X-Git-Url: http://git.shiar.nl/sheet.git/blobdiff_plain/504c0879e49d9ffb9575ad6549c21d3990dc27d0..4e051d54eec34ed62a4fa8982c2b34493a61e016:/charset.plp

diff --git a/charset.plp b/charset.plp
index ebcbc32..73b3740 100644
--- a/charset.plp
+++ b/charset.plp
@@ -19,24 +19,23 @@ Html({
 <h1>Character encoding</h1>
 
 <:
+use POSIX qw( ceil );
 use Shiar_Sheet::FormatChar;
 my $glyphs = Shiar_Sheet::FormatChar->new;
-my @nibble = (0..9, 'A'..'F');
-my $nibsize = 1;
+my $cols = 16;  # columns
 
-use Encode qw(decode resolve_alias);
 # generate character table(s)
 # (~16x faster than decoding in loop;
 #  substr strings is twice as fast as splitting to an array)
 my %ALIAS = (
 #	default => [qw(unicode utf-8 iso-8859-1 cp437 -cp1252- --iso-8859-15- -koi8-f)],
 	default => [qw(unicode- utf-8 iso-8859-1 -cp1252- --iso-8859-15- cp437 -cp850)],
-	0 => [qw(cp437 cp863)],
-	1 => [qw(iso-8859-1 cp1252 MacRoman cp850)],
-	2 => [qw(iso-8859-2 cp1250 cp852 MacCentralEurRoman MacCroatian MacRumanian)],
-	5 => [qw(koi8-f iso-8859-5 cp1251 MacCyrillic cp855 cp866)],
-	7 => [qw(iso-8859-7 cp1253 MacGreek cp737 cp869)],
-	8 => [qw(iso-8859-8 cp1255 MacHebrew cp862)],
+	0 => [qw(cp437 -cp863)],
+	1 => [qw(iso-8859-1 -cp1252 -MacRoman -cp850)],
+	2 => [qw(iso-8859-2 -cp1250 -cp852 -MacCentralEurRoman -MacCroatian -MacRumanian)],
+	5 => [qw(koi8-f -iso-8859-5 -cp1251 -MacCyrillic -cp855 -cp866)],
+	7 => [qw(iso-8859-7 -cp1253 -MacGreek -cp737 -cp869)],
+	8 => [qw(iso-8859-8 -cp1255 -MacHebrew -cp862)],
 );
 my @request = map {
 	if (my $input = $_) {
@@ -67,37 +66,36 @@ my @request = map {
 			$row{table} = ' ' x 1024;
 			$row{set} = 'Unicode planes';
 			$row{cell} = do 'charset-ucplanes.inc.pl'
-				or printf "<p class=error>Table data could not be read: <em>%s</em>.</p>\n", $@ || $!;
-			@nibble = (map { $_.0, $_.8 } 0 .. 7);
-			$nibsize = 8;
+				or Alert('Table data could not be read', $@ || $!);
+			$cols *= 2;
 		}
-		elsif ($row{set} = resolve_alias($input)) {
+		elsif ($row{set} = Encode::resolve_alias($input)) {
 			if ($row{set} eq 'Internal') {
 				$row{table} = ' ' x ($endpoint < 255 ? 640 : 8192);
 				$row{set} = 'Unicode BMP';
 				$row{cell} = do 'charset-unicode.inc.pl'
-					or printf "<p class=error>Table data could not be read: <em>%s</em>.</p>\n", $@ || $!;
+					or Alert('Table data could not be read', $@ || $!);
 			}
 			elsif ($row{set} eq 'utf-8-strict') {
 				$row{table} = undef;
 				$row{set} = 'UTF-8';
 				$row{cell} = do 'charset-utf8.inc.pl'
-					or printf "<p class=error>Table data could not be read: <em>%s</em>.</p>\n", $@ || $!;
+					or Alert('Table data could not be read', $@ || $!);
 			}
 			else {
-				$row{table} = decode($row{set}, pack 'C*', $row{offset} .. $endpoint);
+				$row{table} = Encode::decode($row{set}, pack 'C*', $row{offset} .. $endpoint);
 			}
 		}
 		else {
-			print "<p>Encoding $input unknown</p>\n";
+			Alert("Encoding <q>$input</q> unknown");
 		}
-		\%row;
+		$row{set} ? \%row : ();
 	}
 	else {
 		();
 	}
 } map { defined $ALIAS{$_} ? @{ $ALIAS{$_} } : $_ }
-	$ENV{PATH_INFO} =~ /\w/ ? split(m{[/+\s]}, $ENV{PATH_INFO}) : 'default';
+	$Request =~ /\w/ ? split(m{[/+\s]}, $Request) : 'default';
 my $NOCHAR = chr 0xFFFD;
 
 for my $cp437 (grep {$request[$_]->{set} eq 'cp437'} 0 .. $#request) {
@@ -113,31 +111,42 @@ sub range_cell {
 	my $def = $table->{$offset} or return;
 	my ($len, $class, $name, $title) = @{$def};
 
