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     [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", "EUC-JP" ],
  83     [ 28591, "ISO8859/8859-1.TXT",                "ISO 8859-1 Latin 1" ],
  84     [ 28592, "ISO8859/8859-2.TXT",                "ISO 8859-2 Latin 2 (East European)" ],
  85     [ 28593, "ISO8859/8859-3.TXT",                "ISO 8859-3 Latin 3 (South European)" ],
  86     [ 28594, "ISO8859/8859-4.TXT",                "ISO 8859-4 Latin 4 (Baltic old)" ],
  87     [ 28595, "ISO8859/8859-5.TXT",                "ISO 8859-5 Cyrillic" ],
  88     [ 28596, "ISO8859/8859-6.TXT",                "ISO 8859-6 Arabic" ],
  89     [ 28597, "ISO8859/8859-7.TXT",                "ISO 8859-7 Greek" ],
  90     [ 28598, "ISO8859/8859-8.TXT",                "ISO 8859-8 Hebrew" ],
  91     [ 28599, "ISO8859/8859-9.TXT",                "ISO 8859-9 Latin 5 (Turkish)" ],
  92     [ 28600, "ISO8859/8859-10.TXT",               "ISO 8859-10 Latin 6 (Nordic)" ],
  93     [ 28603, "ISO8859/8859-13.TXT",               "ISO 8859-13 Latin 7 (Baltic)" ],
  94     [ 28604, "ISO8859/8859-14.TXT",               "ISO 8859-14 Latin 8 (Celtic)" ],
  95     [ 28605, "ISO8859/8859-15.TXT",               "ISO 8859-15 Latin 9 (Euro)" ],
  96     [ 28606, "ISO8859/8859-16.TXT",               "ISO 8859-16 Latin 10 (Balkan)" ]
  97 );
  98
  99
 100 %ctype =
 101 (
 102     "upper"  => 0x0001,
 103     "lower"  => 0x0002,
 104     "digit"  => 0x0004,
 105     "space"  => 0x0008,
 106     "punct"  => 0x0010,
 107     "cntrl"  => 0x0020,
 108     "blank"  => 0x0040,
 109     "xdigit" => 0x0080,
 110     "alpha"  => 0x0100
 111 );
 112
 113 %categories =
 114 (
 115     "Lu" => $ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
 116     "Ll" => $ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
 117     "Lt" => $ctype{"alpha"},    # Letter, Titlecase
 118     "Mn" => $ctype{"punct"},    # Mark, Non-Spacing
 119     "Mc" => $ctype{"punct"},    # Mark, Spacing Combining
 120     "Me" => $ctype{"punct"},    # Mark, Enclosing
 121     "Nd" => $ctype{"digit"},    # Number, Decimal Digit
 122     "Nl" => $ctype{"punct"},    # Number, Letter
 123     "No" => $ctype{"punct"},    # Number, Other
 124     "Zs" => $ctype{"space"},    # Separator, Space
 125     "Zl" => 0,                  # Separator, Line
 126     "Zp" => 0,                  # Separator, Paragraph
 127     "Cc" => $ctype{"cntrl"},    # Other, Control
 128     "Cf" => 0,                  # Other, Format
 129     "Cs" => 0,                  # Other, Surrogate
 130     "Co" => 0,                  # Other, Private Use
 131     "Cn" => 0,                  # Other, Not Assigned
 132     "Lm" => $ctype{"punct"},    # Letter, Modifier
 133     "Lo" => $ctype{"alpha"},    # Letter, Other
 134     "Pc" => $ctype{"punct"},    # Punctuation, Connector
 135     "Pd" => $ctype{"punct"},    # Punctuation, Dash
 136     "Ps" => $ctype{"punct"},    # Punctuation, Open
 137     "Pe" => $ctype{"punct"},    # Punctuation, Close
 138     "Pi" => $ctype{"punct"},    # Punctuation, Initial quote
 139     "Pf" => $ctype{"punct"},    # Punctuation, Final quote
 140     "Po" => $ctype{"punct"},    # Punctuation, Other
 141     "Sm" => $ctype{"punct"},    # Symbol, Math
 142     "Sc" => $ctype{"punct"},    # Symbol, Currency
