Unadorned numbers evil.

p4raw-id: //depot/perl@13672

5 files changed, 24 insertions, 19 deletions

diff --git a/pp.c b/pp.c
index 2fd40daa1a..9237a8b219 100644
--- a/pp.c
+++ b/pp.c
@@ -3210,7 +3210,7 @@ PP(pp_ucfirst)
     STRLEN slen;
 
     if (DO_UTF8(sv)) {
-	U8 tmpbuf[UTF8_MAXLEN*2+1];
+	U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 	STRLEN ulen;
 	STRLEN tculen;
 
@@ -3265,7 +3265,7 @@ PP(pp_lcfirst)
 
     if (DO_UTF8(sv) && (s = (U8*)SvPV(sv, slen)) && slen && UTF8_IS_START(*s)) {
 	STRLEN ulen;
-	U8 tmpbuf[UTF8_MAXLEN*2+1];
+	U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 	U8 *tend;
 	UV uv;
 
@@ -3322,7 +3322,7 @@ PP(pp_uc)
 	STRLEN ulen;
 	register U8 *d;
 	U8 *send;
-	U8 tmpbuf[UTF8_MAXLEN*2+1];
+	U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 
 	s = (U8*)SvPV(sv,len);
 	if (!len) {
@@ -3389,7 +3389,7 @@ PP(pp_lc)
 	STRLEN ulen;
 	register U8 *d;
 	U8 *send;
-	U8 tmpbuf[UTF8_MAXLEN*2+1];
+	U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 
 	s = (U8*)SvPV(sv,len);
 	if (!len) {
diff --git a/regcomp.c b/regcomp.c
index a0f32f8c8b..53d89479a5 100644
--- a/regcomp.c
+++ b/regcomp.c
@@ -2988,7 +2988,7 @@ tryagain:
 	    char *oldp, *s;
 	    STRLEN numlen;
 	    STRLEN ulen;
-	    U8 tmpbuf[UTF8_MAXLEN*2+1];
+	    U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 
             parse_start = RExC_parse - 1;
 
diff --git a/regexec.c b/regexec.c
index 4b073d2da4..35a0a6c2b0 100644
--- a/regexec.c
+++ b/regexec.c
@@ -932,8 +932,8 @@ S_find_byclass(pTHX_ regexp * prog, regnode *c, char *s, char *strend, char *sta
 	    ln = STR_LEN(c);
 	    if (UTF) {
 	        STRLEN ulen1, ulen2;
-		U8 tmpbuf1[UTF8_MAXLEN*2+1];
-		U8 tmpbuf2[UTF8_MAXLEN*2+1];
+		U8 tmpbuf1[UTF8_MAXLEN_UCLC+1];
+		U8 tmpbuf2[UTF8_MAXLEN_UCLC+1];
 
 		to_utf8_lower((U8*)m, tmpbuf1, &ulen1);
 		to_utf8_upper((U8*)m, tmpbuf2, &ulen2);
@@ -2285,7 +2285,7 @@ S_regmatch(pTHX_ regnode *prog)
 		char *l = locinput;
 		char *e;
 		STRLEN ulen;
-		U8 tmpbuf[UTF8_MAXLEN*2+1];
+		U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 		e = s + ln;
 		while (s < e) {
 		    if (l >= PL_regeol)
@@ -2558,8 +2558,8 @@ S_regmatch(pTHX_ regnode *prog)
 		 */
 		if (OP(scan) == REFF) {
 		    STRLEN ulen1, ulen2;
-		    U8 tmpbuf1[UTF8_MAXLEN*2+1];
-		    U8 tmpbuf2[UTF8_MAXLEN*2+1];
+		    U8 tmpbuf1[UTF8_MAXLEN_UCLC+1];
+		    U8 tmpbuf2[UTF8_MAXLEN_UCLC+1];
 		    while (s < e) {
 			if (l >= PL_regeol)
 			    sayNO;
@@ -3364,8 +3364,8 @@ S_regmatch(pTHX_ regnode *prog)
 		    else { /* UTF */
 			if (OP(text_node) == EXACTF || OP(text_node) == REFF) {
 			     STRLEN ulen1, ulen2;
-			     U8 tmpbuf1[UTF8_MAXLEN*2+1];
-			     U8 tmpbuf2[UTF8_MAXLEN*2+1];
+			     U8 tmpbuf1[UTF8_MAXLEN_UCLC+1];
+			     U8 tmpbuf2[UTF8_MAXLEN_UCLC+1];
 
