X-Git-Url: http://git.shiar.nl/sheet.git/blobdiff_plain/c5edeb4e1b7c302cc5840cc83d7474b8f1384ab6..8e8998a5c2e51159db372eed7e4e293df83ccc91:/charset.plp

diff --git a/charset.plp b/charset.plp
index 486fece..1b672e9 100644
--- a/charset.plp
+++ b/charset.plp
@@ -12,13 +12,63 @@ Html({
 		reference common overview table
 	'],
 	stylesheet => [qw'light'],
-	data => [qw'charset-unicode.inc.pl charset-utf8.inc.pl'],
+	data => [qw(
+		charset-encoding.inc.pl
+		charset-unicode.inc.pl charset-ucplanes.inc.pl charset-utf8.inc.pl
+	)],
 });
 
-my @tablist = split /[^\w-]+/, $Request || 'default';
+my @tablist = split m{/+}, $Request || 'default';
+
+use List::Util qw( first pairmap pairfirst pairs );
 
 :>
-<h1>Character encoding</h1>
+<h1>Character encodings</h1>
+
+<p>
+<:
+if ($tablist[0] eq 'default') {
+	say "Overview of Unicode allocation and common latin code pages.";
+	say "Compare alternate charsets:";
+}
+else {
+	say "Charset comparison:";
+}
+
+sub optionlink {
+	my ($title, $href, $selected) = @_;
+	return sprintf(
+		$selected ? '<strong>%s</strong>' : '<a href="%2$s">%s</a>',
+		EscapeHTML($title), $href
+	);
+}
+
+print join " •\n", (
+	map {
+		join " ·\n", pairmap {
+			optionlink($b || ucfirst $a, '/charset'.($a && "/$a?compare"), $a eq $Request);
+		} @{$_}
+	}
+	[
+		iso      => 'ISO',
+		win      => 'Windows',
+		dos      => 'DOS',
+		mac      => 'Apple',
+		ebcdic   => 'EBCDIC',
+		$tablist[0] eq 'default' ? () : ('' => 'common'),
+	],
+	[
+		westeur  => 'West',
+		centeur  => 'Central',
+		norteur  => 'North European',
+		turkish  => 0,
+		greek    => 0,
+		cyrillic => 0,
+		hebrew   => 0,
+	],
+);
+:>.
+</p>
 
 <:
 use POSIX qw( ceil );
@@ -26,26 +76,43 @@ use Shiar_Sheet::FormatChar;
 my $glyphs = Shiar_Sheet::FormatChar->new;
 my @request;
 
+my $charsets = do 'charset-encoding.inc.pl'
+	or Alert('Encoding metadata could not be read', $@ || $!);
+
 sub tabinput {
 	# generate character table(s)
 	my $input = shift or return;
 
