charset cheat sheet

Character encodings

<: -my $diinfo = do 'digraphs.inc.pl'; -my %di = map { $diinfo->{$_}->[0] => $_ } grep { ref $diinfo->{$_} } - keys %$diinfo; - -use Encode qw(decode); -# generate character table(s) -# (~16x faster than decoding in loop; -# substr strings is twice as fast as splitting to an array) -my @request = ('iso-8859-1', 'cp437'); -my @tables = map { decode($_, pack 'C*', 0..255) } @request; -my $NOCHAR = chr 0xFFFD; +if ($tablist[0] eq 'default') { + say "Overview of Unicode allocation and common latin code pages."; + say "Compare alternate charsets:"; +} +else { + say "Charset comparison:"; +} -for my $cp437 (grep {$request[$_] eq 'cp437'} 0 .. $#request) { - substr($tables[$cp437], 237, 1) = pack 'U*', 0x3D5; # phi sign - substr($tables[$cp437], 0, 32) = pack 'U*', map {hex} qw( - 2007 263A 263B 2665 2666 2663 2660 2022 25D8 25CB 25D9 2642 2640 266A 266B 263C - 25BA 25C4 2195 203C 00B6 00A7 25AC 21A8 2191 2193 2192 2190 221F 2194 25B2 25BC +sub optionlink { + my ($title, $href, $selected) = @_; + return sprintf( + $selected ? '%s' : '%s', + EscapeHTML($title), $href ); } -sub quote { - local $_ = shift; - s/"/"/g; - s//>/g; - return $_; -} +print join " â¢\n", ( + map { + join " Â·\n", pairmap { + optionlink($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, + ], +); +:>. +

\n"; +<: +use POSIX qw( ceil ); +use Shiar_Sheet::FormatChar; +my $glyphs = Shiar_Sheet::FormatChar->new; +my @request; -my @nibble = (0..9, 'A'..'F'); -for my $tablenum (0 .. $#tables) { - print '

'; - printf '', $request[$tablenum]; - print ''; - for my $section (qw{thead}) { - print "<$section>'; - for my $msb (0 .. $#nibble) { - print '

