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 (CR)',
20 'NEL' => 'NEXT LINE (NEL)',
21 # More convenience. For futher convencience,
22 # it is suggested some way using using the NamesList
23 # aliases is implemented.
24 'ZWNJ' => 'ZERO WIDTH NON-JOINER',
25 'ZWJ' => 'ZERO WIDTH JOINER',
26 'BOM' => 'BYTE ORDER MARK',
30 # Pre-3.2 compatibility (only for the first 256 characters).
31 'HORIZONTAL TABULATION' => 'CHARACTER TABULATION',
32 'VERTICAL TABULATION' => 'LINE TABULATION',
33 'FILE SEPARATOR' => 'INFORMATION SEPARATOR FOUR',
34 'GROUP SEPARATOR' => 'INFORMATION SEPARATOR THREE',
35 'RECORD SEPARATOR' => 'INFORMATION SEPARATOR TWO',
36 'UNIT SEPARATOR' => 'INFORMATION SEPARATOR ONE',
37 'PARTIAL LINE DOWN' => 'PARTIAL LINE FORWARD',
38 'PARTIAL LINE UP' => 'PARTIAL LINE BACKWARD',
43 # This is not optimized in any way yet
48 if (exists $alias1{$name}) {
49 $name = $alias1{$name};
51 if (exists $alias2{$name}) {
53 warnings::warnif('deprecated', qq{Unicode character name "$name" is deprecated, use "$alias2{$name}" instead});
54 $name = $alias2{$name};
61 if ($name eq "BYTE ORDER MARK") {
65 ## Suck in the code/name list as a big string.
67 ## "0052\t\tLATIN CAPITAL LETTER R\n"
68 $txt = do "unicore/Name.pl" unless $txt;
70 ## @off will hold the index into the code/name string of the start and
71 ## end of the name as we find it.
73 ## If :full, look for the the name exactly
74 if ($^H{charnames_full} and $txt =~ /\t\t\Q$name\E$/m) {
75 @off = ($-[0], $+[0]);
78 ## If we didn't get above, and :short allowed, look for the short name.
79 ## The short name is like "greek:Sigma"
81 if ($^H{charnames_short} and $name =~ /^(.+?):(.+)/s) {
82 my ($script, $cname) = ($1,$2);
83 my $case = ( $cname =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
84 if ($txt =~ m/\t\t\U$script\E (?:$case )?LETTER \U\Q$cname\E$/m) {
85 @off = ($-[0], $+[0]);
90 ## If we still don't have it, check for the name among the loaded
94 my $case = ( $name =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
95 for my $script ( @{$^H{charnames_scripts}} )
97 if ($txt =~ m/\t\t$script (?:$case )?LETTER \U\Q$name\E$/m) {
98 @off = ($-[0], $+[0]);
104 ## If we don't have it by now, give up.
106 carp "Unknown charname '$name'";
111 ## Now know where in the string the name starts.
112 ## The code, in hex, is before that.
114 ## The code can be 4-6 characters long, so we've got to sort of
115 ## go look for it, just after the newline that comes before $off[0].
117 ## This would be much easier if unicore/Name.pl had info in
118 ## a name/code order, instead of code/name order.
120 ## The +1 after the rindex() is to skip past the newline we're finding,
121 ## or, if the rindex() fails, to put us to an offset of zero.
123 my $hexstart = rindex($txt, "\n", $off[0]) + 1;
125 ## we know where it starts, so turn into number -
126 ## the ordinal for the char.
127 $ord = hex substr($txt, $hexstart, $off[0] - $hexstart);
130 if ($^H & $bytes::hint_bits) { # "use bytes" in effect?
132 return chr $ord if $ord <= 255;
133 my $hex = sprintf "%04x", $ord;
134 if (not defined $fname) {
135 $fname = substr $txt, $off[0] + 2, $off[1] - $off[0] - 2;
137 croak "Character 0x$hex with name '$fname' is above 0xFF";
140 no warnings 'utf8'; # allow even illegal characters
141 return pack "U", $ord;
146 shift; ## ignore class name
150 carp("`use charnames' needs explicit imports list");
152 $^H |= $charnames::hint_bits;
153 $^H{charnames} = \&charnames ;
156 ## fill %h keys with our @_ args.
