3 # Generate code page .c files from ftp.unicode.org descriptions
5 # Copyright 2000 Alexandre Julliard
7 # This library is free software; you can redistribute it and/or
8 # modify it under the terms of the GNU Lesser General Public
9 # License as published by the Free Software Foundation; either
10 # version 2.1 of the License, or (at your option) any later version.
12 # This library is distributed in the hope that it will be useful,
13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 # Lesser General Public License for more details.
17 # You should have received a copy of the GNU Lesser General Public
18 # License along with this library; if not, write to the Free Software
19 # Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
22 # base directory for ftp.unicode.org files
23 $BASEDIR = "ftp.unicode.org/Public/";
24 $MAPPREFIX = $BASEDIR . "MAPPINGS/";
27 $UNICODEDATA = $BASEDIR . "UNIDATA/UnicodeData.txt";
30 $SORTKEYS = "www.unicode.org/reports/tr10/allkeys.txt";
33 $DEFAULTS = "./defaults";
35 # Default char for undefined mappings
40 [ 37, "VENDORS/MICSFT/EBCDIC/CP037.TXT", "IBM EBCDIC US Canada" ],
41 [ 424, "VENDORS/MISC/CP424.TXT", "IBM EBCDIC Hebrew" ],
42 [ 437, "VENDORS/MICSFT/PC/CP437.TXT", "OEM United States" ],
43 [ 500, "VENDORS/MICSFT/EBCDIC/CP500.TXT", "IBM EBCDIC International" ],
44 [ 737, "VENDORS/MICSFT/PC/CP737.TXT", "OEM Greek 437G" ],
45 [ 775, "VENDORS/MICSFT/PC/CP775.TXT", "OEM Baltic" ],
46 [ 850, "VENDORS/MICSFT/PC/CP850.TXT", "OEM Multilingual Latin 1" ],
47 [ 852, "VENDORS/MICSFT/PC/CP852.TXT", "OEM Slovak Latin 2" ],
48 [ 855, "VENDORS/MICSFT/PC/CP855.TXT", "OEM Cyrillic" ],
49 [ 856, "VENDORS/MISC/CP856.TXT", "Hebrew PC" ],
50 [ 857, "VENDORS/MICSFT/PC/CP857.TXT", "OEM Turkish" ],
51 [ 860, "VENDORS/MICSFT/PC/CP860.TXT", "OEM Portuguese" ],
52 [ 861, "VENDORS/MICSFT/PC/CP861.TXT", "OEM Icelandic" ],
53 [ 862, "VENDORS/MICSFT/PC/CP862.TXT", "OEM Hebrew" ],
54 [ 863, "VENDORS/MICSFT/PC/CP863.TXT", "OEM Canadian French" ],
55 [ 864, "VENDORS/MICSFT/PC/CP864.TXT", "OEM Arabic" ],
56 [ 865, "VENDORS/MICSFT/PC/CP865.TXT", "OEM Nordic" ],
57 [ 866, "VENDORS/MICSFT/PC/CP866.TXT", "OEM Russian" ],
58 [ 869, "VENDORS/MICSFT/PC/CP869.TXT", "OEM Greek" ],
59 [ 874, "VENDORS/MICSFT/PC/CP874.TXT", "ANSI/OEM Thai" ],
60 [ 875, "VENDORS/MICSFT/EBCDIC/CP875.TXT", "IBM EBCDIC Greek" ],
61 [ 878, "VENDORS/MISC/KOI8-R.TXT", "Russian KOI8" ],
62 [ 932, "VENDORS/MICSFT/WINDOWS/CP932.TXT", "ANSI/OEM Japanese Shift-JIS" ],
63 [ 936, "VENDORS/MICSFT/WINDOWS/CP936.TXT", "ANSI/OEM Simplified Chinese GBK" ],
64 [ 949, "VENDORS/MICSFT/WINDOWS/CP949.TXT", "ANSI/OEM Korean Unified Hangul" ],
65 [ 950, "VENDORS/MICSFT/WINDOWS/CP950.TXT", "ANSI/OEM Traditional Chinese Big5" ],
66 [ 1006, "VENDORS/MISC/CP1006.TXT", "IBM Arabic" ],
67 [ 1026, "VENDORS/MICSFT/EBCDIC/CP1026.TXT", "IBM EBCDIC Latin 5 Turkish" ],
68 [ 1250, "VENDORS/MICSFT/WINDOWS/CP1250.TXT", "ANSI Eastern Europe" ],
69 [ 1251, "VENDORS/MICSFT/WINDOWS/CP1251.TXT", "ANSI Cyrillic" ],
