X-Git-Url: http://git.shiar.net/sheet.git/blobdiff_plain/a0ba9298856b2426c5c66b6d2f2b284d98cee594..36f3ac1cd686b98270b6546da7a2ed399336354e:/charset.plp diff --git a/charset.plp b/charset.plp index c9ef2b6..b84816a 100644 --- a/charset.plp +++ b/charset.plp @@ -1,185 +1,554 @@ -<: -use utf8; -use strict; -use warnings; -use open IO => ':utf8'; - -our $VERSION = '1.0'; +<(common.inc.plp)><: -$header{content_type} = 'text/html; charset=utf-8'; +Html({ + title => 'charset cheat sheet', + version => '1.0', + description => [ + "Reference sheet with all glyphs in common character encoding tables,", + "and an overview of Unicode ranges and UTF-8 bytes.", + ], + keywords => [qw' + charset codepage unicode ascii utf8 latin glyph character encoding + reference common overview table + '], + stylesheet => [qw'light'], + data => [qw'charset-unicode.inc.pl charset-ucplanes.inc.pl charset-utf8.inc.pl'], +}); -:> - +my @tablist = split m{/+}, $Request || 'default'; -
-<: -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; - s/>/>/g; - return $_; -} +print join " â¢\n", ( + map { + join " ·\n", pairmap { + optionlink($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, + ], +); +:>. +
+ +<: +use POSIX qw( ceil ); +use Shiar_Sheet::FormatChar; +my $glyphs = Shiar_Sheet::FormatChar->new; +my @request; -print "â±"; - print ' | ', $_ for @nibble; - print "\n"; + state $ALIAS = { + default => [qw( u+0-27F utf-8+realsize iso-8859-1 iso-8859-15 cp1252 cp437 cp850 )], + unicode => [qw( uu+cols=32+realsize u+0-FFF u+1000-18AF u+18B0-1FFF )], + us => [qw( cp437 cp863 gsm0338 AdobeStandardEncoding )], + ebcdic => [qw( cp37 cp500 cp1047 posix-bc cp1026 cp875 )], + iso => [map {"iso-8859-$_"} 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 )], + symbols => [qw( symbol dingbats MacDingbats wingdings wingdings2 wingdings3 webdings )], + + # 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;
+
+ state $visible = {'' => 1}; # all present tables
+ my %row = (offset => 0, cols => 16);
+ my $params = $input =~ s/[+](.*)\z// ? $1 : undef;
+
+ if (not defined $params) {
+ state $INHERIT = {
+ 'ascii' => ['' => '00-7F'],
+ 'cp437' => ['cp850' => 0, 'ascii' => '00-1F+80'], # ascii range overridden later
+ 'gsm0338' => ['ascii' => '00-7F'],
+ 'dingbats' => ['' => '20-7F+A0'],
+ 'MacDingbats' => ['dingbats' => '80-9F'],
+ 'AdobeZdingbat'=> ['MacDingbats' => '80-9F'], # should be identical but maps to private use
+ 'symbol' => ['' => '20-7F+A0'],
+ 'AdobeSymbol' => ['symbol' => '20-7F+A0', '' => '20-7F+A0'], # minor differences, irrelevant except for different 'â¬'
+ 'wingdings' => ['' => '20'],
+ 'wingdings2' => ['' => '20'],
+ 'wingdings3' => ['' => '20'],
+ 'webdings' => ['' => '20'],
+
+ 'iso-8859-2' => ['iso-8859-1' => 'A0'],
+ 'iso-8859-3' => ['iso-8859-1' => 'A0'], #TODO: also apply to iso-8859-9
+ 'iso-8859-4' => ['iso-8859-2' => 'A0'],
+ 'iso-8859-5' => ['iso-8859-1' => 'A0'],
+ 'iso-8859-6' => ['cp1256' => '80', 'iso-8859-1' => 'A0'],
+ 'iso-8859-7' => ['iso-8859-1' => 'A0'],
