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
   8 # base directory for ftp.unicode.org files
   9 $BASEDIR = "ftp.unicode.org/Public/";
  10 $MAPPREFIX = $BASEDIR . "MAPPINGS/";
  11
  12 # UnicodeData file
  13 $UNICODEDATA = $BASEDIR . "UNIDATA/UnicodeData.txt";
  14
  15 # Defaults mapping
  16 $DEFAULTS = "./defaults";
  17
  18 # Default char for undefined mappings
  19 $DEF_CHAR = ord '?';
  20
  21 @allfiles =
  22 (
  23     [ 37,    "VENDORS/MICSFT/EBCDIC/CP037.TXT",   "IBM EBCDIC US Canada" ],
  24     [ 42,    "VENDORS/ADOBE/symbol.txt",          "Symbol" ],
  25     [ 424,   "VENDORS/MISC/CP424.TXT",            "IBM EBCDIC Hebrew" ],
  26     [ 437,   "VENDORS/MICSFT/PC/CP437.TXT",       "OEM United States" ],
  27     [ 500,   "VENDORS/MICSFT/EBCDIC/CP500.TXT",   "IBM EBCDIC International" ],
  28     [ 737,   "VENDORS/MICSFT/PC/CP737.TXT",       "OEM Greek 437G" ],
  29     [ 775,   "VENDORS/MICSFT/PC/CP775.TXT",       "OEM Baltic" ],
  30     [ 850,   "VENDORS/MICSFT/PC/CP850.TXT",       "OEM Multilingual Latin 1" ],
  31     [ 852,   "VENDORS/MICSFT/PC/CP852.TXT",       "OEM Slovak Latin 2" ],
  32     [ 855,   "VENDORS/MICSFT/PC/CP855.TXT",       "OEM Cyrillic" ],
  33     [ 856,   "VENDORS/MISC/CP856.TXT",            "Hebrew PC" ],
  34     [ 857,   "VENDORS/MICSFT/PC/CP857.TXT",       "OEM Turkish" ],
  35     [ 860,   "VENDORS/MICSFT/PC/CP860.TXT",       "OEM Portuguese" ],
  36     [ 861,   "VENDORS/MICSFT/PC/CP861.TXT",       "OEM Icelandic" ],
  37     [ 862,   "VENDORS/MICSFT/PC/CP862.TXT",       "OEM Hebrew" ],
  38     [ 863,   "VENDORS/MICSFT/PC/CP863.TXT",       "OEM Canadian French" ],
  39     [ 864,   "VENDORS/MICSFT/PC/CP864.TXT",       "OEM Arabic" ],
  40     [ 865,   "VENDORS/MICSFT/PC/CP865.TXT",       "OEM Nordic" ],
  41     [ 866,   "VENDORS/MICSFT/PC/CP866.TXT",       "OEM Russian" ],
  42     [ 869,   "VENDORS/MICSFT/PC/CP869.TXT",       "OEM Greek" ],
  43     [ 874,   "VENDORS/MICSFT/PC/CP874.TXT",       "ANSI/OEM Thai" ],
  44     [ 875,   "VENDORS/MICSFT/EBCDIC/CP875.TXT",   "IBM EBCDIC Greek" ],
  45     [ 878,   "VENDORS/MISC/KOI8-R.TXT",           "Russian KOI8" ],
  46     [ 932,   "VENDORS/MICSFT/WINDOWS/CP932.TXT",  "ANSI/OEM Japanese Shift-JIS" ],
  47     [ 936,   "VENDORS/MICSFT/WINDOWS/CP936.TXT",  "ANSI/OEM Simplified Chinese GBK" ],
  48     [ 949,   "VENDORS/MICSFT/WINDOWS/CP949.TXT",  "ANSI/OEM Korean Unified Hangul" ],
  49     [ 950,   "VENDORS/MICSFT/WINDOWS/CP950.TXT",  "ANSI/OEM Traditional Chinese Big5" ],
  50     [ 1006,  "VENDORS/MISC/CP1006.TXT",           "IBM Arabic" ],
  51     [ 1026,  "VENDORS/MICSFT/EBCDIC/CP1026.TXT",  "IBM EBCDIC Latin 5 Turkish" ],
  52     [ 1250,  "VENDORS/MICSFT/WINDOWS/CP1250.TXT", "ANSI Eastern Europe" ],
  53     [ 1251,  "VENDORS/MICSFT/WINDOWS/CP1251.TXT", "ANSI Cyrillic" ],
  54     [ 1252,  "VENDORS/MICSFT/WINDOWS/CP1252.TXT", "ANSI Latin 1" ],
  55     [ 1253,  "VENDORS/MICSFT/WINDOWS/CP1253.TXT", "ANSI Greek" ],
  56     [ 1254,  "VENDORS/MICSFT/WINDOWS/CP1254.TXT", "ANSI Turkish" ],
  57     [ 1255,  "VENDORS/MICSFT/WINDOWS/CP1255.TXT", "ANSI Hebrew" ],
  58     [ 1256,  "VENDORS/MICSFT/WINDOWS/CP1256.TXT", "ANSI Arabic" ],
