libs/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 # Sort keys file
  30 $SORTKEYS = "www.unicode.org/reports/tr10/allkeys.txt";
  31
  32 # Defaults mapping
  33 $DEFAULTS = "./defaults";
  34
  35 # Default char for undefined mappings
  36 $DEF_CHAR = ord '?';
  37
  38 @allfiles =
  39 (
  40     [ 37,    "VENDORS/MICSFT/EBCDIC/CP037.TXT",   "IBM EBCDIC US Canada" ],
  41     [ 42,    "VENDORS/ADOBE/symbol.txt",          "Symbol" ],
  42     [ 424,   "VENDORS/MISC/CP424.TXT",            "IBM EBCDIC Hebrew" ],
  43     [ 437,   "VENDORS/MICSFT/PC/CP437.TXT",       "OEM United States" ],
  44     [ 500,   "VENDORS/MICSFT/EBCDIC/CP500.TXT",   "IBM EBCDIC International" ],
  45     [ 737,   "VENDORS/MICSFT/PC/CP737.TXT",       "OEM Greek 437G" ],
  46     [ 775,   "VENDORS/MICSFT/PC/CP775.TXT",       "OEM Baltic" ],
  47     [ 850,   "VENDORS/MICSFT/PC/CP850.TXT",       "OEM Multilingual Latin 1" ],
  48     [ 852,   "VENDORS/MICSFT/PC/CP852.TXT",       "OEM Slovak Latin 2" ],
  49     [ 855,   "VENDORS/MICSFT/PC/CP855.TXT",       "OEM Cyrillic" ],
  50     [ 856,   "VENDORS/MISC/CP856.TXT",            "Hebrew PC" ],
  51     [ 857,   "VENDORS/MICSFT/PC/CP857.TXT",       "OEM Turkish" ],
  52     [ 860,   "VENDORS/MICSFT/PC/CP860.TXT",       "OEM Portuguese" ],
  53     [ 861,   "VENDORS/MICSFT/PC/CP861.TXT",       "OEM Icelandic" ],
  54     [ 862,   "VENDORS/MICSFT/PC/CP862.TXT",       "OEM Hebrew" ],
  55     [ 863,   "VENDORS/MICSFT/PC/CP863.TXT",       "OEM Canadian French" ],
  56     [ 864,   "VENDORS/MICSFT/PC/CP864.TXT",       "OEM Arabic" ],
  57     [ 865,   "VENDORS/MICSFT/PC/CP865.TXT",       "OEM Nordic" ],
  58     [ 866,   "VENDORS/MICSFT/PC/CP866.TXT",       "OEM Russian" ],
  59     [ 869,   "VENDORS/MICSFT/PC/CP869.TXT",       "OEM Greek" ],
  60     [ 874,   "VENDORS/MICSFT/PC/CP874.TXT",       "ANSI/OEM Thai" ],
  61     [ 875,   "VENDORS/MICSFT/EBCDIC/CP875.TXT",   "IBM EBCDIC Greek" ],
  62     [ 878,   "VENDORS/MISC/KOI8-R.TXT",           "Russian KOI8" ],
  63     [ 932,   "VENDORS/MICSFT/WINDOWS/CP932.TXT",  "ANSI/OEM Japanese Shift-JIS" ],
  64     [ 936,   "VENDORS/MICSFT/WINDOWS/CP936.TXT",  "ANSI/OEM Simplified Chinese GBK" ],
  65     [ 949,   "VENDORS/MICSFT/WINDOWS/CP949.TXT",  "ANSI/OEM Korean Unified Hangul" ],
  66     [ 950,   "VENDORS/MICSFT/WINDOWS/CP950.TXT",  "ANSI/OEM Traditional Chinese Big5" ],
  67     [ 1006,  "VENDORS/MISC/CP1006.TXT",           "IBM Arabic" ],
  68     [ 1026,  "VENDORS/MICSFT/EBCDIC/CP1026.TXT",  "IBM EBCDIC Latin 5 Turkish" ],
  69     [ 1250,  "VENDORS/MICSFT/WINDOWS/CP1250.TXT", "ANSI Eastern Europe" ],
  70     [ 1251,  "VENDORS/MICSFT/WINDOWS/CP1251.TXT", "ANSI Cyrillic" ],
  71     [ 1252,  "VENDORS/MICSFT/WINDOWS/CP1252.TXT", "ANSI Latin 1" ],
  72     [ 1253,  "VENDORS/MICSFT/WINDOWS/CP1253.TXT", "ANSI Greek" ],
  73     [ 1254,  "VENDORS/MICSFT/WINDOWS/CP1254.TXT", "ANSI Turkish" ],
  74     [ 1255,  "VENDORS/MICSFT/WINDOWS/CP1255.TXT", "ANSI Hebrew" ],
  75     [ 1256,  "VENDORS/MICSFT/WINDOWS/CP1256.TXT", "ANSI Arabic" ],
  76     [ 1257,  "VENDORS/MICSFT/WINDOWS/CP1257.TXT", "ANSI Baltic" ],
  77     [ 1258,  "VENDORS/MICSFT/WINDOWS/CP1258.TXT", "ANSI/OEM Viet Nam" ],
  78     [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT",      "Mac Roman" ],
  79     [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT",      "Mac Greek" ],
  80     [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT",   "Mac Cyrillic" ],
  81     [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT",     "Mac Latin 2" ],
  82     [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT",    "Mac Icelandic" ],
  83     [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT",    "Mac Turkish" ],
  84     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           "Russian KOI8" ],
  85     [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", "EUC-JP" ],
  86     [ 28591, "ISO8859/8859-1.TXT",                "ISO 8859-1 Latin 1" ],
  87     [ 28592, "ISO8859/8859-2.TXT",                "ISO 8859-2 Latin 2 (East European)" ],
  88     [ 28593, "ISO8859/8859-3.TXT",                "ISO 8859-3 Latin 3 (South European)" ],
  89     [ 28594, "ISO8859/8859-4.TXT",                "ISO 8859-4 Latin 4 (Baltic old)" ],
  90     [ 28595, "ISO8859/8859-5.TXT",                "ISO 8859-5 Cyrillic" ],
  91     [ 28596, "ISO8859/8859-6.TXT",                "ISO 8859-6 Arabic" ],
  92     [ 28597, "ISO8859/8859-7.TXT",                "ISO 8859-7 Greek" ],
  93     [ 28598, "ISO8859/8859-8.TXT",                "ISO 8859-8 Hebrew" ],
  94     [ 28599, "ISO8859/8859-9.TXT",                "ISO 8859-9 Latin 5 (Turkish)" ],
  95     [ 28600, "ISO8859/8859-10.TXT",               "ISO 8859-10 Latin 6 (Nordic)" ],
  96     [ 28603, "ISO8859/8859-13.TXT",               "ISO 8859-13 Latin 7 (Baltic)" ],
  97     [ 28604, "ISO8859/8859-14.TXT",               "ISO 8859-14 Latin 8 (Celtic)" ],
  98     [ 28605, "ISO8859/8859-15.TXT",               "ISO 8859-15 Latin 9 (Euro)" ],
  99     [ 28606, "ISO8859/8859-16.TXT",               "ISO 8859-16 Latin 10 (Balkan)" ]
 100 );
 101
 102
 103 %ctype =
 104 (
 105     "upper"  => 0x0001,
 106     "lower"  => 0x0002,
 107     "digit"  => 0x0004,
 108     "space"  => 0x0008,
 109     "punct"  => 0x0010,
 110     "cntrl"  => 0x0020,
 111     "blank"  => 0x0040,
 112     "xdigit" => 0x0080,
 113     "alpha"  => 0x0100
 114 );
 115
 116 %categories =
 117 (
 118     "Lu" => $ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
