Audit #ifdef EBCDIC and #ifndef ASCIIish, replace latter with former.

Use ASCII_TO_NATIVE and NATIVE_TO_ASCII to avoid some #ifs.

p4raw-id: //depot/perlio@9105

diff --git a/regcomp.c b/regcomp.c
index 227737c..05a48d9 100644 (file)
--- a/regcomp.c
+++ b/regcomp.c
@@ -2837,19 +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':
@@ -3267,13 +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++, '}');
@@ -3417,7 +3404,7 @@ 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 */
@@ -3433,7 +3420,7 @@ 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 */
@@ -3733,7 +3720,7 @@ 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)))
                {

diff --git a/toke.c b/toke.c
index 0d4fc1d..0bc4a53 100644 (file)
--- a/toke.c
+++ b/toke.c
@@ -1265,7 +1265,7 @@ S_scan_const(pTHX_ char *start)
                               (char)min, (char)max);
                 }
 
-#ifndef ASCIIish
+#ifdef EBCDIC
                if ((isLOWER(min) && isLOWER(max)) ||
                    (isUPPER(min) && isUPPER(max))) {
                    if (isLOWER(min)) {
@@ -1450,13 +1450,11 @@ S_scan_const(pTHX_ char *start)
                 * There will always enough room in sv since such
                 * escapes will be longer than any UT-F8 sequence
                 * they can end up as. */
-
-               /* This spot is wrong for EBCDIC.  Characters like
-                * the lowercase letters and digits are >127 in EBCDIC,
-                * so here they would need to be mapped to the Unicode
-                * repertoire.   --jhi */
                
-               if (uv > 127) {
+               /* We need to map to chars to ASCII before doing the tests
+                  to cover EBCDIC
+               */
+               if (NATIVE_TO_ASCII(uv) > 127) {
                    if (!has_utf8 && uv > 255) {
                        /* Might need to recode whatever we have
                         * accumulated so far if it contains any
@@ -1465,11 +1463,11 @@ S_scan_const(pTHX_ char *start)
                         * (Can't we keep track of that and avoid
                         *  this rescan? --jhi)
                         */
-                       int hicount = 0;
+                       int hicount = 0;
                        char *c;
 
                        for (c = SvPVX(sv); c < d; c++) {
-                           if (UTF8_IS_CONTINUED(*c))
+                           if (UTF8_IS_CONTINUED(NATIVE_TO_ASCII(*c)))
                                hicount++;
                        }
                        if (hicount) {
@@ -1485,13 +1483,15 @@ S_scan_const(pTHX_ char *start)
                            dst = d - 1;
 
                            while (src < dst) {
-                               if (UTF8_IS_CONTINUED(*src)) {
-                                   *dst-- = UTF8_EIGHT_BIT_LO(*src);
-                                   *dst-- = UTF8_EIGHT_BIT_HI(*src--);
+                               U8 ch = NATIVE_TO_ASCII(*src);
+                               if (UTF8_IS_CONTINUED(ch)) {
+                                   *dst-- = UTF8_EIGHT_BIT_LO(ch);
+                                   *dst-- = UTF8_EIGHT_BIT_HI(ch);
                                }
                                else {
-                                   *dst-- = *src--;
+                                   *dst-- = ch;
                                }
+                               src--;
                            }
                         }
                     }
@@ -1566,18 +1566,14 @@ S_scan_const(pTHX_ char *start)
            /* \c is a control character */
            case 'c':
                s++;
-#ifdef EBCDIC
-               *d = *s++;
-               if (isLOWER(*d))
-                  *d = toUPPER(*d);
-               *d = toCTRL(*d);
-               d++;
-#else
                {
                    U8 c = *s++;
+#ifdef EBCDIC
+                   if (isLOWER(c))
+                       c = toUPPER(c);
+#endif
                    *d++ = toCTRL(c);
                }
-#endif
                continue;
 
            /* printf-style backslashes, formfeeds, newlines, etc */
@@ -1596,21 +1592,12 @@ S_scan_const(pTHX_ char *start)
            case 't':
                *d++ = '\t';
                break;
-#ifdef EBCDIC
            case 'e':
-               *d++ = '\047';  /* CP 1047 */
+               *d++ = ASCII_TO_NATIVE('\033');
                break;
            case 'a':
-               *d++ = '\057';  /* CP 1047 */
-               break;
-#else
-           case 'e':
-               *d++ = '\033';
+               *d++ = ASCII_TO_NATIVE('\007');
                break;
-           case 'a':
-               *d++ = '\007';
-               break;
-#endif
            } /* end switch */
 
            s++;
@@ -1618,7 +1605,7 @@ S_scan_const(pTHX_ char *start)
        } /* end if (backslash) */
 
     default_action:
-       if (UTF8_IS_CONTINUED(*s) && (this_utf8 || has_utf8)) {
+       if (UTF8_IS_CONTINUED(NATIVE_TO_ASCII(*s)) && (this_utf8 || has_utf8)) {
            STRLEN len = (STRLEN) -1;
            UV uv;
            if (this_utf8) {

diff --git a/utf8.c b/utf8.c
index 450138a..7302bb7 100644 (file)
--- a/utf8.c
+++ b/utf8.c
@@ -253,11 +253,7 @@ Perl_utf8n_to_uvuni(pTHX_ U8* s, STRLEN curlen, STRLEN* retlen, U32 flags)
 {
     UV uv = *s, ouv;
     STRLEN len = 1;
-#ifdef EBCDIC
-    bool dowarn = 0;
-#else
     bool dowarn = ckWARN_d(WARN_UTF8);
-#endif
     STRLEN expectlen = 0;
     U32 warning = 0;