nobu	2007-12-29 22:44:32 +0900 (Sat, 29 Dec 2007)

  New Revision: 14782

  Modified files:
    trunk/ChangeLog
    trunk/re.c
    trunk/string.c

  Log:
    * re.c (rb_reg_regsub): returns the given string itself if nothing
      changed.
    
    * string.c (rb_str_sub_bang): keeps code-range as possible.
    
    * string.c (str_gsub): adjusts code-range.  [ruby-core:14566]


  http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/string.c?r1=14782&r2=14781
  http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/ChangeLog?r1=14782&r2=14781
  http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/re.c?r1=14782&r2=14781

Index: re.c
===================================================================
--- re.c	(revision 14781)
+++ re.c	(revision 14782)
@@ -2856,13 +2856,10 @@
 	}
     }
 
+    if (!val) return str;
     if (p < e) {
-	if (!val) {
-	    val = rb_str_buf_new(e-p);
-	}
         rb_str_buf_cat(val, p, e-p);
     }
-    if (!val) return str;
     rb_enc_associate(val, enc);
 
     return val;
Index: ChangeLog
===================================================================
--- ChangeLog	(revision 14781)
+++ ChangeLog	(revision 14782)
@@ -1,3 +1,12 @@
+Sat Dec 29 22:44:30 2007  Nobuyoshi Nakada  <nobu@r...>
+
+	* re.c (rb_reg_regsub): returns the given string itself if nothing
+	  changed.
+
+	* string.c (rb_str_sub_bang): keeps code-range as possible.
+
+	* string.c (str_gsub): adjusts code-range.  [ruby-core:14566]
+
 Sat Dec 29 21:54:37 2007  Nobuyoshi Nakada  <nobu@r...>
 
 	* common.mk (clean, distclean, realclean): should include clean-enc
Index: string.c
===================================================================
--- string.c	(revision 14781)
+++ string.c	(revision 14782)
@@ -2425,6 +2425,7 @@
     pat = get_pat(argv[0], 1);
     if (rb_reg_search(pat, str, 0, 0) >= 0) {
 	rb_encoding *enc;
+	int cr = ENC_CODERANGE(str);
 
 	match = rb_backref_get();
 	regs = RMATCH(match)->regs;
@@ -2446,6 +2447,10 @@
 	rb_str_modify(str);
 	rb_enc_associate(str, enc);
 	if (OBJ_TAINTED(repl)) tainted = 1;
+	if (ENC_CODERANGE_UNKNOWN < cr && cr < ENC_CODERANGE_BROKEN) {
+	    int cr2 = ENC_CODERANGE(repl);
+	    if (cr2 == ENC_CODERANGE_UNKNOWN || cr2 > cr) cr = cr2;
+	}
 	plen = END(0) - BEG(0);
 	if (RSTRING_LEN(repl) > plen) {
 	    RESIZE_CAPA(str, RSTRING_LEN(str) + RSTRING_LEN(repl) - plen);
@@ -2459,6 +2464,7 @@
 	       RSTRING_PTR(repl), RSTRING_LEN(repl));
 	STR_SET_LEN(str, RSTRING_LEN(str) + RSTRING_LEN(repl) - plen);
 	RSTRING_PTR(str)[RSTRING_LEN(str)] = '\0';
+	ENC_CODERANGE_SET(str, cr);
 	if (tainted) OBJ_TAINT(str);
 
 	return str;
@@ -2516,7 +2522,7 @@
     int iter = 0;
     char *buf, *bp, *sp, *cp;
     int tainted = 0;
-    rb_encoding *enc;
+    int cr;
     
     switch (argc) {
       case 1:
@@ -2533,7 +2539,6 @@
     }
 
     pat = get_pat(argv[0], 1);
-    enc = rb_enc_get(pat);
     offset=0; n=0;
     beg = rb_reg_search(pat, str, 0, 0);
     if (beg < 0) {
@@ -2547,6 +2552,7 @@
     bp = buf;
     sp = cp = RSTRING_PTR(str);
     slen = RSTRING_LEN(str);
+    cr = ENC_CODERANGE(str);
 
     rb_str_locktmp(dest);
     do {
@@ -2573,6 +2579,10 @@
 	}
 	rb_enc_associate(str, enc);
 	if (OBJ_TAINTED(val)) tainted = 1;
+	if (ENC_CODERANGE_UNKNOWN < cr && cr < ENC_CODERANGE_BROKEN) {
+	    int cr2 = ENC_CODERANGE(val);
+	    if (cr2 == ENC_CODERANGE_UNKNOWN || cr2 > cr) cr = cr2;
+	}
 	len = (bp - buf) + (beg - offset) + RSTRING_LEN(val) + 3;
 	if (blen < len) {
 	    while (blen < len) blen *= 2;
@@ -2636,6 +2646,7 @@
     }
     STR_SET_LEN(str, bp - buf);
 
+    ENC_CODERANGE_SET(str, cr);
     if (tainted) OBJ_TAINT(str);
     return str;
 }

--
ML: ruby-changes@q...
Info: http://www.atdot.net/~ko1/quickml