6 no if $] >= 5.018, warnings => 'experimental::smartmatch';
8 use open OUT => ':utf8', ':std';
11 our $VERSION = '1.00';
14 # prepare presentational string for some control(lish) entries
15 "\xAD" => {string => '-'},
16 "\x{200E}" => {string => '→'},
17 "\x{200F}" => {string => '←'},
18 "\x{200B}" => {string => '␣'},
19 "\x{200C}" => {string => '|'}, # ISO-9995-7-081 lookalike (alt: ∣ ⊺ ⟙)
20 "\x{200D}" => {string => '⁀'}, # join (alt: ∤ |ͯ ⨝)
22 $info{chr $_} //= {} for 32 .. 126;
25 my $tables = do 'unicode-table.inc.pl' or die $@ || $!;
26 for (values %$tables) {
29 length $_ == 1 or next; # ignore meta values
36 } or warn "Failed reading unicode tables: $@";
39 require HTML::Entities;
41 HTML::Entities->import('%char2entity');
42 while (my ($char, $entity) = each %char2entity) {
43 $entity =~ /[a-zA-Z]/ or next; # only actual aliases
44 $info{$char}->{html} = substr($entity, 1, -1);
47 } or warn "Failed importing html entities: $@";
50 'digraphs.inc.pl' => 'u-di',
54 my $di = do $_ or die "Error reading digraphs file $_: ", $@ || $!;
55 while (my ($mnem, $cp) = each %$di) {
56 length $mnem == 2 or next; # limit to digraphs
57 my $class = $diinc{$_};
60 $class = 'u-prop' if $cp->[2] and $cp->[2] =~ m/\bXz\b/;
63 $info{$cp}->{di} //= $mnem;
64 $info{$cp}->{class}->{$class}++;
69 # read introducing unicode versions for known characters
70 my $agemap = do 'unicode-age.inc.pl' or die $@ || $!;
71 for my $chr (keys %info) {
72 my $version = $agemap->{ord $chr} or next;
73 $info{$chr}->{class}->{'u-v'.$version}++
76 } or warn "Failed including unicode version data $@";
78 for my $chr (keys %info) {
80 # attempt to get unicode character information
83 Unicode::UCD::charinfo($cp)
84 || { block => '?', category => 'Xn', name => '', script => '' }
87 $info->{$_} = $info{$chr}->{$_} for keys %{ $info{$chr} };
89 # categorise by unicode types and writing script
90 $info->{class}->{$_}++ for $info->{category};
91 $info->{class}->{$_}++ for $info->{script} || ();
93 # add custom categories for certain blocks
94 $info->{class}->{Xa}++ if $info->{block} eq 'Basic Latin';
95 $info->{class}->{Xl}++ if $info->{block} eq 'Latin-1 Supplement';
98 when ($info->{string}) {
99 # keep predefined presentational string
101 when ($info->{combining}) {
102 # overlay combining accents
103 $info->{string} = chr(9676) . $chr;
105 when (($cp & ~0b1001_1111) == 0 or $cp == 127) {
106 # control characters (first 32 chars from 0 and 128)
107 # rename to something more descriptive
108 $info->{name} = $info->{unicode10}
109 ? '<'.$info->{unicode10}.'>' # the old name was much more useful
110 : sprintf('<control U+%04X>', $cp); # at least identify by value
111 # show descriptive symbols instead of control chars themselves
112 $info->{string} = $cp < 32 ? chr($cp + 0x2400) :
113 $cp == 127 ? chr(0x2421) :
121 # output perl code of hash
122 say "# automatically generated by $0";
125 for my $cp (sort keys %info) {
126 $info{$cp}->{classstr} = join(' ', sort keys %{ $info{$cp}->{class} });
127 # convert info hashes into arrays of strings to output in display order
128 my $row = [ map { $info{$cp}->{$_} } qw/classstr name di html string/ ];
129 # strip off trailing missing values (especially string may be unknown)
130 defined $row->[-1] ? last : pop @$row for 1 .. @$row;
131 # final line (assume safe within single quotes)
132 say sprintf '"\x{%X}" => [%s],',
133 ord $cp, join(',', map { escapeq($_) } @$row);
139 return 'undef' if not defined;
148 mkcharinfo - Gather Unicode character details in Perl array
152 mkcharinfo > unicode-char.inc.pl
154 Test by printing the description of U+0041 (latin A):
156 perl -e'$u = do "unicode-char.inc.pl"; print $u->{A}->[1]'
160 Mischa POSLAWSKY <perl@shiar.org>
164 Licensed under the GNU Affero General Public License version 3.