 	state $ALIAS = {
-		default    => [qw( unicode+0-192 utf-8 iso-8859-1 cp1252+128-159 iso-8859-15+160-191 cp437 cp850+128 )],
-		us         => [qw( cp437 cp863+128 AdobeStandardEncoding gsm0338+0-127 )],
-		ebcdic     => [qw( cp37 cp500  cp875  cp1026  cp1047  posix-bc )],
-		westeur    => [qw( iso-8859-1 iso-8859-15+160-191 cp1252+128-159 iso-8859-14+160 cp850+128 MacRoman+128 nextstep+128 hp-roman8+160 )],
-		centeur    => [qw( iso-8859-2 iso-8859-16+160 cp1250+128 cp852+128 MacCentralEurRoman+128 MacCroatian+128 MacRomanian+128 )], # MacRumanian only for DB
-		turkish    => [qw( iso-8859-3 iso-8859-9+128 cp857+128 cp1254+128 MacTurkish+128 )],
-		baltic     => [qw( iso-8859-4 iso-8859-13+160 cp775+128 cp1257+128 )],
-		nordic     => [qw( iso-8859-10 cp865+128 cp861+160-191 MacIcelandic+128 MacSami+160 )],
-		cyrillic   => [qw( koi8-f koi8-r+128-192 koi8-u+160-192 iso-8859-5+128 cp1251+128 MacCyrillic+128 cp855+128 cp866+128 )], # MacUkrainian is broken
-		arabic     => [qw( iso-8859-6 cp1006+160 cp864+128 cp1256+128 MacArabic )], # MacFarsi same as MacArabic?
-		greek      => [qw( iso-8859-7 cp1253+128 MacGreek+128 cp737+128 cp869+128 )],
-		hebrew     => [qw( iso-8859-8 cp1255+128 MacHebrew+128 cp862+128 )],
-		thai       => [qw( iso-8859-11 cp874+128-159 MacThai+128 )],
+		default    => [qw( u+0-27F utf-8+realsize iso-8859-1 iso-8859-15 cp1252 cp437 cp850 )],
+		unicode    => [qw( uu+cols=32+realsize u+0-FFF u+1000-18AF u+18B0-1FFF )],
+		us         => [qw( cp437 cp863 gsm0338 AdobeStandardEncoding )],
+		ebcdic     => [qw( cp37 cp500 cp1047 posix-bc cp1026 cp875 )],
+		iso        => [map {"iso-8859-$_"} 1 .. 11, 13 .. 16],
+		dos        => [qw( cp437 cp865 cp861 cp860 cp863 cp850 cp857 cp852 cp775
+		                   cp737 cp869 cp866 cp855 cp862 cp864 )],
+		aix        => [qw( cp1006 )],
+		win        => [qw( cp1252 cp1250 cp1254 cp1257 cp1258 cp1253 cp1251 cp1255 cp1256 cp874 )],
+		mac        => [qw( MacRoman MacRomanian MacRumanian MacCroatian MacCentralEurRoman MacTurkish MacIcelandic MacSami
+		                   MacGreek MacCyrillic MacHebrew MacArabic MacFarsi MacThai )],
+
+		# languages
+		westeur    => [qw( iso-8859-1 iso-8859-15 cp1252 iso-8859-14 cp850 hp-roman8 nextstep MacRoman )],
+		centeur    => [qw( iso-8859-2 iso-8859-16 cp1250 cp852 MacRomanian MacCroatian MacCentralEurRoman )], # MacRumanian only for DB
+		turkish    => [qw( iso-8859-9 iso-8859-3 cp1254 cp857 MacTurkish )],
+		norteur    => [qw( baltic nordic )],
+		baltic     => [qw( iso-8859-4 iso-8859-13 cp1257 cp775 )],
+		nordic     => [qw( iso-8859-10 cp865 cp861 MacIcelandic MacSami )],
+		cyrillic   => [qw( koi8-r koi8-u koi8-f iso-8859-5 cp1251 MacCyrillic cp866 cp855
+		                   +400 +2DE0 +A640-A69F +500-52F )], # MacUkrainian is broken
+		arabic     => [qw( iso-8859-6 cp1256 MacArabic cp864 cp1006 MacFarsi
+		                   +600 +8A0-8BF+8E0 +750-77F )],
+		greek      => [qw( iso-8859-7 cp1253 MacGreek cp737 cp869  +370 +1F00 )],
+		hebrew     => [qw( iso-8859-8 cp1255 MacHebrew cp862  +590 )],
+		thai       => [qw( iso-8859-11 cp874 MacThai )],
 		vietnamese => [qw( viscii cp1258 MacVietnamese )],
-		symbol     => [qw( symbol dingbats MacDingbats AdobeZdingbat AdobeSymbol )],
+		symbols    => [qw( symbol dingbats MacDingbats wingdings wingdings2 wingdings3 webdings )],
+
 		# iso-code shorthand
 		1 => 'westeur',
 		2 => 'centeur',
@@ -64,61 +131,147 @@ sub tabinput {
 		return;
 	}
 
+	state $visible = {'' => 1};  # all present tables
 	my %row = (offset => 0, cols => 16);
-	my $endpoint = 255;
 	my $params = $input =~ s/[+](.*)\z// ? $1 : undef;
 
