unicode/cpmap.pl

   1 #!/usr/bin/perl
   2 #
   3 # Generate code page .c files from ftp.unicode.org descriptions
   4 #
   5 # Copyright 2000 Alexandre Julliard
   6 #
   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.
  11 #
  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.
  16 #
  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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  20 #
  21
  22 # base directory for ftp.unicode.org files
  23 $BASEDIR = "ftp.unicode.org/Public/";
  24 $MAPPREFIX = $BASEDIR . "MAPPINGS/";
  25
  26 # UnicodeData file
  27 $UNICODEDATA = $BASEDIR . "UNIDATA/UnicodeData.txt";
  28
  29 # Defaults mapping
  30 $DEFAULTS = "./defaults";
  31
  32 # Default char for undefined mappings
  33 $DEF_CHAR = ord '?';
  34
  35 @allfiles =
  36 (
  37     [ 37,    "VENDORS/MICSFT/EBCDIC/CP037.TXT",   "IBM EBCDIC US Canada" ],
  38     [ 42,    "VENDORS/ADOBE/symbol.txt",          "Symbol" ],
  39     [ 424,   "VENDORS/MISC/CP424.TXT",            "IBM EBCDIC Hebrew" ],
  40     [ 437,   "VENDORS/MICSFT/PC/CP437.TXT",       "OEM United States" ],
  41     [ 500,   "VENDORS/MICSFT/EBCDIC/CP500.TXT",   "IBM EBCDIC International" ],
  42     [ 737,   "VENDORS/MICSFT/PC/CP737.TXT",       "OEM Greek 437G" ],
  43     [ 775,   "VENDORS/MICSFT/PC/CP775.TXT",       "OEM Baltic" ],
  44     [ 850,   "VENDORS/MICSFT/PC/CP850.TXT",       "OEM Multilingual Latin 1" ],
  45     [ 852,   "VENDORS/MICSFT/PC/CP852.TXT",       "OEM Slovak Latin 2" ],
  46     [ 855,   "VENDORS/MICSFT/PC/CP855.TXT",       "OEM Cyrillic" ],
  47     [ 856,   "VENDORS/MISC/CP856.TXT",            "Hebrew PC" ],
  48     [ 857,   "VENDORS/MICSFT/PC/CP857.TXT",       "OEM Turkish" ],
  49     [ 860,   "VENDORS/MICSFT/PC/CP860.TXT",       "OEM Portuguese" ],
  50     [ 861,   "VENDORS/MICSFT/PC/CP861.TXT",       "OEM Icelandic" ],
  51     [ 862,   "VENDORS/MICSFT/PC/CP862.TXT",       "OEM Hebrew" ],
  52     [ 863,   "VENDORS/MICSFT/PC/CP863.TXT",       "OEM Canadian French" ],
  53     [ 864,   "VENDORS/MICSFT/PC/CP864.TXT",       "OEM Arabic" ],
  54     [ 865,   "VENDORS/MICSFT/PC/CP865.TXT",       "OEM Nordic" ],
  55     [ 866,   "VENDORS/MICSFT/PC/CP866.TXT",       "OEM Russian" ],
  56     [ 869,   "VENDORS/MICSFT/PC/CP869.TXT",       "OEM Greek" ],
  57     [ 874,   "VENDORS/MICSFT/PC/CP874.TXT",       "ANSI/OEM Thai" ],
  58     [ 875,   "VENDORS/MICSFT/EBCDIC/CP875.TXT",   "IBM EBCDIC Greek" ],
  59     [ 878,   "VENDORS/MISC/KOI8-R.TXT",           "Russian KOI8" ],
  60     [ 932,   "VENDORS/MICSFT/WINDOWS/CP932.TXT",  "ANSI/OEM Japanese Shift-JIS" ],
  61     [ 936,   "VENDORS/MICSFT/WINDOWS/CP936.TXT",  "ANSI/OEM Simplified Chinese GBK" ],
  62     [ 949,   "VENDORS/MICSFT/WINDOWS/CP949.TXT",  "ANSI/OEM Korean Unified Hangul" ],
  63     [ 950,   "VENDORS/MICSFT/WINDOWS/CP950.TXT",  "ANSI/OEM Traditional Chinese Big5" ],
  64     [ 1006,  "VENDORS/MISC/CP1006.TXT",           "IBM Arabic" ],
  65     [ 1026,  "VENDORS/MICSFT/EBCDIC/CP1026.TXT",  "IBM EBCDIC Latin 5 Turkish" ],
  66     [ 1250,  "VENDORS/MICSFT/WINDOWS/CP1250.TXT", "ANSI Eastern Europe" ],
  67     [ 1251,  "VENDORS/MICSFT/WINDOWS/CP1251.TXT", "ANSI Cyrillic" ],
  68     [ 1252,  "VENDORS/MICSFT/WINDOWS/CP1252.TXT", "ANSI Latin 1" ],
  69     [ 1253,  "VENDORS/MICSFT/WINDOWS/CP1253.TXT", "ANSI Greek" ],
  70     [ 1254,  "VENDORS/MICSFT/WINDOWS/CP1254.TXT", "ANSI Turkish" ],
  71     [ 1255,  "VENDORS/MICSFT/WINDOWS/CP1255.TXT", "ANSI Hebrew" ],
  72     [ 1256,  "VENDORS/MICSFT/WINDOWS/CP1256.TXT", "ANSI Arabic" ],