 143     "Sk" => $ctype{"punct"},    # Symbol, Modifier
 144     "So" => $ctype{"punct"}     # Symbol, Other
 145 );
 146
 147 # a few characters need additional categories that cannot be determined automatically
 148 %special_categories =
 149 (
 150     "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
 151                   0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ],
 152     "space"  => [ 0x09..0x0d, 0xfeff ],
 153     "blank"  => [ 0x09, 0x20, 0xa0, 0xfeff ]
 154 );
 155
 156 %directions =
 157 (
 158     "L"   => 1,    # Left-to-Right
 159     "LRE" => 11,   # Left-to-Right Embedding
 160     "LRO" => 11,   # Left-to-Right Override
 161     "R"   => 2,    # Right-to-Left
 162     "AL"  => 2,    # Right-to-Left Arabic
 163     "RLE" => 11,   # Right-to-Left Embedding
 164     "RLO" => 11,   # Right-to-Left Override
 165     "PDF" => 11,   # Pop Directional Format
 166     "EN"  => 3,    # European Number
 167     "ES"  => 4,    # European Number Separator
 168     "ET"  => 5,    # European Number Terminator
 169     "AN"  => 6,    # Arabic Number
 170     "CS"  => 7,    # Common Number Separator
 171     "NSM" => 0,    # Non-Spacing Mark
 172     "BN"  => 0,    # Boundary Neutral
 173     "B"   => 8,    # Paragraph Separator
 174     "S"   => 9,    # Segment Separator
 175     "WS"  => 10,   # Whitespace
 176     "ON"  => 11    # Other Neutrals
 177 );
 178
 179
 180 ################################################################
 181 # main routine
 182
 183 READ_DEFAULTS();
 184 DUMP_CASE_MAPPINGS();
 185 DUMP_COMPOSE_TABLES();
 186 DUMP_CTYPE_TABLES();
 187
 188 foreach $file (@allfiles) { HANDLE_FILE( @$file ); }
 189
 190 OUTPUT_CPTABLE();
 191
 192 exit(0);
 193
 194
 195 ################################################################
 196 # read in the defaults file
 197 sub READ_DEFAULTS
 198 {
 199     @unicode_defaults = ();
 200     @unicode_aliases = ();
 201     @tolower_table = ();
 202     @toupper_table = ();
 203     @category_table = ();
 204     @direction_table = ();
 205     @decomp_table = ();
 206     @compose_table = ();
 207
 208     # first setup a few default mappings
 209
 210     open DEFAULTS or die "Cannot open $DEFAULTS";
 211     print "Loading $DEFAULTS\n";
 212     while (<DEFAULTS>)
 213     {
 214         next if /^\#/;  # skip comments
 215         next if /^$/;  # skip empty lines
 216         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 217         {
 218             my @src = map hex, split /,/,$1;
 219             my $dst = $4;
 220             my $comment = $5;
 221             if ($#src > 0) { push @unicode_aliases, \@src; }
 222             next if ($dst eq "none");
 223             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 224             foreach $src (@src)
 225             {
 226                 die "Duplicate value" if defined($unicode_defaults[$src]);
 227                 $unicode_defaults[$src] = $dst;
 228             }
 229             next;
 230         }
 231         die "Unrecognized line $_\n";
 232     }
 233
 234     # now build mappings from the decomposition field of the Unicode database
 235
 236     open UNICODEDATA or die "Cannot open $UNICODEDATA";
 237     print "Loading $UNICODEDATA\n";
 238     while (<UNICODEDATA>)
 239     {
 240         # Decode the fields ...