70 [ 1252, "VENDORS/MICSFT/WINDOWS/CP1252.TXT", "ANSI Latin 1" ],
71 [ 1253, "VENDORS/MICSFT/WINDOWS/CP1253.TXT", "ANSI Greek" ],
72 [ 1254, "VENDORS/MICSFT/WINDOWS/CP1254.TXT", "ANSI Turkish" ],
73 [ 1255, "VENDORS/MICSFT/WINDOWS/CP1255.TXT", "ANSI Hebrew" ],
74 [ 1256, "VENDORS/MICSFT/WINDOWS/CP1256.TXT", "ANSI Arabic" ],
75 [ 1257, "VENDORS/MICSFT/WINDOWS/CP1257.TXT", "ANSI Baltic" ],
76 [ 1258, "VENDORS/MICSFT/WINDOWS/CP1258.TXT", "ANSI/OEM Viet Nam" ],
77 [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT", "Mac Roman" ],
78 [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT", "Mac Greek" ],
79 [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT", "Mac Cyrillic" ],
80 [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT", "Mac Latin 2" ],
81 [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT", "Mac Icelandic" ],
82 [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT", "Mac Turkish" ],
83 [ 20866, "VENDORS/MISC/KOI8-R.TXT", "Russian KOI8" ],
84 [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", "EUC-JP" ],
85 [ 21866, "VENDORS/MISC/KOI8-U.TXT", "Ukrainian KOI8" ],
86 [ 28591, "ISO8859/8859-1.TXT", "ISO 8859-1 Latin 1" ],
87 [ 28592, "ISO8859/8859-2.TXT", "ISO 8859-2 Latin 2 (East European)" ],
88 [ 28593, "ISO8859/8859-3.TXT", "ISO 8859-3 Latin 3 (South European)" ],
89 [ 28594, "ISO8859/8859-4.TXT", "ISO 8859-4 Latin 4 (Baltic old)" ],
90 [ 28595, "ISO8859/8859-5.TXT", "ISO 8859-5 Cyrillic" ],
91 [ 28596, "ISO8859/8859-6.TXT", "ISO 8859-6 Arabic" ],
92 [ 28597, "ISO8859/8859-7.TXT", "ISO 8859-7 Greek" ],
93 [ 28598, "ISO8859/8859-8.TXT", "ISO 8859-8 Hebrew" ],
94 [ 28599, "ISO8859/8859-9.TXT", "ISO 8859-9 Latin 5 (Turkish)" ],
95 [ 28600, "ISO8859/8859-10.TXT", "ISO 8859-10 Latin 6 (Nordic)" ],
96 [ 28603, "ISO8859/8859-13.TXT", "ISO 8859-13 Latin 7 (Baltic)" ],
97 [ 28604, "ISO8859/8859-14.TXT", "ISO 8859-14 Latin 8 (Celtic)" ],
98 [ 28605, "ISO8859/8859-15.TXT", "ISO 8859-15 Latin 9 (Euro)" ],
99 [ 28606, "ISO8859/8859-16.TXT", "ISO 8859-16 Latin 10 (Balkan)" ]
118 "Lu" => $ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
119 "Ll" => $ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
120 "Lt" => $ctype{"alpha"}, # Letter, Titlecase
121 "Mn" => $ctype{"punct"}, # Mark, Non-Spacing
122 "Mc" => $ctype{"punct"}, # Mark, Spacing Combining
123 "Me" => $ctype{"punct"}, # Mark, Enclosing
124 "Nd" => $ctype{"digit"}, # Number, Decimal Digit
125 "Nl" => $ctype{"punct"}, # Number, Letter
126 "No" => $ctype{"punct"}, # Number, Other
127 "Zs" => $ctype{"space"}, # Separator, Space
128 "Zl" => $ctype{"space"}, # Separator, Line
129 "Zp" => $ctype{"space"}, # Separator, Paragraph
130 "Cc" => $ctype{"cntrl"}, # Other, Control
131 "Cf" => 0, # Other, Format
132 "Cs" => 0, # Other, Surrogate
133 "Co" => 0, # Other, Private Use
134 "Cn" => 0, # Other, Not Assigned
135 "Lm" => $ctype{"punct"}, # Letter, Modifier
136 "Lo" => $ctype{"alpha"}, # Letter, Other
137 "Pc" => $ctype{"punct"}, # Punctuation, Connector
138 "Pd" => $ctype{"punct"}, # Punctuation, Dash
