projects / p5sagit/p5-mst-13.2.git / blobdiff
? search:
summary | shortlog | log | commit | commitdiff | tree
raw
PATCH for [ID 20010305.003]
[p5sagit/p5-mst-13.2.git] / regcomp.c
diff --git a/regcomp.c b/regcomp.c
index 547f756..4638d77 100644 (file)
--- a/regcomp.c
+++ b/regcomp.c
@@ -807,7 +807,7 @@ S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp, I32 *deltap, reg
            if (UTF) {
                U8 *s = (U8*)STRING(scan);
                l = utf8_length(s, s + l);
-               uc = utf8_to_uv_simple(s, NULL);
+               uc = utf8_to_uvchr(s, NULL);
            }
            min += l;
            if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
@@ -862,7 +862,7 @@ S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp, I32 *deltap, reg
            if (UTF) {
                U8 *s = (U8 *)STRING(scan);
                l = utf8_length(s, s + l);
-               uc = utf8_to_uv_simple(s, NULL);
+               uc = utf8_to_uvchr(s, NULL);
            }
            min += l;
            if (data && (flags & SCF_DO_SUBSTR))
@@ -1930,6 +1930,8 @@ Perl_pregcomp(pTHX_ char *exp, char *xend, PMOP *pm)
        r->reganch |= ROPT_LOOKBEHIND_SEEN;
     if (RExC_seen & REG_SEEN_EVAL)
        r->reganch |= ROPT_EVAL_SEEN;
+    if (RExC_seen & REG_SEEN_SANY)
+       r->reganch |= ROPT_SANY_SEEN;
     Newz(1002, r->startp, RExC_npar, I32);
     Newz(1002, r->endp, RExC_npar, I32);
     PL_regdata = r->data; /* for regprop() */
@@ -2638,6 +2640,7 @@ tryagain:
            break;
        case 'C':
            ret = reg_node(pRExC_state, SANY);
+           RExC_seen |= REG_SEEN_SANY;
            *flagp |= HASWIDTH|SIMPLE;
            nextchar(pRExC_state);
            break;
@@ -2645,22 +2648,16 @@ tryagain:
            ret = reg_node(pRExC_state, CLUMP);
            *flagp |= HASWIDTH;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_mark)
-               is_utf8_mark((U8*)"~");         /* preload table */
            break;
        case 'w':
            ret = reg_node(pRExC_state, LOC ? ALNUML     : ALNUM);
            *flagp |= HASWIDTH|SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_alnum)
-               is_utf8_alnum((U8*)"a");        /* preload table */
            break;
        case 'W':
            ret = reg_node(pRExC_state, LOC ? NALNUML     : NALNUM);
            *flagp |= HASWIDTH|SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_alnum)
-               is_utf8_alnum((U8*)"a");        /* preload table */
            break;
        case 'b':
            RExC_seen_zerolen++;
@@ -2668,8 +2665,6 @@ tryagain:
            ret = reg_node(pRExC_state, LOC ? BOUNDL     : BOUND);
            *flagp |= SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_alnum)
-               is_utf8_alnum((U8*)"a");        /* preload table */
            break;
        case 'B':
            RExC_seen_zerolen++;
@@ -2677,36 +2672,26 @@ tryagain:
            ret = reg_node(pRExC_state, LOC ? NBOUNDL     : NBOUND);
            *flagp |= SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_alnum)
-               is_utf8_alnum((U8*)"a");        /* preload table */
            break;
        case 's':
            ret = reg_node(pRExC_state, LOC ? SPACEL     : SPACE);
            *flagp |= HASWIDTH|SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_space)
-               is_utf8_space((U8*)" ");        /* preload table */
            break;
        case 'S':
            ret = reg_node(pRExC_state, LOC ? NSPACEL     : NSPACE);
            *flagp |= HASWIDTH|SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_space)
-               is_utf8_space((U8*)" ");        /* preload table */
            break;
        case 'd':
            ret = reg_node(pRExC_state, DIGIT);
            *flagp |= HASWIDTH|SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_digit)
-               is_utf8_digit((U8*)"1");        /* preload table */
            break;
        case 'D':
            ret = reg_node(pRExC_state, NDIGIT);
            *flagp |= HASWIDTH|SIMPLE;
            nextchar(pRExC_state);
-           if (UTF && !PL_utf8_digit)
-               is_utf8_digit((U8*)"1");        /* preload table */
            break;
        case 'p':
        case 'P':
@@ -2852,19 +2837,11 @@ tryagain:
                        p++;
                        break;
                    case 'e':
-#ifdef ASCIIish
-                         ender = '\033';
-#else
-                         ender = '\047';
-#endif
+                         ender = ASCII_TO_NATIVE('\033');
                        p++;
                        break;
                    case 'a':
-#ifdef ASCIIish
-                         ender = '\007';
-#else
-                         ender = '\057';
-#endif
+                         ender = ASCII_TO_NATIVE('\007');
                        p++;
                        break;
                    case 'x':
@@ -2878,6 +2855,8 @@ tryagain:
                            else {
                                numlen = 1;     /* allow underscores */
                                ender = (UV)scan_hex(p + 1, e - p - 1, &numlen);
+                               if (ender > 0xff)
+                                   RExC_utf8 = 1;
                                /* numlen is generous */
                                if (numlen + len >= 127) {
                                    p--;
@@ -2923,7 +2902,7 @@ tryagain:
                default:
                  normal_default:
                    if (UTF8_IS_START(*p) && UTF) {
-                       ender = utf8_to_uv((U8*)p, RExC_end - p,
+                       ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
                                               &numlen, 0);
                        p += numlen;
                    }
@@ -2935,7 +2914,7 @@ tryagain:
                    p = regwhite(p, RExC_end);
                if (UTF && FOLD) {
                    if (LOC)
-                       ender = toLOWER_LC_uni(ender);
+                       ender = toLOWER_LC_uvchr(ender);
                    else
                        ender = toLOWER_uni(ender);
                }
@@ -2944,7 +2923,7 @@ tryagain:
                        p = oldp;
                    /* ender is a Unicode value so it can be > 0xff --
                     * in other words, do not use UTF8_IS_CONTINUED(). */
-                   else if (ender >= 0x80 && UTF) {
+                   else if (NATIVE_TO_ASCII(ender) >= 0x80 && UTF) {
                        reguni(pRExC_state, ender, s, &numlen);
                        s += numlen;
                        len += numlen;
@@ -2957,7 +2936,7 @@ tryagain:
                }
                /* ender is a Unicode value so it can be > 0xff --
                 * in other words, do not use UTF8_IS_CONTINUED(). */
-               if (ender >= 0x80 && UTF) {
+               if (NATIVE_TO_ASCII(ender) >= 0x80 && UTF) {
                    reguni(pRExC_state, ender, s, &numlen);
                    s += numlen;
                    len += numlen - 1;
@@ -3222,7 +3201,7 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
        if (!range)
            rangebegin = RExC_parse;
        if (UTF) {
-           value = utf8_to_uv((U8*)RExC_parse,
+           value = utf8n_to_uvchr((U8*)RExC_parse,
                               RExC_end - RExC_parse,
                               &numlen, 0);
            RExC_parse += numlen;
@@ -3233,7 +3212,7 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
            namedclass = regpposixcc(pRExC_state, value);
        else if (value == '\\') {
            if (UTF) {
-               value = utf8_to_uv((U8*)RExC_parse,
+               value = utf8n_to_uvchr((U8*)RExC_parse,
                                   RExC_end - RExC_parse,
                                   &numlen, 0);
                RExC_parse += numlen;
@@ -3280,13 +3259,8 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
            case 't':   value = '\t';                   break;
            case 'f':   value = '\f';                   break;
            case 'b':   value = '\b';                   break;
-#ifdef ASCIIish
-           case 'e':   value = '\033';                 break;
-           case 'a':   value = '\007';                 break;
-#else
-           case 'e':   value = '\047';                 break;
-           case 'a':   value = '\057';                 break;
-#endif
+           case 'e':   value = ASCII_TO_NATIVE('\033');break;
+           case 'a':   value = ASCII_TO_NATIVE('\007');break;
            case 'x':
                if (*RExC_parse == '{') {
                    e = strchr(RExC_parse++, '}');
@@ -3346,8 +3320,7 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
                    else {
                        ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
                        Perl_sv_catpvf(aTHX_ listsv,
-                                      /* 0x002D is Unicode for '-' */
-                                      "%04"UVxf"\n002D\n", (UV)lastvalue);
+                                      "%04"UVxf"\n%04"UVxf"\n", (UV)lastvalue, (UV) '-');
                    }
                }
 
