4 our $VERSION = do { my @r = (q$Revision: 1.34 $ =~ /\d+/g); sprintf "%d."."%02d" x $#r, @r };
9 # Public, encouraged API is exported by default
17 our @Alias; # ordered matching list
18 our %Alias; # cached known aliases
24 unless (exists $Alias{$_})
26 $Alias{$_} = undef; # Recursion guard
27 for (my $i=0; $i < @Alias; $i += 2)
29 my $alias = $Alias[$i];
30 my $val = $Alias[$i+1];
32 if (ref($alias) eq 'Regexp' && $_ =~ $alias)
34 $DEBUG and warn "eval $val";
36 # $@ and warn "$val, $@";
38 elsif (ref($alias) eq 'CODE')
40 $DEBUG and warn "$alias", "->", "($val)";
41 $new = $alias->($val);
43 elsif (lc($_) eq lc($alias))
49 next if $new eq $_; # avoid (direct) recursion on bugs
50 $DEBUG and warn "$alias, $new";
51 my $enc = (ref($new)) ? $new : Encode::find_encoding($new);
62 if (my $e = $Alias{$_}){
67 warn "find_alias($class, $_)->name = $name";
76 my ($alias,$name) = splice(@_,0,2);
77 unshift(@Alias, $alias => $name); # newer one has precedence
78 # clear %Alias cache to allow overrides
82 if (ref($alias) eq 'Regexp' && $k =~ $alias)
84 $DEBUG and warn "delete \$Alias\{$k\}";
87 elsif (ref($alias) eq 'CODE')
89 $DEBUG and warn "delete \$Alias\{$k\}";
90 delete $Alias{$alias->($name)};
94 $DEBUG and warn "delete \$Alias\{$alias\}";
95 delete $Alias{$alias};
100 # Allow latin-1 style names as well
101 # 0 1 2 3 4 5 6 7 8 9 10
102 our @Latin2iso = ( 0, 1, 2, 3, 4, 9, 10, 13, 14, 15, 16 );
103 # Allow winlatin1 style names as well
113 'vietnamese' => 1258,
127 # Try all-lower-case version should all else fails
128 define_alias( qr/^(.*)$/ => '"\L$1"' );
131 define_alias( qr/^UCS-?2-?LE$/i => '"UCS-2LE"' );
132 define_alias( qr/^UCS-?2-?(BE)?$/i => '"UCS-2BE"',
133 qr/^UCS-?4-?(BE|LE)?$/i => 'uc("UTF-32$1")',
134 qr/^iso-10646-1$/i => '"UCS-2BE"' );
135 define_alias( qr/^UTF(16|32)-?BE$/i => '"UTF-$1BE"',
136 qr/^UTF(16|32)-?LE$/i => '"UTF-$1LE"',
137 qr/^UTF(16|32)$/i => '"UTF-$1"',
140 define_alias(qr/^(?:US-?)ascii$/i => '"ascii"');
141 define_alias('C' => 'ascii');
142 define_alias(qr/\bISO[-_]?646[-_]?US$/i => '"ascii"');
143 # Allow variants of iso-8859-1 etc.
144 define_alias( qr/\biso[-_]?(\d+)[-_](\d+)$/i => '"iso-$1-$2"' );
146 # At least HP-UX has these.
147 define_alias( qr/\biso8859(\d+)$/i => '"iso-8859-$1"' );
150 define_alias( qr/\b(?:hp-)?(arabic|greek|hebrew|kana|roman|thai|turkish)8$/i => '"${1}8"' );
152 # The Official name of ASCII.
153 define_alias( qr/\bANSI[-_]?X3\.4[-_]?1968$/i => '"ascii"' );
155 # This is a font issue, not an encoding issue.
156 # (The currency symbol of the Latin 1 upper half
157 # has been redefined as the euro symbol.)
158 define_alias( qr/^(.+)\@euro$/i => '"$1"' );
160 define_alias( qr/\b(?:iso[-_]?)?latin[-_]?(\d+)$/i
161 => 'defined $Encode::Alias::Latin2iso[$1] ? "iso-8859-$Encode::Alias::Latin2iso[$1]" : undef' );
163 define_alias( qr/\bwin(latin[12]|cyrillic|baltic|greek|turkish|
164 hebrew|arabic|baltic|vietnamese)$/ix =>
165 '"cp" . $Encode::Alias::Winlatin2cp{lc($1)}' );
167 # Common names for non-latin prefered MIME names
168 define_alias( 'ascii' => 'US-ascii',
169 'cyrillic' => 'iso-8859-5',
170 'arabic' => 'iso-8859-6',
171 'greek' => 'iso-8859-7',
172 'hebrew' => 'iso-8859-8',
173 'thai' => 'iso-8859-11',
174 'tis620' => 'iso-8859-11',
177 # At least AIX has IBM-NNN (surprisingly...) instead of cpNNN.
178 # And Microsoft has their own naming (again, surprisingly).
179 # And windows-* is registered in IANA!
180 define_alias( qr/\b(?:cp|ibm|ms|windows)[-_ ]?(\d{2,4})$/i => '"cp$1"');
182 # Sometimes seen with a leading zero.
183 # define_alias( qr/\bcp037\b/i => '"cp37"');
186 # predefined in *.ucm; unneeded
187 # define_alias( qr/\bmacIcelandic$/i => '"macIceland"');
188 define_alias( qr/^mac_(.*)$/i => '"mac$1"');
189 # Ououououou. gone. They are differente!
