Upgrade to Unicode::Normalize 1.01

p4raw-id: //depot/perl@28406

7 files changed, 116 insertions, 47 deletions

diff --git a/ext/Unicode/Normalize/Changes b/ext/Unicode/Normalize/Changes
index 8d05a3dfd6..b7eb1d53ca 100644
--- a/ext/Unicode/Normalize/Changes
+++ b/ext/Unicode/Normalize/Changes
@@ -1,5 +1,10 @@
 Revision history for Perl extension Unicode::Normalize.
 
+1.01  Tue Jun 13 22:01:53 2006
+    - XSUB: sv_setpvn() needs cast to (char*).
+    - XSUB: avoid double FETCH for tied scalar variables.
+    - added tie.t.
+
 1.00  Thu May 25 20:35:06 2006
     - Pure Perl: compose($not_canonically_reordered) works like that in XSUB,
       where an intervening character with higher combining class blocks
diff --git a/ext/Unicode/Normalize/Normalize.pm b/ext/Unicode/Normalize/Normalize.pm
index 16d7664b33..b51177d922 100644
--- a/ext/Unicode/Normalize/Normalize.pm
+++ b/ext/Unicode/Normalize/Normalize.pm
@@ -13,7 +13,7 @@ use Carp;
 
 no warnings 'utf8';
 
-our $VERSION = '1.00';
+our $VERSION = '1.01';
 our $PACKAGE = __PACKAGE__;
 
 require Exporter;
diff --git a/ext/Unicode/Normalize/Normalize.xs b/ext/Unicode/Normalize/Normalize.xs
index 3dea0f4746..b051e2a0fb 100644
--- a/ext/Unicode/Normalize/Normalize.xs
+++ b/ext/Unicode/Normalize/Normalize.xs
@@ -181,17 +181,17 @@ static U8* pv_cat_decompHangul(U8* d, UV uv)
     return d;
 }
 
-static char * sv_2pvunicode(SV *sv, STRLEN *lp)
+static char* sv_2pvunicode(SV *sv, STRLEN *lp)
 {
     char *s;
     STRLEN len;
-    s = (char*)SvPV(sv,len);
+    s = SvPV(sv,len);
     if (!SvUTF8(sv)) {
-	SV* tmpsv = sv_mortalcopy(sv);
+	SV* tmpsv = sv_2mortal(newSVpvn(s, len));
 	if (!SvPOK(tmpsv))
-	    (void)sv_pvn_force(tmpsv,&len);
+	    s = SvPV_force(tmpsv,len);
 	sv_utf8_upgrade(tmpsv);
-	s = (char*)SvPV(tmpsv,len);
+	s = SvPV(tmpsv,len);
     }
     if (lp)
 	*lp = len;
@@ -292,7 +292,7 @@ U8* pv_utf8_reorder(U8* s, STRLEN slen, U8* d, STRLEN dlen)
 		else {
 		    Renew(seq_ext, seq_max, UNF_cc);
 		}
-		seq_ptr = seq_ext; /* till now use seq_ext */
+		seq_ptr = seq_ext; /* use seq_ext from now */
 	    }
 
 	    seq_ptr[cc_pos].cc  = curCC;
@@ -336,7 +336,7 @@ U8* pv_utf8_compose(U8* s, STRLEN slen, U8* d, STRLEN dlen, bool iscontig)
     U8* e = s + slen;
     U8* dend = d + dlen;
 
-    UV uvS; /* code point of the starter */
+    UV uvS = 0; /* code point of the starter */
     bool valid_uvS = FALSE; /* if FALSE, uvS isn't initialized yet */
     U8 preCC = 0;
 
@@ -464,7 +464,7 @@ decompose(src, compat = &PL_sv_no)
     dst = newSVpvn("", 0);
     dlen = slen;
     New(0, d, dlen+1, U8);
-    dend = pv_utf8_decompose(s, slen, &d, dlen, SvTRUE(compat));
+    dend = pv_utf8_decompose(s, slen, &d, dlen, (bool)SvTRUE(compat));
     sv_setpvn(dst, (char *)d, dend - d);
     SvUTF8_on(dst);
     Safefree(d);
@@ -691,7 +691,7 @@ checkFCD(src)
     for (p = s; p < e; p += retlen) {
 	U8 *sCan;
 	UV uvLead;
-	STRLEN canlen, canret;
+	STRLEN canlen = 0;
 	UV uv = utf8n_to_uvuni(p, e - p, &retlen, AllowAnyUTF);
 	if (!retlen)
 	    croak(ErrRetlenIsZero, "checkFCD or -FCC");
@@ -699,6 +699,7 @@ checkFCD(src)
 	sCan = (U8*) dec_canonical(uv);
 
