regcomp.c: Synthetic start class should include ord >255 folds

Some characters above 255 fold to the < 256 range.  These need to be in
the synthetic start class so the optimizer won't reject them.

This is temporary code which creates false positives, to be
replaced by more precise matching later.

diff --git a/regcomp.c b/regcomp.c
index fa9e492..50b8877 100644 (file)
--- a/regcomp.c
+++ b/regcomp.c
@@ -3105,11 +3105,29 @@ S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
                    && (!(data->start_class->flags & ANYOF_LOC_NONBITMAP_FOLD)
                        || !ANYOF_BITMAP_TEST(data->start_class, PL_fold_latin1[uc])))
                     )
+               {
                    compat = 0;
+               }
                ANYOF_CLASS_ZERO(data->start_class);
                ANYOF_BITMAP_ZERO(data->start_class);
                if (compat)
                    ANYOF_BITMAP_SET(data->start_class, uc);
+               else if (uc >= 0x100) {
+                   int i;
+
+                   /* Some Unicode code points fold to the Latin1 range; as
+                    * XXX temporary code, instead of figuring out if this is
+                    * one, just assume it is and set all the start class bits
+                    * that could be some such above 255 code point's fold
+                    * which will generate fals positives.  As the code
+                    * elsewhere that does compute the fold settles down, it
+                    * can be extracted out and re-used here */
+                   for (i = 0; i < 256; i++){
+                       if (_HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(i)) {
+                           ANYOF_BITMAP_SET(data->start_class, i);
+                       }
+                   }
+               }
                data->start_class->flags &= ~ANYOF_EOS;
                if (uc < 0x100)
                  data->start_class->flags &= ~ANYOF_UNICODE_ALL;
@@ -3170,6 +3188,14 @@ S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
                        ANYOF_BITMAP_SET(data->start_class, PL_fold_latin1[uc]);
                    }
                }
+               else if (uc >= 0x100) {
+                   int i;
+                   for (i = 0; i < 256; i++){
+                       if (_HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(i)) {
+                           ANYOF_BITMAP_SET(data->start_class, i);
+                       }
+                   }
+               }
            }
            else if (flags & SCF_DO_STCLASS_OR) {
                if (data->start_class->flags & ANYOF_LOC_NONBITMAP_FOLD) {