7 use bytes (); # for $bytes::hint_bits
8 $charnames::hint_bits = 0x20000;
11 # Icky 3.2 names with parentheses.
12 'LINE FEED' => 'LINE FEED (LF)',
13 'FORM FEED' => 'FORM FEED (FF)',
14 'CARRIAGE RETURN' => 'CARRIAGE RETURN (CR)',
15 'NEXT LINE' => 'NEXT LINE (NEL)',
17 'LF' => 'LINE FEED (LF)',
18 'FF' => 'FORM FEED (FF)',
19 'CR' => 'CARRIAGE RETURN (LF)',
20 'NEL' => 'NEXT LINE (NEL)',
21 'BOM' => 'BYTE ORDER MARK',
25 # Pre-3.2 compatibility (only for the first 256 characters).
26 'HORIZONTAL TABULATION' => 'CHARACTER TABULATION',
27 'VERTICAL TABULATION' => 'LINE TABULATION',
28 'FILE SEPARATOR' => 'INFORMATION SEPARATOR FOUR',
29 'GROUP SEPARATOR' => 'INFORMATION SEPARATOR THREE',
30 'RECORD SEPARATOR' => 'INFORMATION SEPARATOR TWO',
31 'UNIT SEPARATOR' => 'INFORMATION SEPARATOR ONE',
32 'PARTIAL LINE DOWN' => 'PARTIAL LINE FORWARD',
33 'PARTIAL LINE UP' => 'PARTIAL LINE BACKWARD',
38 # This is not optimized in any way yet
43 if (exists $alias1{$name}) {
44 $name = $alias1{$name};
46 if (exists $alias2{$name}) {
48 warnings::warnif('deprecated', qq{Unicode character name "$name" is deprecated, use "$alias2{$name}" instead});
49 $name = $alias2{$name};
56 if ($name eq "BYTE ORDER MARK") {
60 ## Suck in the code/name list as a big string.
62 ## "0052\t\tLATIN CAPITAL LETTER R\n"
63 $txt = do "unicore/Name.pl" unless $txt;
65 ## @off will hold the index into the code/name string of the start and
66 ## end of the name as we find it.
68 ## If :full, look for the the name exactly
69 if ($^H{charnames_full} and $txt =~ /\t\t\Q$name\E$/m) {
70 @off = ($-[0], $+[0]);
73 ## If we didn't get above, and :short allowed, look for the short name.
74 ## The short name is like "greek:Sigma"
76 if ($^H{charnames_short} and $name =~ /^(.+?):(.+)/s) {
77 my ($script, $cname) = ($1,$2);
78 my $case = ( $cname =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
79 if ($txt =~ m/\t\t\U$script\E (?:$case )?LETTER \U\Q$cname\E$/m) {
80 @off = ($-[0], $+[0]);
85 ## If we still don't have it, check for the name among the loaded
89 my $case = ( $name =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
90 for my $script ( @{$^H{charnames_scripts}} )
92 if ($txt =~ m/\t\t$script (?:$case )?LETTER \U\Q$name\E$/m) {
93 @off = ($-[0], $+[0]);
99 ## If we don't have it by now, give up.
101 carp "Unknown charname '$name'";
106 ## Now know where in the string the name starts.
107 ## The code, in hex, is before that.
109 ## The code can be 4-6 characters long, so we've got to sort of
110 ## go look for it, just after the newline that comes before $off[0].
112 ## This would be much easier if unicore/Name.pl had info in
113 ## a name/code order, instead of code/name order.
115 ## The +1 after the rindex() is to skip past the newline we're finding,
116 ## or, if the rindex() fails, to put us to an offset of zero.
118 my $hexstart = rindex($txt, "\n", $off[0]) + 1;
120 ## we know where it starts, so turn into number -
121 ## the ordinal for the char.
122 $ord = hex substr($txt, $hexstart, $off[0] - $hexstart);
125 if ($^H & $bytes::hint_bits) { # "use bytes" in effect?
127 return chr $ord if $ord <= 255;
128 my $hex = sprintf "%04x", $ord;
129 if (not defined $fname) {
130 $fname = substr $txt, $off[0] + 2, $off[1] - $off[0] - 2;
132 croak "Character 0x$hex with name '$fname' is above 0xFF";
135 no warnings 'utf8'; # allow even illegal characters
136 return pack "U", $ord;
141 shift; ## ignore class name
145 carp("`use charnames' needs explicit imports list");
147 $^H |= $charnames::hint_bits;
148 $^H{charnames} = \&charnames ;
151 ## fill %h keys with our @_ args.
156 $^H{charnames_full} = delete $h{':full'};
157 $^H{charnames_short} = delete $h{':short'};
158 $^H{charnames_scripts} = [map uc, keys %h];
161 ## If utf8? warnings are enabled, and some scripts were given,
162 ## see if at least we can find one letter of each script.