 	if (sCan) {
+	    STRLEN canret;
 	    canlen = (STRLEN)strlen((char *) sCan);
 	    uvLead = utf8n_to_uvuni(sCan, canlen, &canret, AllowAnyUTF);
 	    if (!canret)
@@ -721,6 +722,7 @@ checkFCD(src)
 	}
 
 	if (sCan) {
+	    STRLEN canret;
 	    UV uvTrail;
 	    U8* eCan = sCan + canlen;
 	    U8* pCan = utf8_hop(eCan, -1);
diff --git a/ext/Unicode/Normalize/README b/ext/Unicode/Normalize/README
index e70d7ea037..79a2b5e15f 100644
--- a/ext/Unicode/Normalize/README
+++ b/ext/Unicode/Normalize/README
@@ -1,10 +1,12 @@
-Unicode/Normalize version 1.00
+Unicode/Normalize version 1.01
 ===================================
 
 Unicode::Normalize - Unicode Normalization Forms
 
 SYNOPSIS
 
+(1) using function names exported by default:
+
   use Unicode::Normalize;
 
   $NFD_string  = NFD($string);  # Normalization Form D
@@ -12,7 +14,7 @@ SYNOPSIS
   $NFKD_string = NFKD($string); # Normalization Form KD
   $NFKC_string = NFKC($string); # Normalization Form KC
 
-  or
+(2) using function names exported on request:
 
   use Unicode::Normalize 'normalize';
 
@@ -21,11 +23,10 @@ SYNOPSIS
   $NFKD_string = normalize('KD', $string);  # Normalization Form KD
   $NFKC_string = normalize('KC', $string);  # Normalization Form KC
 
-
 INSTALLATION
 
-Perl 5.6.1 or later.
-(Perl 5.8.0 or later is recommended.)
+Perl 5.6.1 or later is required.
+Perl 5.8.1 or later is recommended.
 
 To install this module (XSUB: needs a C compiler), type the following:
 
@@ -55,42 +56,34 @@ type the following (!! "enableXS" must run before "Makefile.PL" !!):
 
 DEPENDENCIES
 
-This module requires other modules and libraries following:
-
-Carp
-Exporter
-File::Copy
-File::Spec
+This module requires some core modules and pragmas,
+and the following files from Unicode Character Database,
+which are included in recent perl core distributions.
 
-unicore/CombiningClass.pl         (or unicode/CombiningClass.pl)
-unicore/Decomposition.pl          (or unicode/Decomposition.pl)
-unicore/CompositionExclusions.txt (or unicode/CompExcl.txt)
+- unicore/CombiningClass.pl (or unicode/CombiningClass.pl)
+- unicore/Decomposition.pl (or unicode/Decomposition.pl)
+- unicore/CompositionExclusions.txt (or unicode/CompExcl.txt)
 
-CAVEAT
+CAVEATS
 
-(1) In the perl-current, unicore/CompExcl.txt
-  is renamed unicore/CompositionExclusions.txt.
+(1) After these unicore/*.* files are updated:
 
-(2) After these unicore/*.* files are updated.
-
-  In the case of an XS edition:
+In the case of XSUB:
     You must rebuild the module,
     as the data will be compiled on building.
 
-  In the case of a pure Perl edition:
+In the case of pure perl:
     Rebuilding is not necessary,
     as the data will be read on requirement.
 
-(3) Pure Perl edition, Normalize.pmN, may work without any other file
+(2) Normalize.pmN (pure perl module) may work without any other file
     in this distribution (it must be renamed Normalize.pm, though)
 
-COPYRIGHT AND LICENCE
-
-  SADAHIRO Tomoyuki <SADAHIRO@cpan.org>
+COPYRIGHT AND LICENSE
 
-  http://homepage1.nifty.com/nomenclator/perl/
+SADAHIRO Tomoyuki <SADAHIRO@cpan.org>
 
-  Copyright(C) 2001-2006, SADAHIRO Tomoyuki. Japan. All rights reserved.
+Copyright(C) 2001-2006, SADAHIRO Tomoyuki. Japan. All rights reserved.
 
-  This module is free software; you can redistribute it
-  and/or modify it under the same terms as Perl itself.
+This module is free software; you can redistribute it and/or
+modify it under the same terms as Perl itself.
diff --git a/ext/Unicode/Normalize/t/func.t b/ext/Unicode/Normalize/t/func.t
index 81421ce1f3..794fff65bf 100644
--- a/ext/Unicode/Normalize/t/func.t
+++ b/ext/Unicode/Normalize/t/func.t
@@ -101,9 +101,9 @@ sub uprops {
   return $r;
 }
 
