8 use bytes (); # for $bytes::hint_bits
9 $charnames::hint_bits = 0x20000; # HINT_LOCALIZE_HH
12 # Icky 3.2 names with parentheses.
13 'LINE FEED' => 'LINE FEED (LF)',
14 'FORM FEED' => 'FORM FEED (FF)',
15 'CARRIAGE RETURN' => 'CARRIAGE RETURN (CR)',
16 'NEXT LINE' => 'NEXT LINE (NEL)',
18 'LF' => 'LINE FEED (LF)',
19 'FF' => 'FORM FEED (FF)',
20 'CR' => 'CARRIAGE RETURN (CR)',
21 'NEL' => 'NEXT LINE (NEL)',
22 # More convenience. For futher convencience,
23 # it is suggested some way using using the NamesList
24 # aliases is implemented.
25 'ZWNJ' => 'ZERO WIDTH NON-JOINER',
26 'ZWJ' => 'ZERO WIDTH JOINER',
27 'BOM' => 'BYTE ORDER MARK',
31 # Pre-3.2 compatibility (only for the first 256 characters).
32 'HORIZONTAL TABULATION' => 'CHARACTER TABULATION',
33 'VERTICAL TABULATION' => 'LINE TABULATION',
34 'FILE SEPARATOR' => 'INFORMATION SEPARATOR FOUR',
35 'GROUP SEPARATOR' => 'INFORMATION SEPARATOR THREE',
36 'RECORD SEPARATOR' => 'INFORMATION SEPARATOR TWO',
37 'UNIT SEPARATOR' => 'INFORMATION SEPARATOR ONE',
38 'PARTIAL LINE DOWN' => 'PARTIAL LINE FORWARD',
39 'PARTIAL LINE UP' => 'PARTIAL LINE BACKWARD',
43 # User defined aliasses. Even more convenient :)
50 my $alias = ref $_[0] ? $_[0] : { @_ };
51 @alias3{keys %$alias} = values %$alias;
56 my ($arg, $file) = @_;
57 if (-f $arg && File::Spec->file_name_is_absolute ($arg)) {
60 elsif ($arg =~ m/^\w+$/) {
61 $file = "unicore/${arg}_alias.pl";
64 croak "Charnames alias files can only have identifier characters";
66 if (my @alias = do $file) {
67 @alias == 1 && !defined $alias[0] and
68 croak "$file cannot be used as alias file for charnames";
70 croak "$file did not return a (valid) list of alias pairs";
77 # This is not optimized in any way yet
82 if (exists $alias1{$name}) {
83 $name = $alias1{$name};
85 elsif (exists $alias2{$name}) {
87 warnings::warnif('deprecated', qq{Unicode character name "$name" is deprecated, use "$alias2{$name}" instead});
88 $name = $alias2{$name};
90 elsif (exists $alias3{$name}) {
91 $name = $alias3{$name};
98 if ($name eq "BYTE ORDER MARK") {
102 ## Suck in the code/name list as a big string.
104 ## "0052\t\tLATIN CAPITAL LETTER R\n"
105 $txt = do "unicore/Name.pl" unless $txt;
107 ## @off will hold the index into the code/name string of the start and
108 ## end of the name as we find it.
110 ## If :full, look for the name exactly
111 if ($^H{charnames_full} and $txt =~ /\t\t\Q$name\E$/m) {
112 @off = ($-[0], $+[0]);
115 ## If we didn't get above, and :short allowed, look for the short name.
116 ## The short name is like "greek:Sigma"
118 if ($^H{charnames_short} and $name =~ /^(.+?):(.+)/s) {
119 my ($script, $cname) = ($1, $2);
120 my $case = $cname =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL";
121 if ($txt =~ m/\t\t\U$script\E (?:$case )?LETTER \U\Q$cname\E$/m) {
122 @off = ($-[0], $+[0]);
127 ## If we still don't have it, check for the name among the loaded
130 my $case = $name =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL";
131 for my $script (@{$^H{charnames_scripts}}) {
132 if ($txt =~ m/\t\t$script (?:$case )?LETTER \U\Q$name\E$/m) {
133 @off = ($-[0], $+[0]);
139 ## If we don't have it by now, give up.
