make /\S/ match the same things /[\S]/ matches; likewise for

\D (from Rick Delaney <rick@consumercontact.com>)

p4raw-id: //depot/perl@5126

diff --git a/regexec.c b/regexec.c
index cef9887..e63fa6f 100644 (file)
--- a/regexec.c
+++ b/regexec.c
@@ -2084,7 +2084,7 @@ S_regmatch(pTHX_ regnode *prog)
            PL_reg_flags |= RF_tainted;
            /* FALL THROUGH */
        case SPACE:
-           if (!nextchr && locinput >= PL_regeol)
+           if (!nextchr)
                sayNO;
            if (!(OP(scan) == SPACE
                  ? isSPACE(nextchr) : isSPACE_LC(nextchr)))
@@ -2095,11 +2095,11 @@ S_regmatch(pTHX_ regnode *prog)
            PL_reg_flags |= RF_tainted;
            /* FALL THROUGH */
        case SPACEUTF8:
-           if (!nextchr && locinput >= PL_regeol)
+           if (!nextchr)
                sayNO;
            if (nextchr & 0x80) {
                if (!(OP(scan) == SPACEUTF8
-                     ? swash_fetch(PL_utf8_space,(U8*)locinput)
+                     ? swash_fetch(PL_utf8_space, (U8*)locinput)
                      : isSPACE_LC_utf8((U8*)locinput)))
                {
                    sayNO;
@@ -2117,9 +2117,9 @@ S_regmatch(pTHX_ regnode *prog)
            PL_reg_flags |= RF_tainted;
            /* FALL THROUGH */
        case NSPACE:
-           if (!nextchr)
+           if (!nextchr && locinput >= PL_regeol)
                sayNO;
-           if (OP(scan) == SPACE
+           if (OP(scan) == NSPACE
                ? isSPACE(nextchr) : isSPACE_LC(nextchr))
                sayNO;
            nextchr = UCHARAT(++locinput);
@@ -2128,11 +2128,11 @@ S_regmatch(pTHX_ regnode *prog)
            PL_reg_flags |= RF_tainted;
            /* FALL THROUGH */
        case NSPACEUTF8:
-           if (!nextchr)
+           if (!nextchr && locinput >= PL_regeol)
                sayNO;
            if (nextchr & 0x80) {
                if (OP(scan) == NSPACEUTF8
-                   ? swash_fetch(PL_utf8_space,(U8*)locinput)
+                   ? swash_fetch(PL_utf8_space, (U8*)locinput)
                    : isSPACE_LC_utf8((U8*)locinput))
                {
                    sayNO;
@@ -2150,7 +2150,7 @@ S_regmatch(pTHX_ regnode *prog)
            PL_reg_flags |= RF_tainted;
            /* FALL THROUGH */
        case DIGIT:
-           if (!nextchr && locinput >= PL_regeol)
+           if (!nextchr)
                sayNO;
            if (!(OP(scan) == DIGIT
                  ? isDIGIT(nextchr) : isDIGIT_LC(nextchr)))
@@ -2164,9 +2164,9 @@ S_regmatch(pTHX_ regnode *prog)
            if (!nextchr)
                sayNO;
            if (nextchr & 0x80) {
-               if (OP(scan) == NDIGITUTF8
-                   ? swash_fetch(PL_utf8_digit,(U8*)locinput)
-                   : isDIGIT_LC_utf8((U8*)locinput))
+               if (!(OP(scan) == DIGITUTF8
+                     ? swash_fetch(PL_utf8_digit, (U8*)locinput)
+                     : isDIGIT_LC_utf8((U8*)locinput)))
                {
                    sayNO;
                }
@@ -2174,7 +2174,8 @@ S_regmatch(pTHX_ regnode *prog)
                nextchr = UCHARAT(locinput);
                break;
            }
-           if (!isDIGIT(nextchr))
+           if (!(OP(scan) == DIGITUTF8
+                 ? isDIGIT(nextchr) : isDIGIT_LC(nextchr)))
                sayNO;
            nextchr = UCHARAT(++locinput);
            break;
@@ -2182,9 +2183,9 @@ S_regmatch(pTHX_ regnode *prog)
            PL_reg_flags |= RF_tainted;
            /* FALL THROUGH */
        case NDIGIT:
-           if (!nextchr)
+           if (!nextchr && locinput >= PL_regeol)
                sayNO;
-           if (OP(scan) == DIGIT
+           if (OP(scan) == NDIGIT
                ? isDIGIT(nextchr) : isDIGIT_LC(nextchr))
                sayNO;
            nextchr = UCHARAT(++locinput);
@@ -2196,13 +2197,18 @@ S_regmatch(pTHX_ regnode *prog)
            if (!nextchr && locinput >= PL_regeol)
                sayNO;
            if (nextchr & 0x80) {
-               if (swash_fetch(PL_utf8_digit,(U8*)locinput))
+               if (OP(scan) == NDIGITUTF8
+                   ? swash_fetch(PL_utf8_digit, (U8*)locinput)
+                   : isDIGIT_LC_utf8((U8*)locinput))
+               {
                    sayNO;
+               }
                locinput += PL_utf8skip[nextchr];
                nextchr = UCHARAT(locinput);
                break;
            }
-           if (isDIGIT(nextchr))
+           if (OP(scan) == NDIGITUTF8
+               ? isDIGIT(nextchr) : isDIGIT_LC(nextchr))
                sayNO;
            nextchr = UCHARAT(++locinput);
            break;

diff --git a/t/op/pat.t b/t/op/pat.t
index 9f68550..9dc3ac7 100755 (executable)
--- a/t/op/pat.t
+++ b/t/op/pat.t
@@ -4,7 +4,7 @@
 # the format supported by op/regexp.t.  If you want to add a test
 # that does fit that format, add it to op/re_tests, not here.
 
-print "1..195\n";
+print "1..207\n";
 
 BEGIN {
     chdir 't' if -d 't';
@@ -905,3 +905,51 @@ $text = "abc dbf";
 print "ok $test\n";
 $test++;
 
+@a = map chr,0..255;
+print "not " if grep(/\S/,@a) != grep(/[^\s]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\S/,@a) != grep(/[\S]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\s/,@a) != grep(/[^\S]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\s/,@a) != grep(/[\s]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\D/,@a) != grep(/[^\d]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\D/,@a) != grep(/[\D]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\d/,@a) != grep(/[^\D]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\d/,@a) != grep(/[\d]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\W/,@a) != grep(/[^\w]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\W/,@a) != grep(/[\W]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\w/,@a) != grep(/[^\W]/,@a);
+print "ok $test\n";
+$test++;
+
+print "not " if grep(/\w/,@a) != grep(/[\w]/,@a);
+print "ok $test\n";
+$test++;