Regex Bug in 5.003_26 thru 003_99a

According to Tom Christiansen:
> I just *knew* I should have included stuff from my FMTEYEWTK on regexen
> in the regression suites.   You'll note that you've now rendered p71 of
> the Camel a liar.  This is considered Very Naughty.

*sigh*  Well, without a bug report, what I could I do?...

> I'd say this is a must-fix.

Here's a patch.  I suppose this should go into _01.

p5p-msgid: 199705161915.PAA18721@rio.atlantic.net
Signed-off-by: Chip Salzenberg <chip@atlantic.net>

diff --git a/regcomp.c b/regcomp.c
index d3788c8..0e9846c 100644 (file)
--- a/regcomp.c
+++ b/regcomp.c
@@ -903,7 +903,9 @@ tryagain:
                    goto defchar;
                else {
                    regsawback = 1;
-                   ret = reganode(REF, num);
+                   ret = reganode((regflags & PMf_FOLD)
+                                  ? ((regflags & PMf_LOCALE) ? REFFL : REFF)
+                                  : REF, num);
                    *flagp |= HASWIDTH;
                    while (isDIGIT(*regparse))
                        regparse++;
@@ -1667,6 +1669,12 @@ char *op;
     case REF:
        sv_catpvf(sv, "REF%d", ARG1(op));
        break;
+    case REFF:
+       sv_catpvf(sv, "REFF%d", ARG1(op));
+       break;
+    case REFFL:
+       sv_catpvf(sv, "REFFL%d", ARG1(op));
+       break;
     case OPEN:
        sv_catpvf(sv, "OPEN%d", ARG1(op));
        break;

diff --git a/regcomp.h b/regcomp.h
index d618066..5915086 100644 (file)
--- a/regcomp.h
+++ b/regcomp.h
@@ -72,25 +72,27 @@
 #define BOUNDL 21      /* no   Match "" at any word boundary */
 #define NBOUND 22      /* no   Match "" at any word non-boundary */
 #define NBOUNDL        23      /* no   Match "" at any word non-boundary */
-#define REF    24      /* num  Match some already matched string */
-#define        OPEN    25      /* num  Mark this point in input as start of #n. */
-#define        CLOSE   26      /* num  Analogous to OPEN. */
-#define MINMOD 27      /* no   Next operator is not greedy. */
-#define GPOS   28      /* no   Matches where last m//g left off. */
-#define IFMATCH        29      /* no   Succeeds if the following matches. */
-#define UNLESSM        30      /* no   Fails if the following matches. */
-#define SUCCEED        31      /* no   Return from a subroutine, basically. */
-#define WHILEM 32      /* no   Do curly processing and see if rest matches. */
-#define ALNUM  33      /* no   Match any alphanumeric character */
-#define ALNUML 34      /* no   Match any alphanumeric char in locale */
-#define NALNUM 35      /* no   Match any non-alphanumeric character */
-#define NALNUML        36      /* no   Match any non-alphanumeric char in locale */
-#define SPACE  37      /* no   Match any whitespace character */
-#define SPACEL 38      /* no   Match any whitespace char in locale */
-#define NSPACE 39      /* no   Match any non-whitespace character */
-#define NSPACEL        40      /* no   Match any non-whitespace char in locale */
-#define DIGIT  41      /* no   Match any numeric character */
-#define NDIGIT 42      /* no   Match any non-numeric character */
+#define REF    24      /* num  Match already matched string */
+#define REFF   25      /* num  Match already matched string, folded */
+#define REFFL  26      /* num  Match already matched string, folded in loc. */
+#define        OPEN    27      /* num  Mark this point in input as start of #n. */
+#define        CLOSE   28      /* num  Analogous to OPEN. */
+#define MINMOD 29      /* no   Next operator is not greedy. */
+#define GPOS   30      /* no   Matches where last m//g left off. */
+#define IFMATCH        31      /* no   Succeeds if the following matches. */
+#define UNLESSM        32      /* no   Fails if the following matches. */
+#define SUCCEED        33      /* no   Return from a subroutine, basically. */
+#define WHILEM 34      /* no   Do curly processing and see if rest matches. */
+#define ALNUM  35      /* no   Match any alphanumeric character */
+#define ALNUML 36      /* no   Match any alphanumeric char in locale */
+#define NALNUM 37      /* no   Match any non-alphanumeric character */
+#define NALNUML        38      /* no   Match any non-alphanumeric char in locale */
+#define SPACE  39      /* no   Match any whitespace character */
+#define SPACEL 40      /* no   Match any whitespace char in locale */
+#define NSPACE 41      /* no   Match any non-whitespace character */
+#define NSPACEL        42      /* no   Match any non-whitespace char in locale */
+#define DIGIT  43      /* no   Match any numeric character */
+#define NDIGIT 44      /* no   Match any non-numeric character */
 
 /*
  * Opcode notes:
@@ -121,7 +123,7 @@ EXT char regarglen[] = {
     0,0,0,0,0,0,0,0,0,0,
     /*CURLY*/ 4, /*CURLYX*/ 4,
     0,0,0,0,0,0,0,0,0,0,0,0,
-    /*REF*/ 2, /*OPEN*/ 2, /*CLOSE*/ 2,
+    /*REF*/ 2, 2, 2, /*OPEN*/ 2, /*CLOSE*/ 2,
     0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
 };
 #endif
@@ -155,6 +157,8 @@ EXT char regkind[] = {
        NBOUND,
        NBOUND,
        REF,
+       REF,
+       REF,
        OPEN,
        CLOSE,
        MINMOD,
@@ -181,7 +185,7 @@ EXT char regkind[] = {
 EXT char varies[];
 #else
 EXT char varies[] = {
-    BRANCH, BACK, STAR, PLUS, CURLY, CURLYX, REF, WHILEM, 0
+    BRANCH, BACK, STAR, PLUS, CURLY, CURLYX, REF, REFF, REFFL, WHILEM, 0
 };
 #endif
 

diff --git a/regexec.c b/regexec.c
index 630b130..7f60a91 100644 (file)
--- a/regexec.c
+++ b/regexec.c
@@ -825,7 +825,11 @@ char *prog;
                sayNO;
            nextchar = UCHARAT(++locinput);
            break;
+       case REFFL:
+           regtainted = TRUE;
+           /* FALL THROUGH */
        case REF:
+       case REFF:
            n = ARG1(scan);  /* which paren pair */
            s = regstartp[n];
            if (!s)
@@ -835,12 +839,19 @@ char *prog;
            if (s == regendp[n])
                break;
            /* Inline the first character, for speed. */
-           if (UCHARAT(s) != nextchar)
+           if (UCHARAT(s) != nextchar &&
+               (OP(scan) == REF ||
+                (UCHARAT(s) != ((OP(scan) == REFF
+                                ? fold : fold_locale)[nextchar]))))
                sayNO;
            ln = regendp[n] - s;
            if (locinput + ln > regeol)
                sayNO;
-           if (ln > 1 && memNE(s, locinput, ln))
+           if (ln > 1 && (OP(scan) == REF
+                          ? memNE(s, locinput, ln)
+                          : (OP(scan) == REFF
+                             ? ibcmp(s, locinput, ln)
+                             : ibcmp_locale(s, locinput, ln))))
                sayNO;
            locinput += ln;
            nextchar = UCHARAT(locinput);