libs/wine/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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, 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     [ 20127, undef,                               "US-ASCII (7bit)" ],
  84     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           "Russian KOI8" ],
  85     [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", "EUC-JP" ],
  86     [ 21866, "VENDORS/MISC/KOI8-U.TXT",           "Ukrainian KOI8" ],
  87     [ 28591, "ISO8859/8859-1.TXT",                "ISO 8859-1 Latin 1" ],
  88     [ 28592, "ISO8859/8859-2.TXT",                "ISO 8859-2 Latin 2 (East European)" ],
  89     [ 28593, "ISO8859/8859-3.TXT",                "ISO 8859-3 Latin 3 (South European)" ],
  90     [ 28594, "ISO8859/8859-4.TXT",                "ISO 8859-4 Latin 4 (Baltic old)" ],
  91     [ 28595, "ISO8859/8859-5.TXT",                "ISO 8859-5 Cyrillic" ],
  92     [ 28596, "ISO8859/8859-6.TXT",                "ISO 8859-6 Arabic" ],
  93     [ 28597, "ISO8859/8859-7.TXT",                "ISO 8859-7 Greek" ],
  94     [ 28598, "ISO8859/8859-8.TXT",                "ISO 8859-8 Hebrew" ],
  95     [ 28599, "ISO8859/8859-9.TXT",                "ISO 8859-9 Latin 5 (Turkish)" ],
  96     [ 28600, "ISO8859/8859-10.TXT",               "ISO 8859-10 Latin 6 (Nordic)" ],
  97     [ 28603, "ISO8859/8859-13.TXT",               "ISO 8859-13 Latin 7 (Baltic)" ],
  98     [ 28604, "ISO8859/8859-14.TXT",               "ISO 8859-14 Latin 8 (Celtic)" ],
  99     [ 28605, "ISO8859/8859-15.TXT",               "ISO 8859-15 Latin 9 (Euro)" ],
 100     [ 28606, "ISO8859/8859-16.TXT",               "ISO 8859-16 Latin 10 (Balkan)" ]
 101 );
 102
 103
 104 %ctype =
 105 (
 106     "upper"  => 0x0001,
 107     "lower"  => 0x0002,
 108     "digit"  => 0x0004,
 109     "space"  => 0x0008,
 110     "punct"  => 0x0010,
 111     "cntrl"  => 0x0020,
 112     "blank"  => 0x0040,
 113     "xdigit" => 0x0080,
 114     "alpha"  => 0x0100
 115 );
 116
 117 %categories =
 118 (
 119     "Lu" => $ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
 120     "Ll" => $ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
 121     "Lt" => $ctype{"alpha"},    # Letter, Titlecase
 122     "Mn" => $ctype{"punct"},    # Mark, Non-Spacing
 123     "Mc" => $ctype{"punct"},    # Mark, Spacing Combining
 124     "Me" => $ctype{"punct"},    # Mark, Enclosing
 125     "Nd" => $ctype{"digit"},    # Number, Decimal Digit
 126     "Nl" => $ctype{"punct"},    # Number, Letter
 127     "No" => $ctype{"punct"},    # Number, Other
 128     "Zs" => $ctype{"space"},    # Separator, Space
 129     "Zl" => $ctype{"space"},    # Separator, Line
 130     "Zp" => $ctype{"space"},    # Separator, Paragraph
 131     "Cc" => $ctype{"cntrl"},    # Other, Control
 132     "Cf" => 0,                  # Other, Format
 133     "Cs" => 0,                  # Other, Surrogate
 134     "Co" => 0,                  # Other, Private Use
 135     "Cn" => 0,                  # Other, Not Assigned
 136     "Lm" => $ctype{"punct"},    # Letter, Modifier
 137     "Lo" => $ctype{"alpha"},    # Letter, Other
 138     "Pc" => $ctype{"punct"},    # Punctuation, Connector
 139     "Pd" => $ctype{"punct"},    # Punctuation, Dash
 140     "Ps" => $ctype{"punct"},    # Punctuation, Open
 141     "Pe" => $ctype{"punct"},    # Punctuation, Close
 142     "Pi" => $ctype{"punct"},    # Punctuation, Initial quote
 143     "Pf" => $ctype{"punct"},    # Punctuation, Final quote
 144     "Po" => $ctype{"punct"},    # Punctuation, Other
 145     "Sm" => $ctype{"punct"},    # Symbol, Math
 146     "Sc" => $ctype{"punct"},    # Symbol, Currency
 147     "Sk" => $ctype{"punct"},    # Symbol, Modifier