139 "Ps" => $ctype{"punct"}, # Punctuation, Open
140 "Pe" => $ctype{"punct"}, # Punctuation, Close
141 "Pi" => $ctype{"punct"}, # Punctuation, Initial quote
142 "Pf" => $ctype{"punct"}, # Punctuation, Final quote
143 "Po" => $ctype{"punct"}, # Punctuation, Other
144 "Sm" => $ctype{"punct"}, # Symbol, Math
145 "Sc" => $ctype{"punct"}, # Symbol, Currency
146 "Sk" => $ctype{"punct"}, # Symbol, Modifier
147 "So" => $ctype{"punct"} # Symbol, Other
150 # a few characters need additional categories that cannot be determined automatically
151 %special_categories =
153 "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
154 0xff10..0xff19
, 0xff21..0xff26
, 0xff41..0xff46
],
155 "space" => [ 0x09..0x0d
, 0x85 ],
156 "blank" => [ 0x09, 0x20, 0xa0, 0x3000, 0xfeff ],
157 "cntrl" => [ 0x070f, 0x180b, 0x180c, 0x180d, 0x180e, 0x200c, 0x200d,
158 0x200e, 0x200f, 0x202a, 0x202b, 0x202c, 0x202d, 0x202e,
159 0x206a, 0x206b, 0x206c, 0x206d, 0x206e, 0x206f, 0xfeff,
160 0xfff9, 0xfffa, 0xfffb ]
165 "L" => 1, # Left-to-Right
166 "LRE" => 11, # Left-to-Right Embedding
167 "LRO" => 11, # Left-to-Right Override
168 "R" => 2, # Right-to-Left
169 "AL" => 2, # Right-to-Left Arabic
170 "RLE" => 11, # Right-to-Left Embedding
171 "RLO" => 11, # Right-to-Left Override
172 "PDF" => 11, # Pop Directional Format
173 "EN" => 3, # European Number
174 "ES" => 4, # European Number Separator
175 "ET" => 5, # European Number Terminator
176 "AN" => 6, # Arabic Number
177 "CS" => 7, # Common Number Separator
178 "NSM" => 0, # Non-Spacing Mark
179 "BN" => 0, # Boundary Neutral
180 "B" => 8, # Paragraph Separator
181 "S" => 9, # Segment Separator
182 "WS" => 10, # Whitespace
183 "ON" => 11 # Other Neutrals
187 ################################################################
191 my @sortkeys = READ_SORTKEYS_FILE
();
192 DUMP_CASE_MAPPINGS
();
193 DUMP_SORTKEYS
(@sortkeys);
194 DUMP_COMPOSE_TABLES
();
197 foreach $file (@allfiles) { HANDLE_FILE
( @
$file ); }
204 ################################################################
205 # read in the defaults file
208 @unicode_defaults = ();
209 @unicode_aliases = ();
212 @digitmap_table = ();
213 @compatmap_table = ();
214 @category_table = ();
215 @direction_table = ();
219 # first setup a few default mappings
221 open DEFAULTS
or die "Cannot open $DEFAULTS";
222 print "Loading $DEFAULTS\n";
225 next if /^\#/; # skip comments
226 next if /^$/; # skip empty lines
227 if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
229 my @src = map hex, split /,/,$1;
232 if ($#src > 0) { push @unicode_aliases, \
@src; }
233 next if ($dst eq "none");
234 $dst = ($dst =~ /\'.\'/) ?
ord substr($dst,1,1) : hex $dst;
237 die "Duplicate value" if defined($unicode_defaults[$src]);
238 $unicode_defaults[$src] = $dst;
242 die "Unrecognized line $_\n";
245 # now build mappings from the decomposition field of the Unicode database
247 open UNICODEDATA
or die "Cannot open $UNICODEDATA";
248 print "Loading $UNICODEDATA\n";
249 while (<UNICODEDATA
>)
251 # Decode the fields ...