141 carp "Unknown charname '$name'";
146 ## Now know where in the string the name starts.
147 ## The code, in hex, is before that.
149 ## The code can be 4-6 characters long, so we've got to sort of
150 ## go look for it, just after the newline that comes before $off[0].
152 ## This would be much easier if unicore/Name.pl had info in
153 ## a name/code order, instead of code/name order.
155 ## The +1 after the rindex() is to skip past the newline we're finding,
156 ## or, if the rindex() fails, to put us to an offset of zero.
158 my $hexstart = rindex($txt, "\n", $off[0]) + 1;
160 ## we know where it starts, so turn into number -
161 ## the ordinal for the char.
162 $ord = hex substr($txt, $hexstart, $off[0] - $hexstart);
165 if ($^H & $bytes::hint_bits) { # "use bytes" in effect?
167 return chr $ord if $ord <= 255;
168 my $hex = sprintf "%04x", $ord;
169 if (not defined $fname) {
170 $fname = substr $txt, $off[0] + 2, $off[1] - $off[0] - 2;
172 croak "Character 0x$hex with name '$fname' is above 0xFF";
175 no warnings 'utf8'; # allow even illegal characters
176 return pack "U", $ord;
181 shift; ## ignore class name
184 carp("`use charnames' needs explicit imports list");
186 $^H |= $charnames::hint_bits;
187 $^H{charnames} = \&charnames ;
190 ## fill %h keys with our @_ args.
192 my ($promote, %h, @args) = (0);
193 while (@_ and $_ = shift) {
194 if ($_ eq ":alias") {
196 croak ":alias needs an argument in charnames";
199 ref $alias eq "HASH" or
200 croak "Only HASH reference supported as argument to :alias";
204 if ($alias =~ m{:(\w+)$}) {
205 $1 eq "full" || $1 eq "short" and
206 croak ":alias cannot use existing pragma :$1 (reversed order?)";
207 alias_file ($1) and $promote = 1;
213 if (m/^:/ and ! ($_ eq ":full" || $_ eq ":short")) {
214 warn "unsupported special '$_' in charnames";
219 @args == 0 && $promote and @args = (":full");
220 @h{@args} = (1) x @args;
222 $^H{charnames_full} = delete $h{':full'};
223 $^H{charnames_short} = delete $h{':short'};
224 $^H{charnames_scripts} = [map uc, keys %h];
227 ## If utf8? warnings are enabled, and some scripts were given,
228 ## see if at least we can find one letter of each script.
230 if (warnings::enabled('utf8') && @{$^H{charnames_scripts}}) {
231 $txt = do "unicore/Name.pl" unless $txt;
233 for my $script (@{$^H{charnames_scripts}}) {
234 if (not $txt =~ m/\t\t$script (?:CAPITAL |SMALL )?LETTER /) {
235 warnings::warn('utf8', "No such script: '$script'");
241 require Unicode::UCD; # for Unicode::UCD::_getcode()
248 carp "charnames::viacode() expects one argument";
253 my $code = Unicode::UCD::_getcode($arg);
258 $hex = sprintf "%04X", $arg;
260 carp("unexpected arg \"$arg\" to charnames::viacode()");
264 if ($code > 0x10FFFF) {
265 carp sprintf "Unicode characters only allocated up to U+10FFFF (you asked for U+%X)", $hex;
269 return $viacode{$hex} if exists $viacode{$hex};
271 $txt = do "unicore/Name.pl" unless $txt;
273 if ($txt =~ m/^$hex\t\t(.+)/m) {
274 return $viacode{$hex} = $1;
285 carp "charnames::vianame() expects one name argument";
291 return chr hex $1 if $arg =~ /^U\+([0-9a-fA-F]+)$/;
293 return $vianame{$arg} if exists $vianame{$arg};
295 $txt = do "unicore/Name.pl" unless $txt;
297 my $pos = index $txt, "\t\t$arg\n";
299 my $posLF = rindex $txt, "\n", $pos;
300 (my $code = substr $txt, $posLF + 1, 6) =~ tr/\t//d;
301 return $vianame{$arg} = hex $code;
303 # If $pos is at the 1st line, $posLF must be $[ - 1 (not found);
304 # then $posLF + 1 equals to $[ (at the beginning of $txt).