 148     "So" => $ctype{"punct"}     # Symbol, Other
 149 );
 150
 151 # a few characters need additional categories that cannot be determined automatically
 152 %special_categories =
 153 (
 154     "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
 155                   0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ],
 156     "space"  => [ 0x09..0x0d, 0x85 ],
 157     "blank"  => [ 0x09, 0x20, 0xa0, 0x3000, 0xfeff ],
 158     "cntrl"  => [ 0x070f, 0x180b, 0x180c, 0x180d, 0x180e, 0x200c, 0x200d,
 159                   0x200e, 0x200f, 0x202a, 0x202b, 0x202c, 0x202d, 0x202e,
 160                   0x206a, 0x206b, 0x206c, 0x206d, 0x206e, 0x206f, 0xfeff,
 161                   0xfff9, 0xfffa, 0xfffb ]
 162 );
 163
 164 %directions =
 165 (
 166     "L"   => 1,    # Left-to-Right
 167     "LRE" => 11,   # Left-to-Right Embedding
 168     "LRO" => 11,   # Left-to-Right Override
 169     "R"   => 2,    # Right-to-Left
 170     "AL"  => 2,    # Right-to-Left Arabic
 171     "RLE" => 11,   # Right-to-Left Embedding
 172     "RLO" => 11,   # Right-to-Left Override
 173     "PDF" => 11,   # Pop Directional Format
 174     "EN"  => 3,    # European Number
 175     "ES"  => 4,    # European Number Separator
 176     "ET"  => 5,    # European Number Terminator
 177     "AN"  => 6,    # Arabic Number
 178     "CS"  => 7,    # Common Number Separator
 179     "NSM" => 0,    # Non-Spacing Mark
 180     "BN"  => 0,    # Boundary Neutral
 181     "B"   => 8,    # Paragraph Separator
 182     "S"   => 9,    # Segment Separator
 183     "WS"  => 10,   # Whitespace
 184     "ON"  => 11    # Other Neutrals
 185 );
 186
 187
 188 ################################################################
 189 # main routine
 190
 191 READ_DEFAULTS();
 192 my @sortkeys = READ_SORTKEYS_FILE();
 193 DUMP_CASE_MAPPINGS();
 194 DUMP_SORTKEYS(@sortkeys);
 195 DUMP_COMPOSE_TABLES();
 196 DUMP_CTYPE_TABLES();
 197
 198 foreach $file (@allfiles) { HANDLE_FILE( @$file ); }
 199
 200 OUTPUT_CPTABLE();
 201
 202 exit(0);
 203
 204
 205 ################################################################
 206 # read in the defaults file
 207 sub READ_DEFAULTS
 208 {
 209     @unicode_defaults = ();
 210     @unicode_aliases = ();
 211     @tolower_table = ();
 212     @toupper_table = ();
 213     @digitmap_table = ();
 214     @compatmap_table = ();
 215     @category_table = ();
 216     @direction_table = ();
 217     @decomp_table = ();
 218     @compose_table = ();
 219
 220     # first setup a few default mappings
 221
 222     open DEFAULTS or die "Cannot open $DEFAULTS";
 223     print "Loading $DEFAULTS\n";
 224     while (<DEFAULTS>)
 225     {
 226         next if /^\#/;  # skip comments
 227         next if /^$/;  # skip empty lines
 228         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 229         {
 230             my @src = map hex, split /,/,$1;
 231             my $dst = $4;
 232             my $comment = $5;
 233             if ($#src > 0) { push @unicode_aliases, \@src; }
 234             next if ($dst eq "none");
 235             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 236             foreach $src (@src)
 237             {
 238                 die "Duplicate value" if defined($unicode_defaults[$src]);
 239                 $unicode_defaults[$src] = $dst;
 240             }
 241             next;
 242         }
 243         die "Unrecognized line $_\n";
 244     }
 245
 246     # now build mappings from the decomposition field of the Unicode database
 247
 248     open UNICODEDATA or die "Cannot open $UNICODEDATA";
 249     print "Loading $UNICODEDATA\n";
 250     while (<UNICODEDATA>)
 251     {
 252         # Decode the fields ...