  73     [ 1257,  "VENDORS/MICSFT/WINDOWS/CP1257.TXT", "ANSI Baltic" ],
  74     [ 1258,  "VENDORS/MICSFT/WINDOWS/CP1258.TXT", "ANSI/OEM Viet Nam" ],
  75     [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT",      "Mac Roman" ],
  76     [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT",      "Mac Greek" ],
  77     [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT",   "Mac Cyrillic" ],
  78     [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT",     "Mac Latin 2" ],
  79     [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT",    "Mac Icelandic" ],
  80     [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT",    "Mac Turkish" ],
  81     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           "Russian KOI8" ],
  82     [ 28591, "ISO8859/8859-1.TXT",                "ISO 8859-1 Latin 1" ],
  83     [ 28592, "ISO8859/8859-2.TXT",                "ISO 8859-2 Latin 2 (East European)" ],
  84     [ 28593, "ISO8859/8859-3.TXT",                "ISO 8859-3 Latin 3 (South European)" ],
  85     [ 28594, "ISO8859/8859-4.TXT",                "ISO 8859-4 Latin 4 (Baltic old)" ],
  86     [ 28595, "ISO8859/8859-5.TXT",                "ISO 8859-5 Cyrillic" ],
  87     [ 28596, "ISO8859/8859-6.TXT",                "ISO 8859-6 Arabic" ],
  88     [ 28597, "ISO8859/8859-7.TXT",                "ISO 8859-7 Greek" ],
  89     [ 28598, "ISO8859/8859-8.TXT",                "ISO 8859-8 Hebrew" ],
  90     [ 28599, "ISO8859/8859-9.TXT",                "ISO 8859-9 Latin 5 (Turkish)" ],
  91     [ 28600, "ISO8859/8859-10.TXT",               "ISO 8859-10 Latin 6 (Nordic)" ],
  92     [ 28603, "ISO8859/8859-13.TXT",               "ISO 8859-13 Latin 7 (Baltic)" ],
  93     [ 28604, "ISO8859/8859-14.TXT",               "ISO 8859-14 Latin 8 (Celtic)" ],
  94     [ 28605, "ISO8859/8859-15.TXT",               "ISO 8859-15 Latin 9 (Euro)" ]
  95 );
  96
  97
  98 %ctype =
  99 (
 100     "upper"  => 0x0001,
 101     "lower"  => 0x0002,
 102     "digit"  => 0x0004,
 103     "space"  => 0x0008,
 104     "punct"  => 0x0010,
 105     "cntrl"  => 0x0020,
 106     "blank"  => 0x0040,
 107     "xdigit" => 0x0080,
 108     "alpha"  => 0x0100
 109 );
 110
 111 %categories =
 112 (
 113     "Lu" => $ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
 114     "Ll" => $ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
 115     "Lt" => $ctype{"alpha"},    # Letter, Titlecase
 116     "Mn" => $ctype{"punct"},    # Mark, Non-Spacing
 117     "Mc" => $ctype{"punct"},    # Mark, Spacing Combining
 118     "Me" => $ctype{"punct"},    # Mark, Enclosing
 119     "Nd" => $ctype{"digit"},    # Number, Decimal Digit
 120     "Nl" => $ctype{"punct"},    # Number, Letter
 121     "No" => $ctype{"punct"},    # Number, Other
 122     "Zs" => $ctype{"space"},    # Separator, Space
 123     "Zl" => 0,                  # Separator, Line
 124     "Zp" => 0,                  # Separator, Paragraph
 125     "Cc" => $ctype{"cntrl"},    # Other, Control
 126     "Cf" => 0,                  # Other, Format
 127     "Cs" => 0,                  # Other, Surrogate
 128     "Co" => 0,                  # Other, Private Use
 129     "Cn" => 0,                  # Other, Not Assigned
 130     "Lm" => $ctype{"punct"},    # Letter, Modifier
 131     "Lo" => $ctype{"alpha"},    # Letter, Other
 132     "Pc" => $ctype{"punct"},    # Punctuation, Connector
 133     "Pd" => $ctype{"punct"},    # Punctuation, Dash
 134     "Ps" => $ctype{"punct"},    # Punctuation, Open
 135     "Pe" => $ctype{"punct"},    # Punctuation, Close
 136     "Pi" => $ctype{"punct"},    # Punctuation, Initial quote