 241         ($code, $name, $cat, $comb, $bidi,
 242          $decomp, $dec, $dig, $num, $mirror,
 243          $oldname, $comment, $upper, $lower, $title) = split /;/;
 244
 245         my $src = hex $code;
 246
 247         die "unknown category $cat" unless defined $categories{$cat};
 248         die "unknown directionality $bidi" unless defined $directions{$bidi};
 249
 250         $uniname[$src] = $name;
 251         $category_table[$src] = $categories{$cat};
 252         $direction_table[$src] = $directions{$bidi};
 253
 254         if ($lower ne "")
 255         {
 256             $tolower_table[$src] = hex $lower;
 257             $category_table[$src] |= $ctype{"upper"}|$ctype{"alpha"};
 258         }
 259         if ($upper ne "")
 260         {
 261             $toupper_table[$src] = hex $upper;
 262             $category_table[$src] |= $ctype{"lower"}|$ctype{"alpha"};
 263         }
 264         if ($dec ne "")
 265         {
 266             $category_table[$src] |= $ctype{"digit"};
 267         }
 268
 269         # copy the category and direction for everything between First/Last pairs
 270         if ($name =~ /, First>/) { $start = $src; }
 271         if ($name =~ /, Last>/)
 272         {
 273             while ($start < $src)
 274             {
 275                 $category_table[$start] = $category_table[$src];
 276                 $direction_table[$start] = $direction_table[$src];
 277                 $start++;
 278             }
 279         }
 280
 281         next if $decomp eq "";  # no decomposition, skip it
 282
 283         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 284         {
 285             # decomposition of the form "<foo> 1234" -> use char if type is known
 286             next unless ($1 eq "font" ||
 287                          $1 eq "noBreak" ||
 288                          $1 eq "circle" ||
 289                          $1 eq "super" ||
 290                          $1 eq "sub" ||
 291                          $1 eq "wide" ||
 292                          $1 eq "narrow" ||
 293                          $1 eq "compat" ||
 294                          $1 eq "small");
 295             $dst = hex $2;
 296         }
 297         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 298         {
 299             # decomposition "<compat> 0020 1234" -> combining accent
 300             $dst = hex $1;
 301         }
 302         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 303         {
 304             # decomposition contains only char values without prefix -> use first char
 305             $dst = hex $1;
 306             $category_table[$src] |= $category_table[$dst];
 307             # store decomposition if it contains two chars
 308             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 309             {
 310                 $decomp_table[$src] = [ hex $1, hex $2 ];
 311                 push @compose_table, [ hex $1, hex $2, $src ];
 312             }
 313         }
 314         else
 315         {
 316             next;
 317         }
 318
 319         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 320
 321         # check for loops
 322         for ($i = $dst; ; $i = $unicode_defaults[$i])
 323         {
 324             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 325             last unless defined($unicode_defaults[$i]);
 326         }
 327         $unicode_defaults[$src] = $dst;
 328     }
 329
 330     # patch the category of some special characters
 331
 332     foreach $cat (keys %special_categories)
 333     {
 334         my $flag = $ctype{$cat};
 335         foreach $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 336     }
 337 }
 338
 339
 340 ################################################################
 341 # parse the input file
 342 sub READ_FILE
 343 {
 344     my $name = shift;
 345     open INPUT,$name or die "Cannot open $name";
 346     @cp2uni = ();
 347     @lead_bytes = ();
 348     @uni2cp = ();
 349
 350     while (<INPUT>)
 351     {
 352         next if /^\#/;  # skip comments
 353         next if /^$/;  # skip empty lines
 354         next if /\x1a/;  # skip ^Z
 355         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 356
 357         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 358         {
 359             $cp = hex $1;
 360             push @lead_bytes,$cp;
 361             $cp2uni[$cp] = 0;
 362             next;
 363         }
 364         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 365         {