 119     "Ll" => $ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
 120     "Lt" => $ctype{"alpha"},    # Letter, Titlecase
 121     "Mn" => $ctype{"punct"},    # Mark, Non-Spacing
 122     "Mc" => $ctype{"punct"},    # Mark, Spacing Combining
 123     "Me" => $ctype{"punct"},    # Mark, Enclosing
 124     "Nd" => $ctype{"digit"},    # Number, Decimal Digit
 125     "Nl" => $ctype{"punct"},    # Number, Letter
 126     "No" => $ctype{"punct"},    # Number, Other
 127     "Zs" => $ctype{"space"},    # Separator, Space
 128     "Zl" => 0,                  # Separator, Line
 129     "Zp" => 0,                  # Separator, Paragraph
 130     "Cc" => $ctype{"cntrl"},    # Other, Control
 131     "Cf" => 0,                  # Other, Format
 132     "Cs" => 0,                  # Other, Surrogate
 133     "Co" => 0,                  # Other, Private Use
 134     "Cn" => 0,                  # Other, Not Assigned
 135     "Lm" => $ctype{"punct"},    # Letter, Modifier
 136     "Lo" => $ctype{"alpha"},    # Letter, Other
 137     "Pc" => $ctype{"punct"},    # Punctuation, Connector
 138     "Pd" => $ctype{"punct"},    # Punctuation, Dash
 139     "Ps" => $ctype{"punct"},    # Punctuation, Open
 140     "Pe" => $ctype{"punct"},    # Punctuation, Close
 141     "Pi" => $ctype{"punct"},    # Punctuation, Initial quote
 142     "Pf" => $ctype{"punct"},    # Punctuation, Final quote
 143     "Po" => $ctype{"punct"},    # Punctuation, Other
 144     "Sm" => $ctype{"punct"},    # Symbol, Math
 145     "Sc" => $ctype{"punct"},    # Symbol, Currency
 146     "Sk" => $ctype{"punct"},    # Symbol, Modifier
 147     "So" => $ctype{"punct"}     # Symbol, Other
 148 );
 149
 150 # a few characters need additional categories that cannot be determined automatically
 151 %special_categories =
 152 (
 153     "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
 154                   0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ],
 155     "space"  => [ 0x09..0x0d, 0xfeff ],
 156     "blank"  => [ 0x09, 0x20, 0xa0, 0xfeff ]
 157 );
 158
 159 %directions =
 160 (
 161     "L"   => 1,    # Left-to-Right
 162     "LRE" => 11,   # Left-to-Right Embedding
 163     "LRO" => 11,   # Left-to-Right Override
 164     "R"   => 2,    # Right-to-Left
 165     "AL"  => 2,    # Right-to-Left Arabic
 166     "RLE" => 11,   # Right-to-Left Embedding
 167     "RLO" => 11,   # Right-to-Left Override
 168     "PDF" => 11,   # Pop Directional Format
 169     "EN"  => 3,    # European Number
 170     "ES"  => 4,    # European Number Separator
 171     "ET"  => 5,    # European Number Terminator
 172     "AN"  => 6,    # Arabic Number
 173     "CS"  => 7,    # Common Number Separator
 174     "NSM" => 0,    # Non-Spacing Mark
 175     "BN"  => 0,    # Boundary Neutral
 176     "B"   => 8,    # Paragraph Separator
 177     "S"   => 9,    # Segment Separator
 178     "WS"  => 10,   # Whitespace
 179     "ON"  => 11    # Other Neutrals
 180 );
 181
 182
 183 ################################################################
 184 # main routine
 185
 186 READ_DEFAULTS();
 187 my @sortkeys = READ_SORTKEYS_FILE();
 188 DUMP_CASE_MAPPINGS();
 189 DUMP_SORTKEYS(@sortkeys);
 190 DUMP_COMPOSE_TABLES();
 191 DUMP_CTYPE_TABLES();
 192
 193 foreach $file (@allfiles) { HANDLE_FILE( @$file ); }
 194
 195 OUTPUT_CPTABLE();
 196
 197 exit(0);
 198
 199
 200 ################################################################
 201 # read in the defaults file
 202 sub READ_DEFAULTS
 203 {
 204     @unicode_defaults = ();
 205     @unicode_aliases = ();
 206     @tolower_table = ();
 207     @toupper_table = ();
 208     @digitmap_table = ();
 209     @compatmap_table = ();
 210     @category_table = ();
 211     @direction_table = ();
 212     @decomp_table = ();
 213     @compose_table = ();
 214
 215     # first setup a few default mappings
 216
 217     open DEFAULTS or die "Cannot open $DEFAULTS";
 218     print "Loading $DEFAULTS\n";
 219     while (<DEFAULTS>)
 220     {
 221         next if /^\#/;  # skip comments
 222         next if /^$/;  # skip empty lines
 223         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 224         {
 225             my @src = map hex, split /,/,$1;
 226             my $dst = $4;
 227             my $comment = $5;
 228             if ($#src > 0) { push @unicode_aliases, \@src; }
 229             next if ($dst eq "none");
 230             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 231             foreach $src (@src)
 232             {
 233                 die "Duplicate value" if defined($unicode_defaults[$src]);
 234                 $unicode_defaults[$src] = $dst;
 235             }
 236             next;
 237         }
 238         die "Unrecognized line $_\n";
 239     }
 240
 241     # now build mappings from the decomposition field of the Unicode database
 242
 243     open UNICODEDATA or die "Cannot open $UNICODEDATA";
 244     print "Loading $UNICODEDATA\n";
 245     while (<UNICODEDATA>)
 246     {
 247         # Decode the fields ...
 248         ($code, $name, $cat, $comb, $bidi,
 249          $decomp, $dec, $dig, $num, $mirror,
 250          $oldname, $comment, $upper, $lower, $title) = split /;/;
 251
 252         my $src = hex $code;
 253
 254         die "unknown category $cat" unless defined $categories{$cat};
 255         die "unknown directionality $bidi" unless defined $directions{$bidi};
 256
 257         $uniname[$src] = $name;
 258         $category_table[$src] = $categories{$cat};
 259         $direction_table[$src] = $directions{$bidi};
 260
