libs/unicode/cpmap.pl

   1 #!/usr/bin/perl
   2 #
   3 # Generate code page .c files from ftp.unicode.org descriptions
   4 #
   5 # Copyright 2000 Alexandre Julliard
   6 #
   7 # This library is free software; you can redistribute it and/or
   8 # modify it under the terms of the GNU Lesser General Public
   9 # License as published by the Free Software Foundation; either
  10 # version 2.1 of the License, or (at your option) any later version.
  11 #
  12 # This library is distributed in the hope that it will be useful,
  13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15 # Lesser General Public License for more details.
  16 #
  17 # You should have received a copy of the GNU Lesser General Public
  18 # License along with this library; if not, write to the Free Software
  19 # Foundation, Inc., 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, "VENDORS/MISC/US-ASCII-QUOTES.TXT",  "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 $name";
 373     @cp2uni = ();
 374     @lead_bytes = ();
 375     @uni2cp = ();
 376
 377     while (<INPUT>)
 378     {
 379         next if /^\#/;  # skip comments
 380         next if /^$/;  # skip empty lines
 381         next if /\x1a/;  # skip ^Z
 382         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 383
 384         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 385         {
 386             $cp = hex $1;
 387             push @lead_bytes,$cp;
 388             $cp2uni[$cp] = 0;
 389             next;
 390         }
 391         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 392         {
 393             $cp = hex $1;
 394             $uni = hex $2;
 395             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 396             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 397             next;
 398         }
 399         die "$name: Unrecognized line $_\n";
 400     }
 401 }
 402
 403
 404 ################################################################
 405 # build EUC-JP table from the JIS 0208 file
 406 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 407 # but this would require 3-byte DBCS characters
 408 sub READ_JIS0208_FILE
 409 {
 410     my $name = shift;
 411     @cp2uni = ();
 412     @lead_bytes = ();
 413     @uni2cp = ();
 414
 415     # ASCII chars
 416     for ($i = 0x00; $i <= 0x7f; $i++)
 417     {
 418         $cp2uni[$i] = $i;
 419         $uni2cp[$i] = $i;
 420     }
 421
 422     # JIS X 0201 right plane
 423     for ($i = 0xa1; $i <= 0xdf; $i++)
 424     {
 425         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 426         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 427     }
 428
 429     # lead bytes
 430     foreach $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 431     {
 432         push @lead_bytes,$i;
 433         $cp2uni[$i] = 0;
 434     }
 435
 436     # undefined chars
 437     foreach $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 438     {
 439         $cp2uni[$i] = $DEF_CHAR;
 440     }
 441
 442     # Shift-JIS compatibility
 443     $uni2cp[0x00a5] = 0x5c;
 444     $uni2cp[0x203e] = 0x7e;
 445
 446     # Fix backslash conversion
 447     $cp2uni[0xa1c0] = 0xff3c;
 448     $uni2cp[0xff3c] = 0xa1c0;
 449
 450     open INPUT, "$name" or die "Cannot open $name";
 451     while (<INPUT>)
 452     {
 453         next if /^\#/;  # skip comments
 454         next if /^$/;  # skip empty lines
 455         next if /\x1a/;  # skip ^Z
 456         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 457         {
 458             $cp = 0x8080 + hex $1;
 459             $uni = hex $2;
 460             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 461             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 462             next;
 463         }
 464         die "$name: Unrecognized line $_\n";
 465     }
 466 }
 467
 468
 469 ################################################################
 470 # build the sort keys table
 471 sub READ_SORTKEYS_FILE
 472 {
 473     my @sortkeys = ();
 474     for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
 475
 476     open INPUT, "$SORTKEYS" or die "Cannot open $SORTKEYS";
 477     print "Loading $SORTKEYS\n";
 478     while (<INPUT>)
 479     {
 480         next if /^\#/;  # skip comments
 481         next if /^$/;  # skip empty lines
 482         next if /\x1a/;  # skip ^Z
 483         next if /^\@version/;  # skip @version header
 484         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 485         {
 486             my ($uni,$variable) = (hex $1, $2);
 487             next if $uni > 65535;
 488             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 489             next;
 490         }
 491         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 492         {
 493             # multiple character sequence, ignored for now
 494             next;
 495         }
 496         die "$SORTKEYS: Unrecognized line $_\n";
 497     }
 498     close INPUT;
