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",   0, "IBM EBCDIC US Canada" ],
  41     [ 424,   "VENDORS/MISC/CP424.TXT",            0, "IBM EBCDIC Hebrew" ],
  42     [ 437,   "VENDORS/MICSFT/PC/CP437.TXT",       1, "OEM United States" ],
  43     [ 500,   "VENDORS/MICSFT/EBCDIC/CP500.TXT",   0, "IBM EBCDIC International" ],
  44     [ 737,   "VENDORS/MICSFT/PC/CP737.TXT",       1, "OEM Greek 437G" ],
  45     [ 775,   "VENDORS/MICSFT/PC/CP775.TXT",       1, "OEM Baltic" ],
  46     [ 850,   "VENDORS/MICSFT/PC/CP850.TXT",       1, "OEM Multilingual Latin 1" ],
  47     [ 852,   "VENDORS/MICSFT/PC/CP852.TXT",       1, "OEM Slovak Latin 2" ],
  48     [ 855,   "VENDORS/MICSFT/PC/CP855.TXT",       1, "OEM Cyrillic" ],
  49     [ 856,   "VENDORS/MISC/CP856.TXT",            0, "Hebrew PC" ],
  50     [ 857,   "VENDORS/MICSFT/PC/CP857.TXT",       1, "OEM Turkish" ],
  51     [ 860,   "VENDORS/MICSFT/PC/CP860.TXT",       1, "OEM Portuguese" ],
  52     [ 861,   "VENDORS/MICSFT/PC/CP861.TXT",       1, "OEM Icelandic" ],
  53     [ 862,   "VENDORS/MICSFT/PC/CP862.TXT",       1, "OEM Hebrew" ],
  54     [ 863,   "VENDORS/MICSFT/PC/CP863.TXT",       1, "OEM Canadian French" ],
  55     [ 864,   "VENDORS/MICSFT/PC/CP864.TXT",       0, "OEM Arabic" ],
  56     [ 865,   "VENDORS/MICSFT/PC/CP865.TXT",       1, "OEM Nordic" ],
  57     [ 866,   "VENDORS/MICSFT/PC/CP866.TXT",       1, "OEM Russian" ],
  58     [ 869,   "VENDORS/MICSFT/PC/CP869.TXT",       1, "OEM Greek" ],
  59     [ 874,   "VENDORS/MICSFT/PC/CP874.TXT",       1, "ANSI/OEM Thai" ],
  60     [ 875,   "VENDORS/MICSFT/EBCDIC/CP875.TXT",   0, "IBM EBCDIC Greek" ],
  61     [ 878,   "VENDORS/MISC/KOI8-R.TXT",           0, "Russian KOI8" ],
  62     [ 932,   "VENDORS/MICSFT/WINDOWS/CP932.TXT",  0, "ANSI/OEM Japanese Shift-JIS" ],
  63     [ 936,   "VENDORS/MICSFT/WINDOWS/CP936.TXT",  0, "ANSI/OEM Simplified Chinese GBK" ],
  64     [ 949,   "VENDORS/MICSFT/WINDOWS/CP949.TXT",  0, "ANSI/OEM Korean Unified Hangul" ],
  65     [ 950,   "VENDORS/MICSFT/WINDOWS/CP950.TXT",  0, "ANSI/OEM Traditional Chinese Big5" ],
  66     [ 1006,  "VENDORS/MISC/CP1006.TXT",           0, "IBM Arabic" ],
  67     [ 1026,  "VENDORS/MICSFT/EBCDIC/CP1026.TXT",  0, "IBM EBCDIC Latin 5 Turkish" ],
  68     [ 1250,  "VENDORS/MICSFT/WINDOWS/CP1250.TXT", 0, "ANSI Eastern Europe" ],
  69     [ 1251,  "VENDORS/MICSFT/WINDOWS/CP1251.TXT", 0, "ANSI Cyrillic" ],
  70     [ 1252,  "VENDORS/MICSFT/WINDOWS/CP1252.TXT", 0, "ANSI Latin 1" ],
  71     [ 1253,  "VENDORS/MICSFT/WINDOWS/CP1253.TXT", 0, "ANSI Greek" ],
  72     [ 1254,  "VENDORS/MICSFT/WINDOWS/CP1254.TXT", 0, "ANSI Turkish" ],
  73     [ 1255,  "VENDORS/MICSFT/WINDOWS/CP1255.TXT", 0, "ANSI Hebrew" ],
  74     [ 1256,  "VENDORS/MICSFT/WINDOWS/CP1256.TXT", 0, "ANSI Arabic" ],
  75     [ 1257,  "VENDORS/MICSFT/WINDOWS/CP1257.TXT", 0, "ANSI Baltic" ],
  76     [ 1258,  "VENDORS/MICSFT/WINDOWS/CP1258.TXT", 0, "ANSI/OEM Viet Nam" ],
  77     [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT",      0, "Mac Roman" ],
  78     [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT",      0, "Mac Greek" ],
  79     [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT",   0, "Mac Cyrillic" ],
  80     [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT",     0, "Mac Latin 2" ],
  81     [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT",    0, "Mac Icelandic" ],
  82     [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT",    0, "Mac Turkish" ],
  83     [ 20127, undef,                               0, "US-ASCII (7bit)" ],
  84     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           0, "Russian KOI8" ],
  85     [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", 0, "EUC-JP" ],
  86     [ 21866, "VENDORS/MISC/KOI8-U.TXT",           0, "Ukrainian KOI8" ],
  87     [ 28591, "ISO8859/8859-1.TXT",                0, "ISO 8859-1 Latin 1" ],
  88     [ 28592, "ISO8859/8859-2.TXT",                0, "ISO 8859-2 Latin 2 (East European)" ],
  89     [ 28593, "ISO8859/8859-3.TXT",                0, "ISO 8859-3 Latin 3 (South European)" ],
  90     [ 28594, "ISO8859/8859-4.TXT",                0, "ISO 8859-4 Latin 4 (Baltic old)" ],
  91     [ 28595, "ISO8859/8859-5.TXT",                0, "ISO 8859-5 Cyrillic" ],
  92     [ 28596, "ISO8859/8859-6.TXT",                0, "ISO 8859-6 Arabic" ],
  93     [ 28597, "ISO8859/8859-7.TXT",                0, "ISO 8859-7 Greek" ],
