charset cheat sheet

Character encodings

<: -my $diinfo = do 'digraphs.inc.pl'; -my %di = map { $diinfo->{$_}->[0] => $_ } grep { ref $diinfo->{$_} } - keys %$diinfo; - -use Encode qw(decode resolve_alias); -# generate character table(s) -# (~16x faster than decoding in loop; -# substr strings is twice as fast as splitting to an array) -my %ALIAS = ( -# default => [qw(unicode utf-8 iso-8859-1 cp437 -cp1252- --iso-8859-15- -koi8-f)], - default => [qw(unicode- utf-8 iso-8859-1 -cp1252- --iso-8859-15- cp437 -cp850)], - 0 => [qw(cp437 cp863)], - 1 => [qw(iso-8859-1 cp1252 MacRoman cp850)], - 2 => [qw(iso-8859-2 cp1250 cp852 MacCentralEurRoman MacCroatian MacRumanian)], - 5 => [qw(koi8-f iso-8859-5 cp1251 MacCyrillic cp855 cp866)], - 7 => [qw(iso-8859-7 cp1253 MacGreek cp737 cp869)], - 8 => [qw(iso-8859-8 cp1255 MacHebrew cp862)], +if ($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( + $selected ? '%s' : '%s', + EscapeHTML($title), $href + ); +} + +print join " â¢\n", ( + map { + join " Â·\n", pairmap { + optionlink($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, + ], ); -my @request = map { - if (my $input = $_) { - my %row = (offset => 0); - 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} ? $row{offset} < 160 ? 159 : 191 : 127; - } - - if ($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 512; - $row{set} = 'Unicode planes'; - $row{cell} = do 'charset-ucplanes.inc.pl'; - } - elsif ($row{set} = resolve_alias($input)) { - if ($row{set} eq 'Internal') { - $row{table} = ' ' x ($endpoint < 255 ? 640 : 4096); - $row{set} = 'Unicode BMP'; - $row{cell} = do 'charset-unicode.inc.pl'; - } - elsif ($row{set} eq 'utf-8-strict') { - $row{table} = undef; - $row{set} = 'UTF-8'; - $row{cell} = do 'charset-utf8.inc.pl'; +:>. +

\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} - $row{startpoint} }++; + } } else { - $row{table} = decode($row{set}, pack 'C*', $row{offset} .. $endpoint); + $row{startpoint} = hex $+{start}; } + $row{endpoint} = hex($+{end} || 0); } else { - print "

Encoding $input unknown

\n"; + Alert("Unknown option $param for charset $input"); + } + } + + if ($input =~ m{ \A (?:wing|web)dings \d* \z }ix) { + eval "require Encode::\u$input"; + } + + if ($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{endpoint} -= $row{offset}; + + if (defined $block) { + $row{set} = sprintf 'Unicode block U+%02Xxx', $block; + $row{offset} %= 0x100; } - \%row; + } + elsif (lc $input eq 'uu') { + $row{cell} = do 'charset-ucplanes.inc.pl' + or Alert('Table data could not be read', $@ || $!); + $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)) { + $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 { + # ~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 - $row{offset}, 1) = 'Ï'; + } + if ($row{offset} < 0x20) { + # replace control characters by visible variants + my $sub = substr 'ââºâ»â¥â¦â£â â¢ââââââªâ«â¼âºâââ¼Â¶Â§â¬â¨âââââââ²â¼', $row{offset}; + substr($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 - $row{offset}, 1) = 'â'; + } + if ($row{offset} <= 0xBD and $row{endpoint} >= 0xFF) { + substr($row{table}, 0xBD - $row{offset}, 2) = 'ââ¯'; # arrow extenders + substr($row{table}, 0xD2 - $row{offset}, 3) = 'Â®Â©â¢'; # serif variants + substr($row{table}, 0xE0 - $row{offset}, 1) = 'â'; # replace lookalike, should match AdobeSymbol + substr($row{table}, 0xE2 - $row{offset}, 3) = 'Â®Â©â¢'; # sans-serif variants + substr($row{table}, 0xE6 - $row{offset}, 10) = 'ââââ¡â¢â£â§â¨â©âª'; + substr($row{table}, 0xF0 - $row{offset}, 1) = 'â¬'; + substr($row{table}, 0xF4 - $row{offset}, 11) = 'â®â¡âââ â¤â¥â¦â«â¬â'; + } + } + + $row{endpoint} -= $row{offset}; + + $visible->{ascii} = # assume common base + $visible->{ $row{set} } = 1; } else { - (); + Alert("Encoding $input unknown"); + return; } -} map { defined $ALIAS{$_} ? @{ $ALIAS{$_} } : $_ } - $ENV{PATH_INFO} =~ /\w/ ? split(m{[/+\s]}, $ENV{PATH_INFO}) : 'default'; + push @request, \%row; +} +tabinput($_) for @tablist; + 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[$cp437]->{table}, 0, 32) = pack 'U*', map {hex} qw( - 2007 263A 263B 2665 2666 2663 2660 2022 25D8 25CB 25D9 2642 2640 266A 266B 263C - 25BA 25C4 2195 203C 00B6 00A7 25AC 21A8 2191 2193 2192 2190 221F 2194 25B2 25BC - ); -} +sub range_cell { + my ($info, $offset) = @_; + my $table = $info->{cell} or return; + my $def = $table->{$offset} or return; + my ($len, $class, $name, $title) = @{$def}; -sub quote { - local $_ = shift; - s/"/"/g; - s//>/g; - return $_; + my $cols = $info->{cols}; + my $colsize = $table->{colsize} || 1; + my $attr = ''; + $len /= $colsize; + $name //= $len <= 2 ? 'res' : 'reserved'; + + 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 = $separate && 'â¦'; + } + else { + # minority on next row + push @next, $separate && '"', $title || $name; + } + $table->{$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}++; + $rows--; + } + if ($rows > 2) { + $info->{skip}->{$offset += $rowsize} = 0; + } + } + + $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"; } -my @nibble = (0..9, 'A'..'F'); 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} * $colsize || 0; + printf '