  59     [ 1257,  "VENDORS/MICSFT/WINDOWS/CP1257.TXT", "ANSI Baltic" ],
  60     [ 1258,  "VENDORS/MICSFT/WINDOWS/CP1258.TXT", "ANSI/OEM Viet Nam" ],
  61     [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT",      "Mac Roman" ],
  62     [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT",      "Mac Greek" ],
  63     [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT",   "Mac Cyrillic" ],
  64     [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT",     "Mac Latin 2" ],
  65     [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT",    "Mac Icelandic" ],
  66     [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT",    "Mac Turkish" ],
  67     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           "Russian KOI8" ],
  68     [ 28591, "ISO8859/8859-1.TXT",                "ISO 8859-1 Latin 1" ],
  69     [ 28592, "ISO8859/8859-2.TXT",                "ISO 8859-2 Eastern Europe" ],
  70     [ 28593, "ISO8859/8859-3.TXT",                "ISO 8859-3 Turkish" ],
  71     [ 28594, "ISO8859/8859-4.TXT",                "ISO 8859-4 Baltic" ],
  72     [ 28595, "ISO8859/8859-5.TXT",                "ISO 8859-5 Cyrillic" ],
  73     [ 28596, "ISO8859/8859-6.TXT",                "ISO 8859-6 Arabic" ],
  74     [ 28597, "ISO8859/8859-7.TXT",                "ISO 8859-7 Greek" ],
  75     [ 28598, "ISO8859/8859-8.TXT",                "ISO 8859-8 Hebrew" ],
  76     [ 28599, "ISO8859/8859-9.TXT",                "ISO 8859-9 Latin 5" ]
  77 );
  78
  79 ################################################################
  80 # main routine
  81
  82 READ_DEFAULTS();
  83 DUMP_CASE_MAPPINGS();
  84
  85 foreach $file (@allfiles) { HANDLE_FILE( @$file ); }
  86
  87 OUTPUT_CPTABLE();
  88
  89 exit(0);
  90
  91
  92 ################################################################
  93 # read in the defaults file
  94 sub READ_DEFAULTS
  95 {
  96     @unicode_defaults = ();
  97     @unicode_aliases = ();
  98     @tolower_table = ();
  99     @toupper_table = ();
 100
 101     # first setup a few default mappings
 102
 103     open DEFAULTS or die "Cannot open $DEFAULTS";
 104     print "Loading $DEFAULTS\n";
 105     while (<DEFAULTS>)
 106     {
 107         next if /^\#/;  # skip comments
 108         next if /^$/;  # skip empty lines
 109         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 110         {
 111             my @src = map hex, split /,/,$1;
 112             my $dst = $4;
 113             my $comment = $5;
 114             if ($#src > 0) { push @unicode_aliases, \@src; }
 115             next if ($dst eq "none");
 116             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 117             foreach $src (@src)
 118             {
 119                 die "Duplicate value" if defined($unicode_defaults[$src]);
 120                 $unicode_defaults[$src] = $dst;
 121             }
 122             next;
 123         }
 124         die "Unrecognized line $_\n";
 125     }
 126
 127     # now build mappings from the decomposition field of the Unicode database
 128
 129     open UNICODEDATA or die "Cannot open $UNICODEDATA";
 130     print "Loading $UNICODEDATA\n";
 131     while (<UNICODEDATA>)
 132     {
 133         # Decode the fields ...
