From: Mischa POSLAWSKY <perl@shiar.org>
Date: Sat, 8 Apr 2017 11:38:23 +0000 (+0200)
Subject: charset: replace table lengths by specific endpoints
X-Git-Tag: v1.10~213
X-Git-Url: http://git.shiar.nl/sheet.git/commitdiff_plain/42d8bbb170d20da856db097f3cec280e37578c31

charset: replace table lengths by specific endpoints

Prepare for optional overrides.
---

diff --git a/charset.plp b/charset.plp
index feb793b..c3477f1 100644
--- a/charset.plp
+++ b/charset.plp
@@ -31,7 +31,7 @@ sub tabinput {
 	my $input = shift or return;
 
 	state $ALIAS = {
-		default    => [qw( unicode+0-192 utf-8 iso-8859-1 iso-8859-15 cp1252 cp437 cp850 )],
+		default    => [qw( unicode+0-639 utf-8 iso-8859-1 iso-8859-15 cp1252 cp437 cp850 )],
 		us         => [qw( cp437 cp863 AdobeStandardEncoding gsm0338 )],
 		ebcdic     => [qw( cp37 cp500 cp875 cp1026 cp1047 posix-bc )],
 		iso        => [map {"iso-8859-$_"} 1 .. 11, 13 .. 16],
@@ -165,28 +165,28 @@ sub tabinput {
 		my $end = $2 ? hex($2) << 4 : $start + 240;
 		$row{table} = join '', map { chr } $start .. $end+15;
 		utf8::upgrade($row{table});  # prevent latin1 output
-		$row{len} = length $row{table};
+		$row{endpoint} = $end + 14 - $start;
 		$row{set} = sprintf 'Unicode block U+%02Xxx', $start >> 8;
 	}
 	elsif ($input eq 'U') {
-		$row{len} = 1024;
 		$row{set} = 'Unicode planes';
 		$row{cell} = do 'charset-ucplanes.inc.pl'
 			or Alert('Table data could not be read', $@ || $!);
 		$row{cols} *= 2;
+		$row{endpoint} = 1023 * $row{cell}->{colsize};
 	}
 	elsif ($row{set} = Encode::resolve_alias($input)) {
 		if ($row{set} eq 'Internal') {
-			$row{len} = $endpoint < 255 ? 640 : 8192;
 			$row{set} = 'Unicode BMP';
 			$row{cell} = do 'charset-unicode.inc.pl'
 				or Alert('Table data could not be read', $@ || $!);
+			$row{endpoint} = ($endpoint || 8191) * $row{cell}->{colsize};
 		}
 		elsif ($row{set} eq 'utf-8-strict') {
-			$row{len} = 256;
 			$row{set} = 'UTF-8';
 			$row{cell} = do 'charset-utf8.inc.pl'
 				or Alert('Table data could not be read', $@ || $!);
+			$row{endpoint} = 255;
 		}
 		else {
 			if ($row{set} eq 'MacHebrew') {
@@ -200,7 +200,7 @@ sub tabinput {
 				# substr strings is twice as fast as splitting to an array
 				$row{table} = Encode::decode($row{set}, pack 'C*', $row{offset} .. $endpoint);
 			}
-			$row{len} = $endpoint - $row{offset};
+			$row{endpoint} = $endpoint - $row{offset};
 
 			if ($row{set} eq 'cp437' and !$row{offset}) {
 				substr($row{table}, 237, 1) = pack 'U*', 0x3D5; # phi sign
@@ -285,7 +285,6 @@ for my $row (@request) {
 	my $rowdiv = 16 ** $coldigits;  # row divide for column digits
 	$rowdiv = 1 if $rowdiv != $cols * $colsize;  # divide only if all columns are matched
 	my $offset = 0;
-	my $endpoint = $offset + $row->{len} * $colsize;
 
 	printf '<div class="section"><table class="glyphs%s">', !$row->{cell} && ' charmap';
 	my $title = $row->{set};
@@ -299,7 +298,7 @@ for my $row (@request) {
 		print "\n";
 	}
 	print '<tbody>';
-	while ($offset < $endpoint - 1) {
+	while ($offset < $row->{endpoint}) {
 		print '<tr><th>';
 		{
 			if (my $rowmod = $offset % $rowdiv) {