tools/make_unicode

   1 #!/usr/bin/perl -w
   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 use strict;
  23
  24 # base URLs for www.unicode.org files
  25 my $MAPPINGS = "http://www.unicode.org/Public/MAPPINGS";
  26 my $UNIDATA = "http://www.unicode.org/Public/7.0.0/ucd";
  27 my $REPORTS = "http://www.unicode.org/reports";
  28 my $RFCS = "http://www.rfc-editor.org/rfc";
  29 my $VERTICALDATA = "http://www.unicode.org/Public/vertical/revision-11";
  30
  31 # Sort keys file
  32 my $SORTKEYS = "tr10/allkeys.txt";
  33
  34 # RFC3454 (stringprep data)
  35 my $STRINGPREP = "rfc3454.txt";
  36
  37 # Defaults mapping
  38 my $DEFAULTS = "tools/unicode-defaults";
  39
  40 # Default char for undefined mappings
  41 my $DEF_CHAR = ord '?';
  42
  43 my @allfiles =
  44 (
  45     [ 37,    "VENDORS/MICSFT/EBCDIC/CP037.TXT",   0, "IBM EBCDIC US Canada", 0, 0x6f ],
  46     [ 424,   "VENDORS/MISC/CP424.TXT",            0, "IBM EBCDIC Hebrew" ],
  47     [ 437,   "VENDORS/MICSFT/PC/CP437.TXT",       1, "OEM United States" ],
  48     [ 500,   "VENDORS/MICSFT/EBCDIC/CP500.TXT",   0, "IBM EBCDIC International", 0, 0x6f ],
  49     [ 737,   "VENDORS/MICSFT/PC/CP737.TXT",       1, "OEM Greek 437G" ],
  50     [ 775,   "VENDORS/MICSFT/PC/CP775.TXT",       1, "OEM Baltic" ],
  51     [ 850,   "VENDORS/MICSFT/PC/CP850.TXT",       1, "OEM Multilingual Latin 1" ],
  52     [ 852,   "VENDORS/MICSFT/PC/CP852.TXT",       1, "OEM Slovak Latin 2" ],
  53     [ 855,   "VENDORS/MICSFT/PC/CP855.TXT",       1, "OEM Cyrillic" ],
  54     [ 856,   "VENDORS/MISC/CP856.TXT",            0, "Hebrew PC" ],
  55     [ 857,   "VENDORS/MICSFT/PC/CP857.TXT",       1, "OEM Turkish", 0xf8bb ],
  56     [ 860,   "VENDORS/MICSFT/PC/CP860.TXT",       1, "OEM Portuguese" ],
  57     [ 861,   "VENDORS/MICSFT/PC/CP861.TXT",       1, "OEM Icelandic" ],
  58     [ 862,   "VENDORS/MICSFT/PC/CP862.TXT",       1, "OEM Hebrew" ],
  59     [ 863,   "VENDORS/MICSFT/PC/CP863.TXT",       1, "OEM Canadian French" ],
  60     [ 864,   "VENDORS/MICSFT/PC/CP864.TXT",       0, "OEM Arabic", 0xf8bb ],
  61     [ 865,   "VENDORS/MICSFT/PC/CP865.TXT",       1, "OEM Nordic" ],
  62     [ 866,   "VENDORS/MICSFT/PC/CP866.TXT",       1, "OEM Russian" ],
  63     [ 869,   "VENDORS/MICSFT/PC/CP869.TXT",       1, "OEM Greek" ],
  64     [ 874,   "VENDORS/MICSFT/WindowsBestFit/bestfit874.txt",  1, "ANSI/OEM Thai" ],
  65     [ 875,   "VENDORS/MICSFT/EBCDIC/CP875.TXT",               0, "IBM EBCDIC Greek", 0, 0x6f ],
  66     [ 878,   "VENDORS/MISC/KOI8-R.TXT",                       0, "Russian KOI8" ],
  67     [ 932,   "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt",  0, "ANSI/OEM Japanese Shift-JIS" ],
  68     [ 936,   "VENDORS/MICSFT/WindowsBestFit/bestfit936.txt",  0, "ANSI/OEM Simplified Chinese GBK" ],
  69     [ 949,   "VENDORS/MICSFT/WindowsBestFit/bestfit949.txt",  0, "ANSI/OEM Korean Unified Hangul" ],
  70     [ 950,   "VENDORS/MICSFT/WindowsBestFit/bestfit950.txt",  0, "ANSI/OEM Traditional Chinese Big5" ],
  71     [ 1006,  "VENDORS/MISC/CP1006.TXT",                       0, "IBM Arabic" ],
  72     [ 1026,  "VENDORS/MICSFT/EBCDIC/CP1026.TXT",              0, "IBM EBCDIC Latin 5 Turkish", 0, 0x6f ],
  73     [ 1250,  "VENDORS/MICSFT/WindowsBestFit/bestfit1250.txt", 0, "ANSI Eastern Europe" ],
  74     [ 1251,  "VENDORS/MICSFT/WindowsBestFit/bestfit1251.txt", 0, "ANSI Cyrillic" ],
  75     [ 1252,  "VENDORS/MICSFT/WindowsBestFit/bestfit1252.txt", 0, "ANSI Latin 1" ],
  76     [ 1253,  "VENDORS/MICSFT/WindowsBestFit/bestfit1253.txt", 0, "ANSI Greek" ],
  77     [ 1254,  "VENDORS/MICSFT/WindowsBestFit/bestfit1254.txt", 0, "ANSI Turkish" ],
  78     [ 1255,  "VENDORS/MICSFT/WindowsBestFit/bestfit1255.txt", 0, "ANSI Hebrew" ],
  79     [ 1256,  "VENDORS/MICSFT/WindowsBestFit/bestfit1256.txt", 0, "ANSI Arabic" ],
  80     [ 1257,  "VENDORS/MICSFT/WindowsBestFit/bestfit1257.txt", 0, "ANSI Baltic" ],
  81     [ 1258,  "VENDORS/MICSFT/WindowsBestFit/bestfit1258.txt", 0, "ANSI/OEM Viet Nam" ],
  82     [ 1361,  "VENDORS/MICSFT/WindowsBestFit/bestfit1361.txt", 0, "Korean Johab" ],
  83     [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT",      0, "Mac Roman" ],
  84     [ 10001, "VENDORS/APPLE/JAPANESE.TXT",        0, "Mac Japanese" ],
  85     [ 10002, "VENDORS/APPLE/CHINTRAD.TXT",        0, "Mac Traditional Chinese" ],
  86     [ 10003, "VENDORS/APPLE/KOREAN.TXT",          0, "Mac Korean", 0xf8e5 ],
  87     [ 10004, "VENDORS/APPLE/ARABIC.TXT",          0, "Mac Arabic", 0xf826 ],
  88     [ 10005, "VENDORS/APPLE/HEBREW.TXT",          0, "Mac Hebrew", 0xf7fa ],
  89     [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT",      0, "Mac Greek" ],
  90     [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT",   0, "Mac Cyrillic" ],
  91     [ 10008, "VENDORS/APPLE/CHINSIMP.TXT",        0, "Mac Simplified Chinese" ],
  92     [ 10010, "VENDORS/APPLE/ROMANIAN.TXT",        0, "Mac Romanian" ],
  93     [ 10017, "VENDORS/APPLE/CYRILLIC.TXT",        0, "Mac Ukrainian" ],
  94     [ 10021, "VENDORS/APPLE/THAI.TXT",            0, "Mac Thai" ],
  95     [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT",     0, "Mac Latin 2" ],
  96     [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT",    0, "Mac Icelandic" ],
  97     [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT",    0, "Mac Turkish", 0xf89f ],
  98     [ 10082, "VENDORS/APPLE/CROATIAN.TXT",        0, "Mac Croatian" ],
  99     [ 20127, undef,                               0, "US-ASCII (7bit)" ],
 100     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           0, "Russian KOI8" ],
 101     [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", 0, "EUC-JP", 0, 0x3f, 0x30fb  ],