 261         if ($lower ne "")
 262         {
 263             $tolower_table[$src] = hex $lower;
 264             $category_table[$src] |= $ctype{"upper"}|$ctype{"alpha"};
 265         }
 266         if ($upper ne "")
 267         {
 268             $toupper_table[$src] = hex $upper;
 269             $category_table[$src] |= $ctype{"lower"}|$ctype{"alpha"};
 270         }
 271         if ($dec ne "")
 272         {
 273             $category_table[$src] |= $ctype{"digit"};
 274         }
 275         if ($dig ne "")
 276         {
 277             $digitmap_table[$src] = ord $dig;
 278         }
 279
 280         # copy the category and direction for everything between First/Last pairs
 281         if ($name =~ /, First>/) { $start = $src; }
 282         if ($name =~ /, Last>/)
 283         {
 284             while ($start < $src)
 285             {
 286                 $category_table[$start] = $category_table[$src];
 287                 $direction_table[$start] = $direction_table[$src];
 288                 $start++;
 289             }
 290         }
 291
 292         next if $decomp eq "";  # no decomposition, skip it
 293
 294         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 295         {
 296             # decomposition of the form "<foo> 1234" -> use char if type is known
 297             if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd))
 298             {
 299                 # Single char decomposition in the compatibility range
 300                 $compatmap_table[$src] = hex $2;
 301             }
 302             next unless ($1 eq "font" ||
 303                          $1 eq "noBreak" ||
 304                          $1 eq "circle" ||
 305                          $1 eq "super" ||
 306                          $1 eq "sub" ||
 307                          $1 eq "wide" ||
 308                          $1 eq "narrow" ||
 309                          $1 eq "compat" ||
 310                          $1 eq "small");
 311             $dst = hex $2;
 312         }
 313         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 314         {
 315             # decomposition "<compat> 0020 1234" -> combining accent
 316             $dst = hex $1;
 317         }
 318         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 319         {
 320             # decomposition contains only char values without prefix -> use first char
 321             $dst = hex $1;
 322             $category_table[$src] |= $category_table[$dst];
 323             # store decomposition if it contains two chars
 324             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 325             {
 326                 $decomp_table[$src] = [ hex $1, hex $2 ];
 327                 push @compose_table, [ hex $1, hex $2, $src ];
 328             }
 329             elsif ($decomp =~ /^(<[a-z]+>\s)*([0-9a-fA-F]+)$/ &&
 330                    (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)))
 331             {
 332                 # Single char decomposition in the compatibility range
 333                 $compatmap_table[$src] = hex $2;
 334             }
 335         }
 336         else
 337         {
 338             next;
 339         }
 340
 341         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 342
 343         # check for loops
 344         for ($i = $dst; ; $i = $unicode_defaults[$i])
 345         {
 346             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 347             last unless defined($unicode_defaults[$i]);
 348         }
 349         $unicode_defaults[$src] = $dst;
 350     }
 351
 352     # patch the category of some special characters
 353
 354     foreach $cat (keys %special_categories)
 355     {
 356         my $flag = $ctype{$cat};
 357         foreach $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 358     }
 359 }
 360
 361
 362 ################################################################
 363 # parse the input file
 364 sub READ_FILE
 365 {
 366     my $name = shift;
 367     open INPUT,$name or die "Cannot open $name";
 368     @cp2uni = ();
 369     @lead_bytes = ();
 370     @uni2cp = ();
 371
 372     while (<INPUT>)
 373     {
 374         next if /^\#/;  # skip comments
 375         next if /^$/;  # skip empty lines
 376         next if /\x1a/;  # skip ^Z
 377         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 378
 379         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 380         {
 381             $cp = hex $1;
 382             push @lead_bytes,$cp;
 383             $cp2uni[$cp] = 0;
 384             next;
 385         }
 386         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 387         {
 388             $cp = hex $1;
 389             $uni = hex $2;
 390             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 391             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 392             next;
 393         }
 394         die "$name: Unrecognized line $_\n";
 395     }
 396 }
 397
 398
 399 ################################################################
 400 # parse the symbol.txt file, since its syntax is different from the other ones
 401 sub READ_SYMBOL_FILE
 402 {
 403     my $name = shift;
 404     open INPUT,$name or die "Cannot open $name";
 405     @cp2uni = ();
 406     @lead_bytes = ();
 407     @uni2cp = ();
 408
 409     while (<INPUT>)
 410     {
 411         next if /^\#/;  # skip comments
 412         next if /^$/;  # skip empty lines
 413         next if /\x1a/;  # skip ^Z
