From: Jarkko Hietaniemi <jhi@iki.fi>
Date: Tue, 23 Oct 2001 12:26:12 +0000 (+0000)
Subject: Fix multicharacter titlecase (ucfirst).
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=44bc797b2bc92bd45d512cde050d525bd921cf16;p=p5sagit%2Fp5-mst-13.2.git

Fix multicharacter titlecase (ucfirst).

p4raw-id: //depot/perl@12601
---

diff --git a/pod/perlunicode.pod b/pod/perlunicode.pod
index 9b4d2e3..37e2f22 100644
--- a/pod/perlunicode.pod
+++ b/pod/perlunicode.pod
@@ -505,10 +505,11 @@ pack('C0', ...).
 =item *
 
 Case translation operators use the Unicode case translation tables
-when provided character input.  Note that C<uc()> translates to
-uppercase, while C<ucfirst> translates to titlecase (for languages
-that make the distinction).  Naturally the corresponding backslash
-sequences have the same semantics.
+when provided character input.  Note that C<uc()> (also known as C<\U>
+in doublequoted strings) translates to uppercase, while C<ucfirst>
+(also known as C<\u> in doublequoted strings) translates to titlecase
+(for languages that make the distinction).  Naturally the
+corresponding backslash sequences have the same semantics.
 
 =item *
 
diff --git a/pp.c b/pp.c
index 134f243..3d93f75 100644
--- a/pp.c
+++ b/pp.c
@@ -3149,27 +3149,27 @@ PP(pp_ucfirst)
     register U8 *s;
     STRLEN slen;
 
-    if (DO_UTF8(sv) && (s = (U8*)SvPV(sv, slen)) && slen && UTF8_IS_START(*s)) {
-	STRLEN ulen;
+    if (DO_UTF8(sv)) {
 	U8 tmpbuf[UTF8_MAXLEN*2+1];
-	U8 *tend;
-	UV uv;
+	STRLEN ulen;
+	STRLEN tculen;
 
-	toTITLE_utf8(s, tmpbuf, &ulen); /* XXX --jhi */
-	uv = utf8_to_uvchr(tmpbuf, 0);
-	
-	tend = uvchr_to_utf8(tmpbuf, uv);
+	s = (U8*)SvPV(sv, slen);
+	utf8_to_uvchr(s, &ulen);
 
-	if (!SvPADTMP(sv) || tend - tmpbuf != ulen || SvREADONLY(sv)) {
+	toTITLE_utf8(s, tmpbuf, &tculen);
+	utf8_to_uvchr(tmpbuf, 0);
+
+	if (!SvPADTMP(sv) || SvREADONLY(sv)) {
 	    dTARGET;
-	    sv_setpvn(TARG, (char*)tmpbuf, tend - tmpbuf);
+	    sv_setpvn(TARG, (char*)tmpbuf, tculen);
 	    sv_catpvn(TARG, (char*)(s + ulen), slen - ulen);
 	    SvUTF8_on(TARG);
 	    SETs(TARG);
 	}
 	else {
 	    s = (U8*)SvPV_force(sv, slen);
-	    Copy(tmpbuf, s, ulen, U8);
+	    Copy(tmpbuf, s, tculen, U8);
 	}
     }
     else {
@@ -3209,7 +3209,7 @@ PP(pp_lcfirst)
 	U8 *tend;
 	UV uv;
 
-	toLOWER_utf8(s, tmpbuf, &ulen); /* XXX --jhi */
+	toLOWER_utf8(s, tmpbuf, &ulen);
 	uv = utf8_to_uvchr(tmpbuf, 0);
 	
 	tend = uvchr_to_utf8(tmpbuf, uv);
@@ -3277,7 +3277,7 @@ PP(pp_uc)
 	    d = (U8*)SvPVX(TARG);
 	    send = s + len;
 	    while (s < send) {
-	        toUPPER_utf8(s, tmpbuf, &ulen); /* XXX --jhi */
+	        toUPPER_utf8(s, tmpbuf, &ulen);
 		Copy(tmpbuf, d, ulen, U8);
 		d += ulen;
 		s += UTF8SKIP(s);
@@ -3344,7 +3344,7 @@ PP(pp_lc)
 	    d = (U8*)SvPVX(TARG);
 	    send = s + len;
 	    while (s < send) {
-	        toLOWER_utf8(s, tmpbuf, &ulen); /* XXX --jhi */
+	        toLOWER_utf8(s, tmpbuf, &ulen);
 		Copy(tmpbuf, d, ulen, U8);
 		d += ulen;
 		s += UTF8SKIP(s);
diff --git a/t/op/lc.t b/t/op/lc.t
index 9333c6c..5c43248 100644
--- a/t/op/lc.t
+++ b/t/op/lc.t
@@ -91,3 +91,18 @@ ok("\U\x{DF}ab\x{149}cd" eq "\x{53}\x{53}AB\x{2BC}\x{4E}CD",
 ok("\L\x{DF}AB\x{149}CD" eq "\x{DF}ab\x{149}cd",
    "multicharacter lowercase");
 
+# titlecase is used for \u / ucfirst.
+
+# \x{587} is ARMENIAN SMALL LIGATURE ECH YIWN and its titlecase is
+# \x{535}\x{582} ARMENIAN CAPITAL LETTER ECH + ARMENIAN SMALL LETTER YIWN
+# while its lowercase is 
+# \x{587} itself
+# and its uppercase is
+# \x{535}\x{552} ARMENIAN CAPITAL LETTER ECH + ARMENIAN CAPITAL LETTER YIWN
+
+$a = "\x{587}";
+
+ok("\L\x{587}" eq "\x{587}",        "ligature lowercase");
+ok("\u\x{587}" eq "\x{535}\x{582}", "ligature titlecase");
+ok("\U\x{587}" eq "\x{535}\x{552}", "ligature uppercase");
+