 499
 500     # compress the keys to 32 bit:
 501     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 502
 503     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 504                        ${$a}[2] <=> ${$b}[2] or
 505                        ${$a}[3] <=> ${$b}[3] or
 506                        ${$a}[4] <=> ${$b}[4] or
 507                        $a cmp $b; } @sortkeys;
 508
 509     my ($n2, $n3) = (1, 1);
 510     my @keys = (-1, -1, -1, -1, -1 );
 511     my @flatkeys = ();
 512
 513     for (my $i = 0; $i < 65536; $i++)
 514     {
 515         my @current = @{$sortkeys[$i]};
 516         next if $current[0] == -1;
 517         if ($current[1] == $keys[1])
 518         {
 519             if ($current[2] == $keys[2])
 520             {
 521                 if ($current[3] == $keys[3])
 522                 {
 523                     # nothing
 524                 }
 525                 else
 526                 {
 527                     $keys[3] = $current[3];
 528                     $n3++;
 529                     die if ($n3 >= 16);
 530                 }
 531             }
 532             else
 533             {
 534                 $keys[2] = $current[2];
 535                 $keys[3] = $current[3];
 536                 $n2++;
 537                 $n3 = 1;
 538                 die if ($n2 >= 256);
 539             }
 540         }
 541         else
 542         {
 543             $keys[1] = $current[1];
 544             $keys[2] = $current[2];
 545             $keys[3] = $current[3];
 546             $n2 = 1;
 547             $n3 = 1;
 548         }
 549
 550         if ($current[2]) { $current[2] = $n2; }
 551         if ($current[3]) { $current[3] = $n3; }
 552         if ($current[4]) { $current[4] = 1; }
 553
 554         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 555     }
 556     return @flatkeys;
 557 }
 558
 559
 560 ################################################################
 561 # build the sort keys table
 562 sub DUMP_SORTKEYS
 563 {
 564     my @keys = @_;
 565
 566     # count the number of 256-key ranges that contain something
 567
 568     my @offsets = ();
 569     my $ranges = 2;
 570     for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
 571     for (my $i = 0; $i < 65536; $i++)
 572     {
 573         next unless defined $keys[$i];
 574         $offsets[$i >> 8] = $ranges * 256;
 575         $ranges++;
 576         $i |= 255;
 577     }
 578
 579     # output the range offsets
 580
 581     open OUTPUT,">collation.c.new" or die "Cannot create collation.c";
 582     printf "Building collation.c\n";
 583     printf OUTPUT "/* Unicode collation element table */\n";
 584     printf OUTPUT "/* generated from %s */\n", $SORTKEYS;
 585     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 586
 587     printf OUTPUT "const unsigned int collation_table[%d] =\n{\n", $ranges*256;
 588     printf OUTPUT "    /* index */\n";
 589     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%08x", 0, @offsets );
 590
 591     # output the default values
 592
 593     printf OUTPUT "    /* defaults */\n";
 594     printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0, (-1) x 256 );
 595
 596     # output all the key ranges
 597
 598     for (my $i = 0; $i < 256; $i++)
 599     {
 600         next if $offsets[$i] == 256;
 601         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 602         printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", -1, @keys[($i<<8) .. ($i<<8)+255] );
 603     }
 604     printf OUTPUT "\n};\n";
 605     close OUTPUT;
 606     save_file("collation.c");
 607 }
 608
 609
 610 ################################################################
 611 # add default mappings once the file had been read
 612 sub ADD_DEFAULT_MAPPINGS
 613 {
 614     # Apply aliases
 615
 616     foreach $alias (@unicode_aliases)
 617     {
 618         my $target = undef;
 619         foreach $src (@$alias)
 620         {
 621             if (defined($uni2cp[$src]))
 622             {
 623                 $target = $uni2cp[$src];
 624                 last;
 625             }
 626         }
 627         next unless defined($target);
 628
 629         # At least one char of the alias set is defined, set the others to the same value
 630         foreach $src (@$alias)
 631         {
 632             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 633         }
 634     }
 635
 636     # For every src -> target mapping in the defaults table,
 637     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 638
 639     for ($src = 0; $src < 65536; $src++)
 640     {
 641         next if defined($uni2cp[$src]);  # source has a definition already
 642         next unless defined($unicode_defaults[$src]);  # no default for this char
 643         my $target = $unicode_defaults[$src];
 644
 645         # do a recursive mapping until we find a target char that is defined
 646         while (!defined($uni2cp[$target]) &&