 253         ($code, $name, $cat, $comb, $bidi,
 254          $decomp, $dec, $dig, $num, $mirror,
 255          $oldname, $comment, $upper, $lower, $title) = split /;/;
 256
 257         my $src = hex $code;
 258
 259         die "unknown category $cat" unless defined $categories{$cat};
 260         die "unknown directionality $bidi" unless defined $directions{$bidi};
 261
 262         $uniname[$src] = $name;
 263         $category_table[$src] = $categories{$cat};
 264         $direction_table[$src] = $directions{$bidi};
 265
 266         if ($lower ne "")
 267         {
 268             $tolower_table[$src] = hex $lower;
 269             $category_table[$src] |= $ctype{"upper"}|$ctype{"alpha"};
 270         }
 271         if ($upper ne "")
 272         {
 273             $toupper_table[$src] = hex $upper;
 274             $category_table[$src] |= $ctype{"lower"}|$ctype{"alpha"};
 275         }
 276         if ($dec ne "")
 277         {
 278             $category_table[$src] |= $ctype{"digit"};
 279         }
 280         if ($dig ne "")
 281         {
 282             $digitmap_table[$src] = ord $dig;
 283         }
 284
 285         # copy the category and direction for everything between First/Last pairs
 286         if ($name =~ /, First>/) { $start = $src; }
 287         if ($name =~ /, Last>/)
 288         {
 289             while ($start < $src)
 290             {
 291                 $category_table[$start] = $category_table[$src];
 292                 $direction_table[$start] = $direction_table[$src];
 293                 $start++;
 294             }
 295         }
 296
 297         next if $decomp eq "";  # no decomposition, skip it
 298
 299         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 300         {
 301             # decomposition of the form "<foo> 1234" -> use char if type is known
 302             if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd))
 303             {
 304                 # Single char decomposition in the compatibility range
 305                 $compatmap_table[$src] = hex $2;
 306             }
 307             next unless ($1 eq "font" ||
 308                          $1 eq "noBreak" ||
 309                          $1 eq "circle" ||
 310                          $1 eq "super" ||
 311                          $1 eq "sub" ||
 312                          $1 eq "wide" ||
 313                          $1 eq "narrow" ||
 314                          $1 eq "compat" ||
 315                          $1 eq "small");
 316             $dst = hex $2;
 317         }
 318         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 319         {
 320             # decomposition "<compat> 0020 1234" -> combining accent
 321             $dst = hex $1;
 322         }
 323         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 324         {
 325             # decomposition contains only char values without prefix -> use first char
 326             $dst = hex $1;
 327             $category_table[$src] |= $category_table[$dst];
 328             # store decomposition if it contains two chars
 329             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 330             {
 331                 $decomp_table[$src] = [ hex $1, hex $2 ];
 332                 push @compose_table, [ hex $1, hex $2, $src ];
 333             }
 334             elsif ($decomp =~ /^(<[a-z]+>\s)*([0-9a-fA-F]+)$/ &&
 335                    (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)))
 336             {
 337                 # Single char decomposition in the compatibility range
 338                 $compatmap_table[$src] = hex $2;
 339             }
 340         }
 341         else
 342         {
 343             next;
 344         }
 345
 346         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 347
 348         # check for loops
 349         for ($i = $dst; ; $i = $unicode_defaults[$i])
 350         {
 351             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 352             last unless defined($unicode_defaults[$i]);
 353         }
 354         $unicode_defaults[$src] = $dst;
 355     }
 356
 357     # patch the category of some special characters
 358
 359     foreach $cat (keys %special_categories)
 360     {
 361         my $flag = $ctype{$cat};
 362         foreach $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 363     }
 364 }
 365
 366
 367 ################################################################
 368 # parse the input file
 369 sub READ_FILE
 370 {
 371     my $name = shift;
 372     open INPUT,$name or die "Cannot open $name";
 373
 374     while (<INPUT>)
 375     {
 376         next if /^\#/;  # skip comments
 377         next if /^$/;  # skip empty lines
 378         next if /\x1a/;  # skip ^Z
 379         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 380
 381         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 382         {
 383             $cp = hex $1;
 384             push @lead_bytes,$cp;
 385             $cp2uni[$cp] = 0;
 386             next;
 387         }
 388         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 389         {
 390             $cp = hex $1;
 391             $uni = hex $2;
 392             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 393             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 394             next;
 395         }
 396         die "$name: Unrecognized line $_\n";
 397     }
 398 }
 399
 400
 401 ################################################################
 402 # fill input data for the 20127 (us-ascii) codepage
 403 sub fill_20127_codepage()