 137     "Pf" => $ctype{"punct"},    # Punctuation, Final quote
 138     "Po" => $ctype{"punct"},    # Punctuation, Other
 139     "Sm" => $ctype{"punct"},    # Symbol, Math
 140     "Sc" => $ctype{"punct"},    # Symbol, Currency
 141     "Sk" => $ctype{"punct"},    # Symbol, Modifier
 142     "So" => $ctype{"punct"}     # Symbol, Other
 143 );
 144
 145 # a few characters need additional categories that cannot be determined automatically
 146 %special_categories =
 147 (
 148     "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
 149                   0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ],
 150     "space"  => [ 0x09..0x0d, 0xfeff ],
 151     "blank"  => [ 0x09, 0x20, 0xa0, 0xfeff ]
 152 );
 153
 154 %directions =
 155 (
 156     "L"   => 1,    # Left-to-Right
 157     "LRE" => 11,   # Left-to-Right Embedding
 158     "LRO" => 11,   # Left-to-Right Override
 159     "R"   => 2,    # Right-to-Left
 160     "AL"  => 2,    # Right-to-Left Arabic
 161     "RLE" => 11,   # Right-to-Left Embedding
 162     "RLO" => 11,   # Right-to-Left Override
 163     "PDF" => 11,   # Pop Directional Format
 164     "EN"  => 3,    # European Number
 165     "ES"  => 4,    # European Number Separator
 166     "ET"  => 5,    # European Number Terminator
 167     "AN"  => 6,    # Arabic Number
 168     "CS"  => 7,    # Common Number Separator
 169     "NSM" => 0,    # Non-Spacing Mark
 170     "BN"  => 0,    # Boundary Neutral
 171     "B"   => 8,    # Paragraph Separator
 172     "S"   => 9,    # Segment Separator
 173     "WS"  => 10,   # Whitespace
 174     "ON"  => 11    # Other Neutrals
 175 );
 176
 177
 178 ################################################################
 179 # main routine
 180
 181 READ_DEFAULTS();
 182 DUMP_CASE_MAPPINGS();
 183 DUMP_COMPOSE_TABLES();
 184 DUMP_CTYPE_TABLES();
 185
 186 foreach $file (@allfiles) { HANDLE_FILE( @$file ); }
 187
 188 OUTPUT_CPTABLE();
 189
 190 exit(0);
 191
 192
 193 ################################################################
 194 # read in the defaults file
 195 sub READ_DEFAULTS
 196 {
 197     @unicode_defaults = ();
 198     @unicode_aliases = ();
 199     @tolower_table = ();
 200     @toupper_table = ();
 201     @category_table = ();
 202     @direction_table = ();
 203     @decomp_table = ();
 204     @compose_table = ();
 205
 206     # first setup a few default mappings
 207
 208     open DEFAULTS or die "Cannot open $DEFAULTS";
 209     print "Loading $DEFAULTS\n";
 210     while (<DEFAULTS>)
 211     {
 212         next if /^\#/;  # skip comments
 213         next if /^$/;  # skip empty lines
 214         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 215         {
 216             my @src = map hex, split /,/,$1;
 217             my $dst = $4;
 218             my $comment = $5;
 219             if ($#src > 0) { push @unicode_aliases, \@src; }
 220             next if ($dst eq "none");
 221             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 222             foreach $src (@src)
 223             {
 224                 die "Duplicate value" if defined($unicode_defaults[$src]);
 225                 $unicode_defaults[$src] = $dst;
 226             }
 227             next;
 228         }
 229         die "Unrecognized line $_\n";
 230     }
 231
 232     # now build mappings from the decomposition field of the Unicode database
 233
 234     open UNICODEDATA or die "Cannot open $UNICODEDATA";
 235     print "Loading $UNICODEDATA\n";
 236     while (<UNICODEDATA>)
 237     {
 238         # Decode the fields ...