 414         if (/^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)\s+(\#.*)?/)
 415         {
 416             $uni = hex $1;
 417             $cp = hex $2;
 418             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 419             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 420             next;
 421         }
 422         die "$name: Unrecognized line $_\n";
 423     }
 424 }
 425
 426
 427 ################################################################
 428 # build EUC-JP table from the JIS 0208 file
 429 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 430 # but this would require 3-byte DBCS characters
 431 sub READ_JIS0208_FILE
 432 {
 433     my $name = shift;
 434     @cp2uni = ();
 435     @lead_bytes = ();
 436     @uni2cp = ();
 437
 438     # ASCII chars
 439     for ($i = 0x00; $i <= 0x7f; $i++)
 440     {
 441         $cp2uni[$i] = $i;
 442         $uni2cp[$i] = $i;
 443     }
 444
 445     # JIS X 0201 right plane
 446     for ($i = 0xa1; $i <= 0xdf; $i++)
 447     {
 448         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 449         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 450     }
 451
 452     # lead bytes
 453     foreach $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 454     {
 455         push @lead_bytes,$i;
 456         $cp2uni[$i] = 0;
 457     }
 458
 459     # undefined chars
 460     foreach $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 461     {
 462         $cp2uni[$i] = $DEF_CHAR;
 463     }
 464
 465     # Shift-JIS compatibility
 466     $uni2cp[0x00a5] = 0x5c;
 467     $uni2cp[0x203e] = 0x7e;
 468
 469     # Fix backslash conversion
 470     $cp2uni[0xa1c0] = 0xff3c;
 471     $uni2cp[0xff3c] = 0xa1c0;
 472
 473     open INPUT, "$name" or die "Cannot open $name";
 474     while (<INPUT>)
 475     {
 476         next if /^\#/;  # skip comments
 477         next if /^$/;  # skip empty lines
 478         next if /\x1a/;  # skip ^Z
 479         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 480         {
 481             $cp = 0x8080 + hex $1;
 482             $uni = hex $2;
 483             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 484             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 485             next;
 486         }
 487         die "$name: Unrecognized line $_\n";
 488     }
 489 }
 490
 491
 492 ################################################################
 493 # build the sort keys table
 494 sub READ_SORTKEYS_FILE
 495 {
 496     my @sortkeys = ();
 497     for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
 498
 499     open INPUT, "$SORTKEYS" or die "Cannot open $SORTKEYS";
 500     print "Loading $SORTKEYS\n";
 501     while (<INPUT>)
 502     {
 503         next if /^\#/;  # skip comments
 504         next if /^$/;  # skip empty lines
 505         next if /\x1a/;  # skip ^Z
 506         next if /^\@version/;  # skip @version header
 507         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 508         {
 509             my ($uni,$variable) = (hex $1, $2);
 510             next if $uni > 65535;
 511             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 512             next;
 513         }
 514         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 515         {
 516             # multiple character sequence, ignored for now
 517             next;
 518         }
 519         die "$SORTKEYS: Unrecognized line $_\n";
 520     }
 521     close INPUT;
 522
 523     # compress the keys to 32 bit:
 524     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 525
 526     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 527                        ${$a}[2] <=> ${$b}[2] or
 528                        ${$a}[3] <=> ${$b}[3] or
 529                        ${$a}[4] <=> ${$b}[4] or
 530                        $a cmp $b; } @sortkeys;
 531
 532     my ($n2, $n3) = (1, 1);
 533     my @keys = (-1, -1, -1, -1, -1 );
 534     my @flatkeys = ();
 535
 536     for (my $i = 0; $i < 65536; $i++)
 537     {
 538         my @current = @{$sortkeys[$i]};
 539         next if $current[0] == -1;
 540         if ($current[1] == $keys[1])
 541         {
 542             if ($current[2] == $keys[2])
 543             {
 544                 if ($current[3] == $keys[3])
 545                 {
 546                     # nothing
 547                 }
 548                 else
 549                 {
 550                     $keys[3] = $current[3];
 551                     $n3++;
 552                     die if ($n3 >= 16);
 553                 }
 554             }
 555             else
 556             {
 557                 $keys[2] = $current[2];
 558                 $keys[3] = $current[3];
 559                 $n2++;
 560                 $n3 = 1;
 561                 die if ($n2 >= 256);
 562             }
 563         }
 564         else
 565         {
 566             $keys[1] = $current[1];
 567             $keys[2] = $current[2];
 568             $keys[3] = $current[3];
 569             $n2 = 1;
 570             $n3 = 1;
 571         }
 572
 573         if ($current[2]) { $current[2] = $n2; }
 574         if ($current[3]) { $current[3] = $n3; }
 575         if ($current[4]) { $current[4] = 1; }
 576
 577         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 578     }
 579     return @flatkeys;
 580 }
 581
 582
 583 ################################################################
 584 # build the sort keys table
 585 sub DUMP_SORTKEYS
 586 {
 587     my @keys = @_;
 588
