digraphs: reuse unicode character details

[sheet.git] / tools / mkcharinfo

diff --git a/tools/mkcharinfo b/tools/mkcharinfo
index b673105339f79357b741efc19c8e671eb3b24680..b58f0c373f6215e828e4ad787683d1cfc23cd114 100755 (executable)
--- a/tools/mkcharinfo
+++ b/tools/mkcharinfo
@@ -11,9 +11,13 @@ use Data::Dump 'pp';
 our $VERSION = '1.00';
 
 my %info = (
+       # prepare presentational string for some control(lish) entries
        "\xAD"     => {string => '-'},
        "\x{200E}" => {string => '→'},
        "\x{200F}" => {string => '←'},
+       "\x{200B}" => {string => '␣'},
+       "\x{200C}" => {string => '|'}, # ISO-9995-7-081 lookalike (alt: ∣ ⊺ ⟙)
+       "\x{200D}" => {string => '⁀'}, # join (alt: ∤ |ͯ ⨝)
 );
 $info{chr $_} //= {} for 32 .. 126;
 
@@ -43,7 +47,8 @@ eval {
 } or warn "Failed importing html entities: $@";
 
 my %diinc = (
-       'digraphs.inc.pl' => 'u-di',
+       'data/digraphs-rfc.inc.pl' => 'u-di',
+       'data/digraphs-shiar.inc.pl' => 'u-prop Xz',
 );
 for (keys %diinc) {
        -e $_ or next;
@@ -51,16 +56,21 @@ for (keys %diinc) {
        while (my ($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}++;
+               $info{chr $cp}->{di} //= $mnem;
+               $info{chr $cp}->{class}->{$class}++;
        }
 }
 
+eval {
+       # read introducing unicode versions for known characters
+       my $agemap = do 'unicode-age.inc.pl' or die $@ || $!;
+       for my $chr (keys %info) {
+               my $version = $agemap->{ord $chr} or next;
+               $info{$chr}->{class}->{'u-v'.$version}++
+       }
+       1;
+} or warn "Failed including unicode version data $@";
+
 for my $chr (keys %info) {
        my $cp = ord $chr;
        # attempt to get unicode character information
@@ -70,7 +80,7 @@ for my $chr (keys %info) {
                        || { block => '?', category => 'Xn', name => '', script => '' }
        } or next;
 
-       $info->{$_} = $info{$chr}->{$_} for qw(di html class string);
+       $info->{$_} = $info{$chr}->{$_} for keys %{ $info{$chr} };
 
        # categorise by unicode types and writing script
        $info->{class}->{$_}++ for $info->{category};
@@ -95,7 +105,9 @@ for my $chr (keys %info) {
                                ? '<'.$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->{string} = $cp < 32   ? chr($cp + 0x2400) :
+                                         $cp == 127 ? chr(0x2421) :
+                                                      chr(0xFFFD);
                }
        }
 
@@ -103,6 +115,7 @@ for my $chr (keys %info) {
 }
 
 # output perl code of hash
+say "# automatically generated by $0";
 say 'use utf8;';
 say '+{';
 for my $cp (sort keys %info) {