%s
â±"; - print '	', $_ for @nibble; - print "\n"; +sub tabinput { + # generate character table(s) + my $input = shift or return; + + state $ALIAS = { + default => [qw( u+0-639 utf-8+realsize iso-8859-1 iso-8859-15 cp1252 cp437 cp850 )], + unicode => [qw( uu+cols=32+realsize u+0-4095 u+4096-6319 u+6320-8191 )], + us => [qw( cp437 cp863 gsm0338 AdobeStandardEncoding )], + ebcdic => [qw( cp37 cp500 cp1047 posix-bc cp1026 cp875 )], + iso => [map {"iso-8859-$_"} 1 .. 11, 13 .. 16], + 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 + U4 U2DE-2DF UA64-A69 U50-52 )], # MacUkrainian is broken + arabic => [qw( iso-8859-6 cp1256 MacArabic cp864 cp1006 MacFarsi + U6 U8A-8F+0-31+64 U75-77 )], + greek => [qw( iso-8859-7 cp1253 MacGreek cp737 cp869 U37-3F U1F )], + hebrew => [qw( iso-8859-8 cp1255 MacHebrew cp862 U59-5F )], + thai => [qw( iso-8859-11 cp874 MacThai )], + vietnamese => [qw( viscii cp1258 MacVietnamese )], + symbol => [qw( symbol dingbats MacDingbats AdobeZdingbat AdobeSymbol )], + + # iso-code shorthand + 1 => 'westeur', + 2 => 'centeur', + 3 => 'turkish', + 4 => 'baltic', + 5 => 'cyrillic', + 6 => 'arabic', + 7 => 'greek', + 8 => 'hebrew', + 9 => 'turkish', + 10 => 'nordic', + 11 => 'thai', + }; + if (my $follow = $ALIAS->{$input}) { + tabinput($_) for ref $follow ? @{$follow} : $follow; + return; } - print '
', $nibble[$msb]; - for my $lsb (0 .. $#nibble) { - my $glyph = substr $tables[$tablenum], ($msb<<4) + $lsb, 1; - if ($glyph eq $NOCHAR) { - print '	'; - next; - } - my $info = [ord $glyph]; - if (defined (my $mnem = $di{ord $glyph})) { - $info = $diinfo->{$mnem}; - } - my ($codepoint, $name, $prop, $script, $string) = @$info; - $glyph = quote($string \|\| $glyph); - my $desc = sprintf 'U+%04X%s', $codepoint, $name && " ($name)"; - my @class = ('X', grep {$_} $prop, $script); + state $visible = {}; # all present tables + my %row = (offset => 0, cols => 16); + my $params = $input =~ s/[+](.*)\z// ? $1 : undef; + + if (not defined $params) { + state $INHERIT = { + 'cp437' => ['cp850' => 0, 'ascii' => '0-31+128'], # ascii range overridden later + 'gsm0338' => ['ascii' => '0-127'], + + 'iso-8859-2' => ['iso-8859-1' => '160'], + 'iso-8859-3' => ['iso-8859-1' => '160'], #TODO: also apply to iso-8859-9 + 'iso-8859-4' => ['iso-8859-2' => '160'], + 'iso-8859-5' => ['iso-8859-1' => '160'], + 'iso-8859-6' => ['cp1256' => '128', 'iso-8859-1' => '160'], + 'iso-8859-7' => ['iso-8859-1' => '160'], + 'iso-8859-8' => ['iso-8859-1' => '160'], + 'iso-8859-9' => ['iso-8859-1' => '208-223+240'], + 'iso-8859-10' => ['iso-8859-4' => '160'], + 'iso-8859-11' => ['iso-8859-1' => '160'], + 'iso-8859-13' => ['iso-8859-4' => '160'], + 'iso-8859-14' => ['iso-8859-1' => '160'], + 'iso-8859-15' => ['iso-8859-1' => '160-191'], + 'iso-8859-16' => ['iso-8859-2' => '160'], + 'hp-roman8' => ['iso-8859-1' => '160'], + + 'cp1252' => ['iso-8859-1' => '128-159'], + 'cp1250' => ['iso-8859-2' => '128-191', 'cp1252' => '128'], + 'cp1254' => ['iso-8859-9' => '128-159', 'cp1252' => '128-159+208'], + 'cp874' => ['iso-8859-11' => '128-159', 'cp1252' => '128'], # windows-874 actually cp1162 + 'cp1257' => ['iso-8859-13' => '128-159+255', 'cp1252' => '128'], + 'cp1251' => ['cp1252' => '128'], + 'cp1253' => ['cp1252' => '128'], + 'cp1255' => ['iso-8859-8' => '128-223', 'cp1252' => '128'], + 'cp1256' => ['cp1252' => '128'], + 'cp1258' => ['cp1252' => '128-159+192'], + + 'cp850' => ['cp437' => '144'], + 'cp860' => ['cp437' => '128-175'], + 'cp861' => ['cp865' => '128-175'], + 'cp863' => ['cp437' => '128-175'], + 'cp865' => ['cp437' => '144-175'], + 'cp852' => ['cp850' => '128', 'cp437' => '128'], + 'cp857' => ['cp850' => '128-175+208-239', 'cp437' => '128'], + 'cp775' => ['cp850' => '128'], # partial cp437 + 'cp866' => ['cp437' => '128-175+224'], + 'cp855' => ['cp437' => '128'], + 'cp1006' => ['iso-8859-6' => '160', 'cp437' => '128'], + 'cp737' => ['cp437' => '128-175+224'], + 'cp869' => ['cp437' => '128'], + 'cp862' => ['cp437' => '128-159'], + 'cp864' => ['MacArabic' => '128', 'iso-8859-6' => '128', 'cp437' => '128'], #TODO: compare form variants + + 'koi8-u' => ['koi8-r' => '144-191'], + 'koi8-f' => ['koi8-u' => '144-191'], - $glyph = "$glyph" if $prop eq 'Zs'; + 'MacRomanian' => ['MacRoman' => '160-191+208-223'], + 'MacRumanian' => ['MacRomanian' => '160-191+208-223', 'MacRoman' => '160-191+208-223'], + 'MacCroatian' => ['MacRoman' => '160'], + 'MacCentralEurRoman' => ['MacRoman' => '128'], + 'MacIcelandic'=> ['MacRoman' => '160-175+208-239'], + 'MacTurkish' => ['MacRoman' => '208-223'], # F5 is unassigned + 'MacSami' => ['MacIcelandic' => '144-191+208-223+240', 'MacRoman' => '144'], + 'MacGreek' => ['MacRoman' => '128'], + 'MacCyrillic' => ['MacRoman' => '128'], + 'MacHebrew' => ['iso-8859-8' => '128', 'MacRoman' => '128-143+160'], # partial ascii + 'MacArabic' => ['iso-8859-6' => '128', 'cp864' => '128', 'MacRoman' => '128'], #TODO: multiple parents + 'MacFarsi' => ['MacArabic' => '176-191', 'MacRoman' => '128'], - printf "\n".'	%s', - join(' ', @class), quote($desc), $glyph; + 'cp37' => ['posix-bc' => '0'], + 'posix-bc' => ['cp1047' => '64'], + 'cp500' => ['cp37' => '64-95+176-191'], + 'cp1047' => ['cp37' => '16-47+80-95+160-191'], + 'cp1026' => ['cp37' => '64'], + 'cp875' => ['cp37' => '48'], + + }; + + my @parents = @{ $INHERIT->{$input} \|\| [] }; + + if (my ($parent, $part) = pairfirst { defined $visible->{$a} } @parents) { + $row{parent} = $parent; + $params = $part; + $params = 128 unless $visible->{$parent} + or ($input eq 'MacCroatian' and defined $visible->{MacRomanian}); } - print "\n"; + elsif (defined $visible->{ascii}) { + $row{parent} = $parents[0]; + $params = $parents[1] // 128; + $params = 128 if $params >= 128; # ascii offset at most + } + elsif (@parents) { + $row{parent} = $parents[0]; + $params = $parents[1] if $parents[1] == 0; # apply ascii end + } + $visible->{$_} //= 0 for $row{parent} \|\| (); } - print "