 404 {
 405     for (my $i = 0; $i < 128; $i++) { $cp2uni[$i] = $uni2cp[$i] = $i; }
 406     for (my $i = 128; $i < 256; $i++) { $cp2uni[$i] = $i & 0x7f; }
 407 }
 408
 409
 410 ################################################################
 411 # build EUC-JP table from the JIS 0208 file
 412 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 413 # but this would require 3-byte DBCS characters
 414 sub READ_JIS0208_FILE
 415 {
 416     my $name = shift;
 417
 418     # ASCII chars
 419     for ($i = 0x00; $i <= 0x7f; $i++)
 420     {
 421         $cp2uni[$i] = $i;
 422         $uni2cp[$i] = $i;
 423     }
 424
 425     # JIS X 0201 right plane
 426     for ($i = 0xa1; $i <= 0xdf; $i++)
 427     {
 428         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 429         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 430     }
 431
 432     # lead bytes
 433     foreach $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 434     {
 435         push @lead_bytes,$i;
 436         $cp2uni[$i] = 0;
 437     }
 438
 439     # undefined chars
 440     foreach $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 441     {
 442         $cp2uni[$i] = $DEF_CHAR;
 443     }
 444
 445     # Shift-JIS compatibility
 446     $uni2cp[0x00a5] = 0x5c;
 447     $uni2cp[0x203e] = 0x7e;
 448
 449     # Fix backslash conversion
 450     $cp2uni[0xa1c0] = 0xff3c;
 451     $uni2cp[0xff3c] = 0xa1c0;
 452
 453     open INPUT, "$name" or die "Cannot open $name";
 454     while (<INPUT>)
 455     {
 456         next if /^\#/;  # skip comments
 457         next if /^$/;  # skip empty lines
 458         next if /\x1a/;  # skip ^Z
 459         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 460         {
 461             $cp = 0x8080 + hex $1;
 462             $uni = hex $2;
 463             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 464             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 465             next;
 466         }
 467         die "$name: Unrecognized line $_\n";
 468     }
 469 }
 470
 471
 472 ################################################################
 473 # build the sort keys table
 474 sub READ_SORTKEYS_FILE
 475 {
 476     my @sortkeys = ();
 477     for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
 478
 479     open INPUT, "$SORTKEYS" or die "Cannot open $SORTKEYS";
 480     print "Loading $SORTKEYS\n";
 481     while (<INPUT>)
 482     {
 483         next if /^\#/;  # skip comments
 484         next if /^$/;  # skip empty lines
 485         next if /\x1a/;  # skip ^Z
 486         next if /^\@version/;  # skip @version header
 487         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 488         {
 489             my ($uni,$variable) = (hex $1, $2);
 490             next if $uni > 65535;
 491             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 492             next;
 493         }
 494         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 495         {
 496             # multiple character sequence, ignored for now
 497             next;
 498         }
 499         die "$SORTKEYS: Unrecognized line $_\n";
 500     }
 501     close INPUT;
 502
 503     # compress the keys to 32 bit:
 504     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 505
 506     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 507                        ${$a}[2] <=> ${$b}[2] or
 508                        ${$a}[3] <=> ${$b}[3] or
 509                        ${$a}[4] <=> ${$b}[4] or
 510                        $a cmp $b; } @sortkeys;
 511
 512     my ($n2, $n3) = (1, 1);
 513     my @keys = (-1, -1, -1, -1, -1 );
 514     my @flatkeys = ();
 515
 516     for (my $i = 0; $i < 65536; $i++)
 517     {
 518         my @current = @{$sortkeys[$i]};
 519         next if $current[0] == -1;
 520         if ($current[1] == $keys[1])
 521         {
 522             if ($current[2] == $keys[2])
 523             {
 524                 if ($current[3] == $keys[3])
 525                 {
 526                     # nothing
 527                 }
 528                 else
 529                 {
 530                     $keys[3] = $current[3];
 531                     $n3++;
 532                     die if ($n3 >= 16);
 533                 }
 534             }
 535             else
 536             {
 537                 $keys[2] = $current[2];
 538                 $keys[3] = $current[3];
 539                 $n2++;
 540                 $n3 = 1;
 541                 die if ($n2 >= 256);
 542             }
 543         }
 544         else
 545         {
 546             $keys[1] = $current[1];
 547             $keys[2] = $current[2];
 548             $keys[3] = $current[3];
 549             $n2 = 1;
 550             $n3 = 1;
 551         }
 552
 553         if ($current[2]) { $current[2] = $n2; }
 554         if ($current[3]) { $current[3] = $n3; }
 555         if ($current[4]) { $current[4] = 1; }
 556
 557         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 558     }
 559     return @flatkeys;
 560 }
 561
 562
 563 ################################################################
 564 # build the sort keys table
 565 sub DUMP_SORTKEYS
 566 {
 567     my @keys = @_;
 568
 569     # count the number of 256-key ranges that contain something
 570
 571     my @offsets = ();
 572     my $ranges = 2;
 573     for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
 574     for (my $i = 0; $i < 65536; $i++)
 575     {
 576         next unless defined $keys[$i];
 577         $offsets[$i >> 8] = $ranges * 256;
 578         $ranges++;
 579         $i |= 255;
 580     }