252 ($code, $name, $cat, $comb, $bidi,
253 $decomp, $dec, $dig, $num, $mirror,
254 $oldname, $comment, $upper, $lower, $title) = split /;/;
258 die "unknown category $cat" unless defined $categories{$cat};
259 die "unknown directionality $bidi" unless defined $directions{$bidi};
261 $uniname[$src] = $name;
262 $category_table[$src] = $categories{$cat};
263 $direction_table[$src] = $directions{$bidi};
267 $tolower_table[$src] = hex $lower;
268 $category_table[$src] |= $ctype{"upper"}|$ctype{"alpha"};
272 $toupper_table[$src] = hex $upper;
273 $category_table[$src] |= $ctype{"lower"}|$ctype{"alpha"};
277 $category_table[$src] |= $ctype{"digit"};
281 $digitmap_table[$src] = ord $dig;
284 # copy the category and direction for everything between First/Last pairs
285 if ($name =~ /, First>/) { $start = $src; }
286 if ($name =~ /, Last>/)
288 while ($start < $src)
290 $category_table[$start] = $category_table[$src];
291 $direction_table[$start] = $direction_table[$src];
296 next if $decomp eq ""; # no decomposition, skip it
298 if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
300 # decomposition of the form "<foo> 1234" -> use char if type is known
301 if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd))
303 # Single char decomposition in the compatibility range
304 $compatmap_table[$src] = hex $2;
306 next unless ($1 eq "font" ||
317 elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
319 # decomposition "<compat> 0020 1234" -> combining accent
322 elsif ($decomp =~ /^([0-9a-fA-F]+)/)
324 # decomposition contains only char values without prefix -> use first char
326 $category_table[$src] |= $category_table[$dst];
327 # store decomposition if it contains two chars
328 if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
330 $decomp_table[$src] = [ hex $1, hex $2 ];
331 push @compose_table, [ hex $1, hex $2, $src ];
333 elsif ($decomp =~ /^(<[a-z]+>\s)*([0-9a-fA-F]+)$/ &&
334 (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)))
336 # Single char decomposition in the compatibility range
337 $compatmap_table[$src] = hex $2;
345 next if defined($unicode_defaults[$src]); # may have been set in the defaults file
348 for ($i = $dst; ; $i = $unicode_defaults[$i])
350 die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
351 last unless defined($unicode_defaults[$i]);
353 $unicode_defaults[$src] = $dst;
356 # patch the category of some special characters
358 foreach $cat (keys %special_categories)
360 my $flag = $ctype{$cat};
361 foreach $i (@
{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
366 ################################################################
367 # parse the input file
371 open INPUT
,$name or die "Cannot open $name";
378 next if /^\#/; # skip comments
379 next if /^$/; # skip empty lines
380 next if /\x1a/; # skip ^Z
381 next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/); # undefined char
383 if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
386 push @lead_bytes,$cp;
390 if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
394 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
395 $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
398 die "$name: Unrecognized line $_\n";
403 ################################################################
404 # build EUC-JP table from the JIS 0208 file
405 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
406 # but this would require 3-byte DBCS characters
407 sub READ_JIS0208_FILE
415 for ($i = 0x00; $i <= 0x7f; $i++)
421 # JIS X 0201 right plane
422 for ($i = 0xa1; $i <= 0xdf; $i++)
424 $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
425 $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
429 foreach $i (0x8e, 0x8f, 0xa1 .. 0xfe)
436 foreach $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
438 $cp2uni[$i] = $DEF_CHAR;
441 # Shift-JIS compatibility
442 $uni2cp[0x00a5] = 0x5c;
443 $uni2cp[0x203e] = 0x7e;
445 # Fix backslash conversion
446 $cp2uni[0xa1c0] = 0xff3c;
447 $uni2cp[0xff3c] = 0xa1c0;
449 open INPUT
, "$name" or die "Cannot open $name";
452 next if /^\#/; # skip comments
453 next if /^$/; # skip empty lines
454 next if /\x1a/; # skip ^Z
455 if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
457 $cp = 0x8080 + hex $1;
459 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
460 $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
463 die "$name: Unrecognized line $_\n";
468 ################################################################
469 # build the sort keys table
470 sub READ_SORTKEYS_FILE
473 for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
475 open INPUT
, "$SORTKEYS" or die "Cannot open $SORTKEYS";
476 print "Loading $SORTKEYS\n";
479 next if /^\#/; # skip comments
480 next if /^$/; # skip empty lines
481 next if /\x1a/; # skip ^Z
482 next if /^\@version/; # skip @version header
483 if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
485 my ($uni,$variable) = (hex $1, $2);
486 next if $uni > 65535;
487 $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
490 if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
492 # multiple character sequence, ignored for now
495 die "$SORTKEYS: Unrecognized line $_\n";
499 # compress the keys to 32 bit:
500 # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
502 @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
503 ${$a}[2] <=> ${$b}[2] or
504 ${$a}[3] <=> ${$b}[3] or
505 ${$a}[4] <=> ${$b}[4] or
506 $a cmp $b; } @sortkeys;
508 my ($n2, $n3) = (1, 1);
509 my @keys = (-1, -1, -1, -1, -1 );
512 for (my $i = 0; $i < 65536; $i++)
514 my @current = @
{$sortkeys[$i]};
515 next if $current[0] == -1;
516 if ($current[1] == $keys[1])
518 if ($current[2] == $keys[2])
520 if ($current[3] == $keys[3])
526 $keys[3] = $current[3];
533 $keys[2] = $current[2];
534 $keys[3] = $current[3];
542 $keys[1] = $current[1];
543 $keys[2] = $current[2];
544 $keys[3] = $current[3];
549 if ($current[2]) { $current[2] = $n2; }
