X-Git-Url: http://git.shiar.nl/sheet.git/blobdiff_plain/b4c97cd198613dd4631b970ab4e55915237787f7..HEAD:/charset.plp diff --git a/charset.plp b/charset.plp index 8dfa91e..bbab1d6 100644 --- a/charset.plp +++ b/charset.plp @@ -1,8 +1,12 @@ <(common.inc.plp)><: +my $mode = exists $get{compare}; +my @tablist = split m{/+}, $Request || 'default'; + Html({ title => 'charset cheat sheet', - version => '1.0', + version => '1.3', + canonical => "/charset/$Request" . ($mode && '?compare'), description => [ "Reference sheet with all glyphs in common character encoding tables,", "and an overview of Unicode ranges and UTF-8 bytes.", @@ -11,99 +15,206 @@ Html({ charset codepage unicode ascii utf8 latin glyph character encoding reference common overview table '], - stylesheet => [qw'light'], - data => [qw'charset-unicode.inc.pl charset-utf8.inc.pl'], + (stylesheet => [qw'light']) x !$mode, + data => [qw( + charset.inc.pl charset-encoding.inc.pl + charset-unicode.inc.pl charset-ucplanes.inc.pl charset-utf8.inc.pl + )], }); +use List::Util qw( first pairmap pairfirst pairs ); + :> -

Character encoding

+

Character encodings

+ +

+<: +if ($tablist[0] eq 'unicode') { + say "Detailed allocation of Unicode blocks."; + print "See ", showlink("charsets", '/charset'), " to compare specific encodings"; +} +else { + +if ($tablist[0] eq 'default') { + say "Overview of Unicode ", showlink("allocation", '/charset/unicode'); + say "and common latin code pages."; + say "Compare alternate charsets:"; +} +else { + say "Charset comparison:"; +} + +print join " •\n", ( + map { + join " ·\n", pairmap { + showlink($b || ucfirst $a, '/charset'.($a && "/$a?compare"), $a eq $Request); + } @{$_} + } + [ + iso => 'ISO', + win => 'Windows', + dos => 'DOS', + mac => 'Apple', + ebcdic => 'EBCDIC', + legacy => 'legacy', + $tablist[0] eq 'default' ? () : ('' => 'common'), + ], + [ + westeur => 'West', + centeur => 'Central', + norteur => 'North European', + turkish => 0, + greek => 0, + cyrillic => 0, + hebrew => 0, + ], +); + +} +:>. +

