X-Git-Url: http://git.shiar.nl/sheet.git/blobdiff_plain/57f9cfa126a76ddeaac15fd55e05277afdcba98f..HEAD:/font.plp

diff --git a/font.plp b/font.plp
index b081ac2..7019573 100644
--- a/font.plp
+++ b/font.plp
@@ -1,144 +1,283 @@
 <(common.inc.plp)><:
 
+my $font = $Request;
+
 Html({
-	title => 'character support sheet',
-	version => 'v1.0',
-	keywords => [qw'
-		unicode glyph char character reference common ipa symbol sign mark table digraph
-	'],
-	stylesheet => [qw'light dark mono circus red'],
-	data => [qw'unicode-table.inc.pl unicode-char.inc.pl'],
+	title => 'font coverage '.($font ? "for $font" : 'sheet'),
+	version => '1.4',
+	keywords => [qw(
+		unicode font glyph char character support overview cover coverage
+		script block symbol sign mark reference table
+	)],
+	stylesheet => [qw( light dark mono circus red )],
+	data => [qw( data/unicode-cover.inc.pl )],
 });
 
-:>
-<h1>Character support</h1>
+if ($font) {
+	my $fontmeta = eval { Data("data/font/$font") }
+		or Abort("Unknown font $font", '404 font not found', ref $@ && $@->[1]);
 
-<p>
-Selected characters from Unicode <a href="/unicode">preset</a>
-or <a href="/charset">range</a>.
-</p>
+	my $map = eval {
+		$get{map} or return;
 
-<div>
+		my $groupinfo = Data('data/unicode-cover');
 
-<:
-use 5.010;
-use Shiar_Sheet::FormatChar;
-my $glyphs = Shiar_Sheet::FormatChar->new;
-
-my %oslist = (
-	win95   => [qw( arial ariuni verdana times )],  # microsoft
-	mac10   => [qw( )],  # apple
-	android => [qw( droidsans )],  # google
-	oss     => [qw( dvsans c2k unifont )],
-);
-my @ossel = qw( win95 oss android );
+		my ($cat, $name) = split m{/}, $get{map}, 2 or die "invalid map\n";
+		if (!$name) {
+			($cat, $name) = ('table', $cat);
+		}
 
-my $tables = do 'unicode-table.inc.pl' or die $@ || $!;
-my (%font, @fontlist);
-for my $os (@ossel) {
-	my $osfonts = $oslist{$os};
-	for my $fontid (@{$osfonts}) {
-		push @fontlist, $fontid;
-		my ($fontmeta, @fontrange) = do "ttfsupport/$fontid.inc.pl";
-		$fontmeta or next;
-		$font{$fontid} = {
-			-id   => $fontmeta->{id} || $fontid,
-			-name => $fontmeta->{name},
-			map { (chr $_ => 1) } @fontrange
-		};
+		my $row = $groupinfo->{$cat}->{$name}
+			or die "unknown character group $cat/$name\n";
+		my $query = $row->{query};
+
+		my @map;
+		for (map { split /[^\d-]/ } $query) {
+			my @range = split /-/, $_, 2;
+			m/^[0-9]+$/ or die "Invalid code point $_ in query $query\n" for @range;
+			push @map, $_ for $range[0] .. ($range[1] // $range[0]);
+		}
+		return \@map;
+	};
+	Abort($@, '404 invalid query') if $@;
+
+	require Unicode::UCD;
+
+	my $pagerows = 0x200;
+	my $pagecols = 32;
+	my $offset = eval {
+		local $_ = $get{q} || 0;
+		return $_ if /\A\d+\z/;  # numeric
+		return hex $1 if /\A (?: 0?x | u\W* ) ([0-9a-f]+) \z/ix;  # hexadecimal
+		return $_->[0]->[0] for Unicode::UCD::charblock(ucfirst) || ();  # block
+		die "Unknown offset query '$_'\n";
+	};
+	Abort($@, '400 invalid offset') if $@;
+
+	say "<h1>Font coverage</h1>";
+	say "<h2>$_</h2>" for EscapeHTML($fontmeta->{name});
+	printf("<p>Version <strong%s>%s</strong> released %s contains %d glyphs.",
+		!!$_->[2] && qq( title="revision $_->[2]"),
+		$_->[1], $_->[0],
+		scalar @{ $fontmeta->{cover} },
+	) for [
+		grep { $_ }
+		($fontmeta->{date} || '?') =~ s/T.*//r,
+		EscapeHTML($fontmeta->{version}),
+		$fontmeta->{revision},
+	];
+	for ($fontmeta->{os}) {
+		say '<br>';
+		print ucfirst join(' ',
+			"distributed",
+			(map { "by $_" } $fontmeta->{oscorp} || "various sources"),
+			(map { "with <em>$_</em>" } $_ || ()),
+			('and published as freeware "Core Web font"') x ($_ eq 'Windows 2000'),
+			(map { "under a $_ license" }
+				map { $fontmeta->{license} ? qq(<a href="$fontmeta->{license}" rel=nofollow>$_</a>) : $_ }
+				$_ && $_ ne 'Android' ? 'proprietary' : 'free',
+			),
+		);
+		print '.';
 	}
-}
+	say '</p>';
+	say "<p>$_</p>" for EscapeHTML($fontmeta->{copyright}) || ();
 