 134         ($code, $name, $cat, $comb, $bidi,
 135          $decomp, $dec, $dig, $num, $mirror,
 136          $oldname, $comment, $upper, $lower, $title) = split /;/;
 137
 138         if ($lower ne "") { $tolower_table[hex $code] = hex $lower; }
 139         if ($upper ne "") { $toupper_table[hex $code] = hex $upper; }
 140
 141         next if $decomp eq "";  # no decomposition, skip it
 142
 143         $src = hex $code;
 144
 145         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 146         {
 147             # decomposition of the form "<foo> 1234" -> use char if type is known
 148             next unless ($1 eq "font" ||
 149                          $1 eq "noBreak" ||
 150                          $1 eq "circle" ||
 151                          $1 eq "super" ||
 152                          $1 eq "sub" ||
 153                          $1 eq "wide" ||
 154                          $1 eq "narrow" ||
 155                          $1 eq "compat" ||
 156                          $1 eq "small");
 157             $dst = hex $2;
 158         }
 159         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 160         {
 161             # decomposition "<compat> 0020 1234" -> combining accent
 162             $dst = hex $1;
 163         }
 164         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 165         {
 166             # decomposition contains only char values without prefix -> use first char
 167             $dst = hex $1;
 168         }
 169         else
 170         {
 171             next;
 172         }
 173
 174         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 175
 176         # check for loops
 177         for ($i = $dst; ; $i = $unicode_defaults[$i])
 178         {
 179             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 180             last unless defined($unicode_defaults[$i]);
 181         }
 182         $unicode_defaults[$src] = $dst;
 183     }
 184 }
 185
 186
 187 ################################################################
 188 # parse the input file
 189 sub READ_FILE
 190 {
 191     my $name = shift;
 192     open INPUT,$name or die "Cannot open $name";
 193     @cp2uni = ();
 194     @lead_bytes = ();
 195     @uni2cp = ();
 196
 197     while (<INPUT>)
 198     {
 199         next if /^\#/;  # skip comments
 200         next if /^$/;  # skip empty lines
 201         next if /\x1a/;  # skip ^Z
 202         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 203
 204         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 205         {
 206             $cp = hex $1;
 207             push @lead_bytes,$cp;
 208             $cp2uni[$cp] = 0;
 209             next;
 210         }
 211         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 212         {
 213             $cp = hex $1;
 214             $uni = hex $2;
 215             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 216             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 217             next;
 218         }
 219         die "$name: Unrecognized line $_\n";
 220     }
 221 }
 222
 223
 224 ################################################################
 225 # parse the symbol.txt file, since its syntax is different from the other ones
 226 sub READ_SYMBOL_FILE
 227 {
 228     my $name = shift;
 229     open INPUT,$name or die "Cannot open $name";
 230     @cp2uni = ();
 231     @lead_bytes = ();
 232     @uni2cp = ();
 233
 234     while (<INPUT>)
 235     {
 236         next if /^\#/;  # skip comments
 237         next if /^$/;  # skip empty lines
 238         next if /\x1a/;  # skip ^Z
 239         if (/^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)\s+(\#.*)?/)
 240         {
 241             $uni = hex $1;
 242             $cp = hex $2;
 243             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 244             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 245             next;
 246         }
 247         die "$name: Unrecognized line $_\n";
 248     }
 249 }
 250
 251
 252 ################################################################
 253 # add default mappings once the file had been read
 254 sub ADD_DEFAULT_MAPPINGS
 255 {
 256     # Apply aliases
 257
 258     foreach $alias (@unicode_aliases)
 259     {
 260         my $target = undef;
 261         foreach $src (@$alias)