 			     to_utf8_lower((U8*)s, tmpbuf1, &ulen1);
 			     to_utf8_upper((U8*)s, tmpbuf2, &ulen2);
@@ -4110,7 +4110,7 @@ S_reginclass(pTHX_ register regnode *n, register U8* p, register bool do_utf8)
 		    match = TRUE;
 		else if (flags & ANYOF_FOLD) {
 		    STRLEN ulen;
-		    U8 tmpbuf[UTF8_MAXLEN*2+1];
+		    U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 
 		    toLOWER_utf8(p, tmpbuf, &ulen);
 		    if (swash_fetch(sw, tmpbuf, do_utf8))
diff --git a/utf8.c b/utf8.c
index 747b78cfe6..3a8c13a193 100644
--- a/utf8.c
+++ b/utf8.c
@@ -903,7 +903,7 @@ Perl_is_uni_punct(pTHX_ UV c)
 bool
 Perl_is_uni_xdigit(pTHX_ UV c)
 {
-    U8 tmpbuf[UTF8_MAXLEN*2+1];
+    U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
     uvchr_to_utf8(tmpbuf, (UV)c);
     return is_utf8_xdigit(tmpbuf);
 }
@@ -911,7 +911,7 @@ Perl_is_uni_xdigit(pTHX_ UV c)
 UV
 Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp)
 {
-    U8 tmpbuf[UTF8_MAXLEN*2+1];
+    U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
     uvchr_to_utf8(tmpbuf, (UV)c);
     return to_utf8_upper(tmpbuf, p, lenp);
 }
@@ -919,7 +919,7 @@ Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp)
 UV
 Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp)
 {
-    U8 tmpbuf[UTF8_MAXLEN*2+1];
+    U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
     uvchr_to_utf8(tmpbuf, (UV)c);
     return to_utf8_title(tmpbuf, p, lenp);
 }
@@ -1596,8 +1596,8 @@ Perl_ibcmp_utf8(pTHX_ const char *s1, bool u1, register I32 len1, const char *s2
      STRLEN la, lb;
      UV ca, cb;
      STRLEN ulen1, ulen2;
-     U8 tmpbuf1[UTF8_MAXLEN*3+1];
-     U8 tmpbuf2[UTF8_MAXLEN*3+1];
+     U8 tmpbuf1[UTF8_MAXLEN_FOLD+1];
+     U8 tmpbuf2[UTF8_MAXLEN_FOLD+1];
      
      while (a < ae && b < be) {
 	  if (u1) {
diff --git a/utf8.h b/utf8.h
index 01c6199f93..1c2243e2da 100644
--- a/utf8.h
+++ b/utf8.h
@@ -135,7 +135,12 @@ END_EXTERN_C
 #define isIDFIRST_lazy(p)	isIDFIRST_lazy_if(p,1)
 #define isALNUM_lazy(p)		isALNUM_lazy_if(p,1)
 
-#define UTF8_MAXLEN 13 /* how wide can a single UTF8 encoded character become */
+/* how wide can a single UTF8 encoded character become */
+#define UTF8_MAXLEN 13
+/* how wide a character can become when upper/lowercased */
+#define UTF8_MAXLEN_UCLC (UTF8_MAXLEN*2)
+/* how wide a character can become when casefolded */
+#define UTF8_MAXLEN_FOLD (UTF8_MAXLEN*3)
 
 #define IN_BYTES (PL_curcop->op_private & HINT_BYTES)
 #define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)