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