  94     [ 28598, "ISO8859/8859-8.TXT",                0, "ISO 8859-8 Hebrew" ],
  95     [ 28599, "ISO8859/8859-9.TXT",                0, "ISO 8859-9 Latin 5 (Turkish)" ],
  96     [ 28600, "ISO8859/8859-10.TXT",               0, "ISO 8859-10 Latin 6 (Nordic)" ],
  97     [ 28603, "ISO8859/8859-13.TXT",               0, "ISO 8859-13 Latin 7 (Baltic)" ],
  98     [ 28604, "ISO8859/8859-14.TXT",               0, "ISO 8859-14 Latin 8 (Celtic)" ],
  99     [ 28605, "ISO8859/8859-15.TXT",               0, "ISO 8859-15 Latin 9 (Euro)" ],
 100     [ 28606, "ISO8859/8859-16.TXT",               0, "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 # get a mapping including glyph chars for MB_USEGLYPHCHARS
 411
 412 sub get_glyphs_mapping(@)
 413 {
 414     $_[0x01] = 0x263a;  # (WHITE SMILING FACE)
 415     $_[0x02] = 0x263b;  # (BLACK SMILING FACE)
 416     $_[0x03] = 0x2665;  # (BLACK HEART SUIT)
 417     $_[0x04] = 0x2666;  # (BLACK DIAMOND SUIT)
 418     $_[0x05] = 0x2663;  # (BLACK CLUB SUIT)
 419     $_[0x06] = 0x2660;  # (BLACK SPADE SUIT)
 420     $_[0x07] = 0x2022;  # (BULLET)
 421     $_[0x08] = 0x25d8;  # (INVERSE BULLET)
 422     $_[0x09] = 0x25cb;  # (WHITE CIRCLE)
 423     $_[0x0a] = 0x25d9;  # (INVERSE WHITE CIRCLE)
 424     $_[0x0b] = 0x2642;  # (MALE SIGN)
 425     $_[0x0c] = 0x2640;  # (FEMALE SIGN)
 426     $_[0x0d] = 0x266a;  # (EIGHTH NOTE)
 427     $_[0x0e] = 0x266b;  # (BEAMED EIGHTH NOTES)
 428     $_[0x0f] = 0x263c;  # (WHITE SUN WITH RAYS)
 429     $_[0x10] = 0x25ba;  # (BLACK RIGHT-POINTING POINTER)
 430     $_[0x11] = 0x25c4;  # (BLACK LEFT-POINTING POINTER)
 431     $_[0x12] = 0x2195;  # (UP DOWN ARROW)
 432     $_[0x13] = 0x203c;  # (DOUBLE EXCLAMATION MARK)
 433     $_[0x14] = 0x00b6;  # (PILCROW SIGN)
 434     $_[0x15] = 0x00a7;  # (SECTION SIGN)
 435     $_[0x16] = 0x25ac;  # (BLACK RECTANGLE)
 436     $_[0x17] = 0x21a8;  # (UP DOWN ARROW WITH BASE)
 437     $_[0x18] = 0x2191;  # (UPWARDS ARROW)
 438     $_[0x19] = 0x2193;  # (DOWNWARDS ARROW)
 439     $_[0x1a] = 0x2192;  # (RIGHTWARDS ARROW)
 440     $_[0x1b] = 0x2190;  # (LEFTWARDS ARROW)
 441     $_[0x1c] = 0x221f;  # (RIGHT ANGLE)
 442     $_[0x1d] = 0x2194;  # (LEFT RIGHT ARROW)
 443     $_[0x1e] = 0x25b2;  # (BLACK UP-POINTING TRIANGLE)
 444     $_[0x1f] = 0x25bc;  # (BLACK DOWN-POINTING TRIANGLE)
 445     $_[0x7f] = 0x2302;  # (HOUSE)
 446     return @_;
 447 }
 448
 449 ################################################################
 450 # build EUC-JP table from the JIS 0208 file
 451 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 452 # but this would require 3-byte DBCS characters
 453 sub READ_JIS0208_FILE
 454 {
 455     my $name = shift;
 456
 457     # ASCII chars
 458     for ($i = 0x00; $i <= 0x7f; $i++)
 459     {
 460         $cp2uni[$i] = $i;
 461         $uni2cp[$i] = $i;
 462     }
 463
 464     # JIS X 0201 right plane
 465     for ($i = 0xa1; $i <= 0xdf; $i++)
 466     {
 467         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 468         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 469     }
 470
 471     # lead bytes
 472     foreach $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 473     {
 474         push @lead_bytes,$i;
 475         $cp2uni[$i] = 0;
 476     }
 477
 478     # undefined chars
 479     foreach $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 480     {
 481         $cp2uni[$i] = $DEF_CHAR;
 482     }
 483
 484     # Shift-JIS compatibility
 485     $uni2cp[0x00a5] = 0x5c;
 486     $uni2cp[0x203e] = 0x7e;
 487
 488     # Fix backslash conversion
 489     $cp2uni[0xa1c0] = 0xff3c;
 490     $uni2cp[0xff3c] = 0xa1c0;
 491
 492     open INPUT, "$name" or die "Cannot open $name";
 493     while (<INPUT>)
 494     {
 495         next if /^\#/;  # skip comments
 496         next if /^$/;  # skip empty lines
 497         next if /\x1a/;  # skip ^Z
 498         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 499         {
 500             $cp = 0x8080 + hex $1;
 501             $uni = hex $2;
 502             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 503             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 504             next;
 505         }
 506         die "$name: Unrecognized line $_\n";
 507     }
 508 }
 509
 510
 511 ################################################################
 512 # build the sort keys table
 513 sub READ_SORTKEYS_FILE
 514 {