550 if ($current[3]) { $current[3] = $n3; }
551 if ($current[4]) { $current[4] = 1; }
553 $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
559 ################################################################
560 # build the sort keys table
565 # count the number of 256-key ranges that contain something
569 for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
570 for (my $i = 0; $i < 65536; $i++)
572 next unless defined $keys[$i];
573 $offsets[$i >> 8] = $ranges * 256;
578 # output the range offsets
580 open OUTPUT
,">collation.c.new" or die "Cannot create collation.c";
581 printf "Building collation.c\n";
582 printf OUTPUT
"/* Unicode collation element table */\n";
583 printf OUTPUT
"/* generated from %s */\n", $SORTKEYS;
584 printf OUTPUT
"/* DO NOT EDIT!! */\n\n";
586 printf OUTPUT
"const unsigned int collation_table[%d] =\n{\n", $ranges*256;
587 printf OUTPUT
" /* index */\n";
588 printf OUTPUT
"%s,\n", DUMP_ARRAY
( "0x%08x", 0, @offsets );
590 # output the default values
592 printf OUTPUT
" /* defaults */\n";
593 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%08x", 0, (-1) x
256 );
595 # output all the key ranges
597 for (my $i = 0; $i < 256; $i++)
599 next if $offsets[$i] == 256;
600 printf OUTPUT
",\n /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
601 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%08x", -1, @keys[($i<<8) .. ($i<<8)+255] );
603 printf OUTPUT
"\n};\n";
605 save_file
("collation.c");
609 ################################################################
610 # add default mappings once the file had been read
611 sub ADD_DEFAULT_MAPPINGS
615 foreach $alias (@unicode_aliases)
618 foreach $src (@
$alias)
620 if (defined($uni2cp[$src]))
622 $target = $uni2cp[$src];
626 next unless defined($target);
628 # At least one char of the alias set is defined, set the others to the same value
629 foreach $src (@
$alias)
631 $uni2cp[$src] = $target unless defined($uni2cp[$src]);
635 # For every src -> target mapping in the defaults table,
636 # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
638 for ($src = 0; $src < 65536; $src++)
640 next if defined($uni2cp[$src]); # source has a definition already
641 next unless defined($unicode_defaults[$src]); # no default for this char
642 my $target = $unicode_defaults[$src];
644 # do a recursive mapping until we find a target char that is defined
645 while (!defined($uni2cp[$target]) &&
646 defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
648 if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
651 # Add an identity mapping for all undefined chars
653 for ($i = 0; $i < 256; $i++)
655 next if defined($cp2uni[$i]);
656 next if defined($uni2cp[$i]);
657 $cp2uni[$i] = $uni2cp[$i] = $i;
661 ################################################################
662 # dump an array of integers
665 my ($format,$default,@array) = @_;
667 for ($i = 0; $i < $#array; $i++)
669 $ret .= sprintf($format, defined $array[$i] ?
$array[$i] : $default);
670 $ret .= (($i % 8) != 7) ?
", " : ",\n ";
672 $ret .= sprintf($format, defined $array[$i] ?
$array[$i] : $default);
676 ################################################################
677 # dump an SBCS mapping table
680 my ($codepage, $name) = @_;
683 # output the ascii->unicode table
685 printf OUTPUT
"static const WCHAR cp2uni[256] =\n";
686 printf OUTPUT
"{\n%s\n};\n\n", DUMP_ARRAY
( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
688 # count the number of unicode->ascii subtables that contain something
692 for ($i = 0; $i < 65536; $i++)
694 next unless defined $uni2cp[$i];
695 $filled[$i >> 8] = 1;
700 # output all the subtables into a single array
702 printf OUTPUT
"static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
703 for ($i = 0; $i < 256; $i++)
705 next unless $filled[$i];
706 printf OUTPUT
" /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
707 printf OUTPUT
"%s,\n", DUMP_ARRAY
( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
709 printf OUTPUT
" /* defaults */\n";
710 printf OUTPUT
"%s\n};\n\n", DUMP_ARRAY
( "0x%02x", 0, ($DEF_CHAR) x
256 );
712 # output a table of the offsets of the subtables in the previous array
716 for ($i = 0; $i < 256; $i++)
718 if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
719 else { push @offsets, ($subtables-1) * 256; }
721 printf OUTPUT
"static const unsigned short uni2cp_high[256] =\n";
722 printf OUTPUT
"{\n%s\n};\n\n", DUMP_ARRAY
( "0x%04x", 0, @offsets );
724 # output the code page descriptor
726 printf OUTPUT
"const struct sbcs_table cptable_%03d =\n{\n", $codepage;
727 printf OUTPUT
" { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
728 $codepage, $DEF_CHAR, $DEF_CHAR, $name;
729 printf OUTPUT
" cp2uni,\n";
730 printf OUTPUT
" uni2cp_low,\n";
731 printf OUTPUT
" uni2cp_high\n};\n";
735 ################################################################
736 # dump a DBCS mapping table
739 my ($codepage, $name) = @_;
742 # build a list of lead bytes that are actually used
745 LBLOOP
: for ($y = 0; $y <= $#lead_bytes; $y++)
747 my $base = $lead_bytes[$y] << 8;
748 for ($x = 0; $x < 256; $x++)
750 if (defined $cp2uni[$base+$x])
752 push @lblist,$lead_bytes[$y];
757 my $unused = ($#lead_bytes > $#lblist);
759 # output the ascii->unicode table for the single byte chars
761 printf OUTPUT
"static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
762 printf OUTPUT
"{\n%s,\n", DUMP_ARRAY
( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
764 # output the default table for unused lead bytes
768 printf OUTPUT
" /* unused lead bytes */\n";
769 printf OUTPUT
"%s,\n", DUMP_ARRAY
( "0x%04x", 0, ($DEF_CHAR) x
256 );
772 # output the ascii->unicode table for each DBCS lead byte
774 for ($y = 0; $y <= $#lblist; $y++)
776 my $base = $lblist[$y] << 8;
777 printf OUTPUT
" /* lead byte %02x */\n", $lblist[$y];
778 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
779 printf OUTPUT
($y < $#lblist) ?