 589     # count the number of 256-key ranges that contain something
 590
 591     my @offsets = ();
 592     my $ranges = 2;
 593     for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
 594     for (my $i = 0; $i < 65536; $i++)
 595     {
 596         next unless defined $keys[$i];
 597         $offsets[$i >> 8] = $ranges * 256;
 598         $ranges++;
 599         $i |= 255;
 600     }
 601
 602     # output the range offsets
 603
 604     open OUTPUT,">collation.c" or die "Cannot create collation.c";
 605     printf "Building collation.c\n";
 606     printf OUTPUT "/* Unicode collation element table */\n";
 607     printf OUTPUT "/* generated from %s */\n", $SORTKEYS;
 608     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 609
 610     printf OUTPUT "const unsigned int collation_table[%d] =\n{\n", $ranges*256;
 611     printf OUTPUT "    /* index */\n";
 612     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%08x", 0, @offsets );
 613
 614     # output the default values
 615
 616     printf OUTPUT "    /* defaults */\n";
 617     printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0, (-1) x 256 );
 618
 619     # output all the key ranges
 620
 621     for (my $i = 0; $i < 256; $i++)
 622     {
 623         next if $offsets[$i] == 256;
 624         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 625         printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", -1, @keys[($i<<8) .. ($i<<8)+255] );
 626     }
 627     printf OUTPUT "\n};\n";
 628     close OUTPUT;
 629 }
 630
 631
 632 ################################################################
 633 # add default mappings once the file had been read
 634 sub ADD_DEFAULT_MAPPINGS
 635 {
 636     # Apply aliases
 637
 638     foreach $alias (@unicode_aliases)
 639     {
 640         my $target = undef;
 641         foreach $src (@$alias)
 642         {
 643             if (defined($uni2cp[$src]))
 644             {
 645                 $target = $uni2cp[$src];
 646                 last;
 647             }
 648         }
 649         next unless defined($target);
 650
 651         # At least one char of the alias set is defined, set the others to the same value
 652         foreach $src (@$alias)
 653         {
 654             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 655         }
 656     }
 657
 658     # For every src -> target mapping in the defaults table,
 659     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 660
 661     for ($src = 0; $src < 65536; $src++)
 662     {
 663         next if defined($uni2cp[$src]);  # source has a definition already
 664         next unless defined($unicode_defaults[$src]);  # no default for this char
 665         my $target = $unicode_defaults[$src];
 666
 667         # do a recursive mapping until we find a target char that is defined
 668         while (!defined($uni2cp[$target]) &&
 669                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 670
 671         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 672     }
 673
 674     # Add an identity mapping for all undefined chars
 675
 676     for ($i = 0; $i < 256; $i++)
 677     {
 678         next if defined($cp2uni[$i]);
 679         next if defined($uni2cp[$i]);
 680         $cp2uni[$i] = $uni2cp[$i] = $i;
 681     }
 682 }
 683
 684 ################################################################
 685 # dump an array of integers
 686 sub DUMP_ARRAY
 687 {
 688     my ($format,$default,@array) = @_;
 689     my $i, $ret = "    ";
 690     for ($i = 0; $i < $#array; $i++)
 691     {
 692         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 693         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 694     }
 695     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 696     return $ret;
 697 }
 698
 699 ################################################################
 700 # dump an SBCS mapping table
 701 sub DUMP_SBCS_TABLE
 702 {
 703     my ($codepage, $name) = @_;
 704     my $i;
 705
 706     # output the ascii->unicode table
 707
 708     printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 709     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 710
 711     # count the number of unicode->ascii subtables that contain something
 712
 713     my @filled = ();
 714     my $subtables = 1;
 715     for ($i = 0; $i < 65536; $i++)
 716     {
 717         next unless defined $uni2cp[$i];
 718         $filled[$i >> 8] = 1;
 719         $subtables++;
 720         $i |= 255;
 721     }
 722
 723     # output all the subtables into a single array
 724
 725     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 726     for ($i = 0; $i < 256; $i++)
 727     {
 728         next unless $filled[$i];
 729         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 730         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
 731     }
 732     printf OUTPUT "    /* defaults */\n";
 733     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($DEF_CHAR) x 256 );
 734
 735     # output a table of the offsets of the subtables in the previous array
 736
 737     my $pos = 0;
 738     my @offsets = ();
 739     for ($i = 0; $i < 256; $i++)
 740     {
 741         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 742         else { push @offsets, ($subtables-1) * 256; }
 743     }
 744     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 745     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 746
 747     # output the code page descriptor
 748
