add a minimal pod/perlrepository.pod

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

diff --git a/regcomp.c b/regcomp.c
index fba4260..3ad1f3b 100644 (file)
--- a/regcomp.c
+++ b/regcomp.c
@@ -7804,6 +7804,22 @@ case ANYOF_N##NAME:                                     \
     what = WORD;                                        \
     break
 
+/* 
+   We dont use PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS as the direct test
+   so that it is possible to override the option here without having to 
+   rebuild the entire core. as we are required to do if we change regcomp.h
+   which is where PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS is defined.
+*/
+#if PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS
+#define BROKEN_UNICODE_CHARCLASS_MAPPINGS
+#endif
+
+#ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
+#define POSIX_CC_UNI_NAME(CCNAME) CCNAME
+#else
+#define POSIX_CC_UNI_NAME(CCNAME) "Posix" CCNAME
+#endif
+
 /*
    parse a class specification and produce either an ANYOF node that
    matches the pattern or if the pattern matches a single char only and
@@ -8092,18 +8108,24 @@ parseit:
                 * A similar issue a little earlier when switching on value.
                 * --jhi */
                switch ((I32)namedclass) {
+               
+               case _C_C_T_(ALNUMC, isALNUMC(value), POSIX_CC_UNI_NAME("Alnum"));
+               case _C_C_T_(ALPHA, isALPHA(value), POSIX_CC_UNI_NAME("Alpha"));
+               case _C_C_T_(BLANK, isBLANK(value), POSIX_CC_UNI_NAME("Blank"));
+               case _C_C_T_(CNTRL, isCNTRL(value), POSIX_CC_UNI_NAME("Cntrl"));
+               case _C_C_T_(GRAPH, isGRAPH(value), POSIX_CC_UNI_NAME("Graph"));
+               case _C_C_T_(LOWER, isLOWER(value), POSIX_CC_UNI_NAME("Lower"));
+               case _C_C_T_(PRINT, isPRINT(value), POSIX_CC_UNI_NAME("Print"));
+               case _C_C_T_(PSXSPC, isPSXSPC(value), POSIX_CC_UNI_NAME("Space"));
+               case _C_C_T_(PUNCT, isPUNCT(value), POSIX_CC_UNI_NAME("Punct"));
+               case _C_C_T_(UPPER, isUPPER(value), POSIX_CC_UNI_NAME("Upper"));
+#ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
                case _C_C_T_(ALNUM, isALNUM(value), "Word");
-               case _C_C_T_(ALNUMC, isALNUMC(value), "Alnum");
-               case _C_C_T_(ALPHA, isALPHA(value), "Alpha");
-               case _C_C_T_(BLANK, isBLANK(value), "Blank");
-               case _C_C_T_(CNTRL, isCNTRL(value), "Cntrl");
-               case _C_C_T_(GRAPH, isGRAPH(value), "Graph");
-               case _C_C_T_(LOWER, isLOWER(value), "Lower");
-               case _C_C_T_(PRINT, isPRINT(value), "Print");
-               case _C_C_T_(PSXSPC, isPSXSPC(value), "Space");
-               case _C_C_T_(PUNCT, isPUNCT(value), "Punct");
                case _C_C_T_(SPACE, isSPACE(value), "SpacePerl");
-               case _C_C_T_(UPPER, isUPPER(value), "Upper");
+#else
+               case _C_C_T_(SPACE, isSPACE(value), "PerlSpace");
+               case _C_C_T_(ALNUM, isALNUM(value), "PerlWord");
+#endif         
                case _C_C_T_(XDIGIT, isXDIGIT(value), "XDigit");
                case _C_C_T_NOLOC_(VERTWS, is_VERTWS_latin1(&value), "VertSpace");
                case _C_C_T_NOLOC_(HORIZWS, is_HORIZWS_latin1(&value), "HorizSpace");
@@ -8150,7 +8172,7 @@ parseit:
                            ANYOF_BITMAP_SET(ret, value);
                    }
                    yesno = '+';
-                   what = "Digit";
+                   what = POSIX_CC_UNI_NAME("Digit");
                    break;
                case ANYOF_NDIGIT:
                    if (LOC)
@@ -8163,7 +8185,7 @@ parseit:
                            ANYOF_BITMAP_SET(ret, value);
                    }
                    yesno = '!';
-                   what = "Digit";
+                   what = POSIX_CC_UNI_NAME("Digit");
                    break;              
                case ANYOF_MAX:
                    /* this is to handle \p and \P */
@@ -8974,6 +8996,17 @@ Perl_regdump(pTHX_ const regexp *r)
 /*
 - regprop - printable representation of opcode
 */
+#define EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags) \
+STMT_START { \
+        if (do_sep) {                           \
+            Perl_sv_catpvf(aTHX_ sv,"%s][%s",PL_colors[1],PL_colors[0]); \
+            if (flags & ANYOF_INVERT)           \
+                /*make sure the invert info is in each */ \
+                sv_catpvs(sv, "^");             \
+            do_sep = 0;                         \
+        }                                       \
+} STMT_END
+
 void
 Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
 {
@@ -9142,6 +9175,8 @@ Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
        Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
        if (flags & ANYOF_INVERT)
            sv_catpvs(sv, "^");
+       
+       /* output what the standard cp 0-255 bitmap matches */
        for (i = 0; i <= 256; i++) {
            if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
                if (rangestart == -1)
@@ -9159,11 +9194,9 @@ Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
                rangestart = -1;
            }
        }
-        if (do_sep) {
-            sv_catpvs(sv,"][");
-            do_sep = 0;
-        }
-            
+        
+        EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
+        /* output any special charclass tests (used mostly under use locale) */
        if (o->flags & ANYOF_CLASS)
            for (i = 0; i < (int)(sizeof(anyofs)/sizeof(char*)); i++)
                if (ANYOF_CLASS_TEST(o,i)) {
@@ -9171,11 +9204,9 @@ Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
                    do_sep = 1;
                }
         
-        if (do_sep) {
-            sv_catpvs(sv,"][");
-            do_sep = 0;
-        }
-
+        EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
+        
+        /* output information about the unicode matching */
        if (flags & ANYOF_UNICODE)
            sv_catpvs(sv, "{unicode}");
        else if (flags & ANYOF_UNICODE_ALL)