+	my $colsize = $table->{colsize} || 1;
 	my $attr = '';
-	$len /= $nibsize;
+	$len /= $colsize;
 	$name //= $len <= 2 ? 'res' : 'reserved';
 
-	if (my $part = $offset/$nibsize % 16) {
+	if (my $part = $offset/$colsize % $cols) {
 		# continued row
-		my $cols = 16 - $part;  # remaining
-		$cols = $len if $len < $cols; #TODO: optimise
-		if ($len -= $cols) {
+		my $rest = $cols - $part;  # remaining
+		$rest = $len if $len < $rest; #TODO: optimise
+		if ($len -= $rest) {
 			# continued on new row
-			$table->{$offset + $nibsize*$cols} = [$len*$nibsize, "$class joinu", $name, $title];
-			$name = '';
+			my @next = ($len * $colsize, "$class joinu");
+			if ($len > $rest) {
+				# minority remains
+				push @next, $name, $title;
+				$title ||= $name;
+				$name = '';
+			}
+			else {
+				# minority on next row
+				push @next, '"', $title || $name;
+			}
+			$table->{$offset + $colsize*$rest} //= \@next;
 			$class .= ' joind';
 		}
-		$len = $cols;
+		$len = $rest;
 	}
-	elsif (my $rows = $len >> 4) {
+	elsif (my $rows = int($len / $cols)) {
 		# multiple full rows
-		if ($len -= $rows << 4) {
+		if ($len -= $rows * $cols) {
 			# partial row remains
-			$table->{$offset + $nibsize*$rows * 16} = [$len*$nibsize, "$class joinu", '', $title];
+			$table->{$offset + $colsize*$rows * $cols} //= [$len*$colsize, "$class joinu", '', $title];
 			$class .= ' joind';
 		}
 		$attr .= sprintf ' rowspan=%d', $rows;
-		$len = 16;
+		$len = $cols;
 	}
 
 	$attr .= sprintf ' colspan=%d', $len unless $len == 1;
@@ -148,26 +157,37 @@ sub range_cell {
 }
 
 for my $row (@request) {
+	my $colsize = $row->{cell} && $row->{cell}->{colsize} || 1;
+	my $coldigits = ceil(log($colsize * $cols) / log(16));  # uniform length of hexadecimal header
+	my $rowdiv = 16 ** $coldigits;  # divider of row headers
+
 	printf '<div class="section"><table class="glyphs%s">', !$row->{cell} && ' charmap';
 	my $title = $row->{set};
 	$title .= " <aside>($_)</aside>" for $row->{setnote} // ();
 	printf '<caption>%s</caption>', $title;
-	print '<col>' x 17;
+	print '<col>' x ($cols + 1);
 	for my $section (qw{thead}) {
 		print "<$section><tr><th>↱";
-		print '<th>', $_ for @nibble;
+		printf '<th>%0*X', $coldigits, $_ * $colsize for 0 .. $cols - 1;
 		print "\n";
 	}
 	print '<tbody>';
-	for my $msb (0 .. (length($row->{table}) || 256) - 1 >> 4) {
-		printf '<tr><th>%X', ($msb + ($row->{offset} >> 4)) * $nibsize;
-		for my $lsb (0 .. $#nibble) {
-			my $val = ( ($msb<<4) + $lsb ) * $nibsize;
+	for my $msb (0 .. ((length($row->{table}) || 256) - 1) / $cols) {
+		print '<tr><th>';
+		{
+			my $rowlabel = ($msb + int($row->{offset} / $cols)) * $cols * $colsize;
+			if (my $rowmod = $rowlabel % $rowdiv) {
+				# offset in column units
+				printf '<small>+%X</small>', $rowmod;
+			}
+			else {
+				# divided row offset
+				printf '%X', $rowlabel / $rowdiv;
+			}
+		}
+		for my $lsb (0 .. $cols - 1) {
+			my $val = ( ($msb * $cols) + $lsb ) * $colsize;
 			if ($row->{cell}) {
-				if (ref $row->{cell} eq 'CODE') {
-					print $row->{cell}->($val);
-					next;
-				}
 				print range_cell($row->{cell}, $val);
 				next;
 			}
@@ -182,7 +202,7 @@ for my $row (@request) {
 		}
 		print "\n";
 	}
-	print "</table></div>\n";
+	say '</table></div>';
 }
 
 :>