 102     [ 21866, "VENDORS/MISC/KOI8-U.TXT",           0, "Ukrainian KOI8" ],
 103     [ 28591, "ISO8859/8859-1.TXT",                0, "ISO 8859-1 Latin 1" ],
 104     [ 28592, "ISO8859/8859-2.TXT",                0, "ISO 8859-2 Latin 2 (East European)" ],
 105     [ 28593, "ISO8859/8859-3.TXT",                0, "ISO 8859-3 Latin 3 (South European)", 0xf7f5 ],
 106     [ 28594, "ISO8859/8859-4.TXT",                0, "ISO 8859-4 Latin 4 (Baltic old)" ],
 107     [ 28595, "ISO8859/8859-5.TXT",                0, "ISO 8859-5 Cyrillic" ],
 108     [ 28596, "ISO8859/8859-6.TXT",                0, "ISO 8859-6 Arabic", 0xf7c8 ],
 109     [ 28597, "ISO8859/8859-7.TXT",                0, "ISO 8859-7 Greek", 0xf7c5 ],
 110     [ 28598, "ISO8859/8859-8.TXT",                0, "ISO 8859-8 Hebrew", 0xf79c ],
 111     [ 28599, "ISO8859/8859-9.TXT",                0, "ISO 8859-9 Latin 5 (Turkish)" ],
 112     [ 28600, "ISO8859/8859-10.TXT",               0, "ISO 8859-10 Latin 6 (Nordic)" ],
 113     [ 28603, "ISO8859/8859-13.TXT",               0, "ISO 8859-13 Latin 7 (Baltic)" ],
 114     [ 28604, "ISO8859/8859-14.TXT",               0, "ISO 8859-14 Latin 8 (Celtic)" ],
 115     [ 28605, "ISO8859/8859-15.TXT",               0, "ISO 8859-15 Latin 9 (Euro)" ],
 116     [ 28606, "ISO8859/8859-16.TXT",               0, "ISO 8859-16 Latin 10 (Balkan)" ]
 117 );
 118
 119
 120 my %ctype =
 121 (
 122     "upper"  => 0x0001,
 123     "lower"  => 0x0002,
 124     "digit"  => 0x0004,
 125     "space"  => 0x0008,
 126     "punct"  => 0x0010,
 127     "cntrl"  => 0x0020,
 128     "blank"  => 0x0040,
 129     "xdigit" => 0x0080,
 130     "alpha"  => 0x0100,
 131     "defin"  => 0x0200
 132 );
 133
 134 my %bracket_types =
 135 (
 136     "o" => 0x0000,
 137     "c" => 0x0001,
 138 );
 139
 140 my %indic_types =
 141 (
 142     "Other"    => 0x0000,
 143     "Bindu"    => 0x0001,
 144     "Visarga"  => 0x0002,
 145     "Avagraha" => 0x0003,
 146     "Nukta"    => 0x0004,
 147     "Virama"   => 0x0005,
 148     "Vowel_Independent"  => 0x0006,
 149     "Vowel_Dependent"  => 0x0007,
 150     "Vowel"  => 0x0008,
 151     "Consonant_Placeholder"  => 0x0009,
 152     "Consonant"  => 0x000a,
 153     "Consonant_Dead"  => 0x000b,
 154     "Consonant_Succeeding_Repha" => 0x000c,
 155     "Consonant_Subjoined"  => 0x000d,
 156     "Consonant_Medial"  => 0x000e,
 157     "Consonant_Final"  => 0x000f,
 158     "Consonant_Head_Letter"  => 0x0010,
 159     "Modifying_Letter"  => 0x0011,
 160     "Tone_Letter"  => 0x0012,
 161     "Tone_Mark"  => 0x0013,
 162     "Register_Shifter"  => 0x0014,
 163     "Consonant_Preceding_Repha" => 0x0015,
 164     "Pure_Killer" => 0x0016,
 165     "Invisible_Stacker" => 0x0017,
 166     "Gemination_Mark" => 0x0018,
 167     "Cantillation_Mark" => 0x0019,
 168     "Non_Joiner" => 0x001a,
 169     "Joiner" => 0x001b,
 170     "Number_Joiner" => 0x001c,
 171     "Number" => 0x001d,
 172     "Brahmi_Joining_Number" => 0x001e
 173 );
 174
 175 my %matra_types =
 176 (
 177     "Right"    => 0x01,
 178     "Left"  => 0x02,
 179     "Visual_Order_Left" => 0x03,
 180     "Left_And_Right"    => 0x04,
 181     "Top"   => 0x05,
 182     "Bottom"  => 0x06,
 183     "Top_And_Bottom"  => 0x07,
 184     "Top_And_Right"  => 0x08,
 185     "Top_And_Left"  => 0x09,
 186     "Top_And_Left_And_Right"  => 0x0a,
 187     "Bottom_And_Right"  => 0x0b,
 188     "Top_And_Bottom_And_Right"  => 0x0c,
 189     "Overstruck"  => 0x0d,
 190     "Invisible"  => 0x0e
 191 );
 192
 193 my %nameprep_flags =
 194 (
 195     "unassigned" => 0x01,
 196     "prohibited" => 0x02,
 197     "bidi_ral"   => 0x04,
 198     "bidi_l"     => 0x08
 199 );
 200
 201 my %break_types =
 202 (
 203     "BK"  => 0x0001,
 204     "CR"  => 0x0002,
 205     "LF"  => 0x0003,
 206     "CM"  => 0x0004,
 207     "SG"  => 0x0005,
 208     "GL"  => 0x0006,
 209     "CB"  => 0x0007,
 210     "SP"  => 0x0008,
 211     "ZW"  => 0x0009,
 212     "NL"  => 0x000a,
 213     "WJ"  => 0x000b,
 214     "JL"  => 0x000c,
 215     "JV"  => 0x000d,
 216     "JT"  => 0x000e,
 217     "H2"  => 0x000f,
 218     "H3"  => 0x0010,
 219     "XX"  => 0x0011,
 220     "OP"  => 0x0012,
 221     "CL"  => 0x0013,
 222     "CP"  => 0x0014,
 223     "QU"  => 0x0015,
 224     "NS"  => 0x0016,
 225     "EX"  => 0x0017,
 226     "SY"  => 0x0018,
 227     "IS"  => 0x0019,
 228     "PR"  => 0x001a,
 229     "PO"  => 0x001b,
 230     "NU"  => 0x001c,
 231     "AL"  => 0x001d,
 232     "ID"  => 0x001e,
 233     "IN"  => 0x001f,
 234     "HY"  => 0x0020,
 235     "BB"  => 0x0021,
 236     "BA"  => 0x0022,
 237     "SA"  => 0x0023,
 238     "AI"  => 0x0024,
 239     "B2"  => 0x0025,
 240     "HL"  => 0x0026,
 241     "CJ"  => 0x0027,
 242     "RI"  => 0x0028,
 243 );
 244
 245 my %vertical_types =
 246 (
 247     "R"  => 0x0000,
 248     "U"  => 0x0001,
 249     "Tr" => 0x0002,
 250     "Tu" => 0x0003,
 251 );
 252
 253 my %categories =
 254 (
 255     "Lu" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
 256     "Ll" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
 257     "Lt" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}|$ctype{"lower"},    # Letter, Titlecase
 258     "Mn" => $ctype{"defin"},                    # Mark, Non-Spacing
 259     "Mc" => $ctype{"defin"},                    # Mark, Spacing Combining
 260     "Me" => $ctype{"defin"},                    # Mark, Enclosing
 261     "Nd" => $ctype{"defin"}|$ctype{"digit"},    # Number, Decimal Digit
 262     "Nl" => $ctype{"defin"}|$ctype{"alpha"},    # Number, Letter
 263     "No" => $ctype{"defin"},                    # Number, Other
 264     "Zs" => $ctype{"defin"}|$ctype{"space"},    # Separator, Space
 265     "Zl" => $ctype{"defin"}|$ctype{"space"},    # Separator, Line