 366             $cp = hex $1;
 367             $uni = hex $2;
 368             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 369             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 370             next;
 371         }
 372         die "$name: Unrecognized line $_\n";
 373     }
 374 }
 375
 376
 377 ################################################################
 378 # parse the symbol.txt file, since its syntax is different from the other ones
 379 sub READ_SYMBOL_FILE
 380 {
 381     my $name = shift;
 382     open INPUT,$name or die "Cannot open $name";
 383     @cp2uni = ();
 384     @lead_bytes = ();
 385     @uni2cp = ();
 386
 387     while (<INPUT>)
 388     {
 389         next if /^\#/;  # skip comments
 390         next if /^$/;  # skip empty lines
 391         next if /\x1a/;  # skip ^Z
 392         if (/^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)\s+(\#.*)?/)
 393         {
 394             $uni = hex $1;
 395             $cp = hex $2;
 396             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 397             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 398             next;
 399         }
 400         die "$name: Unrecognized line $_\n";
 401     }
 402 }
 403
 404
 405 ################################################################
 406 # build EUC-JP table from the JIS 0208 file
 407 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 408 # but this would require 3-byte DBCS characters
 409 sub READ_JIS0208_FILE
 410 {
 411     my $name = shift;
 412     @cp2uni = ();
 413     @lead_bytes = ();
 414     @uni2cp = ();
 415
 416     # ASCII chars
 417     for ($i = 0x00; $i <= 0x7f; $i++)
 418     {
 419         $cp2uni[$i] = $i;
 420         $uni2cp[$i] = $i;
 421     }
 422
 423     # JIS X 0201 right plane
 424     for ($i = 0xa1; $i <= 0xdf; $i++)
 425     {
 426         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 427         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 428     }
 429
 430     # lead bytes
 431     foreach $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 432     {
 433         push @lead_bytes,$i;
 434         $cp2uni[$i] = 0;
 435     }
 436
 437     # undefined chars
 438     foreach $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 439     {
 440         $cp2uni[$i] = $DEF_CHAR;
 441     }
 442
 443     # Shift-JIS compatibility
 444     $uni2cp[0x00a5] = 0x5c;
 445     $uni2cp[0x203e] = 0x7e;
 446
 447     # Fix backslash conversion
 448     $cp2uni[0xa1c0] = 0xff3c;
 449     $uni2cp[0xff3c] = 0xa1c0;
 450
 451     open INPUT, "$name" or die "Cannot open $name";
 452     while (<INPUT>)
 453     {
 454         next if /^\#/;  # skip comments
 455         next if /^$/;  # skip empty lines
 456         next if /\x1a/;  # skip ^Z
 457         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 458         {
 459             $cp = 0x8080 + hex $1;
 460             $uni = hex $2;
 461             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 462             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 463             next;
 464         }
 465         die "$name: Unrecognized line $_\n";
 466     }
 467 }
 468
 469
 470 ################################################################
 471 # add default mappings once the file had been read
 472 sub ADD_DEFAULT_MAPPINGS
 473 {
 474     # Apply aliases
 475
 476     foreach $alias (@unicode_aliases)
 477     {
 478         my $target = undef;
 479         foreach $src (@$alias)
 480         {
 481             if (defined($uni2cp[$src]))
 482             {
 483                 $target = $uni2cp[$src];
 484                 last;
 485             }
 486         }
 487         next unless defined($target);
 488
 489         # At least one char of the alias set is defined, set the others to the same value
 490         foreach $src (@$alias)
 491         {
 492             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 493         }
 494     }
 495
 496     # For every src -> target mapping in the defaults table,
 497     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 498
 499     for ($src = 0; $src < 65536; $src++)
 500     {
 501         next if defined($uni2cp[$src]);  # source has a definition already
 502         next unless defined($unicode_defaults[$src]);  # no default for this char
 503         my $target = $unicode_defaults[$src];
 504
 505         # do a recursive mapping until we find a target char that is defined
