browser: truncate maximum usage score to 99

[sheet.git] / tools / mkdigraphlist

diff --git a/tools/mkdigraphlist b/tools/mkdigraphlist
index 4af35901004c4fb1d57d681d70bc7dc1ee26ce3f..a0c5ab23488f8e331a7d876a20a57c4ee2202c1d 100755 (executable)
--- a/tools/mkdigraphlist
+++ b/tools/mkdigraphlist
@@ -5,41 +5,45 @@ use warnings;
 use utf8;
 
 use open OUT => ':utf8', ':std';
-use Data::Dumper;
 
-our $VERSION = '1.03';
+our $VERSION = '1.06';
 
-# create a hash of desired input
-my $di = do 'data/digraphs-rfc.inc.pl'
+# import and combine various digraph data
+my $rfc = do 'data/digraphs-rfc.inc.pl'
        or die "error reading digraphs include: ", $@ // $!;
-
-# personal addendums
 my $extra = do 'data/digraphs-shiar.inc.pl'
        or warn "could not include shiar proposals: ", $@ // $!;
-$di = { %{$di}, %{$extra // {}} };
+my $vim = do 'data/digraphs-vim.inc.pl'
+       or warn "could not include vim extensions ", $@ // $!;
+my $vimold = do 'data/digraphs-vim-74.inc.pl'
+       or warn "could not include vim compatibility ", $@ // $!;
+my $di = { %{$vim // {}}, %{$rfc}, %{$extra // {}} };
 
 # optionally get unicode character information
 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 $row (values %{$uninfo}) {
-       my ($class, $name, $di, $html, $string) = @{$row};
-       $row = [$name, $class];
-       push @{$row}, '', $string if defined $string;
-}
-
 # output perl code of hash
 # (assume no backslashes or curlies, so we can just q{} w/o escaping)
 print "# automatically generated by $0\n";
 print "use utf8;\n";
 print "+{\n";
 printf '(map {$_=>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
-       (map {"'$_'"} @{ $uninfo->{ chr $di->{$_} } // [] }),  # optional additional arguments
+printf "q{%s}=>[%s],\n", s/(?=[\\}])/\\/gr, join(',',
+       ord $di->{$_},   # original code point
+       map {"'$_'"}
+       $uninfo->{ $di->{$_} }->[1] // '',  # name
+       join(' ',
+               $rfc->{$_}
+                       ? $vim->{$_} ? 'l4' : 'l1'  # vim+rfc or rfc only
+                       : $vim->{$_} ? 'l3' : 'l2', # vim only or neither
+               $vimold && $vim->{$_} && !$vimold->{$_} ? 'ex' : (), # new vim feature
+       ),
+       ($uninfo->{ $di->{$_} }->[0] // '') =~ s/ u-di| u-prop| ex//gr,  # class
+       $uninfo->{ $di->{$_} }->[4] // (),  # string
 ) for sort keys %{$di};
 print "}\n";
 
@@ -56,21 +60,19 @@ mkdigraphlist - Output character list of combined digraph data
 
 =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),
 or an array ref containing at least the resulting character's
 Unicode code point value.  If available, the following UCD data
-is appended: character name, category, script, and output string.
+is appended: character name, usage classes, unicode classes,
+and replacement output string.
 For example:
 
  +{
-   AE => [198, 'LATIN CAPITAL LETTER AE', 'Lu Xl', 'Latin'],
+   AE => [198, 'LATIN CAPITAL LETTER AE', 'u-di', 'Latin Lu Xl u-v11'],
    EA => 'AE',
   }