164 if (warnings::enabled('utf8') && @{$^H{charnames_scripts}})
166 $txt = do "unicore/Name.pl" unless $txt;
168 for my $script (@{$^H{charnames_scripts}})
170 if (not $txt =~ m/\t\t$script (?:CAPITAL |SMALL )?LETTER /) {
171 warnings::warn('utf8', "No such script: '$script'");
177 require Unicode::UCD; # for Unicode::UCD::_getcode()
184 carp "charnames::viacode() expects one argument";
189 my $code = Unicode::UCD::_getcode($arg);
194 $hex = sprintf "%04X", $arg;
196 carp("unexpected arg \"$arg\" to charnames::viacode()");
200 if ($code > 0x10FFFF) {
201 carp "Unicode characters only allocated up to 0x10FFFF (you asked for $hex)";
205 return $viacode{$hex} if exists $viacode{$hex};
207 $txt = do "unicore/Name.pl" unless $txt;
209 if ($txt =~ m/^$hex\t\t(.+)/m) {
210 return $viacode{$hex} = $1;
212 carp "Unknown charcode '$hex'";
222 carp "charnames::vianame() expects one name argument";
228 return $vianame{$arg} if exists $vianame{$arg};
230 $txt = do "unicore/Name.pl" unless $txt;
232 if ($txt =~ m/^([0-9A-F]+)\t\t($arg)/m) {
233 return $vianame{$arg} = hex $1;
245 charnames - define character names for C<\N{named}> string literal escapes
249 use charnames ':full';
250 print "\N{GREEK SMALL LETTER SIGMA} is called sigma.\n";
252 use charnames ':short';
253 print "\N{greek:Sigma} is an upper-case sigma.\n";
255 use charnames qw(cyrillic greek);
256 print "\N{sigma} is Greek sigma, and \N{be} is Cyrillic b.\n";
258 print charnames::viacode(0x1234); # prints "ETHIOPIC SYLLABLE SEE"
259 printf "%04X", charnames::vianame("GOTHIC LETTER AHSA"); # prints "10330"
263 Pragma C<use charnames> supports arguments C<:full>, C<:short> and
264 script names. If C<:full> is present, for expansion of
265 C<\N{CHARNAME}}> string C<CHARNAME> is first looked in the list of
266 standard Unicode names of chars. If C<:short> is present, and
267 C<CHARNAME> has the form C<SCRIPT:CNAME>, then C<CNAME> is looked up
268 as a letter in script C<SCRIPT>. If pragma C<use charnames> is used
269 with script name arguments, then for C<\N{CHARNAME}}> the name
270 C<CHARNAME> is looked up as a letter in the given scripts (in the
273 For lookup of C<CHARNAME> inside a given script C<SCRIPTNAME>
274 this pragma looks for the names
276 SCRIPTNAME CAPITAL LETTER CHARNAME
277 SCRIPTNAME SMALL LETTER CHARNAME
278 SCRIPTNAME LETTER CHARNAME
280 in the table of standard Unicode names. If C<CHARNAME> is lowercase,
281 then the C<CAPITAL> variant is ignored, otherwise the C<SMALL> variant
284 Note that C<\N{...}> is compile-time, it's a special form of string
285 constant used inside double-quoted strings: in other words, you cannot
286 use variables inside the C<\N{...}>. If you want similar run-time
287 functionality, use charnames::vianame().
289 For the C0 and C1 control characters (U+0000..U+001F, U+0080..U+009F)
290 as of Unicode 3.1, there are no official Unicode names but you can
291 use instead the ISO 6429 names (LINE FEED, ESCAPE, and so forth).
292 In Unicode 3.2 some naming changes will happen since ISO 6429 has been
293 updated. Also note that the U+UU80, U+0081, U+0084, and U+0099
294 do not have names even in ISO 6429.
296 =head1 CUSTOM TRANSLATORS
298 The mechanism of translation of C<\N{...}> escapes is general and not
299 hardwired into F<charnames.pm>. A module can install custom
300 translations (inside the scope which C<use>s the module) with the
301 following magic incantation:
303 use charnames (); # for $charnames::hint_bits
306 $^H |= $charnames::hint_bits;
307 $^H{charnames} = \&translator;
310 Here translator() is a subroutine which takes C<CHARNAME> as an
311 argument, and returns text to insert into the string instead of the
312 C<\N{CHARNAME}> escape. Since the text to insert should be different
313 in C<bytes> mode and out of it, the function should check the current
314 state of C<bytes>-flag as in:
316 use bytes (); # for $bytes::hint_bits
318 if ($^H & $bytes::hint_bits) {
319 return bytes_translator(@_);
322 return utf8_translator(@_);
326 =head1 charnames::viacode(code)
328 Returns the full name of the character indicated by the numeric code.
331 print charnames::viacode(0x2722);
333 prints "FOUR TEARDROP-SPOKED ASTERISK".
335 Returns undef if no name is known for the code.
337 This works only for the standard names, and does not yet apply
338 to custom translators.
340 Notice that the name returned for of U+FEFF is "ZERO WIDTH NO-BREAK
341 SPACE", not "BYTE ORDER MARK".
343 =head1 charnames::vianame(code)
345 Returns the code point indicated by the name.
348 printf "%04X", charnames::vianame("FOUR TEARDROP-SPOKED ASTERISK");
352 Returns undef if no name is known for the name.
354 This works only for the standard names, and does not yet aply
355 to custom translators.
359 A few aliases have been defined for convenience: instead of having
360 to use the official names
367 (yes, with parentheses) one can use
383 though that is of course not a legal character as such.
385 For backward compatibility one can use the old names for
386 certain C0 and C1 controls
390 HORIZONTAL TABULATION CHARACTER TABULATION
391 VERTICAL TABULATION LINE TABULATION
392 FILE SEPARATOR INFORMATION SEPARATOR FOUR
393 GROUP SEPARATOR INFORMATION SEPARATOR THREE
394 RECORD SEPARATOR INFORMATION SEPARATOR TWO
395 UNIT SEPARATOR INFORMATION SEPARATOR ONE
396 PARTIAL LINE DOWN PARTIAL LINE FORWARD
397 PARTIAL LINE UP PARTIAL LINE BACKWARD
399 but the old names in addition to giving the character
400 will also give a warning about being deprecated.
402 =head1 ILLEGAL CHARACTERS
404 If you ask for a character that does not exist, a warning is given
405 and the Unicode I<replacement character> "\x{FFFD}" is returned.
409 Since evaluation of the translation function happens in a middle of
410 compilation (of a string literal), the translation function should not
411 do any C<eval>s or C<require>s. This restriction should be lifted in
412 a future version of Perl.