 506         while (!defined($uni2cp[$target]) &&
 507                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 508
 509         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 510     }
 511
 512     # Add an identity mapping for all undefined chars
 513
 514     for ($i = 0; $i < 256; $i++)
 515     {
 516         next if defined($cp2uni[$i]);
 517         next if defined($uni2cp[$i]);
 518         $cp2uni[$i] = $uni2cp[$i] = $i;
 519     }
 520 }
 521
 522 ################################################################
 523 # dump an array of integers
 524 sub DUMP_ARRAY
 525 {
 526     my ($format,$default,@array) = @_;
 527     my $i, $ret = "    ";
 528     for ($i = 0; $i < $#array; $i++)
 529     {
 530         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 531         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 532     }
 533     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 534     return $ret;
 535 }
 536
 537 ################################################################
 538 # dump an SBCS mapping table
 539 sub DUMP_SBCS_TABLE
 540 {
 541     my ($codepage, $name) = @_;
 542     my $i;
 543
 544     # output the ascii->unicode table
 545
 546     printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 547     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 548
 549     # count the number of unicode->ascii subtables that contain something
 550
 551     my @filled = ();
 552     my $subtables = 1;
 553     for ($i = 0; $i < 65536; $i++)
 554     {
 555         next unless defined $uni2cp[$i];
 556         $filled[$i >> 8] = 1;
 557         $subtables++;
 558         $i |= 255;
 559     }
 560
 561     # output all the subtables into a single array
 562
 563     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 564     for ($i = 0; $i < 256; $i++)
 565     {
 566         next unless $filled[$i];
 567         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 568         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
 569     }
 570     printf OUTPUT "    /* defaults */\n";
 571     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($DEF_CHAR) x 256 );
 572
 573     # output a table of the offsets of the subtables in the previous array
 574
 575     my $pos = 0;
 576     my @offsets = ();
 577     for ($i = 0; $i < 256; $i++)
 578     {
 579         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 580         else { push @offsets, ($subtables-1) * 256; }
 581     }
 582     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 583     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 584
 585     # output the code page descriptor
 586
 587     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 588     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 589                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 590     printf OUTPUT "    cp2uni,\n";
 591     printf OUTPUT "    uni2cp_low,\n";
 592     printf OUTPUT "    uni2cp_high\n};\n";
 593 }
 594
 595
 596 ################################################################
 597 # dump a DBCS mapping table
 598 sub DUMP_DBCS_TABLE
 599 {
 600     my ($codepage, $name) = @_;
 601     my $i, $x, $y;
 602
 603     # build a list of lead bytes that are actually used
 604
 605     my @lblist = ();
 606     LBLOOP: for ($y = 0; $y <= $#lead_bytes; $y++)
 607     {
 608         my $base = $lead_bytes[$y] << 8;
 609         for ($x = 0; $x < 256; $x++)
 610         {
 611             if (defined $cp2uni[$base+$x])
 612             {
 613                 push @lblist,$lead_bytes[$y];
 614                 next LBLOOP;
 615             }
 616         }
 617     }
 618     my $unused = ($#lead_bytes > $#lblist);
 619
 620     # output the ascii->unicode table for the single byte chars
 621
 622     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 623     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 624
 625     # output the default table for unused lead bytes
 626
 627     if ($unused)
 628     {
 629         printf OUTPUT "    /* unused lead bytes */\n";
 630         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 631     }
 632
 633     # output the ascii->unicode table for each DBCS lead byte
 634
 635     for ($y = 0; $y <= $#lblist; $y++)
 636     {
 637         my $base = $lblist[$y] << 8;
 638         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
 639         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
 640         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