 262         {
 263             if (defined($uni2cp[$src]))
 264             {
 265                 $target = $uni2cp[$src];
 266                 last;
 267             }
 268         }
 269         next unless defined($target);
 270
 271         # At least one char of the alias set is defined, set the others to the same value
 272         foreach $src (@$alias)
 273         {
 274             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 275         }
 276     }
 277
 278     # For every src -> target mapping in the defaults table,
 279     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 280
 281     for ($src = 0; $src < 65536; $src++)
 282     {
 283         next if defined($uni2cp[$src]);  # source has a definition already
 284         next unless defined($unicode_defaults[$src]);  # no default for this char
 285         my $target = $unicode_defaults[$src];
 286
 287         # do a recursive mapping until we find a target char that is defined
 288         while (!defined($uni2cp[$target]) &&
 289                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 290
 291         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 292     }
 293
 294     # Add an identity mapping for all undefined chars
 295
 296     for ($i = 0; $i < 256; $i++)
 297     {
 298         next if defined($cp2uni[$i]);
 299         next if defined($uni2cp[$i]);
 300         $cp2uni[$i] = $uni2cp[$i] = $i;
 301     }
 302 }
 303
 304 ################################################################
 305 # dump an array of integers
 306 sub DUMP_ARRAY
 307 {
 308     my ($format,$default,@array) = @_;
 309     my $i, $ret = "    ";
 310     for ($i = 0; $i < $#array; $i++)
 311     {
 312         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 313         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 314     }
 315     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 316     return $ret;
 317 }
 318
 319 ################################################################
 320 # dump an SBCS mapping table
 321 sub DUMP_SBCS_TABLE
 322 {
 323     my ($codepage, $name) = @_;
 324     my $i;
 325
 326     # output the ascii->unicode table
 327
 328     printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 329     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 330
 331     # count the number of unicode->ascii subtables that contain something
 332
 333     my @filled = ();
 334     my $subtables = 1;
 335     for ($i = 0; $i < 65536; $i++)
 336     {
 337         next unless defined $uni2cp[$i];
 338         $filled[$i >> 8] = 1;
 339         $subtables++;
 340         $i = ($i & ~255) + 256;
 341     }
 342
 343     # output all the subtables into a single array
 344
 345     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 346     for ($i = 0; $i < 256; $i++)
 347     {
 348         next unless $filled[$i];
 349         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 350         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
 351     }
 352     printf OUTPUT "    /* defaults */\n";
 353     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($DEF_CHAR) x 256 );
 354
 355     # output a table of the offsets of the subtables in the previous array
 356
 357     my $pos = 0;
 358     my @offsets = ();
 359     for ($i = 0; $i < 256; $i++)
 360     {
 361         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 362         else { push @offsets, ($subtables-1) * 256; }
 363     }
 364     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 365     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 366
 367     # output the code page descriptor
 368
 369     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 370     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 371                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 372     printf OUTPUT "    cp2uni,\n";
 373     printf OUTPUT "    uni2cp_low,\n";
 374     printf OUTPUT "    uni2cp_high\n};\n";
 375 }
 376
 377
 378 ################################################################
 379 # dump a DBCS mapping table
 380 sub DUMP_DBCS_TABLE
 381 {
 382     my ($codepage, $name) = @_;
 383     my $i, $x, $y;
 384
 385     # build a list of lead bytes that are actually used
 386
 387     my @lblist = ();
 388     LBLOOP: for ($y = 0; $y <= $#lead_bytes; $y++)