@@ -3430,7 +3403,7 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
                    if (LOC)
                        ANYOF_CLASS_SET(ret, ANYOF_ASCII);
                    else {
-#ifdef ASCIIish
+#ifndef EBCDIC
                        for (value = 0; value < 128; value++)
                            ANYOF_BITMAP_SET(ret, value);
 #else  /* EBCDIC */
@@ -3446,7 +3419,7 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
                    if (LOC)
                        ANYOF_CLASS_SET(ret, ANYOF_NASCII);
                    else {
-#ifdef ASCIIish
+#ifndef EBCDIC
                        for (value = 128; value < 256; value++)
                            ANYOF_BITMAP_SET(ret, value);
 #else  /* EBCDIC */
@@ -3746,7 +3719,7 @@ S_regclass(pTHX_ RExC_state_t *pRExC_state)
        /* now is the next time */
        if (!SIZE_ONLY) {
            if (lastvalue < 256 && value < 256) {
-#ifndef ASCIIish /* EBCDIC, for example. */
+#ifdef EBCDIC /* EBCDIC, for example. */
                if ((isLOWER(lastvalue) && isLOWER(value)) ||
                    (isUPPER(lastvalue) && isUPPER(value)))
                {
@@ -3909,7 +3882,7 @@ S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
 STATIC void
 S_reguni(pTHX_ RExC_state_t *pRExC_state, UV uv, char* s, STRLEN* lenp)
 {
-    *lenp = SIZE_ONLY ? UNISKIP(uv) : (uv_to_utf8((U8*)s, uv) - (U8*)s);
+    *lenp = SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
 }
 
