[ruby-talk:15162] Regexp (a\1)

 Someone can explain me this result :

pigeon% ruby -e '/(a\1)/ =~ "a\001"; p $&'
"a"
pigeon% 

 With this comment, I expect a match on "a\001"

        /* Can't back reference to a subexpression if inside of it.  */

pigeon% diff -u regex.c~ regex.c
--- regex.c~    Sat Feb 10 18:33:35 2001
+++ regex.c     Tue May 15 15:48:34 2001
@@ -2248,6 +2248,7 @@
          if (!ISDIGIT(c)) PATUNFETCH;
 
          if (c1 >= regnum) {
+         need_to_get_octal:
            /* need to get octal */
            p = p_save;
            c = scan_oct(p_save, 3, &numlen) & 0xff;
@@ -2259,9 +2260,24 @@
        }
 
        /* Can't back reference to a subexpression if inside of it.  */
-       for (stackt = stackp - 2;  stackt > stackb;  stackt -= 5)
-         if (*stackt == c1)
-           goto normal_char;
+       for (stackt = stackp - 2;  stackt > stackb;  ) {
+           switch (*stackt) {
+           case '(':
+               if (stackt[-2] == c1)
+                   goto need_to_get_octal;
+               stackt -= 5;
+               break;
+           case '!':
+               stackt--;
+           case '=':
+           case '>':
+               stackt -= 4;
+               break;
+           default:
+               stackt -= 3;
+               break;
+           }
+       }
        laststart = b;
        BUFPUSH(duplicate);
        BUFPUSH(c1);
pigeon% 

pigeon% ./ruby -e '/(a\1)/ =~ "a\001"; p $&'
"a\001"
pigeon% 


Guy Decoux