 239         ($code, $name, $cat, $comb, $bidi,
 240          $decomp, $dec, $dig, $num, $mirror,
 241          $oldname, $comment, $upper, $lower, $title) = split /;/;
 242
 243         my $src = hex $code;
 244
 245         die "unknown category $cat" unless defined $categories{$cat};
 246         die "unknown directionality $bidi" unless defined $directions{$bidi};
 247
 248         $uniname[$src] = $name;
 249         $category_table[$src] = $categories{$cat};
 250         $direction_table[$src] = $directions{$bidi};
 251
 252         if ($lower ne "")
 253         {
 254             $tolower_table[$src] = hex $lower;
 255             $category_table[$src] |= $ctype{"upper"}|$ctype{"alpha"};
 256         }
 257         if ($upper ne "")
 258         {
 259             $toupper_table[$src] = hex $upper;
 260             $category_table[$src] |= $ctype{"lower"}|$ctype{"alpha"};
 261         }
 262         if ($dec ne "")
 263         {
 264             $category_table[$src] |= $ctype{"digit"};
 265         }
 266
 267         # copy the category and direction for everything between First/Last pairs
 268         if ($name =~ /, First>/) { $start = $src; }
 269         if ($name =~ /, Last>/)
 270         {
 271             while ($start < $src)
 272             {
 273                 $category_table[$start] = $category_table[$src];
 274                 $direction_table[$start] = $direction_table[$src];
 275                 $start++;
 276             }
 277         }
 278
 279         next if $decomp eq "";  # no decomposition, skip it
 280
 281         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 282         {
 283             # decomposition of the form "<foo> 1234" -> use char if type is known
 284             next unless ($1 eq "font" ||
 285                          $1 eq "noBreak" ||
 286                          $1 eq "circle" ||
 287                          $1 eq "super" ||
 288                          $1 eq "sub" ||
 289                          $1 eq "wide" ||
 290                          $1 eq "narrow" ||
 291                          $1 eq "compat" ||
 292                          $1 eq "small");
 293             $dst = hex $2;
 294         }
 295         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 296         {
 297             # decomposition "<compat> 0020 1234" -> combining accent
 298             $dst = hex $1;
 299         }
 300         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 301         {
 302             # decomposition contains only char values without prefix -> use first char
 303             $dst = hex $1;
 304             $category_table[$src] |= $category_table[$dst];
 305             # store decomposition if it contains two chars
 306             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 307             {
 308                 $decomp_table[$src] = [ hex $1, hex $2 ];
 309                 push @compose_table, [ hex $1, hex $2, $src ];
 310             }
 311         }
 312         else
 313         {
 314             next;
 315         }
 316
 317         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 318
 319         # check for loops
 320         for ($i = $dst; ; $i = $unicode_defaults[$i])
 321         {
 322             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 323             last unless defined($unicode_defaults[$i]);
 324         }
 325         $unicode_defaults[$src] = $dst;
 326     }
 327
 328     # patch the category of some special characters
 329
 330     foreach $cat (keys %special_categories)
 331     {
 332         my $flag = $ctype{$cat};
 333         foreach $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 334     }
 335 }
 336
 337
 338 ################################################################
 339 # parse the input file
 340 sub READ_FILE
 341 {
 342     my $name = shift;
 343     open INPUT,$name or die "Cannot open $name";
 344     @cp2uni = ();
 345     @lead_bytes = ();
 346     @uni2cp = ();
 347
 348     while (<INPUT>)
 349     {
 350         next if /^\#/;  # skip comments
 351         next if /^$/;  # skip empty lines
 352         next if /\x1a/;  # skip ^Z
 353         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 354
 355         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 356         {
 357             $cp = hex $1;
 358             push @lead_bytes,$cp;
 359             $cp2uni[$cp] = 0;
 360             next;
 361         }
 362         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 363         {
 364             $cp = hex $1;
 365             $uni = hex $2;
 366             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 367             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 368             next;
 369         }
 370         die "$name: Unrecognized line $_\n";
 371     }
 372 }
 373
 374
 375 ################################################################
 376 # parse the symbol.txt file, since its syntax is different from the other ones
