
+
+ def test_match_regexp
+ r = /./
+ m = r.match("a")
+ assert_equal(r, m.regexp)
+ end
end
--
ML: ruby-changes@q...
Info: http://www.atdot.net/~ko1/quickml