 581
 582     # output the range offsets
 583
 584     open OUTPUT,">collation.c.new" or die "Cannot create collation.c";
 585     printf "Building collation.c\n";
 586     printf OUTPUT "/* Unicode collation element table */\n";
 587     printf OUTPUT "/* generated from %s */\n", $SORTKEYS;
 588     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 589
 590     printf OUTPUT "const unsigned int collation_table[%d] =\n{\n", $ranges*256;
 591     printf OUTPUT "    /* index */\n";
 592     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%08x", 0, @offsets );
 593
 594     # output the default values
 595
 596     printf OUTPUT "    /* defaults */\n";
 597     printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0, (-1) x 256 );
 598
 599     # output all the key ranges
 600
 601     for (my $i = 0; $i < 256; $i++)
 602     {
 603         next if $offsets[$i] == 256;
 604         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 605         printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", -1, @keys[($i<<8) .. ($i<<8)+255] );
 606     }
 607     printf OUTPUT "\n};\n";
 608     close OUTPUT;
 609     save_file("collation.c");
 610 }
 611
 612
 613 ################################################################
 614 # add default mappings once the file had been read
 615 sub ADD_DEFAULT_MAPPINGS
 616 {
 617     # Apply aliases
 618
 619     foreach $alias (@unicode_aliases)
 620     {
 621         my $target = undef;
 622         foreach $src (@$alias)
 623         {
 624             if (defined($uni2cp[$src]))
 625             {
 626                 $target = $uni2cp[$src];
 627                 last;
 628             }
 629         }
 630         next unless defined($target);
 631
 632         # At least one char of the alias set is defined, set the others to the same value
 633         foreach $src (@$alias)
 634         {
 635             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 636         }
 637     }
 638
 639     # For every src -> target mapping in the defaults table,
 640     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 641
 642     for ($src = 0; $src < 65536; $src++)
 643     {
 644         next if defined($uni2cp[$src]);  # source has a definition already
 645         next unless defined($unicode_defaults[$src]);  # no default for this char
 646         my $target = $unicode_defaults[$src];
 647
 648         # do a recursive mapping until we find a target char that is defined
 649         while (!defined($uni2cp[$target]) &&
 650                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 651
 652         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 653     }
 654
 655     # Add an identity mapping for all undefined chars
 656
 657     for ($i = 0; $i < 256; $i++)
 658     {
 659         next if defined($cp2uni[$i]);
 660         next if defined($uni2cp[$i]);
 661         $cp2uni[$i] = $uni2cp[$i] = $i;
 662     }
 663 }
 664
 665 ################################################################
 666 # dump an array of integers
 667 sub DUMP_ARRAY
 668 {
 669     my ($format,$default,@array) = @_;
 670     my $i, $ret = "    ";
 671     for ($i = 0; $i < $#array; $i++)
 672     {
 673         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 674         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 675     }
 676     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 677     return $ret;
 678 }
 679
 680 ################################################################
 681 # dump an SBCS mapping table
 682 sub DUMP_SBCS_TABLE
 683 {
 684     my ($codepage, $name) = @_;
 685     my $i;
 686
 687     # output the ascii->unicode table
 688
 689     printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 690     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 691
 692     # count the number of unicode->ascii subtables that contain something
 693
 694     my @filled = ();
 695     my $subtables = 1;
 696     for ($i = 0; $i < 65536; $i++)
 697     {
 698         next unless defined $uni2cp[$i];
 699         $filled[$i >> 8] = 1;
 700         $subtables++;
 701         $i |= 255;
 702     }
 703
 704     # output all the subtables into a single array
 705
 706     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 707     for ($i = 0; $i < 256; $i++)
 708     {
 709         next unless $filled[$i];
 710         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 711         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
 712     }
 713     printf OUTPUT "    /* defaults */\n";
 714     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($DEF_CHAR) x 256 );
 715
 716     # output a table of the offsets of the subtables in the previous array
 717
 718     my $pos = 0;
 719     my @offsets = ();
 720     for ($i = 0; $i < 256; $i++)
 721     {
 722         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 723         else { push @offsets, ($subtables-1) * 256; }
 724     }
 725     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 726     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 727
 728     # output the code page descriptor
 729
 730     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 731     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 732                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 733     printf OUTPUT "    cp2uni,\n";
 734     printf OUTPUT "    uni2cp_low,\n";
 735     printf OUTPUT "    uni2cp_high\n};\n";
 736 }
 737
 738
 739 ################################################################
 740 # dump a DBCS mapping table
 741 sub DUMP_DBCS_TABLE
