X-Git-Url: http://git.shiar.nl/sheet.git/blobdiff_plain/0d71f602a3268f8059dc48042280fb25c1abb4ba..b5b3537710ed9f73e1c867e0cc27d50439eaf4cd:/tools/convert-unicode.pl

diff --git a/tools/convert-unicode.pl b/tools/convert-unicode.pl
deleted file mode 100755
index 51e27cd..0000000
--- a/tools/convert-unicode.pl
+++ /dev/null
@@ -1,145 +0,0 @@
-#!/usr/bin/env perl
-use 5.010;
-use strict;
-use warnings;
-use utf8;
-
-use open OUT => ':utf8', ':std';
-use Data::Dump 'pp';
-
-our $VERSION = '1.00';
-
-my %info = (
-	"\xAD"     => {string => '-'},
-	"\x{200E}" => {string => '→'},
-	"\x{200F}" => {string => '←'},
-);
-$info{chr $_} //= {} for 32 .. 126;
-
-eval {
-	my $tables = do 'unicode-table.inc.pl' or die $@ || $!;
-	for (values %$tables) {
-		for (values %$_) {
-			for (@$_) {
-				length $_ == 1 or next;  # ignore meta values
-				s/\\//;  # unescape
-				$info{$_} //= {};
-			}
-		}
-	}
-	1;
-} or warn "Failed reading unicode tables: $@";
-
-eval {
-	require HTML::Entities;
-	while (my ($char, $entity) = each %HTML::Entities::char2entity) {
-		$entity =~ /[a-zA-Z]/ or next;  # only actual aliases
-		$info{$char}->{html} = substr($entity, 1, -1);
-	}
-	1;
-} or warn "Failed importing html entities: $@";
-
-my %diinc = (
-	'digraphs.inc.pl' => 'u-di',
-);
-for (keys %diinc) {
-	-e $_ or next;
-	my $di = do $_ or die "Error reading digraphs file $_: ", $@ || $!;
-	while (my ($mnem, $cp) = each %$di) {
-		length $mnem == 2 or next;  # limit to digraphs
-		my $class = $diinc{$_};
-		if (ref $cp) {
-			# old style array
-			$class = 'u-prop' if $cp->[2] and $cp->[2] =~ m/\bXz\b/;
-			$cp = chr $cp->[0];
-		}
-		$info{$cp}->{di} //= $mnem;
-		$info{$cp}->{class}->{$class}++;
-	}
-}
-
-for my $chr (keys %info) {
-	my $cp = ord $chr;
-	# attempt to get unicode character information
-	my $info = eval {
-		require Unicode::UCD;
-		Unicode::UCD::charinfo($cp)
-			|| { block => '?', category => 'Xn', name => '', script => '' }
-	} or next;
-
-	$info->{$_} = $info{$chr}->{$_} for qw(di html class string);
-
-	# categorise by unicode types and writing script
-	$info->{class}->{$_}++ for $info->{category};
-	$info->{class}->{$_}++ for $info->{script} || ();
-
-	# add custom categories for certain blocks
-	$info->{class}->{Xa}++ if $info->{block} eq 'Basic Latin';
-	$info->{class}->{Xl}++ if $info->{block} eq 'Latin-1 Supplement';
-
-	given ($cp) {
-		when ($info->{string}) {
-			# keep predefined presentational string
-		}
-		when ($info->{combining}) {
-			# overlay combining accents
-			$info->{string} = chr(9676) . $chr;
-		}
-		when (($cp & ~0b1001_1111) == 0 or $cp == 127) {
-			# control characters (first 32 chars from 0 and 128)
-			# rename to something more descriptive
-			$info->{name} = $info->{unicode10}
-				? '<'.$info->{unicode10}.'>'  # the old name was much more useful
-				: sprintf('<control U+%04X>', $cp);  # at least identify by value
-			# show descriptive symbols instead of control chars themselves
-			$info->{string} = $cp < 32 ? chr($cp + 0x2400) : chr(0xFFFD);
-		}
-	}
-
-	$info{$chr} = $info;
-}
-
-# output perl code of hash
-say 'use utf8;';
-say '+{';
-for my $cp (sort keys %info) {
-	$info{$cp}->{classstr} = join(' ', sort keys %{ $info{$cp}->{class} });
-	# convert info hashes into arrays of strings to output in display order
-	my $row = [ map { $info{$cp}->{$_} } qw/classstr name di html string/ ];
-	# strip off trailing missing values (especially string may be unknown)
-	defined $row->[-1] ? last : pop @$row for 1 .. @$row;
-	# final line (assume safe within single quotes)
-	say sprintf '"\x{%X}" => [%s],',
-		ord $cp, join(',', map { escapeq($_) } @$row);
-}
-say '}';
-
-sub escapeq {
-	my $_ = shift;
-	return 'undef' if not defined;
-	s/(['\\])/\\$1/g;
-	return "'$_'";
-}
-
-__END__
-
-=head1 NAME
-
-convert-unicode.pl - Gather Unicode character details in Perl array
-
-=head1 SYNOPSIS
-
-    convert-unicode > unicode-char.inc.pl
-
-Test by printing the description of U+0041 (latin A):
-
-    perl -e'$u = do "unicode-char.inc.pl"; print $u->{A}->[1]'
-
-=head1 AUTHOR
-
-Mischa POSLAWSKY <perl@shiar.org>
-
-=head1 LICENSE
-
-Licensed under the GNU Affero General Public License version 3.
-