 647                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 648
 649         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 650     }
 651
 652     # Add an identity mapping for all undefined chars
 653
 654     for ($i = 0; $i < 256; $i++)
 655     {
 656         next if defined($cp2uni[$i]);
 657         next if defined($uni2cp[$i]);
 658         $cp2uni[$i] = $uni2cp[$i] = $i;
 659     }
 660 }
 661
 662 ################################################################
 663 # dump an array of integers
 664 sub DUMP_ARRAY
 665 {
 666     my ($format,$default,@array) = @_;
 667     my $i, $ret = "    ";
 668     for ($i = 0; $i < $#array; $i++)
 669     {
 670         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 671         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 672     }
 673     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 674     return $ret;
 675 }
 676
 677 ################################################################
 678 # dump an SBCS mapping table
 679 sub DUMP_SBCS_TABLE
 680 {
 681     my ($codepage, $name) = @_;
 682     my $i;
 683
 684     # output the ascii->unicode table
 685
 686     printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 687     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 688
 689     # count the number of unicode->ascii subtables that contain something
 690
 691     my @filled = ();
 692     my $subtables = 1;
 693     for ($i = 0; $i < 65536; $i++)
 694     {
 695         next unless defined $uni2cp[$i];
 696         $filled[$i >> 8] = 1;
 697         $subtables++;
 698         $i |= 255;
 699     }
 700
 701     # output all the subtables into a single array
 702
 703     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 704     for ($i = 0; $i < 256; $i++)
 705     {
 706         next unless $filled[$i];
 707         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 708         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $DEF_CHAR, @uni2cp[($i<<8) .. ($i<<8)+255] );
 709     }
 710     printf OUTPUT "    /* defaults */\n";
 711     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($DEF_CHAR) x 256 );
 712
 713     # output a table of the offsets of the subtables in the previous array
 714
 715     my $pos = 0;
 716     my @offsets = ();
 717     for ($i = 0; $i < 256; $i++)
 718     {
 719         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 720         else { push @offsets, ($subtables-1) * 256; }
 721     }
 722     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 723     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 724
 725     # output the code page descriptor
 726
 727     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 728     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 729                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 730     printf OUTPUT "    cp2uni,\n";
 731     printf OUTPUT "    uni2cp_low,\n";
 732     printf OUTPUT "    uni2cp_high\n};\n";
 733 }
 734
 735
 736 ################################################################
 737 # dump a DBCS mapping table
 738 sub DUMP_DBCS_TABLE
 739 {
 740     my ($codepage, $name) = @_;
 741     my $i, $x, $y;
 742
 743     # build a list of lead bytes that are actually used
 744
 745     my @lblist = ();
 746     LBLOOP: for ($y = 0; $y <= $#lead_bytes; $y++)
 747     {
 748         my $base = $lead_bytes[$y] << 8;
 749         for ($x = 0; $x < 256; $x++)
 750         {
 751             if (defined $cp2uni[$base+$x])
 752             {
 753                 push @lblist,$lead_bytes[$y];
 754                 next LBLOOP;
 755             }
 756         }
 757     }
 758     my $unused = ($#lead_bytes > $#lblist);
 759
 760     # output the ascii->unicode table for the single byte chars
 761
 762     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 763     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[0 .. 255] );
 764
 765     # output the default table for unused lead bytes
 766
 767     if ($unused)
 768     {
 769         printf OUTPUT "    /* unused lead bytes */\n";
 770         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 771     }
 772
 773     # output the ascii->unicode table for each DBCS lead byte
 774
 775     for ($y = 0; $y <= $#lblist; $y++)
 776     {
 777         my $base = $lblist[$y] << 8;
 778         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
 779         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @cp2uni[$base .. $base+255] );
 780         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
 781     }
 782
 783     # output the lead byte subtables offsets
 784
 785     my @offsets = ();
 786     for ($x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
 787     for ($x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
 788     if ($unused)
 789     {
 790         # increment all lead bytes offset to take into account the unused table
 791         for ($x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