305 # Otherwise $posLF is the position of "\n";
306 # then $posLF + 1 must be the position of the next to "\n"
307 # (the beginning of the line).
308 # substr($txt, $posLF + 1, 6) may be "0000\t\t", "00A1\t\t",
309 # "10300\t", "100000", etc. So we can get the code via removing TAB.
321 charnames - define character names for C<\N{named}> string literal escapes
325 use charnames ':full';
326 print "\N{GREEK SMALL LETTER SIGMA} is called sigma.\n";
328 use charnames ':short';
329 print "\N{greek:Sigma} is an upper-case sigma.\n";
331 use charnames qw(cyrillic greek);
332 print "\N{sigma} is Greek sigma, and \N{be} is Cyrillic b.\n";
334 use charnames ":full", ":alias" => {
335 e_ACUTE => "LATIN SMALL LETTER E WITH ACUTE",
337 print "\N{e_ACUTE} is a small letter e with an acute.\n";
340 print charnames::viacode(0x1234); # prints "ETHIOPIC SYLLABLE SEE"
341 printf "%04X", charnames::vianame("GOTHIC LETTER AHSA"); # prints "10330"
345 Pragma C<use charnames> supports arguments C<:full>, C<:short>, script
346 names and customized aliases. If C<:full> is present, for expansion of
347 C<\N{CHARNAME}>, the string C<CHARNAME> is first looked up in the list of
348 standard Unicode character names. If C<:short> is present, and
349 C<CHARNAME> has the form C<SCRIPT:CNAME>, then C<CNAME> is looked up
350 as a letter in script C<SCRIPT>. If pragma C<use charnames> is used
351 with script name arguments, then for C<\N{CHARNAME}> the name
352 C<CHARNAME> is looked up as a letter in the given scripts (in the
353 specified order). Customized aliases are explained in L</CUSTOM ALIASES>.
355 For lookup of C<CHARNAME> inside a given script C<SCRIPTNAME>
356 this pragma looks for the names
358 SCRIPTNAME CAPITAL LETTER CHARNAME
359 SCRIPTNAME SMALL LETTER CHARNAME
360 SCRIPTNAME LETTER CHARNAME
362 in the table of standard Unicode names. If C<CHARNAME> is lowercase,
363 then the C<CAPITAL> variant is ignored, otherwise the C<SMALL> variant
366 Note that C<\N{...}> is compile-time, it's a special form of string
367 constant used inside double-quoted strings: in other words, you cannot
368 use variables inside the C<\N{...}>. If you want similar run-time
369 functionality, use charnames::vianame().
371 For the C0 and C1 control characters (U+0000..U+001F, U+0080..U+009F)
372 as of Unicode 3.1, there are no official Unicode names but you can use
373 instead the ISO 6429 names (LINE FEED, ESCAPE, and so forth). In
374 Unicode 3.2 (as of Perl 5.8) some naming changes take place ISO 6429
375 has been updated, see L</ALIASES>. Also note that the U+UU80, U+0081,
376 U+0084, and U+0099 do not have names even in ISO 6429.
378 Since the Unicode standard uses "U+HHHH", so can you: "\N{U+263a}"
379 is the Unicode smiley face, or "\N{WHITE SMILING FACE}".