\n"; -} -{ - print '

'; - for my $section (qw{thead}) { - print "<$section>'; - print ''; - for my $msb (0 .. $#nibble) { - print '

UTF-8

â±"; - print '

', $_ for @nibble; - print "\n"; - } - print '

', $nibble[$msb]; - for my $lsb (0 .. $#nibble) { - my $value = ($msb<<4) + $lsb; - if ($value <= 0x7F) { - print '

Single byte ASCII' - if $value == 0; - } - elsif ($value <= 0xBF) { - print '

Multi-byte continuation' - if $value == 0x80; + 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 (? \d+) (?: [-] (? \d+) )? \z }x) { + if (defined $row{endpoint}) { + # extend earlier range + my $skip = int(($row{endpoint} || $row{offset}) / $row{cols}); + for ($skip + 1 .. ($+{offset} / $row{cols}) - 1) { + $row{skip}->{ $_ * $row{cols} - $row{offset} }++; + } } - elsif ($value <= 0xC1) { - print '

(Overl.)' - if $value == 0xC0; + else { + $row{offset} = $+{offset}; } - elsif ($value <= 0xDF) { - print '

2-byte sequence start' - if $value == 0xC2; - print '

' - if $value == 0xD0; + $row{endpoint} = $+{endpoint} || 0; + } + else { + Alert("Unknown option $param for charset $input"); + } + } + + if ($input =~ m{ \A u ([0-9a-f]+) (?:-([0-9a-f]+))? \z }ix) { + my $start = hex($1) << ($2 ? 4 : 8); + my $end = $2 ? (hex($2) << 4) + $row{cols} - 1 : $start + 255; + $row{table} = join '', map { chr } $start .. $end; + utf8::upgrade($row{table}); # prevent latin1 output + $row{endpoint} = $end - $start; + $row{set} = sprintf 'Unicode block U+%02Xxx', $start >> 8; + $row{offset} = $start % 256; + } + elsif (lc $input eq 'uu') { + $row{set} = 'Unicode planes'; + $row{cell} = do 'charset-ucplanes.inc.pl' + or Alert('Table data could not be read', $@ || $!); + $row{endpoint} = 1023 * $row{cell}->{colsize}; + } + elsif (lc $input eq 'u') { + $row{cell} = do 'charset-unicode.inc.pl' + or Alert('Table data could not be read', $@ || $!); + + $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 && $row{endpoint} < 0x20000 ? 'SMP' : + 'allocations' + ); + } + elsif ($input =~ m/^utf-*8$/i) { + $row{set} = 'UTF-8'; + $row{cell} = do 'charset-utf8.inc.pl' + or Alert('Table data could not be read', $@ || $!); + $row{endpoint} = 255; + } + elsif ($row{set} = Encode::resolve_alias($input)) { + $row{endpoint} ||= 255; + if ($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} .. $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} .. $row{endpoint}); + } + $row{endpoint} -= $row{offset}; + + if ($row{set} eq 'cp437') { + for my $phipos (237 - $row{offset}) { + next if $phipos < 0 or $phipos > $row{endpoint}; + # replace phi glyph + substr($row{table}, $phipos, 1) = pack 'U*', 0x3D5; } - elsif ($value <= 0xEF) { - print '