 792     }
 793     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
 794     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
 795
 796     # count the number of unicode->ascii subtables that contain something
 797
 798     my @filled = ();
 799     my $subtables = 1;
 800     for ($i = 0; $i < 65536; $i++)
 801     {
 802         next unless defined $uni2cp[$i];
 803         $filled[$i >> 8] = 1;
 804         $subtables++;
 805         $i |= 255;
 806     }
 807
 808     # output all the subtables into a single array
 809
 810     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
 811     for ($y = 0; $y < 256; $y++)
 812     {
 813         next unless $filled[$y];
 814         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
 815         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $DEF_CHAR, @uni2cp[($y<<8) .. ($y<<8)+255] );
 816     }
 817     printf OUTPUT "    /* defaults */\n";
 818     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($DEF_CHAR) x 256 );
 819
 820     # output a table of the offsets of the subtables in the previous array
 821
 822     my $pos = 0;
 823     my @offsets = ();
 824     for ($y = 0; $y < 256; $y++)
 825     {
 826         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
 827         else { push @offsets, ($subtables-1) * 256; }
 828     }
 829     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 830     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 831
 832     # output the code page descriptor
 833
 834     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
 835     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
 836                   $codepage, $DEF_CHAR, $DEF_CHAR, $name;
 837     printf OUTPUT "    cp2uni,\n";
 838     printf OUTPUT "    cp2uni_leadbytes,\n";
 839     printf OUTPUT "    uni2cp_low,\n";
 840     printf OUTPUT "    uni2cp_high,\n";
 841     DUMP_LB_RANGES();
 842     printf OUTPUT "};\n";
 843 }
 844
 845
 846 ################################################################
 847 # dump the list of defined lead byte ranges
 848 sub DUMP_LB_RANGES
 849 {
 850     my @list = ();
 851     my $i = 0;
 852     foreach $i (@lead_bytes) { $list[$i] = 1; }
 853     my $on = 0;
 854     printf OUTPUT "    { ";
 855     for ($i = 0; $i < 256; $i++)
 856     {
 857         if ($on)
 858         {
 859             if (!defined $list[$i]) { printf OUTPUT "0x%02x, ", $i-1; $on = 0; }
 860         }
 861         else
 862         {
 863             if ($list[$i]) { printf OUTPUT "0x%02x, ", $i; $on = 1; }
 864         }
 865     }
 866     if ($on) { printf OUTPUT "0xff, "; }
 867     printf OUTPUT "0x00, 0x00 }\n";
 868 }
 869
 870
 871 ################################################################
 872 # dump the case mapping tables
 873 sub DUMP_CASE_MAPPINGS
 874 {
 875     open OUTPUT,">casemap.c.new" or die "Cannot create casemap.c";
 876     printf "Building casemap.c\n";
 877     printf OUTPUT "/* Unicode case mappings */\n";
 878     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 879     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 880
 881     DUMP_CASE_TABLE( "wine_casemap_lower", @tolower_table );
 882     DUMP_CASE_TABLE( "wine_casemap_upper", @toupper_table );
 883     DUMP_CASE_TABLE( "wine_digitmap",  @digitmap_table );
 884     DUMP_CASE_TABLE( "wine_compatmap", @compatmap_table );
 885     close OUTPUT;
 886     save_file("casemap.c");
 887 }
 888
 889
 890 ################################################################
 891 # dump a case mapping table
 892 sub DUMP_CASE_TABLE
 893 {
 894     my ($name,@table) = @_;
 895
 896     # count the number of sub tables that contain something
 897     # also compute the low and upper populated bounds
 898
 899     my @lowerbounds = ( 0, 0 );
 900     my @upperbounds = ( 0, 255 );
 901     my $index = 0;
 902     my @filled = ();
 903     for ($i = 0; $i < 65536; $i++)
 904     {
 905         next unless defined $table[$i];
 906         if (!defined $filled[$i >> 8])
 907         {
 908           $lowerbounds[$index] = $i & 0xff;
 909           $upperbounds[$index] = 0xff - $lowerbounds[$index];
 910           $filled[$i >> 8] = $index * 256 + 512;
 911           $index++;
 912         }
 913         else
 914         {
 915           $upperbounds[$index-1] = 0xff - ($i & 0xff);
 916         }
 917         $table[$i] = ($table[$i] - $i) & 0xffff;
 918     }
 919
 920     # Collapse blocks upwards if possible
 921     my $removed = 0;
 922     $index = 0;
 923     for ($i = 0; $i < 256; $i++)
 924     {
 925         next unless defined $filled[$i];
 926         if ($upperbounds[$index - 1] > $lowerbounds[$index])
 927         {
 928            $removed = $removed + $lowerbounds[$index];
 929         }
 930         else
 931         {
 932            $removed = $removed + $upperbounds[$index - 1];
 933            $lowerbounds[$index] = $upperbounds[$index - 1];
 934         }