 515     my @sortkeys = ();
 516     for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
 517
 518     open INPUT, "$SORTKEYS" or die "Cannot open $SORTKEYS";
 519     print "Loading $SORTKEYS\n";
 520     while (<INPUT>)
 521     {
 522         next if /^\#/;  # skip comments
 523         next if /^$/;  # skip empty lines
 524         next if /\x1a/;  # skip ^Z
 525         next if /^\@version/;  # skip @version header
 526         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 527         {
 528             my ($uni,$variable) = (hex $1, $2);
 529             next if $uni > 65535;
 530             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 531             next;
 532         }
 533         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 534         {
 535             # multiple character sequence, ignored for now
 536             next;
 537         }
 538         die "$SORTKEYS: Unrecognized line $_\n";
 539     }
 540     close INPUT;
 541
 542     # compress the keys to 32 bit:
 543     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 544
 545     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 546                        ${$a}[2] <=> ${$b}[2] or
 547                        ${$a}[3] <=> ${$b}[3] or
 548                        ${$a}[4] <=> ${$b}[4] or
 549                        $a cmp $b; } @sortkeys;
 550
 551     my ($n2, $n3) = (1, 1);
 552     my @keys = (-1, -1, -1, -1, -1 );
 553     my @flatkeys = ();
 554
 555     for (my $i = 0; $i < 65536; $i++)
 556     {
 557         my @current = @{$sortkeys[$i]};
 558         next if $current[0] == -1;
 559         if ($current[1] == $keys[1])
 560         {
 561             if ($current[2] == $keys[2])
 562             {
 563                 if ($current[3] == $keys[3])
 564                 {
 565                     # nothing
 566                 }
 567                 else
 568                 {
 569                     $keys[3] = $current[3];
 570                     $n3++;
 571                     die if ($n3 >= 16);
 572                 }
 573             }
 574             else
 575             {
 576                 $keys[2] = $current[2];
 577                 $keys[3] = $current[3];
 578                 $n2++;
 579                 $n3 = 1;
 580                 die if ($n2 >= 256);
 581             }
 582         }
 583         else
 584         {
 585             $keys[1] = $current[1];
 586             $keys[2] = $current[2];
 587             $keys[3] = $current[3];
 588             $n2 = 1;
 589             $n3 = 1;
 590         }
 591
 592         if ($current[2]) { $current[2] = $n2; }
 593         if ($current[3]) { $current[3] = $n3; }
 594         if ($current[4]) { $current[4] = 1; }
 595
 596         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 597     }
 598     return @flatkeys;
 599 }
 600
 601
 602 ################################################################
 603 # build the sort keys table
 604 sub DUMP_SORTKEYS
 605 {
 606     my @keys = @_;
 607
 608     # count the number of 256-key ranges that contain something
 609
 610     my @offsets = ();
 611     my $ranges = 2;
 612     for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
 613     for (my $i = 0; $i < 65536; $i++)
 614     {
 615         next unless defined $keys[$i];
 616         $offsets[$i >> 8] = $ranges * 256;
 617         $ranges++;
 618         $i |= 255;
 619     }
 620
 621     # output the range offsets
 622
 623     open OUTPUT,">collation.c.new" or die "Cannot create collation.c";
 624     printf "Building collation.c\n";
 625     printf OUTPUT "/* Unicode collation element table */\n";
 626     printf OUTPUT "/* generated from %s */\n", $SORTKEYS;
 627     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 628
 629     printf OUTPUT "const unsigned int collation_table[%d] =\n{\n", $ranges*256;
 630     printf OUTPUT "    /* index */\n";
 631     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%08x", 0, @offsets );
 632
 633     # output the default values
 634
 635     printf OUTPUT "    /* defaults */\n";
 636     printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0, (-1) x 256 );
 637
 638     # output all the key ranges
 639
 640     for (my $i = 0; $i < 256; $i++)
 641     {
 642         next if $offsets[$i] == 256;
 643         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 644         printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", -1, @keys[($i<<8) .. ($i<<8)+255] );
