charset: avoid intermediate endpoint variable

[sheet.git] / charset.plp

diff --git a/charset.plp b/charset.plp
index f26b7bdce3a7d495bb61d69c5d6e205735fcb573..b1167ea8a1a7056646941013be51188927bc9c14 100644 (file)
--- a/charset.plp
+++ b/charset.plp
@@ -61,6 +61,7 @@ print join " •\n", (
                cyrillic => 0,
                hebrew   => 0,
        ],
+       [ uc => 'Unicode' ],
 );
 :>.
 </p>
@@ -77,6 +78,7 @@ sub tabinput {
 
        state $ALIAS = {
                default    => [qw( unicode+0-639 utf-8 iso-8859-1 iso-8859-15 cp1252 cp437 cp850 )],
+               uc         => [qw( U++ unicode+0-4095 unicode+4096-6319  unicode+6320-8191 )],
                us         => [qw( cp437 cp863 gsm0338 AdobeStandardEncoding )],
                ebcdic     => [qw( cp37 cp500 cp1047 posix-bc cp1026 cp875 )],
                iso        => [map {"iso-8859-$_"} 1 .. 11, 13 .. 16],
@@ -124,7 +126,6 @@ sub tabinput {
 
        state $visible = {};  # all present tables
        my %row = (offset => 0, cols => 16);
-       my $endpoint = 255;
        my $params = $input =~ s/[+](.*)\z// ? $1 : undef;
 
        if (not defined $params) {
@@ -219,7 +220,8 @@ sub tabinput {
                }
                $visible->{$_} //= 0 for $row{parent} || ();
        }
-       else {
+
+       if (length $params) {
                # manual option to double table width
                $row{cols} *= 2 if $params =~ s/[+]\z//;
        }
@@ -235,7 +237,7 @@ sub tabinput {
                ]);
 
                $row{offset} = $+{offset};
-               $endpoint = $+{endpoint} if $+{endpoint};
+               $row{endpoint} = $+{endpoint} if $+{endpoint};
                if (my $restart = $+{restart}) {
                        my $skip = int(($+{stop} || $row{offset}) / $row{cols});
                        for ($skip + 1 .. ($restart / $row{cols}) - 1) {
@@ -257,15 +259,22 @@ sub tabinput {
                $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{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};
+
+                       $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 && $row{endpoint} < 0x20000 ? 'SMP' :
+                               'allocations'
+                       );
                }
                elsif ($row{set} eq 'utf-8-strict') {
                        $row{set} = 'UTF-8';
@@ -274,18 +283,19 @@ sub tabinput {
                        $row{endpoint} = 255;
                }
                else {
+                       $row{endpoint} ||= 255;
                        if ($row{set} eq 'MacHebrew' or $row{set} eq 'MacThai') {
                                # array of possibly multiple characters per code point
                                $row{table} = [
-                                       map { Encode::decode($row{set}, pack 'C*', $_) } $row{offset} .. $endpoint
+                                       map { Encode::decode($row{set}, pack 'C*', $_) } $row{offset} .. $row{endpoint}
                                ];
                        }
                        else {
                                # ~16x faster than decoding in loop;
                                # substr strings is twice as fast as splitting to an array
-                               $row{table} = Encode::decode($row{set}, pack 'C*', $row{offset} .. $endpoint);
+                               $row{table} = Encode::decode($row{set}, pack 'C*', $row{offset} .. $row{endpoint});
                        }
-                       $row{endpoint} = $endpoint - $row{offset};
+                       $row{endpoint} -= $row{offset};
 
                        if ($row{set} eq 'cp437' and !$row{offset}) {
                                substr($row{table}, 237, 1) = pack 'U*', 0x3D5; # phi sign
@@ -380,7 +390,7 @@ 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 = 0;
+       my $offset = $row->{startpoint} || 0;
 
        printf '<div class="section"><table class="glyphs%s">', !$row->{cell} && ' charmap';
        my $title = $row->{set};