 749     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 750     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 751                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 752     printf OUTPUT "    cp2uni,\n";
 753     printf OUTPUT "    uni2cp_low,\n";
 754     printf OUTPUT "    uni2cp_high\n};\n";
 755 }
 756
 757
 758 ################################################################
 759 # dump a DBCS mapping table
 760 sub DUMP_DBCS_TABLE
 761 {
 762     my ($codepage, $name) = @_;
 763     my $i, $x, $y;
 764
 765     # build a list of lead bytes that are actually used
 766
 767     my @lblist = ();
 768     LBLOOP: for ($y = 0; $y <= $#lead_bytes; $y++)
 769     {
 770         my $base = $lead_bytes[$y] << 8;
 771         for ($x = 0; $x < 256; $x++)
 772         {
 773             if (defined $cp2uni[$base+$x])
 774             {
 775                 push @lblist,$lead_bytes[$y];
 776                 next LBLOOP;
 777             }
 778         }
 779     }
 780     my $unused = ($#lead_bytes > $#lblist);
 781
 782     # output the ascii->unicode table for the single byte chars
 783
 784     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 785     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 786
 787     # output the default table for unused lead bytes
 788
 789     if ($unused)
 790     {
 791         printf OUTPUT "    /* unused lead bytes */\n";
 792         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 793     }
 794
 795     # output the ascii->unicode table for each DBCS lead byte
 796
 797     for ($y = 0; $y <= $#lblist; $y++)
 798     {
 799         my $base = $lblist[$y] << 8;
 800         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
 801         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
 802         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
 803     }
 804
 805     # output the lead byte subtables offsets
 806
 807     my @offsets = ();
 808     for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
 809     for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
 810     if ($unused)
 811     {
 812         # increment all lead bytes offset to take into account the unused table
 813         for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
 814     }
 815     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
 816     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
 817
 818     # count the number of unicode->ascii subtables that contain something
 819
 820     my @filled = ();
 821     my $subtables = 1;
 822     for ($i = 0; $i < 65536; $i++)
 823     {
 824         next unless defined $uni2cp[$i];
 825         $filled[$i >> 8] = 1;
 826         $subtables++;
 827         $i |= 255;
 828     }
 829
 830     # output all the subtables into a single array
 831
 832     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
 833     for ($y = 0; $y < 256; $y++)
 834     {
 835         next unless $filled[$y];
 836         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
 837         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
 838     }
 839     printf OUTPUT "    /* defaults */\n";
 840     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 841
 842     # output a table of the offsets of the subtables in the previous array
 843
 844     my $pos = 0;
 845     my @offsets = ();
 846     for ($y = 0; $y < 256; $y++)
 847     {
 848         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
 849         else { push @offsets, ($subtables-1) * 256; }
 850     }
 851     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 852     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 853
 854     # output the code page descriptor
 855
 856     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
 857     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
 858                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 859     printf OUTPUT "    cp2uni,\n";
 860     printf OUTPUT "    cp2uni_leadbytes,\n";
 861     printf OUTPUT "    uni2cp_low,\n";
 862     printf OUTPUT "    uni2cp_high,\n";
 863     DUMP_LB_RANGES();
 864     printf OUTPUT "};\n";
 865 }
 866
 867
 868 ################################################################
 869 # dump the list of defined lead byte ranges
 870 sub DUMP_LB_RANGES
 871 {
 872     my @list = ();
 873     my $i = 0;
 874     foreach $i (@lead_bytes) { $list[$i] = 1; }
 875     my $on = 0;
 876     printf OUTPUT "    { ";
 877     for ($i = 0; $i < 256; $i++)
 878     {
 879         if ($on)
 880         {
 881             if (!defined $list[$i]) { printf OUTPUT "0x%02x, ", $i-1; $on = 0; }
 882         }
 883         else
 884         {
 885             if ($list[$i]) { printf OUTPUT "0x%02x, ", $i; $on = 1; }
 886         }
 887     }
 888     if ($on) { printf OUTPUT "0xff, "; }
 889     printf OUTPUT "0x00, 0x00 }\n";
 890 }
 891
 892
 893 ################################################################
 894 # dump the case mapping tables
 895 sub DUMP_CASE_MAPPINGS
 896 {
 897     open OUTPUT,">casemap.c" or die "Cannot create casemap.c";
 898     printf "Building casemap.c\n";
 899     printf OUTPUT "/* Unicode case mappings */\n";
 900     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 901     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 902
 903     DUMP_CASE_TABLE( "wine_casemap_lower", @tolower_table );
 904     DUMP_CASE_TABLE( "wine_casemap_upper", @toupper_table );
 905     DUMP_CASE_TABLE( "wine_digitmap",  @digitmap_table );
 906     DUMP_CASE_TABLE( "wine_compatmap", @compatmap_table );