', !$row->{cell} && ' charmap'; - printf '', $row->{set}; - print '' x 17; + my $title = $row->{set}; + $title .= " " + for $row->{parent} || (); + printf '', $title; + print '' x ($cols + 1); for my $section (qw{thead}) { - print "<$section>'; - for my $msb (0 .. (length($row->{table}) || 256) - 1 >> 4) { - printf ''; + while ($offset <= $row->{endpoint} * $colsize) { + if ($row->{skip}->{$offset}) { + $offset += $cols * $colsize; + next; + } - my $info = [ord $glyph]; - if (defined (my $mnem = $di{ord $glyph})) { - $info = $diinfo->{$mnem}; + print '

%s
â±"; - print '	', $_ for @nibble; + print "<$section>
", $rowdiv == 1 ? '+' : 'â±'; + printf '	%0X', $coldigits, $_ $colsize for 0 .. $cols - 1; print "\n"; } - print '
%X', $msb + ($row->{offset} >> 4); - for my $lsb (0 .. $#nibble) { - if ($row->{cell}) { - print $row->{cell}->(($msb<<4) + $lsb); - next; - } - my $glyph = substr $row->{table}, ($msb<<4) + $lsb, 1; - if ($glyph eq $NOCHAR) { - print '	'; - next; - } + print '
'; + if (defined $row->{skip}->{$offset}) { + print 'â®'; + } + else { + if (my $rowmod = $offset % $rowdiv) { + # offset in column units + printf '+%X', $rowmod; } else { - require Unicode::UCD; - my $fullinfo = Unicode::UCD::charinfo(ord $glyph); - $info = [@$fullinfo{qw/code name category script string/}] if $fullinfo; + # divided row offset + printf '%X', ($offset + $row->{offset}) / $rowdiv; } - my ($codepoint, $name, $prop, $script, $string) = @$info; + } + say ''; - $glyph = quote($string \|\| $glyph); - my $desc = sprintf 'U+%04X%s', $codepoint, $name && " ($name)"; - my @class = ('X', grep {$_} $prop, $script); + for (1 .. $cols) { + if ($row->{cell}) { + print range_cell($row, $offset); + next; + } - $glyph = "$glyph" if $prop eq 'Zs'; + 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); - printf "\n".'	%s', - join(' ', @class), quote($desc), $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"; } - print "

\n"; + say ''; } :>

- +

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

unicode +

inherited +

existing +

original +

unassigned +<: } else { :>

control

whitespace

diacritic @@ -188,22 +544,11 @@ for my $row (@request) {

unicode 5.0 +

unicode 7.0

proposed

deprecated

unassigned

invalid -

+<: } :>

- - -