 641     }
 642
 643     # output the lead byte subtables offsets
 644
 645     my @offsets = ();
 646     for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
 647     for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
 648     if ($unused)
 649     {
 650         # increment all lead bytes offset to take into account the unused table
 651         for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
 652     }
 653     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
 654     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
 655
 656     # count the number of unicode->ascii subtables that contain something
 657
 658     my @filled = ();
 659     my $subtables = 1;
 660     for ($i = 0; $i < 65536; $i++)
 661     {
 662         next unless defined $uni2cp[$i];
 663         $filled[$i >> 8] = 1;
 664         $subtables++;
 665         $i |= 255;
 666     }
 667
 668     # output all the subtables into a single array
 669
 670     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
 671     for ($y = 0; $y < 256; $y++)
 672     {
 673         next unless $filled[$y];
 674         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
 675         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
 676     }
 677     printf OUTPUT "    /* defaults */\n";
 678     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 679
 680     # output a table of the offsets of the subtables in the previous array
 681
 682     my $pos = 0;
 683     my @offsets = ();
 684     for ($y = 0; $y < 256; $y++)
 685     {
 686         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
 687         else { push @offsets, ($subtables-1) * 256; }
 688     }
 689     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 690     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 691
 692     # output the code page descriptor
 693
 694     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
 695     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
 696                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 697     printf OUTPUT "    cp2uni,\n";
 698     printf OUTPUT "    cp2uni_leadbytes,\n";
 699     printf OUTPUT "    uni2cp_low,\n";
 700     printf OUTPUT "    uni2cp_high,\n";
 701     DUMP_LB_RANGES();
 702     printf OUTPUT "};\n";
 703 }
 704
 705
 706 ################################################################
 707 # dump the list of defined lead byte ranges
 708 sub DUMP_LB_RANGES
 709 {
 710     my @list = ();
 711     my $i = 0;
 712     foreach $i (@lead_bytes) { $list[$i] = 1; }
 713     my $on = 0;
 714     printf OUTPUT "    { ";
 715     for ($i = 0; $i < 256; $i++)
 716     {
 717         if ($on)
 718         {
 719             if (!defined $list[$i]) { printf OUTPUT "0x%02x, ", $i-1; $on = 0; }
 720         }
 721         else
 722         {
 723             if ($list[$i]) { printf OUTPUT "0x%02x, ", $i; $on = 1; }
 724         }
 725     }
 726     if ($on) { printf OUTPUT "0xff, "; }
 727     printf OUTPUT "0x00, 0x00 }\n";
 728 }
 729
 730
 731 ################################################################
 732 # dump the case mapping tables
 733 sub DUMP_CASE_MAPPINGS
 734 {
 735     open OUTPUT,">casemap.c" or die "Cannot create casemap.c";
 736     printf "Building casemap.c\n";
 737     printf OUTPUT "/* Unicode case mappings */\n";
 738     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 739     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 740
 741     DUMP_CASE_TABLE( "casemap_lower", @tolower_table );
 742     DUMP_CASE_TABLE( "casemap_upper", @toupper_table );
 743     close OUTPUT;
 744 }
 745
 746
 747 ################################################################
 748 # dump a case mapping table
 749 sub DUMP_CASE_TABLE
 750 {
 751     my ($name,@table) = @_;
 752
 753     # count the number of sub tables that contain something
 754
 755     my @filled = ();
 756     my $pos = 512;
 757     for ($i = 0; $i < 65536; $i++)
 758     {
 759         next unless defined $table[$i];
 760         $filled[$i >> 8] = $pos;
 761         $pos += 256;
 762         $i |= 255;
 763     }
 764     for ($i = 0; $i < 65536; $i++)
 765     {
 766         next unless defined $table[$i];
 767         $table[$i] = ($table[$i] - $i) & 0xffff;
 768     }
 769
 770     # dump the table
 771
 772     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $pos;
 773     printf OUTPUT "{\n    /* index */\n";
 774     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