 389     {
 390         my $base = $lead_bytes[$y] << 8;
 391         for ($x = 0; $x < 256; $x++)
 392         {
 393             if (defined $cp2uni[$base+$x])
 394             {
 395                 push @lblist,$lead_bytes[$y];
 396                 next LBLOOP;
 397             }
 398         }
 399     }
 400     my $unused = ($#lead_bytes > $#lblist);
 401
 402     # output the ascii->unicode table for the single byte chars
 403
 404     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 405     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 406
 407     # output the default table for unused lead bytes
 408
 409     if ($unused)
 410     {
 411         printf OUTPUT "    /* unused lead bytes */\n";
 412         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 413     }
 414
 415     # output the ascii->unicode table for each DBCS lead byte
 416
 417     for ($y = 0; $y <= $#lblist; $y++)
 418     {
 419         my $base = $lblist[$y] << 8;
 420         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
 421         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
 422         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
 423     }
 424
 425     # output the lead byte subtables offsets
 426
 427     my @offsets = ();
 428     for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
 429     for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
 430     if ($unused)
 431     {
 432         # increment all lead bytes offset to take into account the unused table
 433         for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
 434     }
 435     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
 436     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
 437
 438     # count the number of unicode->ascii subtables that contain something
 439
 440     my @filled = ();
 441     my $subtables = 1;
 442     for ($i = 0; $i < 65536; $i++)
 443     {
 444         next unless defined $uni2cp[$i];
 445         $filled[$i >> 8] = 1;
 446         $subtables++;
 447         $i = ($i & ~255) + 256;
 448     }
 449
 450     # output all the subtables into a single array
 451
 452     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
 453     for ($y = 0; $y < 256; $y++)
 454     {
 455         next unless $filled[$y];
 456         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
 457         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
 458     }
 459     printf OUTPUT "    /* defaults */\n";
 460     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 461
 462     # output a table of the offsets of the subtables in the previous array
 463
 464     my $pos = 0;
 465     my @offsets = ();
 466     for ($y = 0; $y < 256; $y++)
 467     {
 468         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
 469         else { push @offsets, ($subtables-1) * 256; }
 470     }
 471     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 472     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 473
 474     # output the code page descriptor
 475
 476     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
 477     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
 478                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 479     printf OUTPUT "    cp2uni,\n";
 480     printf OUTPUT "    cp2uni_leadbytes,\n";
 481     printf OUTPUT "    uni2cp_low,\n";
 482     printf OUTPUT "    uni2cp_high,\n";
 483     DUMP_LB_RANGES();
 484     printf OUTPUT "};\n";
 485 }
 486
 487
 488 ################################################################
 489 # dump the list of defined lead byte ranges
 490 sub DUMP_LB_RANGES
 491 {
 492     my @list = ();
 493     my $i = 0;
 494     foreach $i (@lead_bytes) { $list[$i] = 1; }
 495     my $on = 0;
 496     printf OUTPUT "    { ";
 497     for ($i = 0; $i < 256; $i++)
 498     {
 499         if ($on)
 500         {
 501             if (!defined $list[$i]) { printf OUTPUT "0x%02x, ", $i-1; $on = 0; }
 502         }
 503         else
 504         {
 505             if ($list[$i]) { printf OUTPUT "0x%02x, ", $i; $on = 1; }
 506         }
 507     }
 508     if ($on) { printf OUTPUT "0xff, "; }
 509     printf OUTPUT "0x00, 0x00 }\n";
 510 }
 511
 512
 513 ################################################################
 514 # dump the case mapping tables