 935         $filled[$i] = $filled[$i] - $removed;
 936         $index++;
 937     }
 938
 939     # dump the table
 940
 941     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
 942     printf OUTPUT "{\n    /* index */\n";
 943     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
 944     printf OUTPUT "    /* defaults */\n";
 945     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
 946     $index = 0;
 947     for ($i = 0; $i < 256; $i++)
 948     {
 949         next unless $filled[$i];
 950         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
 951         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0,
 952                       @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
 953         $index++;
 954     }
 955     printf OUTPUT "\n};\n";
 956 }
 957
 958
 959 ################################################################
 960 # dump the ctype tables
 961 sub DUMP_CTYPE_TABLES
 962 {
 963     open OUTPUT,">wctype.c.new" or die "Cannot create wctype.c";
 964     printf "Building wctype.c\n";
 965     printf OUTPUT "/* Unicode ctype tables */\n";
 966     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
 967     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
 968
 969     my $i;
 970     my @array = (0) x 256;
 971
 972     # add the direction in the high 4 bits of the category
 973     for ($i = 0; $i < 65536; $i++)
 974     {
 975         $category_table[$i] |= $direction_table[$i] << 12;
 976     }
 977
 978     # try to merge table rows
 979     for ($row = 0; $row < 256; $row++)
 980     {
 981         my $rowtxt = sprintf "%04x" x 256, @category_table[($row<<8)..($row<<8)+255];
 982         if (defined($sequences{$rowtxt}))
 983         {
 984             # reuse an existing row
 985             $array[$row] = $sequences{$rowtxt};
 986         }
 987         else
 988         {
 989             # create a new row
 990             $sequences{$rowtxt} = $array[$row] = $#array + 1;
 991             push @array, @category_table[($row<<8)..($row<<8)+255];
 992         }
 993     }
 994
 995     printf OUTPUT "const unsigned short wine_wctype_table[%d] =\n{\n", $#array+1;
 996     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
 997     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
 998
 999     close OUTPUT;
1000     save_file("wctype.c");
1001 }
1002
1003
1004 ################################################################
1005 # dump the char composition tables
1006 sub DUMP_COMPOSE_TABLES
1007 {
1008     open OUTPUT,">compose.c.new" or die "Cannot create compose.c";
1009     printf "Building compose.c\n";
1010     printf OUTPUT "/* Unicode char composition */\n";
1011     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1012     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1013
1014     ######### composition table
1015
1016     my @filled = ();
1017     foreach $i (@compose_table)
1018     {
1019         my @comp = @$i;
1020         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
1021     }
1022
1023     # count how many different second chars we have
1024
1025     for ($i = $count = 0; $i < 65536; $i++)
1026     {
1027         next unless defined $filled[$i];
1028         $count++;
1029     }
1030
1031     # build the table of second chars and offsets
1032
1033     my $pos = $count + 1;
1034     for ($i = 0; $i < 65536; $i++)
1035     {
1036         next unless defined $filled[$i];
1037         push @table, $i, $pos;
1038         $pos += @{$filled[$i]};
1039     }
1040     # terminator with last position
1041     push @table, 0, $pos;
1042     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
1043     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
1044
1045     # build the table of first chars and mappings
1046
1047     for ($i = 0; $i < 65536; $i++)
1048     {
1049         next unless defined $filled[$i];
1050         my @table = ();
1051         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
1052         for ($j = 0; $j <= $#list; $j++)
1053         {
1054             push @table, $list[$j][0], $list[$j][1];
1055         }
1056         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
1057     }
1058     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
1059
1060     ######### decomposition table
1061
1062     # first determine all the 16-char subsets that contain something
1063
1064     my @filled = (0) x 4096;
1065     my $pos = 16*2;  # for the null subset
1066     for ($i = 0; $i < 65536; $i++)
1067     {
1068         next unless defined $decomp_table[$i];
1069         $filled[$i >> 4] = $pos;
1070         $pos += 16*2;
1071         $i |= 15;
1072     }
1073     my $total = $pos;
1074
1075     # now count the 256-char subsets that contain something
1076
1077     my @filled_idx = (256) x 256;
1078     $pos = 256 + 16;
1079     for ($i = 0; $i < 4096; $i++)
1080     {
1081         next unless $filled[$i];
1082         $filled_idx[$i >> 4] = $pos;
1083         $pos += 16;