 645     }
 646     printf OUTPUT "\n};\n";
 647     close OUTPUT;
 648     save_file("collation.c");
 649 }
 650
 651
 652 ################################################################
 653 # add default mappings once the file had been read
 654 sub ADD_DEFAULT_MAPPINGS
 655 {
 656     # Apply aliases
 657
 658     foreach $alias (@unicode_aliases)
 659     {
 660         my $target = undef;
 661         foreach $src (@$alias)
 662         {
 663             if (defined($uni2cp[$src]))
 664             {
 665                 $target = $uni2cp[$src];
 666                 last;
 667             }
 668         }
 669         next unless defined($target);
 670
 671         # At least one char of the alias set is defined, set the others to the same value
 672         foreach $src (@$alias)
 673         {
 674             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 675         }
 676     }
 677
 678     # For every src -> target mapping in the defaults table,
 679     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 680
 681     for ($src = 0; $src < 65536; $src++)
 682     {
 683         next if defined($uni2cp[$src]);  # source has a definition already
 684         next unless defined($unicode_defaults[$src]);  # no default for this char
 685         my $target = $unicode_defaults[$src];
 686
 687         # do a recursive mapping until we find a target char that is defined
 688         while (!defined($uni2cp[$target]) &&
 689                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 690
 691         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 692     }
 693
 694     # Add an identity mapping for all undefined chars
 695
 696     for ($i = 0; $i < 256; $i++)
 697     {
 698         next if defined($cp2uni[$i]);
 699         next if defined($uni2cp[$i]);
 700         $cp2uni[$i] = $uni2cp[$i] = $i;
 701     }
 702 }
 703
 704 ################################################################
 705 # dump an array of integers
 706 sub DUMP_ARRAY
 707 {
 708     my ($format,$default,@array) = @_;
 709     my $i, $ret = "    ";
 710     for ($i = 0; $i < $#array; $i++)
 711     {
 712         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 713         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 714     }
 715     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 716     return $ret;
 717 }
 718
 719 ################################################################
 720 # dump an SBCS mapping table
 721 sub DUMP_SBCS_TABLE
 722 {
 723     my ($codepage, $has_glyphs, $name) = @_;
 724     my $i;
 725
 726     # output the ascii->unicode table
 727
 728     if ($has_glyphs)
 729     {
 730         printf OUTPUT "static const WCHAR cp2uni[512] =\n";
 731         printf OUTPUT "{\n%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 732         printf OUTPUT ",\n    /* glyphs */\n%s\n};\n\n",
 733                       DUMP_ARRAY( "0x%04x", $DEF_CHAR, get_glyphs_mapping(@cp2uni[0 .. 255]) );
 734     }
 735     else
 736     {
 737         printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 738         printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 739     }
 740
 741     # count the number of unicode->ascii subtables that contain something
 742
 743     my @filled = ();
 744     my $subtables = 1;
 745     for ($i = 0; $i < 65536; $i++)
 746     {
 747         next unless defined $uni2cp[$i];
 748         $filled[$i >> 8] = 1;
 749         $subtables++;
 750         $i |= 255;
 751     }
 752
 753     # output all the subtables into a single array
 754
 755     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 756     for ($i = 0; $i < 256; $i++)
 757     {
 758         next unless $filled[$i];
 759         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 760         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
 761     }
 762     printf OUTPUT "    /* defaults */\n";
 763     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($DEF_CHAR) x 256 );
 764
 765     # output a table of the offsets of the subtables in the previous array
 766
 767     my $pos = 0;
 768     my @offsets = ();
 769     for ($i = 0; $i < 256; $i++)
 770     {
 771         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 772         else { push @offsets, ($subtables-1) * 256; }
 773     }