",\n" : "\n};\n\n";
782 # output the lead byte subtables offsets
785 for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
786 for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
789 # increment all lead bytes offset to take into account the unused table
790 for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
792 printf OUTPUT
"static const unsigned char cp2uni_leadbytes[256] =\n";
793 printf OUTPUT
"{\n%s\n};\n\n", DUMP_ARRAY
( "0x%02x", 0, @offsets );
795 # count the number of unicode->ascii subtables that contain something
799 for ($i = 0; $i < 65536; $i++)
801 next unless defined $uni2cp[$i];
802 $filled[$i >> 8] = 1;
807 # output all the subtables into a single array
809 printf OUTPUT
"static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
810 for ($y = 0; $y < 256; $y++)
812 next unless $filled[$y];
813 printf OUTPUT
" /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
814 printf OUTPUT
"%s,\n", DUMP_ARRAY
( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
816 printf OUTPUT
" /* defaults */\n";
817 printf OUTPUT
"%s\n};\n\n", DUMP_ARRAY
( "0x%04x", 0, ($DEF_CHAR) x
256 );
819 # output a table of the offsets of the subtables in the previous array
823 for ($y = 0; $y < 256; $y++)
825 if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
826 else { push @offsets, ($subtables-1) * 256; }
828 printf OUTPUT
"static const unsigned short uni2cp_high[256] =\n";
829 printf OUTPUT
"{\n%s\n};\n\n", DUMP_ARRAY
( "0x%04x", 0, @offsets );
831 # output the code page descriptor
833 printf OUTPUT
"const struct dbcs_table cptable_%03d =\n{\n", $codepage;
834 printf OUTPUT
" { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
835 $codepage, $DEF_CHAR, $DEF_CHAR, $name;
836 printf OUTPUT
" cp2uni,\n";
837 printf OUTPUT
" cp2uni_leadbytes,\n";
838 printf OUTPUT
" uni2cp_low,\n";
839 printf OUTPUT
" uni2cp_high,\n";
841 printf OUTPUT
"};\n";
845 ################################################################
846 # dump the list of defined lead byte ranges
851 foreach $i (@lead_bytes) { $list[$i] = 1; }
854 for ($i = 0; $i < 256; $i++)
858 if (!defined $list[$i]) { printf OUTPUT
"0x%02x, ", $i-1; $on = 0; }
862 if ($list[$i]) { printf OUTPUT
"0x%02x, ", $i; $on = 1; }
865 if ($on) { printf OUTPUT
"0xff, "; }
866 printf OUTPUT
"0x00, 0x00 }\n";
870 ################################################################
871 # dump the case mapping tables
872 sub DUMP_CASE_MAPPINGS
874 open OUTPUT
,">casemap.c.new" or die "Cannot create casemap.c";
875 printf "Building casemap.c\n";
876 printf OUTPUT
"/* Unicode case mappings */\n";
877 printf OUTPUT
"/* Automatically generated; DO NOT EDIT!! */\n\n";
878 printf OUTPUT
"#include \"wine/unicode.h\"\n\n";
880 DUMP_CASE_TABLE
( "wine_casemap_lower", @tolower_table );
881 DUMP_CASE_TABLE
( "wine_casemap_upper", @toupper_table );
882 DUMP_CASE_TABLE
( "wine_digitmap", @digitmap_table );
883 DUMP_CASE_TABLE
( "wine_compatmap", @compatmap_table );
885 save_file
("casemap.c");
889 ################################################################
890 # dump a case mapping table
893 my ($name,@table) = @_;
895 # count the number of sub tables that contain something
896 # also compute the low and upper populated bounds
898 my @lowerbounds = ( 0, 0 );
899 my @upperbounds = ( 0, 255 );
902 for ($i = 0; $i < 65536; $i++)
904 next unless defined $table[$i];
905 if (!defined $filled[$i >> 8])
907 $lowerbounds[$index] = $i & 0xff;
908 $upperbounds[$index] = 0xff - $lowerbounds[$index];
909 $filled[$i >> 8] = $index * 256 + 512;
914 $upperbounds[$index-1] = 0xff - ($i & 0xff);
916 $table[$i] = ($table[$i] - $i) & 0xffff;
919 # Collapse blocks upwards if possible
922 for ($i = 0; $i < 256; $i++)
924 next unless defined $filled[$i];
925 if ($upperbounds[$index - 1] > $lowerbounds[$index])
927 $removed = $removed + $lowerbounds[$index];
