use utf8;
use open OUT => ':utf8', ':std';
-use Data::Dumper;
our $VERSION = '1.03';
# personal addendums
my $extra = do 'data/digraphs-shiar.inc.pl'
or warn "could not include shiar proposals: ", $@ // $!;
-$di = { %{$di}, %{$extra // {}} };
-
-$di->{chr $_} = $_ for 32 .. 126;
-$di->{'\\'.$_} = delete $di->{$_} for '{', '}', '\\';
+my $vim = do 'data/digraphs-vim.inc.pl'
+ or warn "could not include vim extensions ", $@ // $!;
+$di = { %{$vim // {}}, %{$di}, %{$extra // {}} };
# optionally get unicode character information
-my %info = eval {
- require Unicode::UCD;
- map {
- $_ => Unicode::UCD::charinfo($di->{$_})
- || { block => '?', category => 'Xn', name => '', script => '' }
- } keys %{$di};
-};
-
-# add custom categories for certain blocks
-for (values %info) {
- $_->{category} .= ' Xa' if $_->{block} eq 'Basic Latin';
- $_->{category} .= ' Xl' if $_->{block} eq 'Latin-1 Supplement';
-}
-
-# mark unofficial extras as such
-$info{$_}->{category} .= ' Xz' for keys %{$extra};
-
-for (keys %{$di}) {
- $info{$_}->{string} = chr(9676) . chr($di->{$_}) if $info{$_}->{combining};
- # find control characters (first 32 chars from 0 and 128)
- next unless ($di->{$_} & ~0b1001_1111) == 0 or $di->{$_} == 127;
- # rename to something more descriptive
- $info{$_}->{name} = $info{$_}->{unicode10}
- ? '<'.$info{$_}->{unicode10}.'>' # the old name was much more useful
- : sprintf('<control U+%04X>', $di->{$_}); # at least identify by value
- # show descriptive symbols instead of control chars themselves
- $info{$_}->{string} = $di->{$_} < 32 ? chr($di->{$_} + 0x2400) : chr(0xFFFD);
-}
-# presentational string for some control(lish) entries
-$info{$_}->{string} = '-' for grep { $di->{$_} == 0x00AD } keys %{$di};
-$info{$_}->{string} = '␣' for grep { $di->{$_} == 0x200B } keys %{$di};
-$info{$_}->{string} = '|' for grep { $di->{$_} == 0x200C } keys %{$di};
-$info{$_}->{string} = '⁀' for grep { $di->{$_} == 0x200D } keys %{$di};
-$info{$_}->{string} = '→' for grep { $di->{$_} == 0x200E } keys %{$di};
-$info{$_}->{string} = '←' for grep { $di->{$_} == 0x200F } keys %{$di};
+my $uninfo = do 'unicode-char.inc.pl'
+ or warn "could not include unicode details: ", $@ // $!;
# convert info hashes into arrays of strings to output in display order
-for my $row (values %info) {
- $row = [ map { $row->{$_} } qw/name category script string/ ];
- # strip off trailing missing values (especially string may be unknown)
- defined $row->[-1] ? last : pop @$row for 1 .. @$row;
+for my $row (values %{$uninfo}) {
+ my ($class, $name, $di, $html, $string) = @{$row};
+ $row = [$name, $class];
+ push @{$row}, '', $string if defined $string;
}
# output perl code of hash
print "use utf8;\n";
print "+{\n";
printf '(map {$_=>0} qw{%s}),'."\n", join(' ',
+ grep { !defined $di->{$_} }
map { substr($_, 1, 1).substr($_, 0, 1) } sort keys %{$di}
);
printf "q{%s}=>[%s],\n", $_, join(',',
$di->{$_}, # original code point
- $info{$_} # optional additional arguments
- ? map {"'$_'"} @{ $info{$_} }
- : ()
+ (map {"'$_'"} @{ $uninfo->{ chr $di->{$_} } // [] }), # optional additional arguments
) for sort keys %{$di};
print "}\n";
=head1 DESCRIPTION
-Parses the official RFC-1345 document, searching the
-'character mnemonic table' for all digraph definitions.
-If successful, Perl code is output resulting in a hash
-with character data keyed by digraph.
-Any errors and warnings are given at STDERR.
+Combines precompiled digraph includes of rfc (1345), vim, and shiar
+and outputs a complete map including character details and usage classes.
The value can either be a scalar string containing another
digraph which can be considered identical (usually inverted),