 774     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 775     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 776
 777     # output the code page descriptor
 778
 779     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 780     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 781                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 782     printf OUTPUT "    cp2uni,\n";
 783     if ($has_glyphs) { printf OUTPUT "    cp2uni + 256,\n"; }
 784     else { printf OUTPUT "    cp2uni,\n"; }
 785     printf OUTPUT "    uni2cp_low,\n";
 786     printf OUTPUT "    uni2cp_high\n};\n";
 787 }
 788
 789
 790 ################################################################
 791 # dump a DBCS mapping table
 792 sub DUMP_DBCS_TABLE
 793 {
 794     my ($codepage, $name) = @_;
 795     my $i, $x, $y;
 796
 797     # build a list of lead bytes that are actually used
 798
 799     my @lblist = ();
 800     LBLOOP: for ($y = 0; $y <= $#lead_bytes; $y++)
 801     {
 802         my $base = $lead_bytes[$y] << 8;
 803         for ($x = 0; $x < 256; $x++)
 804         {
 805             if (defined $cp2uni[$base+$x])
 806             {
 807                 push @lblist,$lead_bytes[$y];
 808                 next LBLOOP;
 809             }
 810         }
 811     }
 812     my $unused = ($#lead_bytes > $#lblist);
 813
 814     # output the ascii->unicode table for the single byte chars
 815
 816     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 817     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 818
 819     # output the default table for unused lead bytes
 820
 821     if ($unused)
 822     {
 823         printf OUTPUT "    /* unused lead bytes */\n";
 824         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 825     }
 826
 827     # output the ascii->unicode table for each DBCS lead byte
 828
 829     for ($y = 0; $y <= $#lblist; $y++)
 830     {
 831         my $base = $lblist[$y] << 8;
 832         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
 833         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
 834         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
 835     }
 836
 837     # output the lead byte subtables offsets
 838
 839     my @offsets = ();
 840     for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
 841     for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
 842     if ($unused)
 843     {
 844         # increment all lead bytes offset to take into account the unused table
 845         for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
 846     }
 847     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
 848     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
 849
 850     # count the number of unicode->ascii subtables that contain something
 851
 852     my @filled = ();
 853     my $subtables = 1;
 854     for ($i = 0; $i < 65536; $i++)
 855     {
 856         next unless defined $uni2cp[$i];
 857         $filled[$i >> 8] = 1;
 858         $subtables++;
 859         $i |= 255;
 860     }
 861
 862     # output all the subtables into a single array
 863
 864     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
 865     for ($y = 0; $y < 256; $y++)
 866     {
 867         next unless $filled[$y];
 868         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
 869         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
 870     }
 871     printf OUTPUT "    /* defaults */\n";
 872     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 873
 874     # output a table of the offsets of the subtables in the previous array
 875
 876     my $pos = 0;
 877     my @offsets = ();
 878     for ($y = 0; $y < 256; $y++)
 879     {
 880         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
 881         else { push @offsets, ($subtables-1) * 256; }
 882     }
 883     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 884     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 885
 886     # output the code page descriptor
 887
 888     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
 889     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
 890                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 891     printf OUTPUT "    cp2uni,\n";
 892     printf OUTPUT "    cp2uni_leadbytes,\n";
 893     printf OUTPUT "    uni2cp_low,\n";
 894     printf OUTPUT "    uni2cp_high,\n";
 895     DUMP_LB_RANGES();
 896     printf OUTPUT "};\n";
 897 }
 898