 515 sub DUMP_CASE_MAPPINGS
 516 {
 517     open OUTPUT,">casemap.c" or die "Cannot create casemap.c";
 518     printf "Building casemap.c\n";
 519     printf OUTPUT "/* Unicode case mappings */\n";
 520     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 521     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 522
 523     DUMP_CASE_TABLE( "casemap_lower", @tolower_table );
 524     DUMP_CASE_TABLE( "casemap_upper", @toupper_table );
 525     close OUTPUT;
 526 }
 527
 528
 529 ################################################################
 530 # dump a case mapping table
 531 sub DUMP_CASE_TABLE
 532 {
 533     my ($name,@table) = @_;
 534
 535     # count the number of sub tables that contain something
 536
 537     my @filled = ();
 538     my $pos = 512;
 539     for ($i = 0; $i < 65536; $i++)
 540     {
 541         next unless defined $table[$i];
 542         $filled[$i >> 8] = $pos;
 543         $pos += 256;
 544         $i = ($i & ~255) + 256;
 545     }
 546     for ($i = 0; $i < 65536; $i++)
 547     {
 548         next unless defined $table[$i];
 549         $table[$i] = ($table[$i] - $i) & 0xffff;
 550     }
 551
 552     # dump the table
 553
 554     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $pos;
 555     printf OUTPUT "{\n    /* index */\n";
 556     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
 557     printf OUTPUT "    /* defaults */\n";
 558     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
 559     for ($i = 0; $i < 256; $i++)
 560     {
 561         next unless $filled[$i];
 562         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 563         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table[($i<<8) .. ($i<<8)+255] );
 564     }
 565     printf OUTPUT "\n};\n";
 566 }
 567
 568
 569 ################################################################
 570 # read an input file and generate the corresponding .c file
 571 sub HANDLE_FILE
 572 {
 573     my ($codepage,$filename,$comment) = @_;
 574
 575     # symbol codepage file is special
 576     if ($codepage == 42) { READ_SYMBOL_FILE($MAPPREFIX . $filename); }
 577     else { READ_FILE($MAPPREFIX . $filename); }
 578
 579     ADD_DEFAULT_MAPPINGS();
 580
 581     my $output = sprintf "c_%03d.c", $codepage;
 582     open OUTPUT,">$output" or die "Cannot create $output";
 583
 584     printf "Building %s from %s (%s)\n", $output, $filename, $comment;
 585
 586     # dump all tables
 587
 588     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
 589     printf OUTPUT "/* generated from %s */\n", $MAPPREFIX . $filename;
 590     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 591     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 592
 593     if ($#lead_bytes == -1) { DUMP_SBCS_TABLE( $codepage, $comment ); }
 594     else { DUMP_DBCS_TABLE( $codepage, $comment ); }
 595     close OUTPUT;
 596 }
 597
 598
 599 ################################################################
 600 # output the list of codepage tables into the cptable.c file
 601 sub OUTPUT_CPTABLE
 602 {
 603     @tables_decl = ();
 604
 605     foreach $file (@allfiles)
 606     {
 607         my ($codepage,$filename,$comment) = @$file;
 608         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
 609     }
 610
 611     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
 612     foreach $file (@allfiles)
 613     {
 614         my ($codepage,$filename,$comment) = @$file;
 615         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
 616     }
 617     push @tables_decl, "};";
 618     REPLACE_IN_FILE( "cptable.c", @tables_decl );
 619 }
 620
 621 ################################################################
 622 # replace the contents of a file between ### cpmap ### marks
 623
 624 sub REPLACE_IN_FILE
 625 {
 626     my $name = shift;
 627     my @data = @_;
 628     my @lines = ();
 629     open(FILE,$name) or die "Can't open $name";
 630     while (<FILE>)
 631     {
 632         push @lines, $_;
 633         last if /\#\#\# cpmap begin \#\#\#/;
 634     }
 635     push @lines, @data;
 636     while (<FILE>)
 637     {
 638         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
 639     }
 640     push @lines, <FILE>;
 641     open(FILE,">$name") or die "Can't modify $name";
 642     print FILE @lines;
 643     close(FILE);
 644 }