* re.c (reg_match_pos): adjust offset based on characters, not

  bytes.  [ruby-dev:38722]

* string.c (rb_str_offset): new function.

* string.c (rb_str_index_m): no call to rb_reg_adjust_startpos().

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@23916 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

4 files changed, 19 insertions, 4 deletions

diff --git a/ChangeLog b/ChangeLog
index e5a535670b..efaca159ae 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,12 @@
+Tue Jun 30 17:44:24 2009  Yukihiro Matsumoto  <matz@ruby-lang.org>
+
+	* re.c (reg_match_pos): adjust offset based on characters, not
+	  bytes.  [ruby-dev:38722]
+
+	* string.c (rb_str_offset): new function.
+
+	* string.c (rb_str_index_m): no call to rb_reg_adjust_startpos().
+
 Tue Jun 30 16:57:07 2009  Nobuyoshi Nakada  <nobu@ruby-lang.org>
 
 	* tool/rbinstall.rb: renamed to get rid of collision agains
diff --git a/include/ruby/intern.h b/include/ruby/intern.h
index ed9e746bf0..f153658d50 100644
--- a/include/ruby/intern.h
+++ b/include/ruby/intern.h
@@ -643,6 +643,7 @@ void rb_str_setter(VALUE, ID, VALUE*);
 VALUE rb_str_intern(VALUE);
 VALUE rb_sym_to_s(VALUE);
 VALUE rb_str_length(VALUE);
+long rb_str_offset(VALUE, long);
 size_t rb_str_capacity(VALUE);
 #if defined __GNUC__
 #define rb_str_new_cstr(str) __extension__ (	\
diff --git a/re.c b/re.c
index 528e94a829..447728da09 100644
--- a/re.c
+++ b/re.c
@@ -2564,7 +2564,7 @@ reg_match_pos(VALUE re, VALUE *strp, long pos)
 		return pos;
 	    }
 	}
-	pos = rb_reg_adjust_startpos(re, str, pos, 0);
+	pos = rb_str_offset(str, pos);
     }
     return rb_reg_search(re, str, pos, 0);
 }
diff --git a/string.c b/string.c
index 7f107c26ad..c9561208aa 100644
--- a/string.c
+++ b/string.c
@@ -1440,6 +1440,13 @@ str_offset(const char *p, const char *e, long nth, rb_encoding *enc, int singleb
     return pp - p;
 }
 
+long
+rb_str_offset(VALUE str, long pos)
+{
+    return str_offset(RSTRING_PTR(str), RSTRING_END(str), pos,
+		      STR_ENC_GET(str), single_byte_optimizable(str));
+}
+
 #ifdef NONASCII_MASK
 static char *
 str_utf8_nth(const char *p, const char *e, long nth)
@@ -2563,9 +2570,8 @@ rb_str_index_m(int argc, VALUE *argv, VALUE str)
     switch (TYPE(sub)) {
       case T_REGEXP:
 	pos = str_offset(RSTRING_PTR(str), RSTRING_END(str), pos,
-		     rb_enc_check(str, sub), single_byte_optimizable(str));
+			 rb_enc_check(str, sub), single_byte_optimizable(str));
 
-	pos = rb_reg_adjust_startpos(sub, str, pos, 0);
 	pos = rb_reg_search(sub, str, pos, 0);
 	pos = rb_str_sublen(str, pos);
 	break;
@@ -2680,7 +2686,6 @@ rb_str_rindex_m(int argc, VALUE *argv, VALUE str)
 			 STR_ENC_GET(str), single_byte_optimizable(str));
 
 	if (!RREGEXP(sub)->ptr || RREGEXP_SRC_LEN(sub)) {
-	    pos = rb_reg_adjust_startpos(sub, str, pos, 1);
 	    pos = rb_reg_search(sub, str, pos, 1);
 	    pos = rb_str_sublen(str, pos);
 	}