 907     close OUTPUT;
 908 }
 909
 910
 911 ################################################################
 912 # dump a case mapping table
 913 sub DUMP_CASE_TABLE
 914 {
 915     my ($name,@table) = @_;
 916
 917     # count the number of sub tables that contain something
 918     # also compute the low and upper populated bounds
 919
 920     my @lowerbounds = ( 0, 0 );
 921     my @upperbounds = ( 0, 255 );
 922     my $index = 0;
 923     my @filled = ();
 924     for ($i = 0; $i < 65536; $i++)
 925     {
 926         next unless defined $table[$i];
 927         if (!defined $filled[$i >> 8])
 928         {
 929           $lowerbounds[$index] = $i & 0xff;
 930           $upperbounds[$index] = 0xff - $lowerbounds[$index];
 931           $filled[$i >> 8] = $index * 256 + 512;
 932           $index++;
 933         }
 934         else
 935         {
 936           $upperbounds[$index-1] = 0xff - ($i & 0xff);
 937         }
 938         $table[$i] = ($table[$i] - $i) & 0xffff;
 939     }
 940
 941     # Collapse blocks upwards if possible
 942     my $removed = 0;
 943     $index = 0;
 944     for ($i = 0; $i < 256; $i++)
 945     {
 946         next unless defined $filled[$i];
 947         if ($upperbounds[$index - 1] > $lowerbounds[$index])
 948         {
 949            $removed = $removed + $lowerbounds[$index];
 950         }
 951         else
 952         {
 953            $removed = $removed + $upperbounds[$index - 1];
 954            $lowerbounds[$index] = $upperbounds[$index - 1];
 955         }
 956         $filled[$i] = $filled[$i] - $removed;
 957         $index++;
 958     }
 959
 960     # dump the table
 961
 962     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
 963     printf OUTPUT "{\n    /* index */\n";
 964     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
 965     printf OUTPUT "    /* defaults */\n";
 966     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
 967     $index = 0;
 968     for ($i = 0; $i < 256; $i++)
 969     {
 970         next unless $filled[$i];
 971         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
 972         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0,
 973                       @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
 974         $index++;
 975     }
 976     printf OUTPUT "\n};\n";
 977 }
 978
 979
 980 ################################################################
 981 # dump the ctype tables
 982 sub DUMP_CTYPE_TABLES
 983 {
 984     open OUTPUT,">wctype.c" or die "Cannot create wctype.c";
 985     printf "Building wctype.c\n";
 986     printf OUTPUT "/* Unicode ctype tables */\n";
 987     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 988     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 989
 990     my $i;
 991     my @array = (0) x 256;
 992
 993     # add the direction in the high 4 bits of the category
 994     for ($i = 0; $i < 65536; $i++)
 995     {
 996         $category_table[$i] |= $direction_table[$i] << 12;
 997     }
 998
 999     # try to merge table rows
1000     for ($row = 0; $row < 256; $row++)
1001     {
1002         my $rowtxt = sprintf "%04x" x 256, @category_table[($row<<8)..($row<<8)+255];
1003         if (defined($sequences{$rowtxt}))
1004         {
1005             # reuse an existing row
1006             $array[$row] = $sequences{$rowtxt};
1007         }
1008         else
1009         {
1010             # create a new row
1011             $sequences{$rowtxt} = $array[$row] = $#array + 1;
1012             push @array, @category_table[($row<<8)..($row<<8)+255];
1013         }
1014     }
1015
1016     printf OUTPUT "const unsigned short wine_wctype_table[%d] =\n{\n", $#array+1;
1017     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1018     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1019
1020     close OUTPUT;
1021 }
1022
1023
1024 ################################################################
1025 # dump the char composition tables
1026 sub DUMP_COMPOSE_TABLES
1027 {
1028     open OUTPUT,">compose.c" or die "Cannot create compose.c";
1029     printf "Building compose.c\n";
1030     printf OUTPUT "/* Unicode char composition */\n";
1031     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1032     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1033
1034     ######### composition table
1035
1036     my @filled = ();
1037     foreach $i (@compose_table)
1038     {
1039         my @comp = @$i;
1040         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
1041     }
1042
1043     # count how many different second chars we have
1044
1045     for ($i = $count = 0; $i < 65536; $i++)
1046     {
1047         next unless defined $filled[$i];
1048         $count++;
1049     }
1050
1051     # build the table of second chars and offsets
1052
1053     my $pos = $count + 1;
1054     for ($i = 0; $i < 65536; $i++)
1055     {
1056         next unless defined $filled[$i];
1057         push @table, $i, $pos;
1058         $pos += @{$filled[$i]};
1059     }
1060     # terminator with last position
1061     push @table, 0, $pos;
1062     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
1063     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
1064
1065     # build the table of first chars and mappings
1066
1067     for ($i = 0; $i < 65536; $i++)
1068     {
1069         next unless defined $filled[$i];
1070         my @table = ();
1071         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
1072         for ($j = 0; $j <= $#list; $j++)
1073         {