 377 sub READ_SYMBOL_FILE
 378 {
 379     my $name = shift;
 380     open INPUT,$name or die "Cannot open $name";
 381     @cp2uni = ();
 382     @lead_bytes = ();
 383     @uni2cp = ();
 384
 385     while (<INPUT>)
 386     {
 387         next if /^\#/;  # skip comments
 388         next if /^$/;  # skip empty lines
 389         next if /\x1a/;  # skip ^Z
 390         if (/^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)\s+(\#.*)?/)
 391         {
 392             $uni = hex $1;
 393             $cp = hex $2;
 394             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 395             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 396             next;
 397         }
 398         die "$name: Unrecognized line $_\n";
 399     }
 400 }
 401
 402
 403 ################################################################
 404 # add default mappings once the file had been read
 405 sub ADD_DEFAULT_MAPPINGS
 406 {
 407     # Apply aliases
 408
 409     foreach $alias (@unicode_aliases)
 410     {
 411         my $target = undef;
 412         foreach $src (@$alias)
 413         {
 414             if (defined($uni2cp[$src]))
 415             {
 416                 $target = $uni2cp[$src];
 417                 last;
 418             }
 419         }
 420         next unless defined($target);
 421
 422         # At least one char of the alias set is defined, set the others to the same value
 423         foreach $src (@$alias)
 424         {
 425             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 426         }
 427     }
 428
 429     # For every src -> target mapping in the defaults table,
 430     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 431
 432     for ($src = 0; $src < 65536; $src++)
 433     {
 434         next if defined($uni2cp[$src]);  # source has a definition already
 435         next unless defined($unicode_defaults[$src]);  # no default for this char
 436         my $target = $unicode_defaults[$src];
 437
 438         # do a recursive mapping until we find a target char that is defined
 439         while (!defined($uni2cp[$target]) &&
 440                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 441
 442         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 443     }
 444
 445     # Add an identity mapping for all undefined chars
 446
 447     for ($i = 0; $i < 256; $i++)
 448     {
 449         next if defined($cp2uni[$i]);
 450         next if defined($uni2cp[$i]);
 451         $cp2uni[$i] = $uni2cp[$i] = $i;
 452     }
 453 }
 454
 455 ################################################################
 456 # dump an array of integers
 457 sub DUMP_ARRAY
 458 {
 459     my ($format,$default,@array) = @_;
 460     my $i, $ret = "    ";
 461     for ($i = 0; $i < $#array; $i++)
 462     {
 463         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 464         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 465     }
 466     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 467     return $ret;
 468 }
 469
 470 ################################################################
 471 # dump an SBCS mapping table
 472 sub DUMP_SBCS_TABLE
 473 {
 474     my ($codepage, $name) = @_;
 475     my $i;
 476
 477     # output the ascii->unicode table
 478
 479     printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 480     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 481
 482     # count the number of unicode->ascii subtables that contain something
 483
 484     my @filled = ();
 485     my $subtables = 1;
 486     for ($i = 0; $i < 65536; $i++)
 487     {
 488         next unless defined $uni2cp[$i];
 489         $filled[$i >> 8] = 1;
 490         $subtables++;
 491         $i |= 255;
 492     }
 493
 494     # output all the subtables into a single array
 495
 496     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 497     for ($i = 0; $i < 256; $i++)
 498     {
 499         next unless $filled[$i];
 500         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 501         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
 502     }
 503     printf OUTPUT "    /* defaults */\n";
 504     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($DEF_CHAR) x 256 );
 505
 506     # output a table of the offsets of the subtables in the previous array
 507
 508     my $pos = 0;
 509     my @offsets = ();
 510     for ($i = 0; $i < 256; $i++)
 511     {
 512         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 513         else { push @offsets, ($subtables-1) * 256; }
 514     }
 515     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 516     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 517
 518     # output the code page descriptor
 519
 520     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 521     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 522                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 523     printf OUTPUT "    cp2uni,\n";
 524     printf OUTPUT "    uni2cp_low,\n";