 899
 900 ################################################################
 901 # dump the list of defined lead byte ranges
 902 sub DUMP_LB_RANGES
 903 {
 904     my @list = ();
 905     my $i = 0;
 906     foreach $i (@lead_bytes) { $list[$i] = 1; }
 907     my $on = 0;
 908     printf OUTPUT "    { ";
 909     for ($i = 0; $i < 256; $i++)
 910     {
 911         if ($on)
 912         {
 913             if (!defined $list[$i]) { printf OUTPUT "0x%02x, ", $i-1; $on = 0; }
 914         }
 915         else
 916         {
 917             if ($list[$i]) { printf OUTPUT "0x%02x, ", $i; $on = 1; }
 918         }
 919     }
 920     if ($on) { printf OUTPUT "0xff, "; }
 921     printf OUTPUT "0x00, 0x00 }\n";
 922 }
 923
 924
 925 ################################################################
 926 # dump the case mapping tables
 927 sub DUMP_CASE_MAPPINGS
 928 {
 929     open OUTPUT,">casemap.c.new" or die "Cannot create casemap.c";
 930     printf "Building casemap.c\n";
 931     printf OUTPUT "/* Unicode case mappings */\n";
 932     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 933     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 934
 935     DUMP_CASE_TABLE( "wine_casemap_lower", @tolower_table );
 936     DUMP_CASE_TABLE( "wine_casemap_upper", @toupper_table );
 937     DUMP_CASE_TABLE( "wine_digitmap",  @digitmap_table );
 938     DUMP_CASE_TABLE( "wine_compatmap", @compatmap_table );
 939     close OUTPUT;
 940     save_file("casemap.c");
 941 }
 942
 943
 944 ################################################################
 945 # dump a case mapping table
 946 sub DUMP_CASE_TABLE
 947 {
 948     my ($name,@table) = @_;
 949
 950     # count the number of sub tables that contain something
 951     # also compute the low and upper populated bounds
 952
 953     my @lowerbounds = ( 0, 0 );
 954     my @upperbounds = ( 0, 255 );
 955     my $index = 0;
 956     my @filled = ();
 957     for ($i = 0; $i < 65536; $i++)
 958     {
 959         next unless defined $table[$i];
 960         if (!defined $filled[$i >> 8])
 961         {
 962           $lowerbounds[$index] = $i & 0xff;
 963           $upperbounds[$index] = 0xff - $lowerbounds[$index];
 964           $filled[$i >> 8] = $index * 256 + 512;
 965           $index++;
 966         }
 967         else
 968         {
 969           $upperbounds[$index-1] = 0xff - ($i & 0xff);
 970         }
 971         $table[$i] = ($table[$i] - $i) & 0xffff;
 972     }
 973
 974     # Collapse blocks upwards if possible
 975     my $removed = 0;
 976     $index = 0;
 977     for ($i = 0; $i < 256; $i++)
 978     {
 979         next unless defined $filled[$i];
 980         if ($upperbounds[$index - 1] > $lowerbounds[$index])
 981         {
 982            $removed = $removed + $lowerbounds[$index];
 983         }
 984         else
 985         {
 986            $removed = $removed + $upperbounds[$index - 1];
 987            $lowerbounds[$index] = $upperbounds[$index - 1];
 988         }
 989         $filled[$i] = $filled[$i] - $removed;
 990         $index++;
 991     }
 992
 993     # dump the table
 994
 995     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
 996     printf OUTPUT "{\n    /* index */\n";
 997     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
 998     printf OUTPUT "    /* defaults */\n";
 999     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
1000     $index = 0;
1001     for ($i = 0; $i < 256; $i++)
1002     {
1003         next unless $filled[$i];
1004         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
1005         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0,
1006                       @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
1007         $index++;
1008     }
1009     printf OUTPUT "\n};\n";
1010 }
1011
1012
1013 ################################################################
1014 # dump the ctype tables
1015 sub DUMP_CTYPE_TABLES
1016 {
1017     open OUTPUT,">wctype.c.new" or die "Cannot create wctype.c";
1018     printf "Building wctype.c\n";
1019     printf OUTPUT "/* Unicode ctype tables */\n";
1020     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1021     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1022
1023     my $i;
1024     my @array = (0) x 256;
1025
1026     # add the direction in the high 4 bits of the category
1027     for ($i = 0; $i < 65536; $i++)
1028     {
