Integrate from the maint-5.8/ branch :

changes 18219, 18236, 18242-3, 18247-8,
18253-5, 18257, 18273-6

p4raw-id: //depot/perl@18280
p4raw-branched: from //depot/maint-5.8/perl@18279 'branch in'
	t/op/lc_user.t
p4raw-integrated: from //depot/maint-5.8/perl@18279 'copy in'
	lib/File/Copy.pm (@17645..) lib/utf8_heavy.pl pod/perlsec.pod
	(@18080..) hints/irix_6.sh (@18173..) t/uni/tr_utf8.t
	(@18197..) pod/perlunicode.pod (@18242..) t/op/pat.t (@18248..)
	t/op/split.t (@18274..) 'edit in' pod/perlguts.pod (@18242..)
	'merge in' pp.c (@18126..) MANIFEST (@18234..)
p4raw-integrated: from //depot/maint-5.8/perl@18254 'merge in'
	pod/perldiag.pod (@18234..)

1 files changed, 17 insertions, 4 deletions

diff --git a/pp.c b/pp.c
index 1a1dceab7b..c78246e7de 100644
--- a/pp.c
+++ b/pp.c
@@ -3324,7 +3324,9 @@ PP(pp_ucfirst)
     STRLEN slen;
 
     SvGETMAGIC(sv);
-    if (DO_UTF8(sv) && (s = (U8*)SvPV_nomg(sv, slen)) && slen && UTF8_IS_START(*s)) {
+    if (DO_UTF8(sv) &&
+	(s = (U8*)SvPV_nomg(sv, slen)) && slen &&
+	UTF8_IS_START(*s)) {
 	U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 	STRLEN ulen;
 	STRLEN tculen;
@@ -3335,8 +3337,16 @@ PP(pp_ucfirst)
 
 	if (!SvPADTMP(sv) || SvREADONLY(sv)) {
 	    dTARGET;
+	    /* slen is the byte length of the whole SV.
+	     * ulen is the byte length of the original Unicode character
+	     * stored as UTF-8 at s.
+	     * tculen is the byte length of the freshly titlecased
+	     * Unicode character stored as UTF-8 at tmpbuf.
+	     * We first set the result to be the titlecased character,
+	     * and then append the rest of the SV data. */
 	    sv_setpvn(TARG, (char*)tmpbuf, tculen);
-	    sv_catpvn(TARG, (char*)(s + ulen), slen - ulen);
+	    if (slen > ulen)
+	        sv_catpvn(TARG, (char*)(s + ulen), slen - ulen);
 	    SvUTF8_on(TARG);
 	    SETs(TARG);
 	}
@@ -3376,7 +3386,9 @@ PP(pp_lcfirst)
     STRLEN slen;
 
     SvGETMAGIC(sv);
-    if (DO_UTF8(sv) && (s = (U8*)SvPV_nomg(sv, slen)) && slen && UTF8_IS_START(*s)) {
+    if (DO_UTF8(sv) &&
+	(s = (U8*)SvPV_nomg(sv, slen)) && slen &&
+	UTF8_IS_START(*s)) {
 	STRLEN ulen;
 	U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
 	U8 *tend;
@@ -3389,7 +3401,8 @@ PP(pp_lcfirst)
 	if (!SvPADTMP(sv) || (STRLEN)(tend - tmpbuf) != ulen || SvREADONLY(sv)) {
 	    dTARGET;
 	    sv_setpvn(TARG, (char*)tmpbuf, tend - tmpbuf);
-	    sv_catpvn(TARG, (char*)(s + ulen), slen - ulen);
+	    if (slen > ulen)
+	        sv_catpvn(TARG, (char*)(s + ulen), slen - ulen);
 	    SvUTF8_on(TARG);
 	    SETs(TARG);
 	}