 525     printf OUTPUT "    uni2cp_high\n};\n";
 526 }
 527
 528
 529 ################################################################
 530 # dump a DBCS mapping table
 531 sub DUMP_DBCS_TABLE
 532 {
 533     my ($codepage, $name) = @_;
 534     my $i, $x, $y;
 535
 536     # build a list of lead bytes that are actually used
 537
 538     my @lblist = ();
 539     LBLOOP: for ($y = 0; $y <= $#lead_bytes; $y++)
 540     {
 541         my $base = $lead_bytes[$y] << 8;
 542         for ($x = 0; $x < 256; $x++)
 543         {
 544             if (defined $cp2uni[$base+$x])
 545             {
 546                 push @lblist,$lead_bytes[$y];
 547                 next LBLOOP;
 548             }
 549         }
 550     }
 551     my $unused = ($#lead_bytes > $#lblist);
 552
 553     # output the ascii->unicode table for the single byte chars
 554
 555     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 556     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 557
 558     # output the default table for unused lead bytes
 559
 560     if ($unused)
 561     {
 562         printf OUTPUT "    /* unused lead bytes */\n";
 563         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 564     }
 565
 566     # output the ascii->unicode table for each DBCS lead byte
 567
 568     for ($y = 0; $y <= $#lblist; $y++)
 569     {
 570         my $base = $lblist[$y] << 8;
 571         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
 572         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
 573         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
 574     }
 575
 576     # output the lead byte subtables offsets
 577
 578     my @offsets = ();
 579     for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
 580     for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
 581     if ($unused)
 582     {
 583         # increment all lead bytes offset to take into account the unused table
 584         for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
 585     }
 586     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
 587     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
 588
 589     # count the number of unicode->ascii subtables that contain something
 590
 591     my @filled = ();
 592     my $subtables = 1;
 593     for ($i = 0; $i < 65536; $i++)
 594     {
 595         next unless defined $uni2cp[$i];
 596         $filled[$i >> 8] = 1;
 597         $subtables++;
 598         $i |= 255;
 599     }
 600
 601     # output all the subtables into a single array
 602
 603     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
 604     for ($y = 0; $y < 256; $y++)
 605     {
 606         next unless $filled[$y];
 607         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
 608         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
 609     }
 610     printf OUTPUT "    /* defaults */\n";
 611     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 612
 613     # output a table of the offsets of the subtables in the previous array
 614
 615     my $pos = 0;
 616     my @offsets = ();
 617     for ($y = 0; $y < 256; $y++)
 618     {
 619         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
 620         else { push @offsets, ($subtables-1) * 256; }
 621     }
 622     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 623     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 624
 625     # output the code page descriptor
 626
 627     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
 628     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
 629                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 630     printf OUTPUT "    cp2uni,\n";
 631     printf OUTPUT "    cp2uni_leadbytes,\n";
 632     printf OUTPUT "    uni2cp_low,\n";
 633     printf OUTPUT "    uni2cp_high,\n";
 634     DUMP_LB_RANGES();
 635     printf OUTPUT "};\n";
 636 }
 637
 638
 639 ################################################################
 640 # dump the list of defined lead byte ranges
 641 sub DUMP_LB_RANGES
 642 {
 643     my @list = ();
 644     my $i = 0;
 645     foreach $i (@lead_bytes) { $list[$i] = 1; }
 646     my $on = 0;
 647     printf OUTPUT "    { ";
 648     for ($i = 0; $i < 256; $i++)
 649     {
 650         if ($on)
 651         {
 652             if (!defined $list[$i]) { printf OUTPUT "0x%02x, ", $i-1; $on = 0; }
 653         }
 654         else
 655         {
 656             if ($list[$i]) { printf OUTPUT "0x%02x, ", $i; $on = 1; }
 657         }
 658     }
 659     if ($on) { printf OUTPUT "0xff, "; }
 660     printf OUTPUT "0x00, 0x00 }\n";
 661 }
 662
 663
 664 ################################################################
 665 # dump the case mapping tables
 666 sub DUMP_CASE_MAPPINGS
 667 {
 668     open OUTPUT,">casemap.c" or die "Cannot create casemap.c";
 669     printf "Building casemap.c\n";
 670     printf OUTPUT "/* Unicode case mappings */\n";
 671     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 672     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 673
