charset: always respect colsize in start- and endpoint

[sheet.git] / charset.plp

diff --git a/charset.plp b/charset.plp
index 5358a37b47e716e6627b72ea80c706cfe8a92058..d0dd864f70345259301537ee719ce1469cef1e45 100644 (file)
--- a/charset.plp
+++ b/charset.plp
@@ -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");
@@ -268,24 +268,22 @@ 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';
                $row{cell} = do 'charset-ucplanes.inc.pl'
                        or Alert('Table data could not be read', $@ || $!);
-               $row{endpoint} = 1023 * $row{cell}->{colsize};
+               $row{endpoint} ||= 1023;
        }
        elsif (lc $input eq 'u') {
                $row{cell} = do 'charset-unicode.inc.pl'
                        or Alert('Table data could not be read', $@ || $!);
 
                $row{endpoint} ||= 8191;
-               $row{endpoint}  *= $row{cell}->{colsize};
-               $row{startpoint} = $row{cell}->{colsize} * $row{offset};
-               $row{offset} = 0;
                $row{set} = 'Unicode ' . (
-                       $row{startpoint} <  0x10000 && $row{endpoint} < 0x10000 ? 'BMP' :
-                       $row{startpoint} >= 0x10000 && $row{endpoint} < 0x20000 ? 'SMP' :
+                       $row{startpoint} <  0x1000 && $row{endpoint} < 0x1000 ? 'BMP' :
+                       $row{startpoint} >= 0x1000 && $row{endpoint} < 0x2000 ? 'SMP' :
                        'allocations'
                );
        }
@@ -296,6 +294,7 @@ 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 +362,7 @@ sub range_cell {
        $len /= $colsize;
        $name //= $len <= 2 ? 'res' : 'reserved';
 
-       if (my $part = $offset/$colsize % $cols) {
+       if (my $part = ($offset/$colsize - $info->{startpoint}) % $cols) {
                # continued row
                my $rest = $cols - $part;  # remaining
                $rest = $len if $len < $rest; #TODO: optimise
@@ -423,7 +422,7 @@ for my $row (@request) {
        my $coldigits = ceil(log($colsize * $cols) / log(16));  # uniform length of hexadecimal header
        my $rowdiv = 16 ** $coldigits;  # row divide for column digits
        $rowdiv = 1 if $rowdiv != $cols * $colsize;  # divide only if all columns are matched
-       my $offset = $row->{startpoint} || 0;
+       my $offset = $row->{startpoint} * $colsize || 0;
 
        printf '<div class="section"><table class="glyphs%s">', !$row->{cell} && ' charmap';
        my $title = $row->{set};
@@ -437,7 +436,7 @@ for my $row (@request) {
                print "\n";
        }
        print '<tbody>';
-       while ($offset < $row->{endpoint}) {
+       while ($offset < $row->{endpoint} * $colsize) {
                if ($row->{skip}->{$offset}) {
                        $offset += $cols * $colsize;
                        next;