-ok(uprops(0x0000), 'xsnfbdmckyg');
-ok(uprops(0x0029), 'xsnfbdmckyg');
-ok(uprops(0x0041), 'xsnfbdmckyg');
+ok(uprops(0x0000), 'xsnfbdmckyg'); # NULL
+ok(uprops(0x0029), 'xsnfbdmckyg'); # RIGHT PARENTHESIS
+ok(uprops(0x0041), 'xsnfbdmckyg'); # LATIN CAPITAL LETTER A
 ok(uprops(0x00A0), 'xsnfbdmcKyG'); # NO-BREAK SPACE
 ok(uprops(0x00C0), 'xsnfbDmcKyg'); # LATIN CAPITAL LETTER A WITH GRAVE
 ok(uprops(0x0300), 'xsnfBdMckYg'); # COMBINING GRAVE ACCENT
@@ -113,7 +113,7 @@ ok(uprops(0x0958), 'XsnFbDmCKyG'); # DEVANAGARI LETTER QA
 ok(uprops(0x0F43), 'XsnFbDmCKyG'); # TIBETAN LETTER GHA
 ok(uprops(0x1100), 'xsnfbdmckyg'); # HANGUL CHOSEONG KIYEOK
 ok(uprops(0x1161), 'xsnfBdMckYg'); # HANGUL JUNGSEONG A
-ok(uprops(0x11AF), 'xsnfBdMckYg'); # HANGUL JONGSEONG RIEU
+ok(uprops(0x11AF), 'xsnfBdMckYg'); # HANGUL JONGSEONG RIEUL
 ok(uprops(0x212B), 'xSnFbDmCKyG'); # ANGSTROM SIGN
 ok(uprops(0xAC00), 'xsnfbDmcKyg'); # HANGUL SYLLABLE GA
 ok(uprops(0xF900), 'xSnFbDmCKyG'); # CJK COMPATIBILITY IDEOGRAPH-F900
diff --git a/ext/Unicode/Normalize/t/illegal.t b/ext/Unicode/Normalize/t/illegal.t
index 9d18aadbb6..976e5097a2 100644
--- a/ext/Unicode/Normalize/t/illegal.t
+++ b/ext/Unicode/Normalize/t/illegal.t
@@ -24,10 +24,6 @@ BEGIN {
 
 #########################
 
-use Test;
-use strict;
-use warnings;
-
 BEGIN {
     use Unicode::Normalize qw(:all);
 
@@ -39,6 +35,10 @@ BEGIN {
     }
 }
 
+use Test;
+use strict;
+use warnings;
+
 BEGIN { plan tests => 112 };
 
 #########################
diff --git a/ext/Unicode/Normalize/t/tie.t b/ext/Unicode/Normalize/t/tie.t
new file mode 100644
index 0000000000..c721491726
--- /dev/null
+++ b/ext/Unicode/Normalize/t/tie.t
@@ -0,0 +1,69 @@
+
+BEGIN {
+    unless ("A" eq pack('U', 0x41)) {
+	print "1..0 # Unicode::Normalize " .
+	    "cannot stringify a Unicode code point\n";
+	exit 0;
+    }
+}
+
+BEGIN {
+    if ($ENV{PERL_CORE}) {
+        chdir('t') if -d 't';
+        @INC = $^O eq 'MacOS' ? qw(::lib) : qw(../lib);
+    }
+}
+
+#########################
+
+BEGIN {
+    use Unicode::Normalize qw(:all);
+
+    unless (exists &Unicode::Normalize::bootstrap or 5.008 <= $]) {
+	print "1..0 # skipped: XSUB, or Perl 5.8.0 or later".
+		" needed for this test\n";
+	print $@;
+	exit;
+    }
+}
+
+use Test;
+use strict;
+use warnings;
+BEGIN { plan tests => 16 };
+
+package tiescalar;
+sub TIESCALAR {
+    my ($class, $instance) = @_;
+    return bless \$instance => $class;
+}
+sub FETCH   { return ${$_[0]}++ }
+sub STORE   { return ${$_[0]} = $_[1] }
+sub DESTROY { undef ${$_[0]} }
+
+#########################
+
+package main;
+
+tie my $tie1, 'tiescalar', "123";
+ok(NFD($tie1),  123);
+ok(NFC($tie1),  124);
+ok(NFKD($tie1), 125);
+ok(NFKC($tie1), 126);
+ok(FCD($tie1),  127);
+ok(FCC($tie1),  128);
+
+tie my $tie2, 'tiescalar', "256";
+ok(normalize('NFD',  $tie2), 256);
+ok(normalize('NFC',  $tie2), 257);
+ok(normalize('NFKD', $tie2), 258);
+ok(normalize('NFKC', $tie2), 259);
+ok(normalize('FCD',  $tie2), 260);
+ok(normalize('FCC',  $tie2), 261);
+
+tie my $tie3, 'tiescalar', "315";
+ok(decompose($tie3),         315);
+ok(reorder($tie3),           316);
+ok(compose($tie3),           317);
+ok(composeContiguous($tie3), 318);
+