 266     "Zp" => $ctype{"defin"}|$ctype{"space"},    # Separator, Paragraph
 267     "Cc" => $ctype{"defin"}|$ctype{"cntrl"},    # Other, Control
 268     "Cf" => $ctype{"defin"}|$ctype{"cntrl"},    # Other, Format
 269     "Cs" => $ctype{"defin"},                    # Other, Surrogate
 270     "Co" => $ctype{"defin"},                    # Other, Private Use
 271     "Cn" => $ctype{"defin"},                    # Other, Not Assigned
 272     "Lm" => $ctype{"defin"}|$ctype{"alpha"},    # Letter, Modifier
 273     "Lo" => $ctype{"defin"}|$ctype{"alpha"},    # Letter, Other
 274     "Pc" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Connector
 275     "Pd" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Dash
 276     "Ps" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Open
 277     "Pe" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Close
 278     "Pi" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Initial quote
 279     "Pf" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Final quote
 280     "Po" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Other
 281     "Sm" => $ctype{"defin"},                    # Symbol, Math
 282     "Sc" => $ctype{"defin"},                    # Symbol, Currency
 283     "Sk" => $ctype{"defin"},                    # Symbol, Modifier
 284     "So" => $ctype{"defin"}                     # Symbol, Other
 285 );
 286
 287 # a few characters need additional categories that cannot be determined automatically
 288 my %special_categories =
 289 (
 290     "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
 291                   0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ],
 292     "space"  => [ 0x09..0x0d, 0x85 ],
 293     "blank"  => [ 0x09, 0x20, 0xa0, 0x3000, 0xfeff ],
 294     "cntrl"  => [ 0x070f, 0x200c, 0x200d,
 295                   0x200e, 0x200f, 0x202a, 0x202b, 0x202c, 0x202d, 0x202e,
 296                   0x206a, 0x206b, 0x206c, 0x206d, 0x206e, 0x206f, 0xfeff,
 297                   0xfff9, 0xfffa, 0xfffb ],
 298     "punct"  => [ 0x24, 0x2b, 0x3c..0x3e, 0x5e, 0x60, 0x7c, 0x7e, 0xa2..0xbe,
 299                   0xd7, 0xf7 ],
 300     "digit"  => [ 0xb2, 0xb3, 0xb9 ],
 301     "lower"  => [ 0x2071, 0x207f ]
 302 );
 303
 304 my %directions =
 305 (
 306     "L"   => 1,    # Left-to-Right
 307     "R"   => 2,    # Right-to-Left
 308     "AL"  => 12,   # Right-to-Left Arabic
 309     "EN"  => 3,    # European Number
 310     "ES"  => 4,    # European Number Separator
 311     "ET"  => 5,    # European Number Terminator
 312     "AN"  => 6,    # Arabic Number
 313     "CS"  => 7,    # Common Number Separator
 314     "NSM" => 13,   # Non-Spacing Mark
 315     "BN"  => 14,   # Boundary Neutral
 316     "B"   => 8,    # Paragraph Separator
 317     "S"   => 9,    # Segment Separator
 318     "WS"  => 10,   # Whitespace
 319     "ON"  => 11,   # Other Neutrals
 320     "LRE" => 15,   # Left-to-Right Embedding
 321     "LRO" => 15,   # Left-to-Right Override
 322     "RLE" => 15,   # Right-to-Left Embedding
 323     "RLO" => 15,   # Right-to-Left Override
 324     "PDF" => 15,   # Pop Directional Format
 325     "LRI" => 15,   # Left-to-Right Isolate
 326     "RLI" => 15,   # Right-to-Left Isolate
 327     "FSI" => 15,   # First Strong Isolate
 328     "PDI" => 15    # Pop Directional Isolate
 329 );
 330
 331 my %joining_types =
 332 (
 333    "U" => 0,    # Non_Joining
 334    "T" => 1,    # Transparent
 335    "R" => 2,    # Right_Joining
 336    "L" => 3,    # Left_Joining
 337    "D" => 4,    # Dual_Joining
 338    "C" => 5,    # Join_Causing
 339 );
 340
 341 my @cp2uni = ();
 342 my @lead_bytes = ();
 343 my @uni2cp = ();
 344 my @unicode_defaults = ();
 345 my @unicode_aliases = ();
 346 my @tolower_table = ();
 347 my @toupper_table = ();
 348 my @digitmap_table = ();
 349 my @compatmap_table = ();
 350 my @category_table = (0) x 65536;
 351 my @joining_table = (0) x 65536;
 352 my @direction_table = ();
 353 my @decomp_table = ();
 354 my @compose_table = ();
 355 my $default_char;
 356 my $default_wchar;
 357
 358 my %joining_forms =
 359 (
 360    "isolated" => [],
 361    "final" => [],
 362    "initial" => [],
 363    "medial" => []
 364 );
 365
 366 ################################################################
 367 # fetch a unicode.org file and open it
 368 sub open_data_file($$)
 369 {
 370     my ($base, $name) = @_;
 371     (my $dir = "data/$name") =~ s/\/[^\/]+$//;
 372     local *FILE;
 373     unless (-f "data/$name")
 374     {
 375         print "Fetching $base/$name...\n";
 376         system "mkdir", "-p", $dir;
 377         !system "wget", "-q", "-O", "data/$name", "$base/$name" or die "cannot fetch $base/$name";
 378     }
 379     open FILE, "<data/$name" or die "cannot open data/$name";
 380     return *FILE;
 381 }
 382
 383 ################################################################
 384 # read in the defaults file
 385 sub READ_DEFAULTS($)
 386 {
 387     my $filename = shift;
 388     my $start;
 389
 390     # first setup a few default mappings
 391
 392     open DEFAULTS, "$filename" or die "Cannot open $filename";
 393     print "Loading $filename\n";
 394     while (<DEFAULTS>)
 395     {
 396         next if /^\#/;  # skip comments
 397         next if /^$/;  # skip empty lines
 398         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 399         {
 400             my @src = map hex, split /,/,$1;
 401             my $dst = $4;
 402             my $comment = $5;
 403             if ($#src > 0) { push @unicode_aliases, \@src; }
 404             next if ($dst eq "none");
 405             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 406             foreach my $src (@src)