161 $^H{charnames_full} = delete $h{':full'};
162 $^H{charnames_short} = delete $h{':short'};
163 $^H{charnames_scripts} = [map uc, keys %h];
166 ## If utf8? warnings are enabled, and some scripts were given,
167 ## see if at least we can find one letter of each script.
169 if (warnings::enabled('utf8') && @{$^H{charnames_scripts}})
171 $txt = do "unicore/Name.pl" unless $txt;
173 for my $script (@{$^H{charnames_scripts}})
175 if (not $txt =~ m/\t\t$script (?:CAPITAL |SMALL )?LETTER /) {
176 warnings::warn('utf8', "No such script: '$script'");
182 require Unicode::UCD; # for Unicode::UCD::_getcode()
189 carp "charnames::viacode() expects one argument";
194 my $code = Unicode::UCD::_getcode($arg);
199 $hex = sprintf "%04X", $arg;
201 carp("unexpected arg \"$arg\" to charnames::viacode()");
205 if ($code > 0x10FFFF) {
206 carp sprintf "Unicode characters only allocated up to U+10FFFF (you asked for U+%X)", $hex;
210 return $viacode{$hex} if exists $viacode{$hex};
212 $txt = do "unicore/Name.pl" unless $txt;
214 if ($txt =~ m/^$hex\t\t(.+)/m) {
215 return $viacode{$hex} = $1;
226 carp "charnames::vianame() expects one name argument";
232 return chr hex $1 if $arg =~ /^U\+([0-9a-fA-F]+)$/;
234 return $vianame{$arg} if exists $vianame{$arg};
236 $txt = do "unicore/Name.pl" unless $txt;
238 my $pos = index $txt, "\t\t$arg\n";
240 my $posLF = rindex $txt, "\n", $pos;
241 (my $code = substr $txt, $posLF + 1, 6) =~ tr/\t//d;
242 return $vianame{$arg} = hex $code;
244 # If $pos is at the 1st line, $posLF must be $[ - 1 (not found);
245 # then $posLF + 1 equals to $[ (at the beginning of $txt).
246 # Otherwise $posLF is the position of "\n";
247 # then $posLF + 1 must be the position of the next to "\n"
248 # (the beginning of the line).
249 # substr($txt, $posLF + 1, 6) may be "0000\t\t", "00A1\t\t",
250 # "10300\t", "100000", etc. So we can get the code via removing TAB.
262 charnames - define character names for C<\N{named}> string literal escapes
266 use charnames ':full';
267 print "\N{GREEK SMALL LETTER SIGMA} is called sigma.\n";
269 use charnames ':short';
270 print "\N{greek:Sigma} is an upper-case sigma.\n";
272 use charnames qw(cyrillic greek);
273 print "\N{sigma} is Greek sigma, and \N{be} is Cyrillic b.\n";
275 print charnames::viacode(0x1234); # prints "ETHIOPIC SYLLABLE SEE"
276 printf "%04X", charnames::vianame("GOTHIC LETTER AHSA"); # prints "10330"
280 Pragma C<use charnames> supports arguments C<:full>, C<:short> and
281 script names. If C<:full> is present, for expansion of
282 C<\N{CHARNAME}> string C<CHARNAME> is first looked in the list of
283 standard Unicode names of chars. If C<:short> is present, and
284 C<CHARNAME> has the form C<SCRIPT:CNAME>, then C<CNAME> is looked up
285 as a letter in script C<SCRIPT>. If pragma C<use charnames> is used
286 with script name arguments, then for C<\N{CHARNAME}> the name
287 C<CHARNAME> is looked up as a letter in the given scripts (in the
290 For lookup of C<CHARNAME> inside a given script C<SCRIPTNAME>
291 this pragma looks for the names
293 SCRIPTNAME CAPITAL LETTER CHARNAME
294 SCRIPTNAME SMALL LETTER CHARNAME
295 SCRIPTNAME LETTER CHARNAME
297 in the table of standard Unicode names. If C<CHARNAME> is lowercase,
298 then the C<CAPITAL> variant is ignored, otherwise the C<SMALL> variant
301 Note that C<\N{...}> is compile-time, it's a special form of string
302 constant used inside double-quoted strings: in other words, you cannot
303 use variables inside the C<\N{...}>. If you want similar run-time
304 functionality, use charnames::vianame().
