string.c: backref substitution

* re.c (rb_reg_regsub): allow nil regexp for string matching.

* string.c (rb_str_sub_bang, str_gsub): make substitution if
  replacement string has backref escapes.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45459 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

re.c

@@ -3402,7 +3402,7 @@ rb_reg_regsub(VALUE str, VALUE src, struct re_registers *regs, VALUE regexp)
 	switch (c) {
 	  case '1': case '2': case '3': case '4':
 	  case '5': case '6': case '7': case '8': case '9':
-            if (onig_noname_group_capture_is_active(RREGEXP(regexp)->ptr)) {
+            if (!NIL_P(regexp) && onig_noname_group_capture_is_active(RREGEXP(regexp)->ptr)) {
                 no = c - '0';
             }
             else {

string.c

@@ -3957,17 +3957,16 @@ rb_str_sub_bang(int argc, VALUE *argv, VALUE str)
 	char *p, *rp;
 	long len, rlen;
 
+	match = rb_backref_get();
+	regs = RMATCH_REGS(match);
 	if (RB_TYPE_P(pat, T_STRING)) {
 	    beg0 = beg;
 	    end0 = beg0 + RSTRING_LEN(pat);
 	    match0 = pat;
 	}
 	else {
-	    match = rb_backref_get();
-	    regs = RMATCH_REGS(match);
 	    beg0 = BEG(0);
 	    end0 = END(0);
-	    if (!iter && NIL_P(hash)) repl = rb_reg_regsub(repl, str, regs, pat);
 	    if (iter) match0 = rb_reg_nth_match(0, match);
 	}
 
@@ -3984,6 +3983,9 @@ rb_str_sub_bang(int argc, VALUE *argv, VALUE str)
 	    str_mod_check(str, p, len);
 	    rb_check_frozen(str);
 	}
+	else {
+	    repl = rb_reg_regsub(repl, str, regs, RB_TYPE_P(pat, T_STRING) ? Qnil : pat);
+	}
 
         enc = rb_enc_compatible(str, repl);
         if (!enc) {
@@ -4086,16 +4088,16 @@ str_gsub(int argc, VALUE *argv, VALUE str, int bang)
     long beg, n;
     long beg0, end0;
     long offset, blen, slen, len, last;
-    int iter = 0;
+    enum {STR, ITER, MAP} mode = STR;
     char *sp, *cp;
     int tainted = 0;
-    int need_backref;
+    int need_backref = -1;
     rb_encoding *str_enc;
 
     switch (argc) {
       case 1:
 	RETURN_ENUMERATOR(str, argc, argv);
-	iter = 1;
+	mode = ITER;
 	break;
       case 2:
 	repl = argv[1];
@@ -4103,6 +4105,9 @@ str_gsub(int argc, VALUE *argv, VALUE str, int bang)
 	if (NIL_P(hash)) {
 	    StringValue(repl);
 	}
+	else {
+	    mode = MAP;
+	}
 	if (OBJ_TAINTED(repl)) tainted = 1;
 	break;
       default:
@@ -4110,7 +4115,6 @@ str_gsub(int argc, VALUE *argv, VALUE str, int bang)
     }
 
     pat = get_pat_quoted(argv[0], 1);
-    need_backref = iter || !NIL_P(hash);
     beg = rb_pat_search(pat, str, 0, need_backref);
     if (beg < 0) {
 	if (bang) return Qnil;	/* no match, no substitution */
@@ -4131,23 +4135,21 @@ str_gsub(int argc, VALUE *argv, VALUE str, int bang)
     do {
 	n++;
 
+	match = rb_backref_get();
+	regs = RMATCH_REGS(match);
 	if (RB_TYPE_P(pat, T_STRING)) {
 	    beg0 = beg;
 	    end0 = beg0 + RSTRING_LEN(pat);
-	    if (!need_backref) val = repl;
 	    match0 = pat;
 	}
 	else {
-	    match = rb_backref_get();
-	    regs = RMATCH_REGS(match);
 	    beg0 = BEG(0);
 	    end0 = END(0);
-	    if (!need_backref) val = rb_reg_regsub(repl, str, regs, pat);
+	    if (mode == ITER) match0 = rb_reg_nth_match(0, match);
-	    if (iter) match0 = rb_reg_nth_match(0, match);
 	}
 
-	if (need_backref) {
+	if (mode) {
-            if (iter) {
+            if (mode == ITER) {
                 val = rb_obj_as_string(rb_yield(match0));
             }
             else {
@@ -4159,6 +4161,16 @@ str_gsub(int argc, VALUE *argv, VALUE str, int bang)
 		rb_raise(rb_eRuntimeError, "block should not cheat");
 	    }
 	}
+	else if (need_backref) {
+	    val = rb_reg_regsub(repl, str, regs, RB_TYPE_P(pat, T_STRING) ? Qnil : pat);
+	    if (need_backref < 0) {
+		need_backref = val != repl;
+	    }
+	}
+	else {
+	    val = repl;
+	}
+
 
 	if (OBJ_TAINTED(val)) tainted = 1;
 

test/ruby/test_string.rb

@@ -808,6 +808,7 @@ class TestString < Test::Unit::TestCase
                  S("hello").gsub(/./) { |s| s[0].to_s + S(' ')})
     assert_equal(S("HELL-o"),
                  S("hello").gsub(/(hell)(.)/) { |s| $1.upcase + S('-') + $2 })
+    assert_equal(S("<>h<>e<>l<>l<>o<>"), S("hello").gsub(S(''), S('<\0>')))
 
     a = S("hello")
     a.taint
@@ -1415,6 +1416,7 @@ class TestString < Test::Unit::TestCase
     assert_equal(S("HELL-o"),   S("hello").sub(/(hell)(.)/) {
                    |s| $1.upcase + S('-') + $2
                    })
+    assert_equal(S("h<e>llo"),  S("hello").sub('e', S('<\0>')))
 
     assert_equal(S("a\\aba"), S("ababa").sub(/b/, '\\'))
     assert_equal(S("ab\\aba"), S("ababa").sub(/(b)/, '\1\\'))