-	if (defined $params and $params =~ m/^ (\d+) (-\d+)? /x) {
-		$row{offset} = $1;
-		$endpoint = -$2 if $2;
+	if (not defined $params) {
+		my $charset = $charsets->{$input} || {};
+		my @parents = @{ $charset->{inherit} || [] };
+
+		if (my ($parent, $part) = pairfirst { defined $visible->{$a} } @parents) {
+			$row{parent} = $parent;
+			$params = $part;
+			$params = 80 unless $visible->{$parent}
+				or ($input eq 'MacCroatian' and defined $visible->{MacRomanian});
+		}
+		elsif (defined $visible->{ascii}) {
+			$row{parent} = $parents[0];
+			$params = $parents[1] // 80;
+			$params = 80 if hex $params >= 0x80;  # ascii offset at most
+		}
+		elsif (@parents) {
+			$row{parent} = $parents[0];
+			$params = $parents[1] if hex $parents[1] == 0;  # apply ascii end
+		}
+		$visible->{$_} //= 0 for $row{parent} || ();
+	}
+
+	for my $param (split /[+]+/, $params // '') {
+		if ($param eq 'realsize') {
+			$row{realsize}++;
+		}
+		elsif ($param =~ m{ \A cols = (\d+) \z }x) {
+			$row{cols} = $1;
+		}
+		elsif ($param =~ m{ \A (?<start> \p{AHex}+) (?: [-] (?<end> \p{AHex}+) )? \z }x) {
+			if (defined $row{endpoint}) {
+				# extend earlier range
+				my $skip = int(($row{endpoint} || $row{startpoint}) / $row{cols});
+				for ($skip + 1 .. (hex($+{start}) / $row{cols}) - 1) {
+					$row{skip}->{ $_ * $row{cols} - $row{startpoint} }++;
+				}
+			}
+			else {
+				$row{startpoint} = hex $+{start};
+			}
+			$row{endpoint} = hex($+{end} || 0);
+		}
+		else {
+			Alert("Unknown option <q>$param</q> for charset $input");
+		}
 	}
-	if ($row{offset}) {
-		$row{setnote} = 'over cp437' if $input eq 'cp850';
-		$row{setnote} = 'over iso-8859-1' if $input =~ /^iso-8859-|^cp125/;
+
+	if ($input =~ m{ \A (?:wing|web)dings \d* \z }ix) {
+		eval "require Encode::\u$input";
 	}
 