 407             {
 408                 die "Duplicate value" if defined($unicode_defaults[$src]);
 409                 $unicode_defaults[$src] = $dst;
 410             }
 411             next;
 412         }
 413         die "Unrecognized line $_\n";
 414     }
 415     close DEFAULTS;
 416
 417     # now build mappings from the decomposition field of the Unicode database
 418
 419     my $UNICODE_DATA = open_data_file( $UNIDATA, "UnicodeData.txt" );
 420     while (<$UNICODE_DATA>)
 421     {
 422         # Decode the fields ...
 423         my ($code, $name, $cat, $comb, $bidi,
 424             $decomp, $dec, $dig, $num, $mirror,
 425             $oldname, $comment, $upper, $lower, $title) = split /;/;
 426         my $dst;
 427         my $src = hex $code;
 428
 429         die "unknown category $cat" unless defined $categories{$cat};
 430         die "unknown directionality $bidi" unless defined $directions{$bidi};
 431
 432         $category_table[$src] = $categories{$cat};
 433         $direction_table[$src] = $directions{$bidi};
 434         $joining_table[$src] = $joining_types{"T"} if $cat eq "Mn" || $cat eq "Me" || $cat eq "Cf";
 435
 436         if ($lower ne "")
 437         {
 438             $tolower_table[$src] = hex $lower;
 439         }
 440         if ($upper ne "")
 441         {
 442             $toupper_table[$src] = hex $upper;
 443         }
 444         if ($dec ne "")
 445         {
 446             $category_table[$src] |= $ctype{"digit"};
 447         }
 448         if ($dig ne "")
 449         {
 450             $digitmap_table[$src] = ord $dig;
 451         }
 452
 453         # copy the category and direction for everything between First/Last pairs
 454         if ($name =~ /, First>/) { $start = $src; }
 455         if ($name =~ /, Last>/)
 456         {
 457             while ($start < $src)
 458             {
 459                 $category_table[$start] = $category_table[$src];
 460                 $direction_table[$start] = $direction_table[$src];
 461                 $start++;
 462             }
 463         }
 464
 465         next if $decomp eq "";  # no decomposition, skip it
 466
 467         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 468         {
 469             # decomposition of the form "<foo> 1234" -> use char if type is known
 470             if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd))
 471             {
 472                 # Single char decomposition in the compatibility range
 473                 $compatmap_table[$src] = hex $2;
 474             }
 475             if ($1 eq "isolated" || $1 eq "final" || $1 eq "initial" || $1 eq "medial")
 476             {
 477                 ${joining_forms{$1}}[hex $2] = $src;
 478                 next;
 479             }
 480             next unless ($1 eq "font" ||
 481                          $1 eq "noBreak" ||
 482                          $1 eq "circle" ||
 483                          $1 eq "super" ||
 484                          $1 eq "sub" ||
 485                          $1 eq "wide" ||
 486                          $1 eq "narrow" ||
 487                          $1 eq "compat" ||
 488                          $1 eq "small");
 489             $dst = hex $2;
 490         }
 491         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 492         {
 493             # decomposition "<compat> 0020 1234" -> combining accent
 494             $dst = hex $1;
 495         }
 496         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 497         {
 498             # decomposition contains only char values without prefix -> use first char
 499             $dst = hex $1;
 500             $category_table[$src] |= $category_table[$dst] if defined $category_table[$dst];
 501             # store decomposition if it contains two chars
 502             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 503             {
 504                 $decomp_table[$src] = [ hex $1, hex $2 ];
 505                 push @compose_table, [ hex $1, hex $2, $src ];
 506             }
 507             elsif ($decomp =~ /^(<[a-z]+>\s)*([0-9a-fA-F]+)$/ &&
 508                    (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)))
 509             {
 510                 # Single char decomposition in the compatibility range
 511                 $compatmap_table[$src] = hex $2;
 512             }
 513         }
 514         else
 515         {
 516             next;
 517         }
 518
 519         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 520
 521         # check for loops
 522         for (my $i = $dst; ; $i = $unicode_defaults[$i])
 523         {
 524             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 525             last unless defined($unicode_defaults[$i]);
 526         }
 527         $unicode_defaults[$src] = $dst;
 528     }
 529     close $UNICODE_DATA;
 530
 531     # patch the category of some special characters
 532
 533     foreach my $cat (keys %special_categories)
 534     {
 535         my $flag = $ctype{$cat};
 536         foreach my $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 537     }
 538 }
 539
 540
 541 ################################################################
 542 # define a new lead byte
 543 sub add_lead_byte($)
 544 {
 545     my $ch = shift;
 546     push @lead_bytes, $ch;
 547     $cp2uni[$ch] = 0;
 548 }
 549
 550 ################################################################
 551 # define a new char mapping
 552 sub add_mapping($$)
 553 {
 554     my ($cp, $uni) = @_;
 555     $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 556     $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 557     if ($cp > 0xff && !defined($cp2uni[$cp >> 8]))
