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