 674     DUMP_CASE_TABLE( "casemap_lower", @tolower_table );
 675     DUMP_CASE_TABLE( "casemap_upper", @toupper_table );
 676     close OUTPUT;
 677 }
 678
 679
 680 ################################################################
 681 # dump a case mapping table
 682 sub DUMP_CASE_TABLE
 683 {
 684     my ($name,@table) = @_;
 685
 686     # count the number of sub tables that contain something
 687
 688     my @filled = ();
 689     my $pos = 512;
 690     for ($i = 0; $i < 65536; $i++)
 691     {
 692         next unless defined $table[$i];
 693         $filled[$i >> 8] = $pos;
 694         $pos += 256;
 695         $i |= 255;
 696     }
 697     for ($i = 0; $i < 65536; $i++)
 698     {
 699         next unless defined $table[$i];
 700         $table[$i] = ($table[$i] - $i) & 0xffff;
 701     }
 702
 703     # dump the table
 704
 705     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $pos;
 706     printf OUTPUT "{\n    /* index */\n";
 707     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
 708     printf OUTPUT "    /* defaults */\n";
 709     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
 710     for ($i = 0; $i < 256; $i++)
 711     {
 712         next unless $filled[$i];
 713         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 714         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table[($i<<8) .. ($i<<8)+255] );
 715     }
 716     printf OUTPUT "\n};\n";
 717 }
 718
 719
 720 ################################################################
 721 # dump the ctype tables
 722 sub DUMP_CTYPE_TABLES
 723 {
 724     open OUTPUT,">wctype.c" or die "Cannot create casemap.c";
 725     printf "Building wctype.c\n";
 726     printf OUTPUT "/* Unicode ctype tables */\n";
 727     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 728     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 729
 730     my $i;
 731     my @array = (0) x 256;
 732
 733     # add the direction in the high 4 bits of the category
 734     for ($i = 0; $i < 65536; $i++)
 735     {
 736         $category_table[$i] |= $direction_table[$i] << 12;
 737     }
 738
 739     # try to merge table rows
 740     for ($row = 0; $row < 256; $row++)
 741     {
 742         my $rowtxt = sprintf "%04x" x 256, @category_table[($row<<8)..($row<<8)+255];
 743         if (defined($sequences{$rowtxt}))
 744         {
 745             # reuse an existing row
 746             $array[$row] = $sequences{$rowtxt};
 747         }
 748         else
 749         {
 750             # create a new row
 751             $sequences{$rowtxt} = $array[$row] = $#array + 1;
 752             push @array, @category_table[($row<<8)..($row<<8)+255];
 753         }
 754     }
 755
 756     printf OUTPUT "const unsigned short wctype_table[%d] =\n{\n", $#array+1;
 757     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
 758     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
 759
 760     close OUTPUT;
 761 }
 762
 763
 764 ################################################################
 765 # dump the char composition tables
 766 sub DUMP_COMPOSE_TABLES
 767 {
 768     open OUTPUT,">compose.c" or die "Cannot create compose.c";
 769     printf "Building compose.c\n";
 770     printf OUTPUT "/* Unicode char composition */\n";
 771     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 772     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 773
 774     ######### composition table
 775
 776     my @filled = ();
 777     foreach $i (@compose_table)
 778     {
 779         my @comp = @$i;
 780         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
 781     }
 782
 783     # count how many different second chars we have
 784
 785     for ($i = $count = 0; $i < 65536; $i++)
 786     {
 787         next unless defined $filled[$i];
 788         $count++;
 789     }
 790
 791     # build the table of second chars and offsets
 792
 793     my $pos = $count + 1;
 794     for ($i = 0; $i < 65536; $i++)
 795     {
 796         next unless defined $filled[$i];
 797         push @table, $i, $pos;
 798         $pos += @{$filled[$i]};
 799     }
 800     # terminator with last position
 801     push @table, 0, $pos;
 802     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
 803     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
 804
 805     # build the table of first chars and mappings
 806
 807     for ($i = 0; $i < 65536; $i++)
 808     {
 809         next unless defined $filled[$i];
 810         my @table = ();
 811         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
 812         for ($j = 0; $j <= $#list; $j++)
 813         {
 814             push @table, $list[$j][0], $list[$j][1];
 815         }
 816         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
 817     }
 818     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
 819
 820     ######### decomposition table
 821
 822     # first determine all the 16-char subsets that contain something