190 # define_alias( qr/\bmacRomanian$/i => '"macRumanian"');
192 # Standardize on the dashed versions.
193 # define_alias( qr/\butf8$/i => 'utf-8' );
194 define_alias( qr/\bkoi8r$/i => 'koi8-r' );
195 define_alias( qr/\bkoi8u$/i => 'koi8-u' );
197 unless ($Encode::ON_EBCDIC){
199 define_alias( qr/\beuc.*cn$/i => '"euc-cn"' );
200 define_alias( qr/\bcn.*euc$/i => '"euc-cn"' );
201 # define_alias( qr/\bGB[- ]?(\d+)$/i => '"euc-cn"' )
202 # CP936 doesn't have vendor-addon for GBK, so they're identical.
203 define_alias( qr/^gbk$/i => '"cp936"');
204 # This fixes gb2312 vs. euc-cn confusion, practically
205 define_alias( qr/\bGB[-_ ]?2312(?:\D.*$|$)/i => '"euc-cn"' );
207 define_alias( qr/\bjis$/i => '"7bit-jis"' );
208 define_alias( qr/\beuc.*jp$/i => '"euc-jp"' );
209 define_alias( qr/\bjp.*euc$/i => '"euc-jp"' );
210 define_alias( qr/\bujis$/i => '"euc-jp"' );
211 define_alias( qr/\bshift.*jis$/i => '"shiftjis"' );
212 define_alias( qr/\bsjis$/i => '"shiftjis"' );
214 define_alias( qr/\beuc.*kr$/i => '"euc-kr"' );
215 define_alias( qr/\bkr.*euc$/i => '"euc-kr"' );
216 # This fixes ksc5601 vs. euc-kr confusion, practically
217 define_alias( qr/(?:x-)?uhc$/i => '"cp949"' );
218 define_alias( qr/(?:x-)?windows-949$/i => '"cp949"' );
219 define_alias( qr/\bks_c_5601-1987$/i => '"cp949"' );
221 define_alias( qr/\bbig-?5$/i => '"big5-eten"' );
222 define_alias( qr/\bbig5-?et(?:en)?$/i => '"big5-eten"' );
223 define_alias( qr/\btca[-_]?big5$/i => '"big5-eten"' );
224 define_alias( qr/\bbig5-?hk(?:scs)?$/i => '"big5-hkscs"' );
225 define_alias( qr/\bhk(?:scs)?[-_]?big5$/i => '"big5-hkscs"' );
228 define_alias( qr/^UTF-8$/i => '"utf8"',);
229 # At last, Map white space and _ to '-'
230 define_alias( qr/^(\S+)[\s_]+(.*)$/i => '"$1-$2"' );
236 # TODO: HP-UX '8' encodings arabic8 greek8 hebrew8 kana8 thai8 turkish8
237 # TODO: HP-UX '15' encodings japanese15 korean15 roi15
238 # TODO: Cyrillic encoding ISO-IR-111 (useful?)
239 # TODO: Armenian encoding ARMSCII-8
240 # TODO: Hebrew encoding ISO-8859-8-1
241 # TODO: Thai encoding TCVN
242 # TODO: Vietnamese encodings VPS
243 # TODO: Mac Asian+African encodings: Arabic Armenian Bengali Burmese
244 # ChineseSimp ChineseTrad Devanagari Ethiopic ExtArabic
245 # Farsi Georgian Gujarati Gurmukhi Hebrew Japanese
246 # Kannada Khmer Korean Laotian Malayalam Mongolian
247 # Oriya Sinhalese Symbol Tamil Telugu Tibetan Vietnamese
251 Encode::Alias - alias definitions to encodings
257 define_alias( newName => ENCODING);
261 Allows newName to be used as an alias for ENCODING. ENCODING may be
262 either the name of an encoding or an encoding object (as described
265 Currently I<newName> can be specified in the following ways:
269 =item As a simple string.
271 =item As a qr// compiled regular expression, e.g.:
273 define_alias( qr/^iso8859-(\d+)$/i => '"iso-8859-$1"' );
275 In this case, if I<ENCODING> is not a reference, it is C<eval>-ed
276 in order to allow C<$1> etc. to be substituted. The example is one
277 way to alias names as used in X11 fonts to the MIME names for the
278 iso-8859-* family. Note the double quotes inside the single quotes.
280 If you are using a regex here, you have to use the quotes as shown or
281 it won't work. Also note that regex handling is tricky even for the
282 experienced. Use it with caution.
284 =item As a code reference, e.g.:
286 define_alias( sub { return /^iso8859-(\d+)$/i ? "iso-8859-$1" : undef } , '');
288 In this case, C<$_> will be set to the name that is being looked up and
289 I<ENCODING> is passed to the sub as its first argument. The example
290 is another way to alias names as used in X11 fonts to the MIME names
291 for the iso-8859-* family.
295 =head2 Alias overloading
297 You can override predefined aliases by simply applying define_alias().
298 The new alias is always evaluated first, and when neccessary,
299 define_alias() flushes the internal cache to make the new definition
302 # redirect SHIFT_JIS to MS/IBM Code Page 932, which is a
303 # superset of SHIFT_JIS
305 define_alias( qr/shift.*jis$/i => '"cp932"' );
306 define_alias( qr/sjis$/i => '"cp932"' );
308 If you want to zap all predefined aliases, you can use
310 Encode::Alias->undef_aliases;
314 Encode::Alias->init_aliases;
316 gets the factory settings back.
320 L<Encode>, L<Encode::Supported>