 742 {
 743     my ($codepage, $name) = @_;
 744     my $i, $x, $y;
 745
 746     # build a list of lead bytes that are actually used
 747
 748     my @lblist = ();
 749     LBLOOP: for ($y = 0; $y <= $#lead_bytes; $y++)
 750     {
 751         my $base = $lead_bytes[$y] << 8;
 752         for ($x = 0; $x < 256; $x++)
 753         {
 754             if (defined $cp2uni[$base+$x])
 755             {
 756                 push @lblist,$lead_bytes[$y];
 757                 next LBLOOP;
 758             }
 759         }
 760     }
 761     my $unused = ($#lead_bytes > $#lblist);
 762
 763     # output the ascii->unicode table for the single byte chars
 764
 765     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 766     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 767
 768     # output the default table for unused lead bytes
 769
 770     if ($unused)
 771     {
 772         printf OUTPUT "    /* unused lead bytes */\n";
 773         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 774     }
 775
 776     # output the ascii->unicode table for each DBCS lead byte
 777
 778     for ($y = 0; $y <= $#lblist; $y++)
 779     {
 780         my $base = $lblist[$y] << 8;
 781         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
 782         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
 783         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
 784     }
 785
 786     # output the lead byte subtables offsets
 787
 788     my @offsets = ();
 789     for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
 790     for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
 791     if ($unused)
 792     {
 793         # increment all lead bytes offset to take into account the unused table
 794         for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
 795     }
 796     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
 797     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
 798
 799     # count the number of unicode->ascii subtables that contain something
 800
 801     my @filled = ();
 802     my $subtables = 1;
 803     for ($i = 0; $i < 65536; $i++)
 804     {
 805         next unless defined $uni2cp[$i];
 806         $filled[$i >> 8] = 1;
 807         $subtables++;
 808         $i |= 255;
 809     }
 810
 811     # output all the subtables into a single array
 812
 813     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
 814     for ($y = 0; $y < 256; $y++)
 815     {
 816         next unless $filled[$y];
 817         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
 818         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
 819     }
 820     printf OUTPUT "    /* defaults */\n";
 821     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 822
 823     # output a table of the offsets of the subtables in the previous array
 824
 825     my $pos = 0;
 826     my @offsets = ();
 827     for ($y = 0; $y < 256; $y++)
 828     {
 829         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
 830         else { push @offsets, ($subtables-1) * 256; }
 831     }
 832     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 833     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 834
 835     # output the code page descriptor
 836
 837     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
 838     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
 839                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 840     printf OUTPUT "    cp2uni,\n";
 841     printf OUTPUT "    cp2uni_leadbytes,\n";
 842     printf OUTPUT "    uni2cp_low,\n";
 843     printf OUTPUT "    uni2cp_high,\n";
 844     DUMP_LB_RANGES();
 845     printf OUTPUT "};\n";
 846 }
 847
 848
 849 ################################################################
 850 # dump the list of defined lead byte ranges
 851 sub DUMP_LB_RANGES
 852 {
 853     my @list = ();
 854     my $i = 0;
 855     foreach $i (@lead_bytes) { $list[$i] = 1; }
 856     my $on = 0;
 857     printf OUTPUT "    { ";
 858     for ($i = 0; $i < 256; $i++)
 859     {
 860         if ($on)
 861         {
 862             if (!defined $list[$i]) { printf OUTPUT "0x%02x, ", $i-1; $on = 0; }
 863         }
 864         else
 865         {
 866             if ($list[$i]) { printf OUTPUT "0x%02x, ", $i; $on = 1; }
 867         }
 868     }
 869     if ($on) { printf OUTPUT "0xff, "; }
 870     printf OUTPUT "0x00, 0x00 }\n";
 871 }
 872
 873
 874 ################################################################
 875 # dump the case mapping tables
 876 sub DUMP_CASE_MAPPINGS
 877 {
 878     open OUTPUT,">casemap.c.new" or die "Cannot create casemap.c";
 879     printf "Building casemap.c\n";
 880     printf OUTPUT "/* Unicode case mappings */\n";
 881     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 882     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 883
 884     DUMP_CASE_TABLE( "wine_casemap_lower", @tolower_table );
 885     DUMP_CASE_TABLE( "wine_casemap_upper", @toupper_table );
 886     DUMP_CASE_TABLE( "wine_digitmap",  @digitmap_table );
 887     DUMP_CASE_TABLE( "wine_compatmap", @compatmap_table );
 888     close OUTPUT;
 889     save_file("casemap.c");
 890 }
 891
 892
 893 ################################################################
 894 # dump a case mapping table
 895 sub DUMP_CASE_TABLE
 896 {
 897     my ($name,@table) = @_;
 898
 899     # count the number of sub tables that contain something
 900     # also compute the low and upper populated bounds
 901
 902     my @lowerbounds = ( 0, 0 );
 903     my @upperbounds = ( 0, 255 );
 904     my $index = 0;