 /*
@@ -4279,7 +4252,7 @@ Perl_regprop(pTHX_ SV *sv, regnode *o)
                    U8 s[UTF8_MAXLEN+1];
                
                    for (i = 0; i <= 256; i++) { /* just the first 256 */
-                       U8 *e = uv_to_utf8(s, i);
+                       U8 *e = uvuni_to_utf8(s, i);
                        
                        if (i < 256 && swash_fetch(sw, s)) {
                            if (rangestart == -1)
@@ -4289,14 +4262,14 @@ Perl_regprop(pTHX_ SV *sv, regnode *o)
                        
                            if (i <= rangestart + 3)
                                for (; rangestart < i; rangestart++) {
-                                   for(e = uv_to_utf8(s, rangestart), p = s; p < e; p++)
+                                   for(e = uvuni_to_utf8(s, rangestart), p = s; p < e; p++)
                                        put_byte(sv, *p);
                                }
                            else {
-                               for (e = uv_to_utf8(s, rangestart), p = s; p < e; p++)
+                               for (e = uvuni_to_utf8(s, rangestart), p = s; p < e; p++)
                                    put_byte(sv, *p);
                                sv_catpv(sv, "-");
-                                   for (e = uv_to_utf8(s, i - 1), p = s; p < e; p++)
+                                   for (e = uvuni_to_utf8(s, i - 1), p = s; p < e; p++)
                                        put_byte(sv, *p);
                                }
                                rangestart = -1;
Unnamed repository; edit this file to name it for gitweb.