1084         $i |= 15;
1085     }
1086     my $null_offset = $pos;  # null mapping
1087     $total += $pos;
1088
1089     # add the index offsets to the subsets positions
1090
1091     for ($i = 0; $i < 4096; $i++)
1092     {
1093         next unless $filled[$i];
1094         $filled[$i] += $null_offset;
1095     }
1096
1097     # dump the main index
1098
1099     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
1100     printf OUTPUT "{\n    /* index */\n";
1101     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1102     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1103
1104     # dump the second-level indexes
1105
1106     for ($i = 0; $i < 256; $i++)
1107     {
1108         next unless ($filled_idx[$i] > 256);
1109         my @table = @filled[($i<<4)..($i<<4)+15];
1110         for ($j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1111         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1112         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1113     }
1114
1115     # dump the 16-char subsets
1116
1117     printf OUTPUT ",\n    /* null mapping */\n";
1118     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
1119
1120     for ($i = 0; $i < 4096; $i++)
1121     {
1122         next unless $filled[$i];
1123         my @table = (0) x 32;
1124         for ($j = 0; $j < 16; $j++)
1125         {
1126             if (defined $decomp_table[($i<<4) + $j])
1127             {
1128                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
1129                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
1130             }
1131         }
1132         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1133         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1134     }
1135
1136     printf OUTPUT "\n};\n";
1137     close OUTPUT;
1138     save_file("compose.c");
1139 }
1140
1141
1142 ################################################################
1143 # read an input file and generate the corresponding .c file
1144 sub HANDLE_FILE
1145 {
1146     my ($codepage,$filename,$comment) = @_;
1147
1148     # symbol codepage file is special
1149     if ($codepage == 20932) { READ_JIS0208_FILE($MAPPREFIX . $filename); }
1150     else { READ_FILE($MAPPREFIX . $filename); }
1151
1152     # hack: 0x00a5 must map to backslash in Shift-JIS
1153     if ($codepage == 932) { $uni2cp[0x00a5] = 0x5c; }
1154
1155     ADD_DEFAULT_MAPPINGS();
1156
1157     my $output = sprintf "c_%03d.c", $codepage;
1158     open OUTPUT,">$output.new" or die "Cannot create $output";
1159
1160     printf "Building %s from %s (%s)\n", $output, $filename, $comment;
1161
1162     # dump all tables
1163
1164     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
1165     printf OUTPUT "/* generated from %s */\n", $MAPPREFIX . $filename;
1166     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
1167     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1168
1169     if ($#lead_bytes == -1) { DUMP_SBCS_TABLE( $codepage, $comment ); }
1170     else { DUMP_DBCS_TABLE( $codepage, $comment ); }
1171     close OUTPUT;
1172     save_file($output);
1173 }
1174
1175
1176 ################################################################
1177 # save a file if modified
1178 sub save_file($)
1179 {
1180     my $file = shift;
1181     if (!system "cmp $file $file.new >/dev/null")
1182     {
1183         unlink "$file.new";
1184     }
1185     else
1186     {
1187         rename "$file.new", "$file";
1188     }
1189 }
1190
1191
1192 ################################################################
1193 # output the list of codepage tables into the cptable.c file
1194 sub OUTPUT_CPTABLE
1195 {
1196     @tables_decl = ();
1197
1198     foreach $file (@allfiles)
1199     {
1200         my ($codepage,$filename,$comment) = @$file;
1201         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
1202     }
1203
1204     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
1205     foreach $file (@allfiles)
1206     {
1207         my ($codepage,$filename,$comment) = @$file;
1208         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
1209     }
1210     push @tables_decl, "};";
1211     REPLACE_IN_FILE( "cptable.c", @tables_decl );
1212 }
1213
1214 ################################################################
1215 # replace the contents of a file between ### cpmap ### marks
1216
1217 sub REPLACE_IN_FILE
1218 {
1219     my $name = shift;
1220     my @data = @_;
1221     my @lines = ();
1222     open(FILE,$name) or die "Can't open $name";
1223     while (<FILE>)
1224     {
1225         push @lines, $_;
1226         last if /\#\#\# cpmap begin \#\#\#/;
1227     }
1228     push @lines, @data;
1229     while (<FILE>)
1230     {
1231         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
1232     }
1233     push @lines, <FILE>;
1234     open(FILE,">$name.new") or die "Can't modify $name";
1235     print FILE @lines;
1236     close(FILE);
1237     save_file($name);
1238 }