 905     my @filled = ();
 906     for ($i = 0; $i < 65536; $i++)
 907     {
 908         next unless defined $table[$i];
 909         if (!defined $filled[$i >> 8])
 910         {
 911           $lowerbounds[$index] = $i & 0xff;
 912           $upperbounds[$index] = 0xff - $lowerbounds[$index];
 913           $filled[$i >> 8] = $index * 256 + 512;
 914           $index++;
 915         }
 916         else
 917         {
 918           $upperbounds[$index-1] = 0xff - ($i & 0xff);
 919         }
 920         $table[$i] = ($table[$i] - $i) & 0xffff;
 921     }
 922
 923     # Collapse blocks upwards if possible
 924     my $removed = 0;
 925     $index = 0;
 926     for ($i = 0; $i < 256; $i++)
 927     {
 928         next unless defined $filled[$i];
 929         if ($upperbounds[$index - 1] > $lowerbounds[$index])
 930         {
 931            $removed = $removed + $lowerbounds[$index];
 932         }
 933         else
 934         {
 935            $removed = $removed + $upperbounds[$index - 1];
 936            $lowerbounds[$index] = $upperbounds[$index - 1];
 937         }
 938         $filled[$i] = $filled[$i] - $removed;
 939         $index++;
 940     }
 941
 942     # dump the table
 943
 944     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
 945     printf OUTPUT "{\n    /* index */\n";
 946     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
 947     printf OUTPUT "    /* defaults */\n";
 948     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
 949     $index = 0;
 950     for ($i = 0; $i < 256; $i++)
 951     {
 952         next unless $filled[$i];
 953         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
 954         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0,
 955                       @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
 956         $index++;
 957     }
 958     printf OUTPUT "\n};\n";
 959 }
 960
 961
 962 ################################################################
 963 # dump the ctype tables
 964 sub DUMP_CTYPE_TABLES
 965 {
 966     open OUTPUT,">wctype.c.new" or die "Cannot create wctype.c";
 967     printf "Building wctype.c\n";
 968     printf OUTPUT "/* Unicode ctype tables */\n";
 969     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 970     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 971
 972     my $i;
 973     my @array = (0) x 256;
 974
 975     # add the direction in the high 4 bits of the category
 976     for ($i = 0; $i < 65536; $i++)
 977     {
 978         $category_table[$i] |= $direction_table[$i] << 12;
 979     }
 980
 981     # try to merge table rows
 982     for ($row = 0; $row < 256; $row++)
 983     {
 984         my $rowtxt = sprintf "%04x" x 256, @category_table[($row<<8)..($row<<8)+255];
 985         if (defined($sequences{$rowtxt}))
 986         {
 987             # reuse an existing row
 988             $array[$row] = $sequences{$rowtxt};
 989         }
 990         else
 991         {
 992             # create a new row
 993             $sequences{$rowtxt} = $array[$row] = $#array + 1;
 994             push @array, @category_table[($row<<8)..($row<<8)+255];
 995         }
 996     }
 997
 998     printf OUTPUT "const unsigned short wine_wctype_table[%d] =\n{\n", $#array+1;
 999     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1000     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1001
1002     close OUTPUT;
1003     save_file("wctype.c");
1004 }
1005
1006
1007 ################################################################
1008 # dump the char composition tables
1009 sub DUMP_COMPOSE_TABLES
1010 {
1011     open OUTPUT,">compose.c.new" or die "Cannot create compose.c";
1012     printf "Building compose.c\n";
1013     printf OUTPUT "/* Unicode char composition */\n";
1014     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1015     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1016
1017     ######### composition table
1018
1019     my @filled = ();
1020     foreach $i (@compose_table)
1021     {
1022         my @comp = @$i;
1023         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
1024     }
1025
1026     # count how many different second chars we have
1027
1028     for ($i = $count = 0; $i < 65536; $i++)
1029     {
1030         next unless defined $filled[$i];
1031         $count++;
1032     }
1033
1034     # build the table of second chars and offsets
1035
1036     my $pos = $count + 1;
1037     for ($i = 0; $i < 65536; $i++)
1038     {
1039         next unless defined $filled[$i];
1040         push @table, $i, $pos;
1041         $pos += @{$filled[$i]};
1042     }
1043     # terminator with last position
1044     push @table, 0, $pos;
1045     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
1046     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
1047
1048     # build the table of first chars and mappings
1049
1050     for ($i = 0; $i < 65536; $i++)
1051     {
1052         next unless defined $filled[$i];
1053         my @table = ();
1054         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
1055         for ($j = 0; $j <= $#list; $j++)
1056         {
1057             push @table, $list[$j][0], $list[$j][1];
1058         }
1059         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
1060     }
1061     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
1062
1063     ######### decomposition table
1064
1065     # first determine all the 16-char subsets that contain something
1066
1067     my @filled = (0) x 4096;
1068     my $pos = 16*2;  # for the null subset
1069     for ($i = 0; $i < 65536; $i++)
1070     {