931 $removed = $removed + $upperbounds[$index - 1];
932 $lowerbounds[$index] = $upperbounds[$index - 1];
934 $filled[$i] = $filled[$i] - $removed;
940 printf OUTPUT
"const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
941 printf OUTPUT
"{\n /* index */\n";
942 printf OUTPUT
"%s,\n", DUMP_ARRAY
( "0x%04x", 256, @filled );
943 printf OUTPUT
" /* defaults */\n";
944 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%04x", 0, (0) x
256 );
946 for ($i = 0; $i < 256; $i++)
948 next unless $filled[$i];
949 printf OUTPUT
",\n /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
950 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%04x", 0,
951 @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
954 printf OUTPUT
"\n};\n";
958 ################################################################
959 # dump the ctype tables
960 sub DUMP_CTYPE_TABLES
962 open OUTPUT
,">wctype.c.new" or die "Cannot create wctype.c";
963 printf "Building wctype.c\n";
964 printf OUTPUT
"/* Unicode ctype tables */\n";
965 printf OUTPUT
"/* Automatically generated; DO NOT EDIT!! */\n\n";
966 printf OUTPUT
"#include \"wine/unicode.h\"\n\n";
969 my @array = (0) x
256;
971 # add the direction in the high 4 bits of the category
972 for ($i = 0; $i < 65536; $i++)
974 $category_table[$i] |= $direction_table[$i] << 12;
977 # try to merge table rows
978 for ($row = 0; $row < 256; $row++)
980 my $rowtxt = sprintf "%04x" x
256, @category_table[($row<<8)..($row<<8)+255];
981 if (defined($sequences{$rowtxt}))
983 # reuse an existing row
984 $array[$row] = $sequences{$rowtxt};
989 $sequences{$rowtxt} = $array[$row] = $#array + 1;
990 push @array, @category_table[($row<<8)..($row<<8)+255];
994 printf OUTPUT
"const unsigned short wine_wctype_table[%d] =\n{\n", $#array+1;
995 printf OUTPUT
" /* offsets */\n%s,\n", DUMP_ARRAY
( "0x%04x", 0, @array[0..255] );
996 printf OUTPUT
" /* values */\n%s\n};\n", DUMP_ARRAY
( "0x%04x", 0, @array[256..$#array] );
999 save_file
("wctype.c");
1003 ################################################################
1004 # dump the char composition tables
1005 sub DUMP_COMPOSE_TABLES
1007 open OUTPUT
,">compose.c.new" or die "Cannot create compose.c";
1008 printf "Building compose.c\n";
1009 printf OUTPUT
"/* Unicode char composition */\n";
1010 printf OUTPUT
"/* Automatically generated; DO NOT EDIT!! */\n\n";
1011 printf OUTPUT
"#include \"wine/unicode.h\"\n\n";
1013 ######### composition table
1016 foreach $i (@compose_table)
1019 push @
{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
1022 # count how many different second chars we have
1024 for ($i = $count = 0; $i < 65536; $i++)
1026 next unless defined $filled[$i];
1030 # build the table of second chars and offsets
1032 my $pos = $count + 1;
1033 for ($i = 0; $i < 65536; $i++)
1035 next unless defined $filled[$i];
1036 push @table, $i, $pos;
1037 $pos += @
{$filled[$i]};
1039 # terminator with last position
1040 push @table, 0, $pos;
1041 printf OUTPUT
"const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
1042 printf OUTPUT
" /* second chars + offsets */\n%s", DUMP_ARRAY
( "0x%04x", 0, @table );
1044 # build the table of first chars and mappings
1046 for ($i = 0; $i < 65536; $i++)
1048 next unless defined $filled[$i];
1050 my @list = sort { $a->[0] <=> $b->[0] } @
{$filled[$i]};
1051 for ($j = 0; $j <= $#list; $j++)
1053 push @table, $list[$j][0], $list[$j][1];
1055 printf OUTPUT
",\n /* 0x%04x */\n%s", $i, DUMP_ARRAY
( "0x%04x", 0, @table );
1057 printf OUTPUT
"\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
1059 ######### decomposition table
1061 # first determine all the 16-char subsets that contain something
1063 my @filled = (0) x
4096;
1064 my $pos = 16*2; # for the null subset
1065 for ($i = 0; $i < 65536; $i++)
1067 next unless defined $decomp_table[$i];
1068 $filled[$i >> 4] = $pos;
1074 # now count the 256-char subsets that contain something
