digraphs: encode data include directly as json

[sheet.git] / tools / mkdigraphlist

diff --git a/tools/mkdigraphlist b/tools/mkdigraphlist
index e472ce0e0d01a41b5e0f27ef5f16f927c51c4d78..535698712117af07f6f162873bbf12eecc7af747 100755 (executable)
--- a/tools/mkdigraphlist
+++ b/tools/mkdigraphlist
@@ -4,43 +4,46 @@ use strict;
 use warnings;
 use utf8;
 
-use open OUT => ':utf8', ':std';
+use open OUT => ':encoding(utf-8)', ':std';
+use JSON ();
 
-our $VERSION = '1.04';
+our $VERSION = '1.07';
 
 # import and combine various digraph data
-my $rfc = do 'data/digraphs-rfc.inc.pl'
+push @INC, 'data';
+my $rfc = do 'digraphs-rfc.inc.pl'
        or die "error reading digraphs include: ", $@ // $!;
-my $extra = do 'data/digraphs-shiar.inc.pl'
+my $extra = do 'digraphs-shiar.inc.pl'
        or warn "could not include shiar proposals: ", $@ // $!;
-my $vim = do 'data/digraphs-vim.inc.pl'
+my $vim = do 'digraphs-vim.inc.pl'
        or warn "could not include vim extensions ", $@ // $!;
+my $vimold = do '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: ", $@ // $!;
 
-# 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(' ',
+# output json map of character info
+my %table;
+$table{$_} = 0 for (
        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 $di->{$_} }->[1],  # name
-       $vim->{$_}
-               ? $rfc->{$_} ? 'u-di' : 'u-prop'  # vim+rfc or vim-only
-               :              'u-prop ex',  # neither
-       $uninfo->{ chr $di->{$_} }->[0] =~ s/ u-di| u-prop| ex//gr,  # class
-       $uninfo->{ chr $di->{$_} }->[4] // (),  # string
-) for sort keys %{$di};
-print "}\n";
+$table{$_} = [
+       ord $di->{$_},   # original code point
+       $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 JSON->new->ascii->canonical->encode(\%table);
 
 __END__
 
@@ -50,8 +53,7 @@ mkdigraphlist - Output character list of combined digraph data
 
 =head1 SYNOPSIS
 
-    mkdigraphlist >digraphs.inc.pl
-    perl -e'$di = do "digraphs.inc.pl"; print chr $di->{DO}->[0]'
+    mkdigraphlist | jq -r '."DO"[0]' | perl -nE 'say chr' # $
 
 =head1 DESCRIPTION
 
@@ -66,9 +68,9 @@ is appended: character name, usage classes, unicode classes,
 and replacement output string.
 For example:
 
- +{
-   AE => [198, 'LATIN CAPITAL LETTER AE', 'u-di', 'Latin Lu Xl u-v11'],
-   EA => 'AE',
+  {
+   "AE" => [198, "LATIN CAPITAL LETTER AE", "u-di", "Latin Lu Xl u-v11"],
+   "EA" => "AE",
   }
 
 =head1 AUTHOR