-# parse input
-
-my @chars;
-my @querydesc;
-
-my $query = $ENV{PATH_INFO} || $get{q} || 'ipa';
-for ($query) {
-	s{^/}{};
-	when (qr{^[a-z]+(?:/|\z)}) {
-		for (split / /) {
-			push @querydesc, "preset group $_";
-			my ($tablegroup, $tablename) = split m{/}, $_, 2;
-			my @tables = $tablename ? $tables->{$tablegroup}->{$tablename}
-			           : sort values %{ $tables->{$tablegroup} };
-			for (@tables) {
-				my $includerows;  # ignore rows before body row
-				for (@{$_}) {
-					$includerows ||= m/^[.]/ or next;
-					next if /^[.-]/;
-					next if $_ eq '>' or $_ eq '=';
-					push @chars, $_;
+	require Shiar_Sheet::FormatChar;
+	my $glyphs = Shiar_Sheet::FormatChar->new;
+
+	my %cover = map { ($_ => 1) } @{ $fontmeta->{cover} };  # lookup map
+
+	say <<"EOT";
+
+<style>
+	.glyphs tbody th[!colspan] { text-align: right }
+	.glyphs tbody td { font-family: "$fontmeta->{name}" }
+	.glyphs tbody td:nth-child(18) { border-left-width: 2px }
+	\@font-face {
+		font-family: "$fontmeta->{name}";
+		src: url(/data/font/$font.ttf);
+	}
+</style>
+EOT
+	say '<table class="glyphs big">';
+
+	my $offsetlink = '?' . join('&amp;',
+		(map { $_ . '=' . EncodeURI($get{$_}) } grep { defined $get{$_} } qw{ map }),
+		'q',
+	);
+	say "<caption>$_</caption>" for join(' ', grep {$_}
+		$offset > $pagerows && sprintf('<a rel="start" href="%s=%d">◄</a>', $offsetlink, 0),
+		$offset > 0 && sprintf(
+			'<a rel="prev" href="%s=%d" title="U+%2$04X">◅</a>',
+			$offsetlink, $offset - $pagerows,
+		),
+		sprintf('U+%04X', $map ? $map->[$offset] : $offset),
+		Unicode::UCD::charblock($map ? $map->[$offset] : $offset),
+		$offset + $pagerows < ($map ? @{$map} : 0x11_0000) && sprintf(
+			'<a rel="next" href="%s=%d" title="U+%2$04X">▻</a>',
+			$offsetlink, $offset + $pagerows,
+		),
+	);
+
+	for my $cp ($offset .. $offset+$pagerows-1) {
+		$cp = $map->[$cp] or next if $map;
+
+		state $colpos;
+		my $block = Unicode::UCD::charblock($cp);
+		if ($block ne (state $sameblock = $block) and $block ne 'No_Block') {
+			print '<tbody>';
+			printf '<tr><th colspan=%d>%s', $pagecols+1, $block
+				unless $block eq 'No_Block';
+			say '';
+			$sameblock = $block;
+			$colpos = 0;
+		}
+
+		if ($map) {
+			# compare previous code point and indicate gaps
+			state $lastcp = 0;
+			if ($cp != ++$lastcp) {
+				if (!$colpos or $colpos++ % $pagecols > $pagecols - 3) {
+					# nearly last column, start new row
+					$colpos = 0;
 				}
+				else {
+					# mark repositioning in existing row
+					printf '<th>%X', $cp;
+				}
+				$lastcp = $cp;
 			}
 		}
-		when ('ipa') {
-			@chars = grep { !m/[a-zA-Z]/ } @chars;
-		}
-	}
-	when (qr{[\d,;\s+-]+}) {
-		push @querydesc, "character codepoints $_";
-		for (map { split /[^\d-]/ } $_) {
-			my ($charnum, $range) = split /-/, $_;
-			push @chars, chr $_ for $charnum .. ($range // $charnum);
-		}
-	}
-	when (qr{[A-Z]}) {
-		push @querydesc, "unicode match $_";
-		eval {
-			my $match = qr/\A\p{$_}\z/;
-			push @chars, grep { m/$match/ } map { chr $_ }
-				0..0xD7FF, 0xE000..0xFDCF, 0xFDF0..0xFFFD;
-		} or die "invalid unicode match: $_\n";
-	}
-	default {
-		die "unknown parameter: $_\n";
+
+		say sprintf '<tr><th>%X', $cp if $colpos++ % $pagecols == 0;
+
+		my $info = $glyphs->glyph_info($cp);
+		my ($class, $name, $mnem, $entity, $string) = @{$info};
+		my $np = $class =~ /\bC\S\b/;  # noprint if control or invalid
+		# display literal character, with placeholder circle if non-spacing/enclosing
+		$string ||= ($class =~ /\bM[ne]\b/ && chr 9676) . chr($cp);
+		my $html = $np ? !!$cover{$cp} && sprintf("&#%d;", $cp) :
+			EscapeHTML($string);
+		say sprintf '<td class="%s" title="U+%04X%s">%s',
+			!$class ? ('l0', $cp, '', '') : (
+			$cover{$cp} ? $np ? 'l2' : 'l5' : $np ? 'Xi' : 'l1',
+			$cp, !!$name && ": $name",
+			$html
+			);
 	}
+	say '</table>';
+
+	exit;
 }
 