<: use POSIX qw( ceil ); use Shiar_Sheet::FormatChar; my $glyphs = Shiar_Sheet::FormatChar->new; +my @request; -# 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)], -); -my @request = map { - if (my $input = $_) { - my %row = (offset => 0, cols => 16); - my $endpoint = 255; - if ($input =~ s/^--//) { - $row{offset} = $endpoint > 160 ? 160 : 48; - } - elsif ($input =~ s/^-//) { - $row{offset} = $endpoint > 128 ? 128 : 32; - } - if ($input =~ s/-$//) { - $endpoint = $row{offset} ? $row{offset} < 160 ? 159 : 191 : 127; - } - if ($row{offset}) { - $row{setnote} = 'over cp437' if $input eq 'cp850'; - $row{setnote} = 'over iso-8859-1' if $input =~ /^iso-8859-|^cp125/; - } - - if ($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 .. $end+15; - utf8::upgrade($row{table}); # prevent latin1 output - $row{set} = sprintf 'Unicode block U+%02Xxx', $start >> 8; - } - elsif ($input eq 'U') { - $row{table} = ' ' x 1024; - $row{set} = 'Unicode planes'; - $row{cell} = do 'charset-ucplanes.inc.pl' - or Alert('Table data could not be read', $@ || $!); - $row{cols} *= 2; - } - elsif ($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', $@ || $!); +my $charsets = Data('charset'); + +sub tabinput { + # generate character table(s) + my $input = shift or return; + my $params = $input =~ s/[+](.*)\z// ? $1 : undef; + my $charset = $charsets->{lc $input} || {}; + + if (ref $charset ne 'HASH') { + $params and Alert("Parameters ignored for $input", + "Cannot apply $params to multiple charsets.", + ); + tabinput($_) for ref $charset ? @{$charset} : $charset; + return; + } + + state $visible = {'' => 1}; # all present tables + my %row = (offset => 0, cols => 16); + $row{$_} = $charset->{$_} for qw( note table ); # copy metadata + + if (not defined $params) { + 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 (? \p{AHex}+) (?: [-] (? \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} }++; + } } else { - $row{table} = Encode::decode($row{set}, pack 'C*', $row{offset} .. $endpoint); + $row{startpoint} = hex $+{start}; } + $row{endpoint} = hex($+{end} || 0); + } + else { + Alert("Unknown option $param for charset $input"); + } + } + + if ($charset->{setup}) { + eval { $charset->{setup}->(\%row) } + or Alert("Incomplete setup of $input", $@); + } + $row{endpoint} ||= 0xFF; + + if (defined $row{table} or defined $row{cell}) { + $row{set} //= $input; + } + elsif ($row{set} = Encode::resolve_alias($charset->{set} // $input)) { + $row{offset} = delete $row{startpoint}; + if ($charset->{varchar}) { + # array of possibly multiple characters per code point + $row{table} = [ + map { Encode::decode($row{set}, pack 'C*', $_) } $row{offset} .. $row{endpoint} + ]; } else { - Alert("Encoding $input unknown"); + # ~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}); } - $row{set} ? \%row : (); + + $row{endpoint} -= $row{offset}; + $visible->{ascii}++; # assume common base + $row{set} = $input if $charset->{set}; # base override } else { - (); + Alert("Encoding $input unknown"); + return; } -} map { defined $ALIAS{$_} ? @{ $ALIAS{$_} } : $_ } - $Request =~ /\w/ ? split(m{[/+\s]}, $Request) : 'default'; -my $NOCHAR = chr 0xFFFD; -for my $cp437 (grep {$request[$_]->{set} eq 'cp437'} 0 .. $#request) { - substr($request[$cp437]->{table}, 237, 1) = pack 'U*', 0x3D5; # phi sign - substr($request[$cp437]->{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 - ); + if (my $replace = $charset->{replace}) { + while (my ($offset, $sub) = each %{$replace}) { + $offset -= $row{offset}; + + if (ref $row{table} eq 'ARRAY') { + $row{table}->[$offset] = $sub + if $offset >= 0 and $offset <= $row{endpoint}; + next; + } + + my $length = length $sub; + + if ($offset < 0) { + $offset > -$length or next; # at least one character after start + # trim leftmost part to start at offset + substr($sub, 0, -$offset) = ''; + $length += $offset; + $offset = 0; + } + + if ((my $excess = $row{endpoint} - $offset - $length + 1) < 0) { + $excess > -$length or next; + # trim rightmost part to prevent overflow + substr($sub, $excess) = ''; + $length += $excess; + } + + substr($row{table}, $offset, $length) = $sub; + } + } + + push @request, \%row; + $visible->{ $row{set} } = 1 if $row{table}; } +tabinput($_) for @tablist; + +my $NOCHAR = chr 0xFFFD; sub range_cell { my ($info, $offset) = @_; @@ -117,22 +228,23 @@ 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'; @@ -141,11 +253,24 @@ 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; } @@ -154,55 +279,83 @@ 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 $title; - return "$name"; + return "$name\n"; } for my $row (@request) { my $cols = $row->{cols}; 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 + 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} * $colsize || 0; printf '
', !$row->{cell} && ' charmap'; my $title = $row->{set}; - $title .= " " for $row->{setnote} // (); + $title .= " " for $row->{parent} || (); + $title .= " " for $row->{note} || (); printf '', $title; print '' x ($cols + 1); for my $section (qw{thead}) { - print "<$section>'; - for my $msb (0 .. ((length($row->{table}) || 256) - 1) / $cols) { + while ($offset <= $row->{endpoint} * $colsize) { + if ($row->{skip}->{$offset + $row->{offset}}) { + $offset += $cols * $colsize; + next; + } + print '
%s
↱"; + print "<$section>
", $rowdiv == 1 ? '+' : '↱'; printf '%0*X', $coldigits, $_ * $colsize for 0 .. $cols - 1; print "\n"; } + print '
'; - { - my $rowlabel = ($msb + int($row->{offset} / $cols)) * $cols * $colsize; - if (my $rowmod = $rowlabel % $rowdiv) { + if (defined $row->{skip}->{$offset + $row->{offset}}) { + print '⋮'; + } + else { + if (my $rowmod = $offset % $rowdiv) { # offset in column units printf '+%X', $rowmod; } else { # divided row offset - printf '%X', $rowlabel / $rowdiv; + printf '%X', ($offset + $row->{offset}) / $rowdiv; } } - for my $lsb (0 .. $cols - 1) { - my $val = ( ($msb * $cols) + $lsb ) * $colsize; + say ''; + + for (1 .. $cols) { if ($row->{cell}) { - print range_cell($row, $val); + print range_cell($row, $offset); next; } - my $glyph = substr $row->{table}, $val, 1; - if ($glyph eq $NOCHAR) { - print ''; - 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) = !defined $glyph || $glyph eq $NOCHAR ? () : + $glyphs->glyph_html($glyph); + + if ($mode) { + state $visible = {}; + $class = ( + $cp == ord $glyph ? 'l4' : + $row->{parent} && $glyph eq + Encode::decode($row->{parent}, pack 'C', $cp) ? 'l3' : + !defined $cell ? undef : + $visible->{$glyph} ? 'l2' : + 'l1' + ); + $visible->{$glyph}++; } - print "\n".$glyphs->glyph_cell($glyph); + printf '%s', $class, $cell) : '>'; + } + continue { + $offset += $colsize; } - print "\n"; } say '
'; } @@ -211,7 +364,13 @@ for my $row (@request) {
- +
<: if ($mode) { :> +
unicode + inherited + existing + original + unassigned +<: } else { :> control whitespace diacritic @@ -243,11 +402,11 @@ for my $row (@request) {
-
unicode 7.0 + unicode 10.0 proposed deprecated unassigned invalid -
+<: } :>