 558     {
 559         add_lead_byte( $cp >> 8 );
 560     }
 561 }
 562
 563 ################################################################
 564 # parse a standard codepage file
 565 sub read_codepage_file($)
 566 {
 567     my $name = shift;
 568     my $INPUT = open_data_file( $MAPPINGS, $name );
 569
 570     while (<$INPUT>)
 571     {
 572         next if /^\#/;  # skip comments
 573         next if /^$/;  # skip empty lines
 574         next if /\x1a/;  # skip ^Z
 575         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 576
 577         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 578         {
 579             add_lead_byte( hex $1 );
 580             next;
 581         }
 582         # 0x12 0x3456
 583         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 584         {
 585             add_mapping( hex $1, hex $2 );
 586             next;
 587         }
 588         # 0x12 0x3456+0x7890
 589         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\+.*\s+(\#.*)?/)
 590         {
 591             add_mapping( hex $1, hex $2 );
 592             next;
 593         }
 594         # 0x12 <LR>+0x3456
 595         if (/^0x([0-9a-fA-F]+)\s+<LR>\+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 596         {
 597             add_mapping( hex $1, hex $2 );
 598             next;
 599         }
 600         # 0x12 <RL>+0x3456
 601         if (/^0x([0-9a-fA-F]+)\s+<RL>\+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 602         {
 603             my $uni = hex $2;
 604             add_mapping( hex $1, $uni ) if $uni >= 0x80;
 605             next;
 606         }
 607         die "$name: Unrecognized line $_\n";
 608     }
 609     close $INPUT;
 610 }
 611
 612
 613 ################################################################
 614 # fill input data for the 20127 (us-ascii) codepage
 615 sub fill_20127_codepage()
 616 {
 617     for (my $i = 0; $i < 128; $i++) { $cp2uni[$i] = $uni2cp[$i] = $i; }
 618     for (my $i = 128; $i < 256; $i++) { $cp2uni[$i] = $i & 0x7f; }
 619 }
 620
 621 ################################################################
 622 # get a mapping including glyph chars for MB_USEGLYPHCHARS
 623
 624 sub get_glyphs_mapping(@)
 625 {
 626     $_[0x01] = 0x263a;  # (WHITE SMILING FACE)
 627     $_[0x02] = 0x263b;  # (BLACK SMILING FACE)
 628     $_[0x03] = 0x2665;  # (BLACK HEART SUIT)
 629     $_[0x04] = 0x2666;  # (BLACK DIAMOND SUIT)
 630     $_[0x05] = 0x2663;  # (BLACK CLUB SUIT)
 631     $_[0x06] = 0x2660;  # (BLACK SPADE SUIT)
 632     $_[0x07] = 0x2022;  # (BULLET)
 633     $_[0x08] = 0x25d8;  # (INVERSE BULLET)
 634     $_[0x09] = 0x25cb;  # (WHITE CIRCLE)
 635     $_[0x0a] = 0x25d9;  # (INVERSE WHITE CIRCLE)
 636     $_[0x0b] = 0x2642;  # (MALE SIGN)
 637     $_[0x0c] = 0x2640;  # (FEMALE SIGN)
 638     $_[0x0d] = 0x266a;  # (EIGHTH NOTE)
 639     $_[0x0e] = 0x266b;  # (BEAMED EIGHTH NOTES)
 640     $_[0x0f] = 0x263c;  # (WHITE SUN WITH RAYS)
 641     $_[0x10] = 0x25ba;  # (BLACK RIGHT-POINTING POINTER)
 642     $_[0x11] = 0x25c4;  # (BLACK LEFT-POINTING POINTER)
 643     $_[0x12] = 0x2195;  # (UP DOWN ARROW)
 644     $_[0x13] = 0x203c;  # (DOUBLE EXCLAMATION MARK)
 645     $_[0x14] = 0x00b6;  # (PILCROW SIGN)
 646     $_[0x15] = 0x00a7;  # (SECTION SIGN)
 647     $_[0x16] = 0x25ac;  # (BLACK RECTANGLE)
 648     $_[0x17] = 0x21a8;  # (UP DOWN ARROW WITH BASE)
 649     $_[0x18] = 0x2191;  # (UPWARDS ARROW)
 650     $_[0x19] = 0x2193;  # (DOWNWARDS ARROW)
 651     $_[0x1a] = 0x2192;  # (RIGHTWARDS ARROW)
 652     $_[0x1b] = 0x2190;  # (LEFTWARDS ARROW)
 653     $_[0x1c] = 0x221f;  # (RIGHT ANGLE)
 654     $_[0x1d] = 0x2194;  # (LEFT RIGHT ARROW)
 655     $_[0x1e] = 0x25b2;  # (BLACK UP-POINTING TRIANGLE)
 656     $_[0x1f] = 0x25bc;  # (BLACK DOWN-POINTING TRIANGLE)
 657     $_[0x7f] = 0x2302;  # (HOUSE)
 658     return @_;
 659 }
 660
 661 ################################################################
 662 # build EUC-JP table from the JIS 0208 file
 663 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 664 # but this would require 3-byte DBCS characters
 665 sub READ_JIS0208_FILE($)
 666 {
 667     my $name = shift;
 668
 669     # ASCII chars
 670     for (my $i = 0x00; $i <= 0x7f; $i++)
 671     {
 672         $cp2uni[$i] = $i;
 673         $uni2cp[$i] = $i;
 674     }
 675
 676     # JIS X 0201 right plane
 677     for (my $i = 0xa1; $i <= 0xdf; $i++)
 678     {
 679         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 680         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 681     }
 682
 683     # lead bytes
 684     foreach my $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 685     {
 686         push @lead_bytes,$i;
 687         $cp2uni[$i] = 0;
 688     }
 689
 690     # undefined chars
 691     foreach my $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 692     {
 693         $cp2uni[$i] = $DEF_CHAR;
 694     }
 695
 696     # Shift-JIS compatibility
 697     $uni2cp[0x00a5] = 0x5c;
 698     $uni2cp[0x203e] = 0x7e;
 699
 700     # Fix backslash conversion
 701     $cp2uni[0xa1c0] = 0xff3c;
 702     $uni2cp[0xff3c] = 0xa1c0;
 703
 704     my $INPUT = open_data_file( $MAPPINGS, $name );
 705     while (<$INPUT>)
 706     {
 707         next if /^\#/;  # skip comments
 708         next if /^$/;  # skip empty lines
 709         next if /\x1a/;  # skip ^Z
 710         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 711         {
 712             my $cp = 0x8080 + hex $1;
 713             my $uni = hex $2;
 714             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 715             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 716             next;
 717         }