1029         $category_table[$i] |= $direction_table[$i] << 12;
1030     }
1031
1032     # try to merge table rows
1033     for ($row = 0; $row < 256; $row++)
1034     {
1035         my $rowtxt = sprintf "%04x" x 256, @category_table[($row<<8)..($row<<8)+255];
1036         if (defined($sequences{$rowtxt}))
1037         {
1038             # reuse an existing row
1039             $array[$row] = $sequences{$rowtxt};
1040         }
1041         else
1042         {
1043             # create a new row
1044             $sequences{$rowtxt} = $array[$row] = $#array + 1;
1045             push @array, @category_table[($row<<8)..($row<<8)+255];
1046         }
1047     }
1048
1049     printf OUTPUT "const unsigned short wine_wctype_table[%d] =\n{\n", $#array+1;
1050     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1051     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1052
1053     close OUTPUT;
1054     save_file("wctype.c");
1055 }
1056
1057
1058 ################################################################
1059 # dump the char composition tables
1060 sub DUMP_COMPOSE_TABLES
1061 {
1062     open OUTPUT,">compose.c.new" or die "Cannot create compose.c";
1063     printf "Building compose.c\n";
1064     printf OUTPUT "/* Unicode char composition */\n";
1065     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1066     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1067
1068     ######### composition table
1069
1070     my @filled = ();
1071     foreach $i (@compose_table)
1072     {
1073         my @comp = @$i;
1074         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
1075     }
1076
1077     # count how many different second chars we have
1078
1079     for ($i = $count = 0; $i < 65536; $i++)
1080     {
1081         next unless defined $filled[$i];
1082         $count++;
1083     }
1084
1085     # build the table of second chars and offsets
1086
1087     my $pos = $count + 1;
1088     for ($i = 0; $i < 65536; $i++)
1089     {
1090         next unless defined $filled[$i];
1091         push @table, $i, $pos;
1092         $pos += @{$filled[$i]};
1093     }
1094     # terminator with last position
1095     push @table, 0, $pos;
1096     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
1097     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
1098
1099     # build the table of first chars and mappings
1100
1101     for ($i = 0; $i < 65536; $i++)
1102     {
1103         next unless defined $filled[$i];
1104         my @table = ();
1105         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
1106         for ($j = 0; $j <= $#list; $j++)
1107         {
1108             push @table, $list[$j][0], $list[$j][1];
1109         }
1110         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
1111     }
1112     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
1113
1114     ######### decomposition table
1115
1116     # first determine all the 16-char subsets that contain something
1117
1118     my @filled = (0) x 4096;
1119     my $pos = 16*2;  # for the null subset
1120     for ($i = 0; $i < 65536; $i++)
1121     {
1122         next unless defined $decomp_table[$i];
1123         $filled[$i >> 4] = $pos;
1124         $pos += 16*2;
1125         $i |= 15;
1126     }
1127     my $total = $pos;
1128
1129     # now count the 256-char subsets that contain something
1130
1131     my @filled_idx = (256) x 256;
1132     $pos = 256 + 16;
1133     for ($i = 0; $i < 4096; $i++)
1134     {
1135         next unless $filled[$i];
1136         $filled_idx[$i >> 4] = $pos;
1137         $pos += 16;
1138         $i |= 15;
1139     }
1140     my $null_offset = $pos;  # null mapping
1141     $total += $pos;
1142
1143     # add the index offsets to the subsets positions
1144
1145     for ($i = 0; $i < 4096; $i++)
1146     {
1147         next unless $filled[$i];
1148         $filled[$i] += $null_offset;
1149     }
1150
1151     # dump the main index
1152
1153     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
1154     printf OUTPUT "{\n    /* index */\n";
1155     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1156     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1157
1158     # dump the second-level indexes
1159
1160     for ($i = 0; $i < 256; $i++)
1161     {
1162         next unless ($filled_idx[$i] > 256);
1163         my @table = @filled[($i<<4)..($i<<4)+15];