306 For the C0 and C1 control characters (U+0000..U+001F, U+0080..U+009F)
307 as of Unicode 3.1, there are no official Unicode names but you can use
308 instead the ISO 6429 names (LINE FEED, ESCAPE, and so forth). In
309 Unicode 3.2 (as of Perl 5.8) some naming changes take place ISO 6429
310 has been updated, see L</ALIASES>. Also note that the U+UU80, U+0081,
311 U+0084, and U+0099 do not have names even in ISO 6429.
313 Since the Unicode standard uses "U+HHHH", so can you: "\N{U+263a}"
314 is the Unicode smiley face, or "\N{WHITE SMILING FACE}".
316 =head1 CUSTOM TRANSLATORS
318 The mechanism of translation of C<\N{...}> escapes is general and not
319 hardwired into F<charnames.pm>. A module can install custom
320 translations (inside the scope which C<use>s the module) with the
321 following magic incantation:
323 use charnames (); # for $charnames::hint_bits
326 $^H |= $charnames::hint_bits;
327 $^H{charnames} = \&translator;
330 Here translator() is a subroutine which takes C<CHARNAME> as an
331 argument, and returns text to insert into the string instead of the
332 C<\N{CHARNAME}> escape. Since the text to insert should be different
333 in C<bytes> mode and out of it, the function should check the current
334 state of C<bytes>-flag as in:
336 use bytes (); # for $bytes::hint_bits
338 if ($^H & $bytes::hint_bits) {
339 return bytes_translator(@_);
342 return utf8_translator(@_);
346 =head1 charnames::viacode(code)
348 Returns the full name of the character indicated by the numeric code.
351 print charnames::viacode(0x2722);
353 prints "FOUR TEARDROP-SPOKED ASTERISK".
355 Returns undef if no name is known for the code.
357 This works only for the standard names, and does not yet apply
358 to custom translators.
360 Notice that the name returned for of U+FEFF is "ZERO WIDTH NO-BREAK
361 SPACE", not "BYTE ORDER MARK".
363 =head1 charnames::vianame(name)
365 Returns the code point indicated by the name.
368 printf "%04X", charnames::vianame("FOUR TEARDROP-SPOKED ASTERISK");
372 Returns undef if the name is unknown.
374 This works only for the standard names, and does not yet apply
375 to custom translators.
379 A few aliases have been defined for convenience: instead of having
380 to use the official names
387 (yes, with parentheses) one can use
408 for ZERO WIDTH NON-JOINER and ZERO WIDTH JOINER.
410 For backward compatibility one can use the old names for
411 certain C0 and C1 controls
415 HORIZONTAL TABULATION CHARACTER TABULATION
416 VERTICAL TABULATION LINE TABULATION
417 FILE SEPARATOR INFORMATION SEPARATOR FOUR
418 GROUP SEPARATOR INFORMATION SEPARATOR THREE
419 RECORD SEPARATOR INFORMATION SEPARATOR TWO
420 UNIT SEPARATOR INFORMATION SEPARATOR ONE
421 PARTIAL LINE DOWN PARTIAL LINE FORWARD
422 PARTIAL LINE UP PARTIAL LINE BACKWARD
424 but the old names in addition to giving the character
425 will also give a warning about being deprecated.
427 =head1 ILLEGAL CHARACTERS
429 If you ask by name for a character that does not exist, a warning is
430 given and the Unicode I<replacement character> "\x{FFFD}" is returned.
432 If you ask by code for a character that does not exist, no warning is
433 given and C<undef> is returned. (Though if you ask for a code point
434 past U+10FFFF you do get a warning.)
438 Since evaluation of the translation function happens in a middle of
439 compilation (of a string literal), the translation function should not
440 do any C<eval>s or C<require>s. This restriction should be lifted in
441 a future version of Perl.