Prefer !UTF8_IS_INVARIANT() over UTF8_IS_CONTINUED() when that

is the sense of the test being done. Avoid some magical 127 and 128 values
by using macros.

p4raw-id: //depot/perlio@9199

diff --git a/doop.c b/doop.c
index bd66b42..e4a516a 100644 (file)
--- a/doop.c
+++ b/doop.c
@@ -317,7 +317,7 @@ S_do_trans_simple_utf8(pTHX_ SV *sv)/* SPC - OK */
     if (!isutf8) {
        U8 *t = s, *e = s + len;
        while (t < e)
-           if ((hibit = UTF8_IS_CONTINUED(*t++)))
+           if ((hibit = !UTF8_IS_INVARIANT(*t++)))
                break;
        if (hibit)
            s = bytes_to_utf8(s, &len);

diff --git a/pp_ctl.c b/pp_ctl.c
index 93b89b1..f08f8bf 100644 (file)
--- a/pp_ctl.c
+++ b/pp_ctl.c
@@ -551,7 +551,13 @@ PP(pp_formline)
            if (item_is_utf) {
                while (arg--) {
                    if (UTF8_IS_CONTINUED(*s)) {
-                       switch (UTF8SKIP(s)) {
+                       STRLEN skip = UTF8SKIP(s);
+                       switch (skip) {
+                       default:
+                           Move(s,t,skip,char);
+                           s += skip;
+                           t += skip;
+                           break;
                        case 7: *t++ = *s++;
                        case 6: *t++ = *s++;
                        case 5: *t++ = *s++;

diff --git a/sv.c b/sv.c
index 4d3181a..18c5ac9 100644 (file)
--- a/sv.c
+++ b/sv.c
@@ -2978,7 +2978,7 @@ Perl_sv_utf8_upgrade(pTHX_ register SV *sv)
     e = (U8 *) SvEND(sv);
     t = s;
     while (t < e) {
-       if ((hibit = UTF8_IS_CONTINUED(NATIVE_TO_ASCII(*t++))))
+       if ((hibit = !UTF8_IS_INVARIANT(*t++)))
            break;
     }
     if (hibit) {
@@ -3097,8 +3097,8 @@ bool
 Perl_sv_utf8_decode(pTHX_ register SV *sv)
 {
     if (SvPOK(sv)) {
-        char *c;
-        char *e;
+        U8 *c;
+        U8 *e;
 
        /* The octets may have got themselves encoded - get them back as bytes */
         if (!sv_utf8_downgrade(sv, TRUE))
@@ -3107,12 +3107,12 @@ Perl_sv_utf8_decode(pTHX_ register SV *sv)
         /* it is actually just a matter of turning the utf8 flag on, but
          * we want to make sure everything inside is valid utf8 first.
          */
-        c = SvPVX(sv);
-       if (!is_utf8_string((U8*)c, SvCUR(sv)+1))
+        c = (U8 *) SvPVX(sv);
+       if (!is_utf8_string(c, SvCUR(sv)+1))
            return FALSE;
-        e = SvEND(sv);
+        e = (U8 *) SvEND(sv);
         while (c < e) {
-            if (UTF8_IS_CONTINUED(*c++)) {
+            if (!UTF8_IS_INVARIANT(*c++)) {
                SvUTF8_on(sv);
                break;
            }
@@ -7127,7 +7127,7 @@ Perl_sv_vcatpvfn(pTHX_ SV *sv, const char *pat, STRLEN patlen, va_list *args, SV
 
        case 'c':
            uv = args ? va_arg(*args, int) : SvIVx(argsv);
-           if ((uv > 255 || (uv > 127 && SvUTF8(sv))) && !IN_BYTE) {
+           if ((uv > 255 || (!UTF8_IS_INVARIANT(uv) && SvUTF8(sv))) && !IN_BYTE) {
                eptr = (char*)utf8buf;
                elen = uvchr_to_utf8((U8*)eptr, uv) - utf8buf;
                is_utf = TRUE;