1076 my @filled_idx = (256) x
256;
1078 for ($i = 0; $i < 4096; $i++)
1080 next unless $filled[$i];
1081 $filled_idx[$i >> 4] = $pos;
1085 my $null_offset = $pos; # null mapping
1088 # add the index offsets to the subsets positions
1090 for ($i = 0; $i < 4096; $i++)
1092 next unless $filled[$i];
1093 $filled[$i] += $null_offset;
1096 # dump the main index
1098 printf OUTPUT
"const WCHAR unicode_decompose_table[%d] =\n", $total;
1099 printf OUTPUT
"{\n /* index */\n";
1100 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%04x", 0, @filled_idx );
1101 printf OUTPUT
",\n /* null sub-index */\n%s", DUMP_ARRAY
( "0x%04x", 0, ($null_offset) x
16 );
1103 # dump the second-level indexes
1105 for ($i = 0; $i < 256; $i++)
1107 next unless ($filled_idx[$i] > 256);
1108 my @table = @filled[($i<<4)..($i<<4)+15];
1109 for ($j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1110 printf OUTPUT
",\n /* sub-index %02x */\n", $i;
1111 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%04x", 0, @table );
1114 # dump the 16-char subsets
1116 printf OUTPUT
",\n /* null mapping */\n";
1117 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%04x", 0, (0) x
32 );
1119 for ($i = 0; $i < 4096; $i++)
1121 next unless $filled[$i];
1122 my @table = (0) x
32;
1123 for ($j = 0; $j < 16; $j++)
1125 if (defined $decomp_table[($i<<4) + $j])
1127 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
1128 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
1131 printf OUTPUT
",\n /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1132 printf OUTPUT
"%s", DUMP_ARRAY
( "0x%04x", 0, @table );
1135 printf OUTPUT
"\n};\n";
1137 save_file
("compose.c");
1141 ################################################################
1142 # read an input file and generate the corresponding .c file
1145 my ($codepage,$filename,$comment) = @_;
1147 # symbol codepage file is special
1148 if ($codepage == 20932) { READ_JIS0208_FILE
($MAPPREFIX . $filename); }
1149 else { READ_FILE
($MAPPREFIX . $filename); }
1151 # hack: 0x00a5 must map to backslash in Shift-JIS
1152 if ($codepage == 932) { $uni2cp[0x00a5] = 0x5c; }
1154 ADD_DEFAULT_MAPPINGS
();
1156 my $output = sprintf "c_%03d.c", $codepage;
1157 open OUTPUT
,">$output.new" or die "Cannot create $output";
1159 printf "Building %s from %s (%s)\n", $output, $filename, $comment;
1163 printf OUTPUT
"/* code page %03d (%s) */\n", $codepage, $comment;
1164 printf OUTPUT
"/* generated from %s */\n", $MAPPREFIX . $filename;
1165 printf OUTPUT
"/* DO NOT EDIT!! */\n\n";
1166 printf OUTPUT
"#include \"wine/unicode.h\"\n\n";
1168 if ($#lead_bytes == -1) { DUMP_SBCS_TABLE
( $codepage, $comment ); }
1169 else { DUMP_DBCS_TABLE
( $codepage, $comment ); }
1175 ################################################################
1176 # save a file if modified
1180 if (!system "cmp $file $file.new >/dev/null")
1186 rename "$file.new", "$file";
1191 ################################################################
1192 # output the list of codepage tables into the cptable.c file
1197 foreach $file (@allfiles)
1199 my ($codepage,$filename,$comment) = @
$file;
1200 push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
1203 push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
1204 foreach $file (@allfiles)
1206 my ($codepage,$filename,$comment) = @
$file;
1207 push @tables_decl, sprintf(" &cptable_%03d,\n", $codepage);
1209 push @tables_decl, "};";
1210 REPLACE_IN_FILE
( "cptable.c", @tables_decl );
1213 ################################################################
1214 # replace the contents of a file between ### cpmap ### marks
1221 open(FILE
,$name) or die "Can't open $name";
1225 last if /\#\#\# cpmap begin \#\#\#/;
1230 if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
1232 push @lines, <FILE
>;
1233 open(FILE
,">$name.new") or die "Can't modify $name";