-	if ($input =~ /^U([0-9a-f]+)(?:-([0-9a-f]+))?/) {
-		my $start = hex($1) << ($2 ? 4 : 8);
-		my $end = $2 ? hex($2) << 4 : $start + 240;
-		$row{table} = join '', map { chr } $start .. $end+15;
+	if ($input eq '') {
+		$row{offset} = delete $row{startpoint};
+		$row{set} = 'Unicode characters';
+		my $block = $row{offset} >> 8;
+		$row{endpoint} ||= ($block + 1 << 8) - 1;
+		$block == ($row{endpoint} >> 8) or undef $block;
+
+		$row{table} = join '', map { chr } $row{offset} .. $row{endpoint};
 		utf8::upgrade($row{table});  # prevent latin1 output
-		$row{set} = sprintf 'Unicode block U+%02Xxx', $start >> 8;
+
+		$row{endpoint} -= $row{offset};
+
+		if (defined $block) {
+			$row{set} = sprintf 'Unicode block U+%02Xxx', $block;
+			$row{offset} %= 0x100;
+		}
 	}
-	elsif ($input eq 'U') {
-		$row{table} = ' ' x 1024;
-		$row{set} = 'Unicode planes';
+	elsif (lc $input eq 'uu') {
 		$row{cell} = do 'charset-ucplanes.inc.pl'
 			or Alert('Table data could not be read', $@ || $!);
-		$row{cols} *= 2;
+		$row{endpoint} ||= 0x3FF;
+		$row{set} = 'Unicode planes';
+	}
+	elsif (lc $input eq 'u') {
+		$row{cell} = do 'charset-unicode.inc.pl'
+			or Alert('Table data could not be read', $@ || $!);
+
+		$row{endpoint} ||= 0x1FFF;
+		$row{set} = 'Unicode ' . (
+			$row{startpoint} <  0x1000 && $row{endpoint} < 0x1000 ? 'BMP' :
+			$row{startpoint} >= 0x1000 && $row{endpoint} < 0x2000 ? 'SMP' :
+			'allocations'
+		);
+	}
+	elsif ($input =~ m/^utf-*8$/i) {
+		$row{set} = 'UTF-8';
+		$row{cell} = do 'charset-utf8.inc.pl'
+			or Alert('Table data could not be read', $@ || $!);
+		$row{endpoint} = 0xFF;
 	}
 	elsif ($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 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 Alert('Table data could not be read', $@ || $!);
+		$row{offset} = delete $row{startpoint};
+		$row{endpoint} ||= 0xFF;
+		if ($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} .. $row{endpoint}
+			];
 		}
 		else {
-			$row{table} = Encode::decode($row{set}, pack 'C*', $row{offset} .. $endpoint);
-				# (~16x faster than decoding in loop;
-				#  substr strings is twice as fast as splitting to an array)
-
-			if ($row{set} eq 'cp437' and !$row{offset}) {
-				substr($row{table}, 237, 1) = pack 'U*', 0x3D5; # phi sign
-				substr($row{table}, 0, 32) = pack 'U*', map {hex} qw(
-					2007 263A 263B 2665 2666 2663 2660 2022
-					25D8 25CB 25D9 2642 2640 266A 266B 263C
-					25BA 25C4 2195 203C 00B6 00A7 25AC 21A8
-					2191 2193 2192 2190 221F 2194 25B2 25BC
-				);
+			# ~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} .. $row{endpoint});
+		}
+
+		if ($row{set} eq 'cp437') {
+			if ($row{offset} <= 0xED and $row{endpoint} >= 0xED) {
+				# replace phi glyph
+				substr($row{table}, 0xED - $row{offset}, 1) = 'ϕ';
+			}
+			if ($row{offset} < 0x20) {
+				# replace control characters by visible variants
+				my $sub = substr ' ☺☻♥♦♣♠•◘○◙♂♀♪♫☼►◄↕‼¶§▬↨↑↓→←∟↔▲▼', $row{offset};
+				substr($row{table}, 0, length $sub) = $sub;
+			}
+		}
+		elsif ($row{set} eq 'symbol') {
+			if ($row{offset} <= 0x60 and $row{endpoint} >= 0x60) {
+				# replace radical extender by closest unicode equivalent
+				substr($row{table}, 0x60 - $row{offset},  1) = '│';
+			}
+			if ($row{offset} <= 0xBD and $row{endpoint} >= 0xFF) {
+				substr($row{table}, 0xBD - $row{offset},  2) = '⏐⎯'; # arrow extenders
+				substr($row{table}, 0xD2 - $row{offset},  3) = '®©™'; # serif variants
+				substr($row{table}, 0xE0 - $row{offset},  1) = '◊'; # replace lookalike, should match AdobeSymbol
+				substr($row{table}, 0xE2 - $row{offset},  3) = '®©™'; # sans-serif variants
+				substr($row{table}, 0xE6 - $row{offset}, 10) = '⎛⎜⎝⎡⎢⎣⎧⎨⎩⎪';
+				substr($row{table}, 0xF0 - $row{offset},  1) = '€';
+				substr($row{table}, 0xF4 - $row{offset}, 11) = '⎮⌡⎞⎟⎠⎤⎥⎦⎫⎬⎭';
 			}
 		}
+
+		$row{endpoint} -= $row{offset};
+
+		$visible->{ascii} =  # assume common base
+		$visible->{ $row{set} } = 1;
 	}
 	else {
 		Alert("Encoding <q>$input</q> unknown");
@@ -142,22 +295,23 @@ 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 < $rest; #TODO: optimise
 		if ($len -= $rest) {
 			# continued on new row
 			my @next = ($len * $colsize, "$class joinu");
+			my $separate = $cols - $len > $rest;  # columns not on next row
 			if ($len > $rest) {
 				# minority remains
 				push @next, $name, $title;
 				$title ||= $name;
-				$name = '';
+				$name = $separate && '…';
 			}
 			else {
 				# minority on next row
-				push @next, '"', $title || $name;
+				push @next, $separate && '"', $title || $name;
 			}
 			$table->{$offset + $colsize*$rest} //= \@next;
 			$class .= ' joind';
@@ -166,11 +320,24 @@ sub range_cell {
 	}
 	elsif (my $rows = int($len / $cols)) {
 		# multiple full rows
+		my $rowsize = $colsize * $cols;
 		if ($len -= $rows * $cols) {
 			# partial row remains
-			$table->{$offset + $colsize*$rows * $cols} //= [$len*$colsize, "$class joinu", '', $title];
+			$table->{$offset + $rowsize * $rows} //= [$len*$colsize, "$class joinu", '', $title];
 			$class .= ' joind';
 		}
+
+		unless ($info->{realsize}) {
+			# coalesce multiple rows
+			while ($rows > 3) {
+				$info->{skip}->{$offset += $rowsize}++;
+				$rows--;
+			}
+			if ($rows > 2) {
+				$info->{skip}->{$offset += $rowsize} = 0;
+			}
+		}
+
 		$attr .= sprintf ' rowspan=%d', $rows;
 		$len = $cols;
 	}
@@ -179,7 +346,7 @@ sub range_cell {
 	$attr .= $1 if $class and $class =~ s/( \w+="[^"]*")//;
 	$attr .= sprintf ' class="%s"', $class if $class;
 	$attr .= sprintf ' title="%s"', EscapeHTML($title) if $title;
-	return "<td$attr>$name";
+	return "<td$attr>$name\n";
 }
 