diff --git a/toke.c b/toke.c
index 3652c11..53159f3 100644 (file)
--- a/toke.c
+++ b/toke.c
@@ -1457,7 +1457,7 @@ S_scan_const(pTHX_ char *start)
                /* We need to map to chars to ASCII before doing the tests
                   to cover EBCDIC
                */
-               if (NATIVE_TO_UNI(uv) > 127) {
+               if (!UTF8_IS_INVARIANT(uv)) {
                    if (!has_utf8 && uv > 255) {
                        /* Might need to recode whatever we have
                         * accumulated so far if it contains any
@@ -1467,13 +1467,13 @@ S_scan_const(pTHX_ char *start)
                         *  this rescan? --jhi)
                         */
                        int hicount = 0;
-                       char *c;
-                       for (c = SvPVX(sv); c < d; c++) {
-                           if (UTF8_IS_CONTINUED(NATIVE_TO_ASCII(*c))) {
+                       U8 *c;
+                       for (c = (U8 *) SvPVX(sv); c < (U8 *)d; c++) {
+                           if (!UTF8_IS_INVARIANT(*c)) {
                                hicount++;
                            }
                        }
-                       if (hicount || NATIVE_TO_ASCII('A') != 'A') {
+                       if (hicount) {
                            STRLEN offset = d - SvPVX(sv);
                            U8 *src, *dst;
                            d = SvGROW(sv, SvLEN(sv) + hicount + 1) + offset;
@@ -1481,13 +1481,13 @@ S_scan_const(pTHX_ char *start)
                            dst = src+hicount;
                            d  += hicount;
                            while (src >= (U8 *)SvPVX(sv)) {
-                               U8 ch = NATIVE_TO_ASCII(*src);
-                               if (UTF8_IS_CONTINUED(ch)) {
+                               if (!UTF8_IS_INVARIANT(*src)) {
+                                   U8 ch = NATIVE_TO_ASCII(*src);
                                    *dst-- = UTF8_EIGHT_BIT_LO(ch);
                                    *dst-- = UTF8_EIGHT_BIT_HI(ch);
                                }
                                else {
-                                   *dst-- = ch;
+                                   *dst-- = *src;
                                }
                                src--;
                            }
@@ -1603,9 +1603,8 @@ S_scan_const(pTHX_ char *start)
        } /* end if (backslash) */
 
     default_action:
-#ifndef EBCDIC
        /* The 'has_utf8' here is very dubious */
-       if (UTF8_IS_CONTINUED(NATIVE_TO_ASCII(*s)) && (this_utf8 || has_utf8)) {
+       if (!UTF8_IS_INVARIANT((U8)(*s)) && (this_utf8 || has_utf8)) {
            STRLEN len = (STRLEN) -1;
            UV uv;
            if (this_utf8) {
@@ -1630,7 +1629,6 @@ S_scan_const(pTHX_ char *start)
           }
            continue;
        }
-#endif
        *d++ = NATIVE_TO_NEED(has_utf8,*s++);
     } /* while loop to process each character */
 
@@ -3726,7 +3724,7 @@ Perl_yylex(pTHX)
            missingterm((char*)0);
        yylval.ival = OP_CONST;
        for (d = SvPV(PL_lex_stuff, len); len; len--, d++) {
-           if (*d == '$' || *d == '@' || *d == '\\' || UTF8_IS_CONTINUED(*d)) {
+           if (*d == '$' || *d == '@' || *d == '\\' || !UTF8_IS_INVARIANT((U8)*d)) {
                yylval.ival = OP_STRINGIFY;
                break;
            }
@@ -6698,7 +6696,7 @@ S_scan_str(pTHX_ char *start, int keep_quoted, int keep_delims)
 
     /* after skipping whitespace, the next character is the terminator */
     term = *s;
-    if (UTF8_IS_CONTINUED(term) && UTF)
+    if (!UTF8_IS_INVARIANT((U8)term) && UTF)
        has_utf8 = TRUE;
 
     /* mark where we are */
@@ -6745,7 +6743,7 @@ S_scan_str(pTHX_ char *start, int keep_quoted, int keep_delims)
                   have found the terminator */
                else if (*s == term)
                    break;
-               else if (!has_utf8 && UTF8_IS_CONTINUED(*s) && UTF)
+               else if (!has_utf8 && !UTF8_IS_INVARIANT((U8)*s) && UTF)
                    has_utf8 = TRUE;
                *to = *s;
            }
@@ -6774,7 +6772,7 @@ S_scan_str(pTHX_ char *start, int keep_quoted, int keep_delims)
                    break;
                else if (*s == PL_multi_open)
                    brackets++;
-               else if (!has_utf8 && UTF8_IS_CONTINUED(*s) && UTF)
+               else if (!has_utf8 && !UTF8_IS_INVARIANT((U8)*s) && UTF)
                    has_utf8 = TRUE;
                *to = *s;
            }