 823
 824     my @filled = (0) x 4096;
 825     my $pos = 16*2;  # for the null subset
 826     for ($i = 0; $i < 65536; $i++)
 827     {
 828         next unless defined $decomp_table[$i];
 829         $filled[$i >> 4] = $pos;
 830         $pos += 16*2;
 831         $i |= 15;
 832     }
 833     my $total = $pos;
 834
 835     # now count the 256-char subsets that contain something
 836
 837     my @filled_idx = (256) x 256;
 838     $pos = 256 + 16;
 839     for ($i = 0; $i < 4096; $i++)
 840     {
 841         next unless $filled[$i];
 842         $filled_idx[$i >> 4] = $pos;
 843         $pos += 16;
 844         $i |= 15;
 845     }
 846     my $null_offset = $pos;  # null mapping
 847     $total += $pos;
 848
 849     # add the index offsets to the subsets positions
 850
 851     for ($i = 0; $i < 4096; $i++)
 852     {
 853         next unless $filled[$i];
 854         $filled[$i] += $null_offset;
 855     }
 856
 857     # dump the main index
 858
 859     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
 860     printf OUTPUT "{\n    /* index */\n";
 861     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
 862     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
 863
 864     # dump the second-level indexes
 865
 866     for ($i = 0; $i < 256; $i++)
 867     {
 868         next unless ($filled_idx[$i] > 256);
 869         my @table = @filled[($i<<4)..($i<<4)+15];
 870         for ($j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
 871         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
 872         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
 873     }
 874
 875     # dump the 16-char subsets
 876
 877     printf OUTPUT ",\n    /* null mapping */\n";
 878     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
 879
 880     for ($i = 0; $i < 4096; $i++)
 881     {
 882         next unless $filled[$i];
 883         my @table = (0) x 32;
 884         for ($j = 0; $j < 16; $j++)
 885         {
 886             if (defined $decomp_table[($i<<4) + $j])
 887             {
 888                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
 889                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
 890             }
 891         }
 892         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
 893         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
 894     }
 895
 896     printf OUTPUT "\n};\n";
 897     close OUTPUT;
 898 }
 899
 900
 901 ################################################################
 902 # read an input file and generate the corresponding .c file
 903 sub HANDLE_FILE
 904 {
 905     my ($codepage,$filename,$comment) = @_;
 906
 907     # symbol codepage file is special
 908     if ($codepage == 42) { READ_SYMBOL_FILE($MAPPREFIX . $filename); }
 909     else { READ_FILE($MAPPREFIX . $filename); }
 910
 911     ADD_DEFAULT_MAPPINGS();
 912
 913     my $output = sprintf "c_%03d.c", $codepage;
 914     open OUTPUT,">$output" or die "Cannot create $output";
 915
 916     printf "Building %s from %s (%s)\n", $output, $filename, $comment;
 917
 918     # dump all tables
 919
 920     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
 921     printf OUTPUT "/* generated from %s */\n", $MAPPREFIX . $filename;
 922     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 923     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 924
 925     if ($#lead_bytes == -1) { DUMP_SBCS_TABLE( $codepage, $comment ); }
 926     else { DUMP_DBCS_TABLE( $codepage, $comment ); }
 927     close OUTPUT;
 928 }
 929
 930
 931 ################################################################
 932 # output the list of codepage tables into the cptable.c file
 933 sub OUTPUT_CPTABLE
 934 {
 935     @tables_decl = ();
 936
 937     foreach $file (@allfiles)
 938     {
 939         my ($codepage,$filename,$comment) = @$file;
 940         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
 941     }
 942
 943     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
 944     foreach $file (@allfiles)
 945     {
 946         my ($codepage,$filename,$comment) = @$file;
 947         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
 948     }
 949     push @tables_decl, "};";
 950     REPLACE_IN_FILE( "cptable.c", @tables_decl );
 951 }
 952
 953 ################################################################
 954 # replace the contents of a file between ### cpmap ### marks
 955
 956 sub REPLACE_IN_FILE
 957 {
 958     my $name = shift;
 959     my @data = @_;
 960     my @lines = ();
 961     open(FILE,$name) or die "Can't open $name";
 962     while (<FILE>)
 963     {
 964         push @lines, $_;
 965         last if /\#\#\# cpmap begin \#\#\#/;
 966     }
 967     push @lines, @data;
 968     while (<FILE>)
 969     {
 970         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
 971     }
 972     push @lines, <FILE>;
 973     open(FILE,">$name") or die "Can't modify $name";
 974     print FILE @lines;
 975     close(FILE);
 976 }