1071         next unless defined $decomp_table[$i];
1072         $filled[$i >> 4] = $pos;
1073         $pos += 16*2;
1074         $i |= 15;
1075     }
1076     my $total = $pos;
1077
1078     # now count the 256-char subsets that contain something
1079
1080     my @filled_idx = (256) x 256;
1081     $pos = 256 + 16;
1082     for ($i = 0; $i < 4096; $i++)
1083     {
1084         next unless $filled[$i];
1085         $filled_idx[$i >> 4] = $pos;
1086         $pos += 16;
1087         $i |= 15;
1088     }
1089     my $null_offset = $pos;  # null mapping
1090     $total += $pos;
1091
1092     # add the index offsets to the subsets positions
1093
1094     for ($i = 0; $i < 4096; $i++)
1095     {
1096         next unless $filled[$i];
1097         $filled[$i] += $null_offset;
1098     }
1099
1100     # dump the main index
1101
1102     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
1103     printf OUTPUT "{\n    /* index */\n";
1104     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1105     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1106
1107     # dump the second-level indexes
1108
1109     for ($i = 0; $i < 256; $i++)
1110     {
1111         next unless ($filled_idx[$i] > 256);
1112         my @table = @filled[($i<<4)..($i<<4)+15];
1113         for ($j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1114         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1115         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1116     }
1117
1118     # dump the 16-char subsets
1119
1120     printf OUTPUT ",\n    /* null mapping */\n";
1121     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
1122
1123     for ($i = 0; $i < 4096; $i++)
1124     {
1125         next unless $filled[$i];
1126         my @table = (0) x 32;
1127         for ($j = 0; $j < 16; $j++)
1128         {
1129             if (defined $decomp_table[($i<<4) + $j])
1130             {
1131                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
1132                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
1133             }
1134         }
1135         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1136         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1137     }
1138
1139     printf OUTPUT "\n};\n";
1140     close OUTPUT;
1141     save_file("compose.c");
1142 }
1143
1144
1145 ################################################################
1146 # read an input file and generate the corresponding .c file
1147 sub HANDLE_FILE
1148 {
1149     my ($codepage,$filename,$comment) = @_;
1150
1151     @cp2uni = ();
1152     @lead_bytes = ();
1153     @uni2cp = ();
1154
1155     # symbol codepage file is special
1156     if ($codepage == 20932) { READ_JIS0208_FILE($MAPPREFIX . $filename); }
1157     elsif ($codepage == 20127) { fill_20127_codepage(); }
1158     else { READ_FILE($MAPPREFIX . $filename); }
1159
1160     # hack: 0x00a5 must map to backslash in Shift-JIS
1161     if ($codepage == 932) { $uni2cp[0x00a5] = 0x5c; }
1162
1163     ADD_DEFAULT_MAPPINGS();
1164
1165     my $output = sprintf "c_%03d.c", $codepage;
1166     open OUTPUT,">$output.new" or die "Cannot create $output";
1167
1168     printf "Building %s from %s (%s)\n", $output, $filename || "hardcoded data", $comment;
1169
1170     # dump all tables
1171
1172     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
1173     printf OUTPUT "/* generated from %s */\n", $MAPPREFIX . $filename if $filename;
1174     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
1175     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1176
1177     if ($#lead_bytes == -1) { DUMP_SBCS_TABLE( $codepage, $comment ); }
1178     else { DUMP_DBCS_TABLE( $codepage, $comment ); }
1179     close OUTPUT;
1180     save_file($output);
1181 }
1182
1183
1184 ################################################################
1185 # save a file if modified
1186 sub save_file($)
1187 {
1188     my $file = shift;
1189     if (!system "cmp $file $file.new >/dev/null")
1190     {
1191         unlink "$file.new";
1192     }
1193     else
1194     {
1195         rename "$file.new", "$file";
1196     }
1197 }
1198
1199
1200 ################################################################
1201 # output the list of codepage tables into the cptable.c file
1202 sub OUTPUT_CPTABLE
1203 {
1204     @tables_decl = ();
1205
1206     foreach $file (@allfiles)
1207     {
1208         my ($codepage,$filename,$comment) = @$file;
1209         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
1210     }
1211
1212     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
1213     foreach $file (@allfiles)
1214     {
1215         my ($codepage,$filename,$comment) = @$file;
1216         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
1217     }
1218     push @tables_decl, "};";
1219     REPLACE_IN_FILE( "cptable.c", @tables_decl );
1220 }
1221
1222 ################################################################
1223 # replace the contents of a file between ### cpmap ### marks
1224
1225 sub REPLACE_IN_FILE
1226 {
1227     my $name = shift;
1228     my @data = @_;
1229     my @lines = ();
1230     open(FILE,$name) or die "Can't open $name";
1231     while (<FILE>)
1232     {
1233         push @lines, $_;
1234         last if /\#\#\# cpmap begin \#\#\#/;
1235     }
1236     push @lines, @data;
1237     while (<FILE>)
1238     {
1239         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
1240     }
1241     push @lines, <FILE>;
1242     open(FILE,">$name.new") or die "Can't modify $name";
1243     print FILE @lines;
1244     close(FILE);
1245     save_file($name);
1246 }