+ 'iso-8859-8' => ['iso-8859-1' => 'A0'],
+ 'iso-8859-9' => ['iso-8859-1' => 'D0-DF+F0'],
+ 'iso-8859-10' => ['iso-8859-4' => 'A0'],
+ 'iso-8859-11' => ['iso-8859-1' => 'A0'],
+ 'iso-8859-13' => ['iso-8859-4' => 'A0'],
+ 'iso-8859-14' => ['iso-8859-1' => 'A0'],
+ 'iso-8859-15' => ['iso-8859-1' => 'A0-BF'],
+ 'iso-8859-16' => ['iso-8859-2' => 'A0'],
+ 'hp-roman8' => ['iso-8859-1' => 'A0'],
+
+ 'cp1252' => ['iso-8859-1' => '80-9F'],
+ 'cp1250' => ['iso-8859-2' => '80-BF', 'cp1252' => '80'],
+ 'cp1254' => ['iso-8859-9' => '80-9F', 'cp1252' => '80-9F+D0'],
+ 'cp874' => ['iso-8859-11' => '80-9F', 'cp1252' => '80'], # windows-874 actually cp1162
+ 'cp1257' => ['iso-8859-13' => '80-9F+FF', 'cp1252' => '80'],
+ 'cp1251' => ['cp1252' => '80'],
+ 'cp1253' => ['cp1252' => '80'],
+ 'cp1255' => ['iso-8859-8' => '80-DF', 'cp1252' => '80'],
+ 'cp1256' => ['cp1252' => '80'],
+ 'cp1258' => ['cp1252' => '80-9F+C0'],
+
+ 'cp850' => ['cp437' => '90'],
+ 'cp860' => ['cp437' => '80-AF'],
+ 'cp861' => ['cp865' => '80-AF'],
+ 'cp863' => ['cp437' => '80-AF'],
+ 'cp865' => ['cp437' => '90-AF'],
+ 'cp852' => ['cp850' => '80', 'cp437' => '80'],
+ 'cp857' => ['cp850' => '80-AF+D0-EF', 'cp437' => '80'],
+ 'cp775' => ['cp850' => '80'], # partial cp437
+ 'cp866' => ['cp437' => '80-AF+E0'],
+ 'cp855' => ['cp437' => '80'],
+ 'cp1006' => ['iso-8859-6' => 'A0', 'cp437' => '80'],
+ 'cp737' => ['cp437' => '80-AF+E0'],
+ 'cp869' => ['cp437' => '80'],
+ 'cp862' => ['cp437' => '80-9F'],
+ 'cp864' => ['MacArabic' => '80', 'iso-8859-6' => '80', 'cp437' => '80'], #TODO: compare form variants
+
+ 'koi8-u' => ['koi8-r' => '90-BF'],
+ 'koi8-f' => ['koi8-u' => '90-BF'],
+
+ 'MacRomanian' => ['MacRoman' => 'A0-BF+D0-DF'],
+ 'MacRumanian' => ['MacRomanian' => 'A0-BF+D0-DF', 'MacRoman' => 'A0-BF+D0-DF'],
+ 'MacCroatian' => ['MacRoman' => 'A0'],
+ 'MacCentralEurRoman' => ['MacRoman' => '80'],
+ 'MacIcelandic'=> ['MacRoman' => 'A0-AF+D0-EF'],
+ 'MacTurkish' => ['MacRoman' => 'D0-DF'], # F5 is unassigned
+ 'MacSami' => ['MacIcelandic' => '90-BF+D0-DF+F0', 'MacRoman' => '90'],
+ 'MacGreek' => ['MacRoman' => '80'],
+ 'MacCyrillic' => ['MacRoman' => '80'],
+ 'MacHebrew' => ['iso-8859-8' => '80', 'MacRoman' => '80-8F+A0'], # partial ascii
+ 'MacArabic' => ['iso-8859-6' => '80', 'cp864' => '80', 'MacRoman' => '80'], #TODO: multiple parents
+ 'MacFarsi' => ['MacArabic' => 'B0-BF', 'MacRoman' => '80'],
+
+ 'cp37' => ['posix-bc' => '00'],
+ 'posix-bc' => ['cp1047' => '40'],
+ 'cp500' => ['cp37' => '40-5F+B0-BF'],
+ 'cp1047' => ['cp37' => '10-2F+50-5F+A0-BF'],
+ 'cp1026' => ['cp37' => '40'],
+ 'cp875' => ['cp37' => '30'],
+
+ };
+
+ my @parents = @{ $INHERIT->{$input} || [] };
+
+ 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 >= 128; # 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 (?$paramfor 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; - $glyph = quote($string || $glyph); - my $desc = sprintf 'U+%04X%s', $codepoint, $name && " ($name)"; - my @class = ('X', grep {$_} $prop, $script); + $row{table} = join '', map { chr } $row{offset} .. $row{endpoint}; + utf8::upgrade($row{table}); # prevent latin1 output - $glyph = "$glyph" if $prop eq 'Zs'; + $row{endpoint} -= $row{offset}; - printf "\n".' %s',