 775     printf OUTPUT "    /* defaults */\n";
 776     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
 777     for ($i = 0; $i < 256; $i++)
 778     {
 779         next unless $filled[$i];
 780         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 781         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table[($i<<8) .. ($i<<8)+255] );
 782     }
 783     printf OUTPUT "\n};\n";
 784 }
 785
 786
 787 ################################################################
 788 # dump the ctype tables
 789 sub DUMP_CTYPE_TABLES
 790 {
 791     open OUTPUT,">wctype.c" or die "Cannot create casemap.c";
 792     printf "Building wctype.c\n";
 793     printf OUTPUT "/* Unicode ctype tables */\n";
 794     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 795     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 796
 797     my $i;
 798     my @array = (0) x 256;
 799
 800     # add the direction in the high 4 bits of the category
 801     for ($i = 0; $i < 65536; $i++)
 802     {
 803         $category_table[$i] |= $direction_table[$i] << 12;
 804     }
 805
 806     # try to merge table rows
 807     for ($row = 0; $row < 256; $row++)
 808     {
 809         my $rowtxt = sprintf "%04x" x 256, @category_table[($row<<8)..($row<<8)+255];
 810         if (defined($sequences{$rowtxt}))
 811         {
 812             # reuse an existing row
 813             $array[$row] = $sequences{$rowtxt};
 814         }
 815         else
 816         {
 817             # create a new row
 818             $sequences{$rowtxt} = $array[$row] = $#array + 1;
 819             push @array, @category_table[($row<<8)..($row<<8)+255];
 820         }
 821     }
 822
 823     printf OUTPUT "const unsigned short wctype_table[%d] =\n{\n", $#array+1;
 824     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
 825     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
 826
 827     close OUTPUT;
 828 }
 829
 830
 831 ################################################################
 832 # dump the char composition tables
 833 sub DUMP_COMPOSE_TABLES
 834 {
 835     open OUTPUT,">compose.c" or die "Cannot create compose.c";
 836     printf "Building compose.c\n";
 837     printf OUTPUT "/* Unicode char composition */\n";
 838     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 839     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 840
 841     ######### composition table
 842
 843     my @filled = ();
 844     foreach $i (@compose_table)
 845     {
 846         my @comp = @$i;
 847         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
 848     }
 849
 850     # count how many different second chars we have
 851
 852     for ($i = $count = 0; $i < 65536; $i++)
 853     {
 854         next unless defined $filled[$i];
 855         $count++;
 856     }
 857
 858     # build the table of second chars and offsets
 859
 860     my $pos = $count + 1;
 861     for ($i = 0; $i < 65536; $i++)
 862     {
 863         next unless defined $filled[$i];
 864         push @table, $i, $pos;
 865         $pos += @{$filled[$i]};
 866     }
 867     # terminator with last position
 868     push @table, 0, $pos;
 869     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
 870     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
 871
 872     # build the table of first chars and mappings
 873
 874     for ($i = 0; $i < 65536; $i++)
 875     {
 876         next unless defined $filled[$i];
 877         my @table = ();
 878         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
 879         for ($j = 0; $j <= $#list; $j++)
 880         {
 881             push @table, $list[$j][0], $list[$j][1];
 882         }
 883         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
 884     }
 885     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
 886
 887     ######### decomposition table
 888
 889     # first determine all the 16-char subsets that contain something
 890
 891     my @filled = (0) x 4096;
 892     my $pos = 16*2;  # for the null subset
 893     for ($i = 0; $i < 65536; $i++)
 894     {
 895         next unless defined $decomp_table[$i];
 896         $filled[$i >> 4] = $pos;
 897         $pos += 16*2;
 898         $i |= 15;
 899     }
 900     my $total = $pos;
 901
 902     # now count the 256-char subsets that contain something
 903
 904     my @filled_idx = (256) x 256;
 905     $pos = 256 + 16;
 906     for ($i = 0; $i < 4096; $i++)
 907     {
 908         next unless $filled[$i];
 909         $filled_idx[$i >> 4] = $pos;