-@chars <= 1500 or die sprintf(
-	'too many matches (%d) for %s'."\n",
-	scalar @chars, join(', ', @querydesc),
+:>
+<h1>Font coverage</h1>
+
+<p>
+Character support of Unicode
+<a href="/charset">blocks</a> and <a href="/unicode">presets</a>.
+</p>
+
+<div>
+
+<:
+
+my $cover = Data('data/unicode-cover');
+
+my @ossel = @{ $cover->{osdefault} };
+my @fontlist = map { @{ $cover->{os}->{$_} } } @ossel;
+
+my @rows = (
+	'version/11',
+	'version/63',
+	'block/Latin-1 Supplement',
+	'block/Latin Extended-A',
+	'block/Latin Extended Additional',
+	'block/Latin Extended-B',
+	'script/Latin',
+	'script/Greek',
+	'script/Cyrillic',
+	'script/Arabic',
+	'script/Hebrew',
+	'script/Devanagari',
+	'script/Thai',
+	'script/Hangul',
+	'table/japanese',
+	'script/Han',
+	'table/ipa',
+	'table/punctuation',
+	'block/Dingbats',
+	'table/symbols',
+	'category/Sc', # currency
+	'table/math',
+	'category/Sm', # mathematical
+	'table/arrows/single',
+	'table/lines/single',
+	'table/block',
+	'table/lines',
+	'table/html',
 );
 