381 =head1 CUSTOM TRANSLATORS
383 The mechanism of translation of C<\N{...}> escapes is general and not
384 hardwired into F<charnames.pm>. A module can install custom
385 translations (inside the scope which C<use>s the module) with the
386 following magic incantation:
388 use charnames (); # for $charnames::hint_bits
391 $^H |= $charnames::hint_bits;
392 $^H{charnames} = \&translator;
395 Here translator() is a subroutine which takes C<CHARNAME> as an
396 argument, and returns text to insert into the string instead of the
397 C<\N{CHARNAME}> escape. Since the text to insert should be different
398 in C<bytes> mode and out of it, the function should check the current
399 state of C<bytes>-flag as in:
401 use bytes (); # for $bytes::hint_bits
403 if ($^H & $bytes::hint_bits) {
404 return bytes_translator(@_);
407 return utf8_translator(@_);
411 =head1 CUSTOM ALIASES
413 This version of charnames supports three mechanisms of adding local
414 or customized aliases to standard Unicode naming conventions (:full)
416 =head2 Anonymous hashes
418 use charnames ":full", ":alias" => {
419 e_ACUTE => "LATIN SMALL LETTER E WITH ACUTE",
421 my $str = "\N{e_ACUTE}";
425 use charnames ":full", ":alias" => "pro";
427 will try to read "unicore/pro_alias.pl" from the @INC path. This
428 file should return a list in plain perl:
431 A_GRAVE => "LATIN CAPITAL LETTER A WITH GRAVE",
432 A_CIRCUM => "LATIN CAPITAL LETTER A WITH CIRCUMFLEX",
433 A_DIAERES => "LATIN CAPITAL LETTER A WITH DIAERESIS",
434 A_TILDE => "LATIN CAPITAL LETTER A WITH TILDE",
435 A_BREVE => "LATIN CAPITAL LETTER A WITH BREVE",
436 A_RING => "LATIN CAPITAL LETTER A WITH RING ABOVE",
437 A_MACRON => "LATIN CAPITAL LETTER A WITH MACRON",
440 =head2 Alias shortcut
442 use charnames ":alias" => ":pro";
444 works exactly the same as the alias pairs, only this time,
445 ":full" is inserted automatically as first argument (if no
446 other argument is given).
448 =head1 charnames::viacode(code)
450 Returns the full name of the character indicated by the numeric code.
453 print charnames::viacode(0x2722);
455 prints "FOUR TEARDROP-SPOKED ASTERISK".
457 Returns undef if no name is known for the code.
459 This works only for the standard names, and does not yet apply
460 to custom translators.
462 Notice that the name returned for of U+FEFF is "ZERO WIDTH NO-BREAK
463 SPACE", not "BYTE ORDER MARK".
465 =head1 charnames::vianame(name)
467 Returns the code point indicated by the name.
470 printf "%04X", charnames::vianame("FOUR TEARDROP-SPOKED ASTERISK");
474 Returns undef if the name is unknown.
476 This works only for the standard names, and does not yet apply
477 to custom translators.
481 A few aliases have been defined for convenience: instead of having
482 to use the official names
489 (yes, with parentheses) one can use
510 for ZERO WIDTH NON-JOINER and ZERO WIDTH JOINER.
512 For backward compatibility one can use the old names for
513 certain C0 and C1 controls
517 HORIZONTAL TABULATION CHARACTER TABULATION
518 VERTICAL TABULATION LINE TABULATION
519 FILE SEPARATOR INFORMATION SEPARATOR FOUR
520 GROUP SEPARATOR INFORMATION SEPARATOR THREE
521 RECORD SEPARATOR INFORMATION SEPARATOR TWO
522 UNIT SEPARATOR INFORMATION SEPARATOR ONE
523 PARTIAL LINE DOWN PARTIAL LINE FORWARD
524 PARTIAL LINE UP PARTIAL LINE BACKWARD
526 but the old names in addition to giving the character
527 will also give a warning about being deprecated.
529 =head1 ILLEGAL CHARACTERS
531 If you ask by name for a character that does not exist, a warning is
532 given and the Unicode I<replacement character> "\x{FFFD}" is returned.
534 If you ask by code for a character that does not exist, no warning is
535 given and C<undef> is returned. (Though if you ask for a code point
536 past U+10FFFF you do get a warning.)
540 Since evaluation of the translation function happens in a middle of
541 compilation (of a string literal), the translation function should not
542 do any C<eval>s or C<require>s. This restriction should be lifted in
543 a future version of Perl.