 910         $pos += 16;
 911         $i |= 15;
 912     }
 913     my $null_offset = $pos;  # null mapping
 914     $total += $pos;
 915
 916     # add the index offsets to the subsets positions
 917
 918     for ($i = 0; $i < 4096; $i++)
 919     {
 920         next unless $filled[$i];
 921         $filled[$i] += $null_offset;
 922     }
 923
 924     # dump the main index
 925
 926     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
 927     printf OUTPUT "{\n    /* index */\n";
 928     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
 929     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
 930
 931     # dump the second-level indexes
 932
 933     for ($i = 0; $i < 256; $i++)
 934     {
 935         next unless ($filled_idx[$i] > 256);
 936         my @table = @filled[($i<<4)..($i<<4)+15];
 937         for ($j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
 938         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
 939         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
 940     }
 941
 942     # dump the 16-char subsets
 943
 944     printf OUTPUT ",\n    /* null mapping */\n";
 945     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
 946
 947     for ($i = 0; $i < 4096; $i++)
 948     {
 949         next unless $filled[$i];
 950         my @table = (0) x 32;
 951         for ($j = 0; $j < 16; $j++)
 952         {
 953             if (defined $decomp_table[($i<<4) + $j])
 954             {
 955                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
 956                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
 957             }
 958         }
 959         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
 960         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
 961     }
 962
 963     printf OUTPUT "\n};\n";
 964     close OUTPUT;
 965 }
 966
 967
 968 ################################################################
 969 # read an input file and generate the corresponding .c file
 970 sub HANDLE_FILE
 971 {
 972     my ($codepage,$filename,$comment) = @_;
 973
 974     # symbol codepage file is special
 975     if ($codepage == 42) { READ_SYMBOL_FILE($MAPPREFIX . $filename); }
 976     elsif ($codepage == 20932) { READ_JIS0208_FILE($MAPPREFIX . $filename); }
 977     else { READ_FILE($MAPPREFIX . $filename); }
 978
 979     ADD_DEFAULT_MAPPINGS();
 980
 981     my $output = sprintf "c_%03d.c", $codepage;
 982     open OUTPUT,">$output" or die "Cannot create $output";
 983
 984     printf "Building %s from %s (%s)\n", $output, $filename, $comment;
 985
 986     # dump all tables
 987
 988     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
 989     printf OUTPUT "/* generated from %s */\n", $MAPPREFIX . $filename;
 990     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 991     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 992
 993     if ($#lead_bytes == -1) { DUMP_SBCS_TABLE( $codepage, $comment ); }
 994     else { DUMP_DBCS_TABLE( $codepage, $comment ); }
 995     close OUTPUT;
 996 }
 997
 998
 999 ################################################################
1000 # output the list of codepage tables into the cptable.c file
1001 sub OUTPUT_CPTABLE
1002 {
1003     @tables_decl = ();
1004
1005     foreach $file (@allfiles)
1006     {
1007         my ($codepage,$filename,$comment) = @$file;
1008         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
1009     }
1010
1011     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
1012     foreach $file (@allfiles)
1013     {
1014         my ($codepage,$filename,$comment) = @$file;
1015         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
1016     }
1017     push @tables_decl, "};";
1018     REPLACE_IN_FILE( "cptable.c", @tables_decl );
1019 }
1020
1021 ################################################################
1022 # replace the contents of a file between ### cpmap ### marks
1023
1024 sub REPLACE_IN_FILE
1025 {
1026     my $name = shift;
1027     my @data = @_;
1028     my @lines = ();
1029     open(FILE,$name) or die "Can't open $name";
1030     while (<FILE>)
1031     {
1032         push @lines, $_;
1033         last if /\#\#\# cpmap begin \#\#\#/;
1034     }
1035     push @lines, @data;
1036     while (<FILE>)
1037     {
1038         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
1039     }
1040     push @lines, <FILE>;
1041     open(FILE,">$name") or die "Can't modify $name";
1042     print FILE @lines;
1043     close(FILE);
1044 }