 718         die "$name: Unrecognized line $_\n";
 719     }
 720     close $INPUT;
 721 }
 722
 723
 724 ################################################################
 725 # build the sort keys table
 726 sub READ_SORTKEYS_FILE()
 727 {
 728     my @sortkeys = ();
 729     for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
 730
 731     my $INPUT = open_data_file( $REPORTS, $SORTKEYS );
 732     while (<$INPUT>)
 733     {
 734         next if /^\#/;  # skip comments
 735         next if /^$/;  # skip empty lines
 736         next if /\x1a/;  # skip ^Z
 737         next if /^\@version/;  # skip @version header
 738         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 739         {
 740             my ($uni,$variable) = (hex $1, $2);
 741             next if $uni > 65535;
 742             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 743             next;
 744         }
 745         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 746         {
 747             # multiple character sequence, ignored for now
 748             next;
 749         }
 750         die "$SORTKEYS: Unrecognized line $_\n";
 751     }
 752     close $INPUT;
 753
 754     # compress the keys to 32 bit:
 755     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 756
 757     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 758                        ${$a}[2] <=> ${$b}[2] or
 759                        ${$a}[3] <=> ${$b}[3] or
 760                        ${$a}[4] <=> ${$b}[4] or
 761                        $a cmp $b; } @sortkeys;
 762
 763     my ($n2, $n3) = (1, 1);
 764     my @keys = (-1, -1, -1, -1, -1 );
 765     my @flatkeys = ();
 766
 767     for (my $i = 0; $i < 65536; $i++)
 768     {
 769         my @current = @{$sortkeys[$i]};
 770         next if $current[0] == -1;
 771         if ($current[1] == $keys[1])
 772         {
 773             if ($current[2] == $keys[2])
 774             {
 775                 if ($current[3] == $keys[3])
 776                 {
 777                     # nothing
 778                 }
 779                 else
 780                 {
 781                     $keys[3] = $current[3];
 782                     $n3++;
 783                     die if ($n3 >= 16);
 784                 }
 785             }
 786             else
 787             {
 788                 $keys[2] = $current[2];
 789                 $keys[3] = $current[3];
 790                 $n2++;
 791                 $n3 = 1;
 792                 die if ($n2 >= 256);
 793             }
 794         }
 795         else
 796         {
 797             $keys[1] = $current[1];
 798             $keys[2] = $current[2];
 799             $keys[3] = $current[3];
 800             $n2 = 1;
 801             $n3 = 1;
 802         }
 803
 804         if ($current[2]) { $current[2] = $n2; }
 805         if ($current[3]) { $current[3] = $n3; }
 806         if ($current[4]) { $current[4] = 1; }
 807
 808         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 809     }
 810     return @flatkeys;
 811 }
 812
 813
 814 ################################################################
 815 # build the sort keys table
 816 sub DUMP_SORTKEYS($@)
 817 {
 818     my ($filename, @keys) = @_;
 819
 820     # count the number of 256-key ranges that contain something
 821
 822     my @offsets = ();
 823     my $ranges = 2;
 824     for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
 825     for (my $i = 0; $i < 65536; $i++)
 826     {
 827         next unless defined $keys[$i];
 828         $offsets[$i >> 8] = $ranges * 256;
 829         $ranges++;
 830         $i |= 255;
 831     }
 832
 833     # output the range offsets
 834
 835     open OUTPUT,">$filename.new" or die "Cannot create $filename";
 836     printf "Building $filename\n";
 837     printf OUTPUT "/* Unicode collation element table */\n";
 838     printf OUTPUT "/* generated from %s */\n", "$REPORTS/$SORTKEYS";
 839     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 840
 841     printf OUTPUT "const unsigned int collation_table[%d] =\n{\n", $ranges*256;
 842     printf OUTPUT "    /* index */\n";
 843     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%08x", 0, @offsets );
 844
 845     # output the default values
 846
 847     printf OUTPUT "    /* defaults */\n";
 848     printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0, (0xffffffff) x 256 );
 849
 850     # output all the key ranges
 851
 852     for (my $i = 0; $i < 256; $i++)
 853     {
 854         next if $offsets[$i] == 256;
 855         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 856         printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0xffffffff, @keys[($i<<8) .. ($i<<8)+255] );
 857     }
 858     printf OUTPUT "\n};\n";
 859     close OUTPUT;
 860     save_file($filename);
 861 }
 862
 863
 864 ################################################################
 865 # add default mappings once the file had been read
 866 sub add_default_mappings($)
 867 {
 868     my $first_private = shift;
 869
 870     # Apply aliases
 871
 872     foreach my $alias (@unicode_aliases)
 873     {
 874         my $target = undef;
 875         foreach my $src (@$alias)
 876         {
 877             if (defined($uni2cp[$src]))
 878             {
 879                 $target = $uni2cp[$src];
 880                 last;
 881             }
 882         }
 883         next unless defined($target);
 884
 885         # At least one char of the alias set is defined, set the others to the same value
 886         foreach my $src (@$alias)
 887         {
 888             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 889         }
 890     }
 891
 892     # For every src -> target mapping in the defaults table,