3-byte sequence start' - if $value == 0xE0; + if ($row{offset} == 0) { + # replace control characters by visible variants + substr($row{table}, 0, 32) = pack 'U*', map {hex} qw( + 2007 263A 263B 2665 2666 2663 2660 2022 + 25D8 25CB 25D9 2642 2640 266A 266B 263C + 25BA 25C4 2195 203C 00B6 00A7 25AC 21A8 + 2191 2193 2192 2190 221F 2194 25B2 25BC + ); } - elsif ($value <= 0xF4) { - print '

4-byte sequence' - if $value == 0xF0; + } + + $visible->{ascii} = # assume common base + $visible->{ $row{set} } = 1; + } + else { + Alert("Encoding $input unknown"); + return; + } + push @request, \%row; +} +tabinput($_) for @tablist; + +my $NOCHAR = chr 0xFFFD; + +sub range_cell { + my ($info, $offset) = @_; + my $table = $info->{cell} or return; + my $def = $table->{$offset} or return; + my ($len, $class, $name, $title) = @{$def}; + + my $cols = $info->{cols}; + my $colsize = $table->{colsize} || 1; + my $attr = ''; + $len /= $colsize; + $name //= $len <= 2 ? 'res' : 'reserved'; + + if (my $part = $offset/$colsize % $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"); + if ($len > $rest) { + # minority remains + push @next, $name, $title; + $title ||= $name; + $name = ''; } - elsif ($value <= 0xF7) { - print '

(Overflow)' - if $value == 0xF5; + else { + # minority on next row + push @next, '"', $title || $name; } - elsif ($value <= 0xFB) { - print '

5-byte' - if $value == 0xF8; + $table->{$offset + $colsize*$rest} //= \@next; + $class .= ' joind'; + } + $len = $rest; + } + elsif (my $rows = int($len / $cols)) { + # multiple full rows + my $rowsize = $colsize * $cols; + if ($len -= $rows * $cols) { + # partial row remains + $table->{$offset + $rowsize * $rows} //= [$len*$colsize, "$class joinu", '', $title]; + $class .= ' joind'; + } + + unless ($info->{realsize}) { + # coalesce multiple rows + while ($rows > 3) { + $info->{skip}->{$offset += $rowsize}++; + $rows--; } - elsif ($value <= 0xFD) { - print '

6-byte' - if $value == 0xFC; + if ($rows > 2) { + $info->{skip}->{$offset += $rowsize} = 0; } - elsif ($value <= 0xFF) { - print '

Invalid' - if $value == 0xFE; + } + + $attr .= sprintf ' rowspan=%d', $rows; + $len = $cols; + } + + $attr .= sprintf ' colspan=%d', $len unless $len == 1; + $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\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; # row divide for column digits + $rowdiv = 1 if $rowdiv != $cols * $colsize; # divide only if all columns are matched + my $offset = $row->{startpoint} || 0; + + printf '

', !$row->{cell} && ' charmap'; + my $title = $row->{set}; + $title .= " " + for grep { $_ ne 'iso-8859-1' } $row->{parent} // (); + printf '', $title; + print '' x ($cols + 1); + for my $section (qw{thead}) { + print "<$section>'; + while ($offset < $row->{endpoint}) { + if ($row->{skip}->{$offset}) { + $offset += $cols * $colsize; + next; + } + + print '

%s
", $rowdiv == 1 ? '+' : 'â±'; + printf '	%0X', $coldigits, $_ $colsize for 0 .. $cols - 1; + print "\n"; + } + print '
'; + if (defined $row->{skip}->{$offset}) { + print 'â®'; + } + else { + if (my $rowmod = $offset % $rowdiv) { + # offset in column units + printf '+%X', $rowmod; } else { - print "\n".'	?'; + # divided row offset + printf '%X', ($offset + $row->{offset}) / $rowdiv; } } - print "\n"; + say ''; + + for (1 .. $cols) { + if ($row->{cell}) { + print range_cell($row, $offset); + 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}++; + } + + say sprintf $class ? '	%s' : '	', + $name, $class, $cell; + } + continue { + $offset += $colsize; + } } - print "

\n"; + say '

'; } -print "

+ <: if (exists $get{compare}) { :> +

unicode +

inherited +

existing +

original +

unassigned +<: } else { :> +

control +

whitespace +

diacritic +

letter +

punctuation +

quote +

symbol +

math +

currency +

numeric +

greek +

latin +

cyrillic +

aramaic +

brahmic +

arabic +

syllabic +

african +

japanese +

cjk +

chinese +

alphabetic +

+ + +

unicode 7.0 +

proposed +

deprecated +

unassigned +

invalid +<: } :>