+if (my $group = $get{q}) {
+	my $grouprows = $cover->{$group}
+		or Abort("Unknown character category $_", 404);
+	@rows = map { "$group/$_" } sort keys %{$grouprows};
+}
+
 # output character list
 
-print '<table class=mapped>';
-say '<caption>'.EscapeHTML(join ', ', @querydesc).'</caption>';
-print '<col>' x 3;
-print "<colgroup span=$_>" for 2, map { scalar @{$oslist{$_}} } @ossel;
+print '<table class="mapped cover">';
+print '<col><col>';
+print "<colgroup span=$_>"
+	for map { scalar @{ $cover->{os}->{$_} } } @ossel;
 
 print '<thead><tr>';
-print '<td colspan=3>character';
-print '<td colspan=2>input';
-printf '<td colspan=%d>%s fonts', scalar @{ $oslist{$_} }, $_
-	for @ossel;
+print '<th colspan=2>';
+for my $os (@ossel) {
+	my $osfonts = $cover->{os}->{$os};
+	my $osfont = $cover->{fonts}->[ $osfonts->[0] ]; # first font
+	printf '<td colspan=%d>%s', scalar @{$osfonts}, $osfont->{os} || ''
+}
 
 print '<tr>';
-print '<td colspan=2>unicode';
-print '<td>name';
-print '<td><a href="/digraphs" title="digraph">di</a><td>html';
-printf '<td title="%s">%s', $font{$_}->{-name}, $font{$_}->{-id} // $_
-	for @fontlist;
+print '<th colspan=2>';
+printf('<td title="%s"><a href="%s">%s</a>', map { EscapeHTML($_) }
+	join("\n", $_->{name}, $_->{description}),
+	"/font/$_->{file}",
+	$_->{abbr},
+) for @{ $cover->{fonts} }[@fontlist];
 say '</thead>';
 
-for my $chr (@chars) {
-	my $codepoint = ord $chr;
-	my $ascii = $codepoint <= 127;
-
-	print "<tr><th>$chr\n";
-	my $info = $glyphs->glyph_info($codepoint);
-	my ($class, $name, $mnem, $html, $string) = @$info;
-	print "<td>$_" for sprintf('%X', $codepoint), EscapeHTML($name || '?');
-	printf '<td class="%s">%s', @$_ for (
-		[$ascii ? 'l0' : defined $mnem ? $class =~ /\bu-di\b/ ? 'l4' : 'l3' : 'l1', $mnem // ''],
-		[$ascii ? 'l0' : defined $html ? 'l4' : 'l1', $html // ''],
-		(map {
-			!$font{$_}->{-id} ? [l0 => '?'] :
-			$font{$_}->{$chr} ? [l4 => '✔'] : [l1 => '✘']
-		} @fontlist),
-	);
+for (@rows) {
+	my ($group, $name) = split m{/}, $_, 2;
+	my $row = $cover->{$group}->{$name};
+
+	print '<tr>';
+	$name = sprintf 'Unicode v%.1f', $name / 10 if $group eq 'version';
+	$name = sprintf '<a href="%s">%s</a>', EncodeURI("/chars/$group/$name"), EscapeHTML($name)
+		if $row->{count} and $row->{count} < 1280;
+	print '<th>', $name;
+	print '<td class=right>', $row->{count};
+	for my $count (@{ $row->{support} }[@fontlist]) {
+		if (not defined $count) {
+			print '<td class="l0">?';
+			next;
+		}
+		if (not $count) {
+			print '<td class="l1">✘';
+			next;
+		}
+		if ($count == $row->{count}) {
+			print '<td class="l5">✔';
+			next;
+		}
+
+		my $rel = $count / $row->{count};
+		my $class = $rel < .5 ? 2 : $rel < .9 ? 3 : 4;
+		printf '<td class="%s">%d', "l$class", $rel*10;
+	}
+	say '</tr>';
 }
 
 say "</table>\n";