EBCDIC: the non-printable characters are different.

[p5sagit/p5-mst-13.2.git] / regcomp.c

diff --git a/regcomp.c b/regcomp.c
index 51b3c5d..7850492 100644 (file)
--- a/regcomp.c
+++ b/regcomp.c
@@ -69,7 +69,7 @@
  *
  ****    Alterations to Henry's code are...
  ****
- ****    Copyright (c) 1991-2001, Larry Wall
+ ****    Copyright (c) 1991-2002, Larry Wall
  ****
  ****    You may distribute under the terms of either the GNU General Public
  ****    License or the Artistic License, as specified in the README file.
@@ -2168,6 +2168,8 @@ S_reg(pTHX_ RExC_state_t *pRExC_state, I32 paren, I32 *flagp)
                /* FALL THROUGH*/
            case '?':           /* (??...) */
                logical = 1;
+               if (*RExC_parse != '{')
+                   goto unknown;
                paren = *RExC_parse++;
                /* FALL THROUGH */
            case '{':           /* (?{...}) */
@@ -4052,22 +4054,25 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
                          ANYOF_BITMAP_SET(ret, i);
          }
          if (value > 255 || UTF) {
+               UV prevnatvalue  = NATIVE_TO_UNI(prevvalue);
+               UV natvalue      = NATIVE_TO_UNI(value);
+
                ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
-               if (prevvalue < value)
+               if (prevnatvalue < natvalue) { /* what about > ? */
                    Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\t%04"UVxf"\n",
-                                  (UV)prevvalue, (UV)value);
-               else if (prevvalue == value) {
-                   Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
-                                  (UV)value);
+                                  prevnatvalue, natvalue);
+               }
+               else if (prevnatvalue == natvalue) {
+                   Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n", natvalue);
                    if (FOLD) {
                         U8 tmpbuf [UTF8_MAXLEN+1];
                         U8 foldbuf[UTF8_MAXLEN_FOLD+1];
                         STRLEN foldlen;
                         UV f;
 
-                        uvchr_to_utf8(tmpbuf, value);
+                        uvchr_to_utf8(tmpbuf, natvalue);
                         to_utf8_fold(tmpbuf, foldbuf, &foldlen);
-                        f = utf8_to_uvchr(foldbuf, 0);
+                        f = UNI_TO_NATIVE(utf8_to_uvchr(foldbuf, 0));
 
                         /* If folding and foldable and a single
                          * character, insert also the folded version