1164         for ($j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1165         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1166         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1167     }
1168
1169     # dump the 16-char subsets
1170
1171     printf OUTPUT ",\n    /* null mapping */\n";
1172     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
1173
1174     for ($i = 0; $i < 4096; $i++)
1175     {
1176         next unless $filled[$i];
1177         my @table = (0) x 32;
1178         for ($j = 0; $j < 16; $j++)
1179         {
1180             if (defined $decomp_table[($i<<4) + $j])
1181             {
1182                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
1183                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
1184             }
1185         }
1186         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1187         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1188     }
1189
1190     printf OUTPUT "\n};\n";
1191     close OUTPUT;
1192     save_file("compose.c");
1193 }
1194
1195
1196 ################################################################
1197 # read an input file and generate the corresponding .c file
1198 sub HANDLE_FILE
1199 {
1200     my ($codepage,$filename,$has_glyphs,$comment) = @_;
1201
1202     @cp2uni = ();
1203     @lead_bytes = ();
1204     @uni2cp = ();
1205
1206     # symbol codepage file is special
1207     if ($codepage == 20932) { READ_JIS0208_FILE($MAPPREFIX . $filename); }
1208     elsif ($codepage == 20127) { fill_20127_codepage(); }
1209     else { READ_FILE($MAPPREFIX . $filename); }
1210
1211     # hack: 0x00a5 must map to backslash in Shift-JIS
1212     if ($codepage == 932) { $uni2cp[0x00a5] = 0x5c; }
1213
1214     ADD_DEFAULT_MAPPINGS();
1215
1216     my $output = sprintf "c_%03d.c", $codepage;
1217     open OUTPUT,">$output.new" or die "Cannot create $output";
1218
1219     printf "Building %s from %s (%s)\n", $output, $filename || "hardcoded data", $comment;
1220
1221     # dump all tables
1222
1223     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
1224     if ($filename)
1225     {
1226         printf OUTPUT "/* generated from %s */\n", $MAPPREFIX . $filename;
1227         printf OUTPUT "/* DO NOT EDIT!! */\n\n";
1228     }
1229     else
1230     {
1231         printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1232     }
1233     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1234
1235     if ($#lead_bytes == -1) { DUMP_SBCS_TABLE( $codepage, $has_glyphs, $comment ); }
1236     else { DUMP_DBCS_TABLE( $codepage, $comment ); }
1237     close OUTPUT;
1238     save_file($output);
1239 }
1240
1241
1242 ################################################################
1243 # save a file if modified
1244 sub save_file($)
1245 {
1246     my $file = shift;
1247     if (!system "cmp $file $file.new >/dev/null")
1248     {
1249         unlink "$file.new";
1250     }
1251     else
1252     {
1253         rename "$file.new", "$file";
1254     }
1255 }
1256
1257
1258 ################################################################
1259 # output the list of codepage tables into the cptable.c file
1260 sub OUTPUT_CPTABLE
1261 {
1262     @tables_decl = ();
1263
1264     foreach $file (@allfiles)
1265     {
1266         my ($codepage,$filename,$comment) = @$file;
1267         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
1268     }
1269
1270     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
1271     foreach $file (@allfiles)
1272     {
1273         my ($codepage,$filename,$comment) = @$file;
1274         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
1275     }
1276     push @tables_decl, "};";
1277     REPLACE_IN_FILE( "cptable.c", @tables_decl );
1278 }
1279
1280 ################################################################
1281 # replace the contents of a file between ### cpmap ### marks
1282
1283 sub REPLACE_IN_FILE
1284 {
1285     my $name = shift;
1286     my @data = @_;
1287     my @lines = ();
1288     open(FILE,$name) or die "Can't open $name";
1289     while (<FILE>)
1290     {
1291         push @lines, $_;
1292         last if /\#\#\# cpmap begin \#\#\#/;
1293     }
1294     push @lines, @data;
1295     while (<FILE>)
1296     {
1297         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
1298     }
1299     push @lines, <FILE>;
1300     open(FILE,">$name.new") or die "Can't modify $name";
1301     print FILE @lines;
1302     close(FILE);
1303     save_file($name);
1304 }