 893     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 894
 895     for (my $src = 0; $src < 65536; $src++)
 896     {
 897         next if defined($uni2cp[$src]);  # source has a definition already
 898         next unless defined($unicode_defaults[$src]);  # no default for this char
 899         my $target = $unicode_defaults[$src];
 900
 901         # do a recursive mapping until we find a target char that is defined
 902         while (!defined($uni2cp[$target]) &&
 903                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 904
 905         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 906     }
 907
 908     # Add a private char mapping for undefined chars
 909
 910     if ($first_private)
 911     {
 912         for (my $i = 32; $i < 256; $i++)
 913         {
 914             next if defined($cp2uni[$i]);
 915             $cp2uni[$i] = $first_private;
 916             $uni2cp[$first_private] = $i;
 917             $first_private++;
 918         }
 919     }
 920
 921     # Add an identity mapping for all undefined chars
 922
 923     for (my $i = 0; $i < 256; $i++)
 924     {
 925         next if defined($cp2uni[$i]);
 926         next if defined($uni2cp[$i]);
 927         $cp2uni[$i] = $uni2cp[$i] = $i;
 928     }
 929 }
 930
 931 ################################################################
 932 # dump an array of integers
 933 sub DUMP_ARRAY($$@)
 934 {
 935     my ($format,$default,@array) = @_;
 936     my $i;
 937     my $ret = "    ";
 938     for ($i = 0; $i < $#array; $i++)
 939     {
 940         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 941         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 942     }
 943     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 944     return $ret;
 945 }
 946
 947
 948 ################################################################
 949 # dump a unicode->ascii mapping table
 950 sub dump_uni2cp_table($$)
 951 {
 952     my ($width, $def) = @_;
 953
 954     # count the number of unicode->ascii subtables that contain something
 955
 956     my @lowerbounds;
 957     my @upperbounds;
 958     my $index = 0;
 959     my @filled = (-1) x 256;
 960     for (my $i = 0; $i < 65536; $i++)
 961     {
 962         next unless defined $uni2cp[$i];
 963         if ($filled[$i >> 8] == -1)
 964         {
 965           $lowerbounds[$index] = $i & 0xff;
 966           $upperbounds[$index] = 0xff - $lowerbounds[$index];
 967           $filled[$i >> 8] = $index * 256;
 968           $index++;
 969         }
 970         else
 971         {
 972           $upperbounds[$index-1] = 0xff - ($i & 0xff);
 973         }
 974     }
 975
 976     # add defaults mapping at the end
 977     $filled[256] = $index * 256;
 978     $lowerbounds[$index] = 255;
 979     $upperbounds[$index] = 0;
 980     $index++;
 981
 982     # collapse blocks upwards if possible
 983     my $removed = 0;
 984     $index = 0;
 985     for (my $i = 0; $i < 257; $i++)
 986     {
 987         next if $filled[$i] == -1;
 988         if ($upperbounds[$index - 1] > $lowerbounds[$index])
 989         {
 990            $removed += $lowerbounds[$index];
 991         }
 992         else
 993         {
 994            $removed += $upperbounds[$index - 1];
 995            $lowerbounds[$index] = $upperbounds[$index - 1];
 996         }
 997         $filled[$i] -= $removed;
 998         $index++;
 999     }
1000
1001     # output all the subtables into a single array
1002
1003     printf OUTPUT "static const unsigned %s uni2cp_low[%d] =\n{\n",
1004                   $width == 8 ? "char" : "short", ($index + 1) * 256 - $removed;
1005
1006     my $format = $width == 8 ? "0x%02x" : "0x%04x";
1007     for (my $i = $index = 0; $i < 257; $i++)
1008     {
1009         next if $filled[$i] == -1;
1010         my $start = ($i << 8) + $lowerbounds[$index];
1011         my $end = ($i << 8) + 255;
1012         if ($i == 256)
1013         {
1014             print OUTPUT "    /* defaults */\n";
1015             printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( $format, 0, ($def) x ($end - $start + 1) );
1016         }
1017         else
1018         {
1019             printf OUTPUT "    /* 0x%04x .. 0x%04x */\n", $start, $end;
1020             printf OUTPUT "%s,\n", DUMP_ARRAY( $format, $def, @uni2cp[$start .. $end] );
1021         }
1022         $index++;
1023     }
1024
1025     # output a table of the offsets of the subtables in the previous array
1026
1027     for (my $i = 0; $i < 256; $i++)
1028     {
1029         if ($filled[$i] == -1) { $filled[$i] = $filled[256]; }
1030     }
1031     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
1032     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @filled[0..255] );
1033 }
1034
1035 ################################################################
1036 # dump an SBCS mapping table
1037 sub dump_sbcs_table($$$$$)
1038 {
1039     my ($codepage, $has_glyphs, $name, $def, $defw) = @_;
1040     my $i;
1041
1042     # output the ascii->unicode table
1043
1044     if ($has_glyphs)
1045     {
1046         printf OUTPUT "static const WCHAR cp2uni[512] =\n";
1047         printf OUTPUT "{\n%s", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
1048         printf OUTPUT ",\n    /* glyphs */\n%s\n};\n\n",
1049                       DUMP_ARRAY( "0x%04x", $defw, get_glyphs_mapping(@cp2uni[0 .. 255]) );
1050     }
1051     else
1052     {
1053         printf OUTPUT "static const WCHAR cp2uni[256] =\n";
1054         printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
1055     }
1056
1057     dump_uni2cp_table( 8, $def );
1058
1059     # output the code page descriptor
1060