1074             push @table, $list[$j][0], $list[$j][1];
1075         }
1076         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
1077     }
1078     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
1079
1080     ######### decomposition table
1081
1082     # first determine all the 16-char subsets that contain something
1083
1084     my @filled = (0) x 4096;
1085     my $pos = 16*2;  # for the null subset
1086     for ($i = 0; $i < 65536; $i++)
1087     {
1088         next unless defined $decomp_table[$i];
1089         $filled[$i >> 4] = $pos;
1090         $pos += 16*2;
1091         $i |= 15;
1092     }
1093     my $total = $pos;
1094
1095     # now count the 256-char subsets that contain something
1096
1097     my @filled_idx = (256) x 256;
1098     $pos = 256 + 16;
1099     for ($i = 0; $i < 4096; $i++)
1100     {
1101         next unless $filled[$i];
1102         $filled_idx[$i >> 4] = $pos;
1103         $pos += 16;
1104         $i |= 15;
1105     }
1106     my $null_offset = $pos;  # null mapping
1107     $total += $pos;
1108
1109     # add the index offsets to the subsets positions
1110
1111     for ($i = 0; $i < 4096; $i++)
1112     {
1113         next unless $filled[$i];
1114         $filled[$i] += $null_offset;
1115     }
1116
1117     # dump the main index
1118
1119     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
1120     printf OUTPUT "{\n    /* index */\n";
1121     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1122     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1123
1124     # dump the second-level indexes
1125
1126     for ($i = 0; $i < 256; $i++)
1127     {
1128         next unless ($filled_idx[$i] > 256);
1129         my @table = @filled[($i<<4)..($i<<4)+15];
1130         for ($j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1131         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1132         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1133     }
1134
1135     # dump the 16-char subsets
1136
1137     printf OUTPUT ",\n    /* null mapping */\n";
1138     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
1139
1140     for ($i = 0; $i < 4096; $i++)
1141     {
1142         next unless $filled[$i];
1143         my @table = (0) x 32;
1144         for ($j = 0; $j < 16; $j++)
1145         {
1146             if (defined $decomp_table[($i<<4) + $j])
1147             {
1148                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
1149                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
1150             }
1151         }
1152         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1153         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1154     }
1155
1156     printf OUTPUT "\n};\n";
1157     close OUTPUT;
1158 }
1159
1160
1161 ################################################################
1162 # read an input file and generate the corresponding .c file
1163 sub HANDLE_FILE
1164 {
1165     my ($codepage,$filename,$comment) = @_;
1166
1167     # symbol codepage file is special
1168     if ($codepage == 42) { READ_SYMBOL_FILE($MAPPREFIX . $filename); }
1169     elsif ($codepage == 20932) { READ_JIS0208_FILE($MAPPREFIX . $filename); }
1170     else { READ_FILE($MAPPREFIX . $filename); }
1171
1172     # hack: 0x00a5 must map to backslash in Shift-JIS
1173     if ($codepage == 932) { $uni2cp[0x00a5] = 0x5c; }
1174
1175     ADD_DEFAULT_MAPPINGS();
1176
1177     my $output = sprintf "c_%03d.c", $codepage;
1178     open OUTPUT,">$output" or die "Cannot create $output";
1179
1180     printf "Building %s from %s (%s)\n", $output, $filename, $comment;
1181
1182     # dump all tables
1183
1184     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
1185     printf OUTPUT "/* generated from %s */\n", $MAPPREFIX . $filename;
1186     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
1187     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1188
1189     if ($#lead_bytes == -1) { DUMP_SBCS_TABLE( $codepage, $comment ); }
1190     else { DUMP_DBCS_TABLE( $codepage, $comment ); }
1191     close OUTPUT;
1192 }
1193
1194
1195 ################################################################
1196 # output the list of codepage tables into the cptable.c file
1197 sub OUTPUT_CPTABLE
1198 {
1199     @tables_decl = ();
1200
1201     foreach $file (@allfiles)
1202     {
1203         my ($codepage,$filename,$comment) = @$file;
1204         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
1205     }
1206
1207     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
1208     foreach $file (@allfiles)
1209     {
1210         my ($codepage,$filename,$comment) = @$file;
1211         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
1212     }
1213     push @tables_decl, "};";
1214     REPLACE_IN_FILE( "cptable.c", @tables_decl );
1215 }
1216
1217 ################################################################
1218 # replace the contents of a file between ### cpmap ### marks
1219
1220 sub REPLACE_IN_FILE
1221 {
1222     my $name = shift;
1223     my @data = @_;
1224     my @lines = ();
1225     open(FILE,$name) or die "Can't open $name";
1226     while (<FILE>)
1227     {
1228         push @lines, $_;
1229         last if /\#\#\# cpmap begin \#\#\#/;
1230     }
1231     push @lines, @data;
1232     while (<FILE>)
1233     {
1234         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
1235     }
1236     push @lines, <FILE>;
1237     open(FILE,">$name") or die "Can't modify $name";
1238     print FILE @lines;
1239     close(FILE);
1240 }