 for my $row (@request) {
@@ -188,12 +355,12 @@ 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 $endpoint = $offset + (length($row->{table}) || 256) * $colsize;
+	my $offset = $row->{startpoint} * $colsize || 0;
 
 	printf '<div class="section"><table class="glyphs%s">', !$row->{cell} && ' charmap';
 	my $title = $row->{set};
-	$title .= " <aside>($_)</aside>" for $row->{setnote} // ();
+	$title .= " <aside>(over $_)</aside>"
+		for $row->{parent} || ();
 	printf '<caption>%s</caption>', $title;
 	print '<col>' x ($cols + 1);
 	for my $section (qw{thead}) {
@@ -201,10 +368,19 @@ for my $row (@request) {
 		printf '<th>%0*X', $coldigits, $_ * $colsize for 0 .. $cols - 1;
 		print "\n";
 	}
+
 	print '<tbody>';
-	while ($offset < $endpoint - 1) {
+	while ($offset <= $row->{endpoint} * $colsize) {
+		if ($row->{skip}->{$offset}) {
+			$offset += $cols * $colsize;
+			next;
+		}
+
 		print '<tr><th>';
-		{
+		if (defined $row->{skip}->{$offset}) {
+			print '⋮';
+		}
+		else {
 			if (my $rowmod = $offset % $rowdiv) {
 				# offset in column units
 				printf '<small>+%X</small>', $rowmod;
@@ -214,24 +390,39 @@ for my $row (@request) {
 				printf '%X', ($offset + $row->{offset}) / $rowdiv;
 			}
 		}
+		say '';
+
 		for (1 .. $cols) {
 			if ($row->{cell}) {
 				print range_cell($row, $offset);
 				next;
 			}
 
-			my $glyph = substr $row->{table}, $offset, 1;
-			if ($glyph eq $NOCHAR) {
-				print '<td>';
-				next;
+			my $cp = $offset + $row->{offset};
+			my $glyph = ref $row->{table} eq 'ARRAY' ? $row->{table}->[$offset] :
+				substr $row->{table}, $offset, 1;
+			my ($cell, $name, $class) = $glyph eq $NOCHAR ? () :
+				$glyphs->glyph_html($glyph);
+
+			if (exists $get{compare}) {
+				state $visible = {};
+				$class = (
+					$cp == ord $glyph ? 'l4' :
+					$row->{parent} && $glyph eq
+						Encode::decode($row->{parent}, pack 'C', $cp) ? 'l3' :
+					!$class ? undef :
+					$visible->{$glyph} ? 'l2' :
+					'l1'
+				);
+				$visible->{$glyph}++;
 			}
 
-			print "\n".$glyphs->glyph_cell($glyph);
+			say sprintf $class ? '<td title="%s" class="X %s">%s' : '<td title="%s">',
+				$name, $class, $cell;
 		}
 		continue {
 			$offset += $colsize;
 		}
-		print "\n";
 	}
 	say '</table></div>';
 }
@@ -240,7 +431,13 @@ for my $row (@request) {
 <hr>
 
 <div class="legend">
-	<table class="glyphs"><tr>
+	<table class="glyphs"><tr><: if (exists $get{compare}) { :>
+	<td class="X l4">unicode
+	<td class="X l3">inherited
+	<td class="X l2">existing
+	<td class="X l1">original
+	<td class="">unassigned
+<: } else { :>
 	<td class="X Cc">control
 	<td class="X Zs"><span>whitespace</span>
 	<td class="X Mn">diacritic<table class="glyphs"><tr>
@@ -277,6 +474,6 @@ for my $row (@request) {
 	<td class="X Xd">deprecated
 	<td class="">unassigned
 	<td class="X Xi">invalid
-	</table>
+<: } :>	</table>
 </div>