charset: offset parameter sets more common startpoint
authorMischa POSLAWSKY <perl@shiar.org>
Fri, 21 Apr 2017 20:19:21 +0000 (22:19 +0200)
committerMischa POSLAWSKY <perl@shiar.org>
Thu, 25 May 2017 20:10:23 +0000 (22:10 +0200)
Only tables are optimised to offset instead of skip.

charset.plp

index 5358a37..82f37f9 100644 (file)
@@ -238,18 +238,18 @@ sub tabinput {
                elsif ($param =~ m{ \A cols = (\d+) \z }x) {
                        $row{cols} = $1;
                }
-               elsif ($param =~ m{ \A (?<offset> \d+) (?: [-] (?<endpoint> \d+) )? \z }x) {
+               elsif ($param =~ m{ \A (?<start> \d+) (?: [-] (?<end> \d+) )? \z }x) {
                        if (defined $row{endpoint}) {
                                # extend earlier range
-                               my $skip = int(($row{endpoint} || $row{offset}) / $row{cols});
-                               for ($skip + 1 .. ($+{offset} / $row{cols}) - 1) {
-                                       $row{skip}->{ $_ * $row{cols} - $row{offset} }++;
+                               my $skip = int(($row{endpoint} || $row{startpoint}) / $row{cols});
+                               for ($skip + 1 .. ($+{start} / $row{cols}) - 1) {
+                                       $row{skip}->{ $_ * $row{cols} - $row{startpoint} }++;
                                }
                        }
                        else {
-                               $row{offset} = $+{offset};
+                               $row{startpoint} = $+{start};
                        }
-                       $row{endpoint} = $+{endpoint} || 0;
+                       $row{endpoint} = $+{end} || 0;
                }
                else {
                        Alert("Unknown option <q>$param</q> for charset $input");
@@ -268,6 +268,7 @@ sub tabinput {
                $row{endpoint} = $end - $start;
                $row{set} = sprintf 'Unicode block U+%02Xxx', $start >> 8;
                $row{offset} = $start % 256;
+               $row{startpoint} = 0;
        }
        elsif (lc $input eq 'uu') {
                $row{set} = 'Unicode planes';
@@ -281,8 +282,7 @@ sub tabinput {
 
                $row{endpoint} ||= 8191;
                $row{endpoint}  *= $row{cell}->{colsize};
-               $row{startpoint} = $row{cell}->{colsize} * $row{offset};
-               $row{offset} = 0;
+               $row{startpoint} *= $row{cell}->{colsize};
                $row{set} = 'Unicode ' . (
                        $row{startpoint} <  0x10000 && $row{endpoint} < 0x10000 ? 'BMP' :
                        $row{startpoint} >= 0x10000 && $row{endpoint} < 0x20000 ? 'SMP' :
@@ -296,6 +296,7 @@ sub tabinput {
                $row{endpoint} = 255;
        }
        elsif ($row{set} = Encode::resolve_alias($input)) {
+               $row{offset} = delete $row{startpoint};
                $row{endpoint} ||= 255;
                if ($row{set} eq 'MacHebrew' or $row{set} eq 'MacThai') {
                        # array of possibly multiple characters per code point
@@ -363,7 +364,7 @@ sub range_cell {
        $len /= $colsize;
        $name //= $len <= 2 ? 'res' : 'reserved';
 
-       if (my $part = $offset/$colsize % $cols) {
+       if (my $part = ($offset - $info->{startpoint})/$colsize % $cols) {
                # continued row
                my $rest = $cols - $part;  # remaining
                $rest = $len if $len < $rest; #TODO: optimise