- join(' ', @class), quote($desc), $glyph;
+ if (defined $block) {
+ $row{set} = sprintf 'Unicode block U+%02Xxx', $block;
+ $row{offset} %= 0x100;
}
- print "\n";
}
- print " | |
â±"; - print ' | ', $_ for @nibble; - print "\n"; + $row{endpoint} ||= 8191; + $row{set} = 'Unicode ' . ( + $row{startpoint} < 0x1000 && $row{endpoint} < 0x1000 ? 'BMP' : + $row{startpoint} >= 0x1000 && $row{endpoint} < 0x2000 ? 'SMP' : + 'allocations' + ); } - 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 $value == 0x80; + elsif ($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{offset} = delete $row{startpoint}; + $row{endpoint} ||= 255; + if ($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} .. $row{endpoint}); + } + + if ($row{set} eq 'cp437') { + if ($row{offset} <= 0xED and $row{endpoint} >= 0xED) { + # replace phi glyph + substr($row{table}, 0xED - $row{offset}, 1) = 'Ï'; } - elsif ($value <= 0xC1) { - print ' | (Overl.)' - if $value == 0xC0; + if ($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) = $sub; } - elsif ($value <= 0xDF) { - print ' | 2-byte sequence start' - if $value == 0xC2; - print ' | ' - if $value == 0xD0; + } + elsif ($row{set} eq 'symbol') { + if ($row{offset} <= 0x60 and $row{endpoint} >= 0x60) { + # replace radical extender by closest unicode equivalent + substr($row{table}, 0x60 - $row{offset}, 1) = 'â'; } - elsif ($value <= 0xEF) { - print ' | 3-byte sequence start' - if $value == 0xE0; + if ($row{offset} <= 0xBD and $row{endpoint} >= 0xFF) { + substr($row{table}, 0xBD - $row{offset}, 2) = 'ââ¯'; # arrow extenders + substr($row{table}, 0xD2 - $row{offset}, 3) = '®©â¢'; # serif variants + substr($row{table}, 0xE0 - $row{offset}, 1) = 'â'; # replace lookalike, should match AdobeSymbol + substr($row{table}, 0xE2 - $row{offset}, 3) = '®©â¢'; # sans-serif variants + substr($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 - $row{offset}, 11) = 'â®â¡âââ â¤â¥â¦â«â¬â'; } - elsif ($value <= 0xF4) { - print ' | 4-byte sequence'
- if $value == 0xF0;
+ }
+
+ $row{endpoint} -= $row{offset};
+
+ $visible->{ascii} = # assume common base
+ $visible->{ $row{set} } = 1;
+ }
+ else {
+ Alert("Encoding $inputunknown"); + return; + } + push @request, \%row; +} +tabinput($_) 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 - $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 = $separate && 'â¦'; } - elsif ($value <= 0xF7) { - print ' | (Overflow)' - if $value == 0xF5; + else { + # minority on next row + push @next, $separate && '"', $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} * $colsize || 0;
+
+ printf '
|
unicode + | inherited + | existing + | original + | unassigned +<: } else { :> + | control + | whitespace + | diacritic
| punctuation
| symbol
| numeric + | greek
| aramaic
| syllabic
| alphabetic + |
unicode 7.0 + | proposed + | deprecated + | unassigned + | invalid +<: } :> |