1061     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
1062     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
1063                   $codepage, $def, $defw, $name;
1064     printf OUTPUT "    cp2uni,\n";
1065     if ($has_glyphs) { printf OUTPUT "    cp2uni + 256,\n"; }
1066     else { printf OUTPUT "    cp2uni,\n"; }
1067     printf OUTPUT "    uni2cp_low,\n";
1068     printf OUTPUT "    uni2cp_high\n};\n";
1069 }
1070
1071
1072 ################################################################
1073 # dump a DBCS mapping table
1074 sub dump_dbcs_table($$$$@)
1075 {
1076     my ($codepage, $name, $def, $defw, @lb_ranges) = @_;
1077
1078     # build a list of lead bytes that are actually used
1079
1080     my @lblist = ();
1081     LBLOOP: for (my $y = 0; $y <= $#lead_bytes; $y++)
1082     {
1083         my $base = $lead_bytes[$y] << 8;
1084         for (my $x = 0; $x < 256; $x++)
1085         {
1086             if (defined $cp2uni[$base+$x])
1087             {
1088                 push @lblist,$lead_bytes[$y];
1089                 next LBLOOP;
1090             }
1091         }
1092     }
1093     my $unused = ($#lead_bytes > $#lblist);
1094
1095     # output the ascii->unicode table for the single byte chars
1096
1097     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
1098     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
1099
1100     # output the default table for unused lead bytes
1101
1102     if ($unused)
1103     {
1104         printf OUTPUT "    /* unused lead bytes */\n";
1105         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($defw) x 256 );
1106     }
1107
1108     # output the ascii->unicode table for each DBCS lead byte
1109
1110     for (my $y = 0; $y <= $#lblist; $y++)
1111     {
1112         my $base = $lblist[$y] << 8;
1113         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
1114         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[$base .. $base+255] );
1115         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
1116     }
1117
1118     # output the lead byte subtables offsets
1119
1120     my @offsets = (0) x 256;
1121     for (my $x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
1122     if ($unused)
1123     {
1124         # increment all lead bytes offset to take into account the unused table
1125         for (my $x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
1126     }
1127     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
1128     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
1129
1130     dump_uni2cp_table( 16, $def );
1131
1132     # output the code page descriptor
1133
1134     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
1135     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
1136                   $codepage, $def, $defw, $name;
1137     printf OUTPUT "    cp2uni,\n";
1138     printf OUTPUT "    cp2uni_leadbytes,\n";
1139     printf OUTPUT "    uni2cp_low,\n";
1140     printf OUTPUT "    uni2cp_high,\n";
1141     printf OUTPUT "    {\n    %s\n    }\n", DUMP_ARRAY( "0x%02x", 0, @lb_ranges, 0, 0 );
1142     printf OUTPUT "};\n";
1143 }
1144
1145
1146 ################################################################
1147 # get the list of defined lead byte ranges
1148 sub get_lb_ranges()
1149 {
1150     my @list = ();
1151     my @ranges = ();
1152     my $i = 0;
1153     foreach $i (@lead_bytes) { $list[$i] = 1; }
1154     my $on = 0;
1155     for (my $i = 0; $i < 256; $i++)
1156     {
1157         if ($on)
1158         {
1159             if (!defined $list[$i]) { push @ranges, $i-1; $on = 0; }
1160         }
1161         else
1162         {
1163             if ($list[$i]) { push @ranges, $i; $on = 1; }
1164         }
1165     }
1166     if ($on) { push @ranges, 0xff; }
1167     return @ranges;
1168 }
1169
1170 ################################################################
1171 # dump the Indic Syllabic Category table
1172 sub dump_indic($)
1173 {
1174     my $filename = shift;
1175     my @indic_table = ($indic_types{'Other'}) x 65536;
1176
1177     my $INPUT = open_data_file( $UNIDATA, "IndicSyllabicCategory.txt" );
1178     while (<$INPUT>)
1179     {
1180         next if /^\#/;  # skip comments
1181         next if /^\s*$/;  # skip empty lines
1182         next if /\x1a/;  # skip ^Z
1183         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*#/)
1184         {
1185             my $type = $2;
1186             die "unknown indic $type" unless defined $indic_types{$type};
1187             if (hex $1 < 65536)
1188             {
1189                 $indic_table[hex $1] = $indic_types{$type};
1190             }
1191             next;
1192         }
1193         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1194         {
1195             my $type = $3;
1196             die "unknown indic $type" unless defined $indic_types{$type};
1197             if (hex $1 < 65536 and hex $2 < 6536)
1198             {
1199                 foreach my $i (hex $1 .. hex $2)
1200                 {
1201                     $indic_table[$i] = $indic_types{$type};
1202                 }
1203             }
1204             next;
1205         }
1206         die "malformed line $_";
1207     }
1208     close $INPUT;
1209
1210     $INPUT = open_data_file( $UNIDATA, "IndicMatraCategory.txt" );
1211     while (<$INPUT>)
1212     {
1213         next if /^\#/;  # skip comments
1214         next if /^\s*$/;  # skip empty lines
1215         next if /\x1a/;  # skip ^Z
1216         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z]+)\s*#/)
1217         {
1218             my $type = $2;
1219             die "unknown matra $type" unless defined $matra_types{$type};
1220             $indic_table[hex $1] += $matra_types{$type} << 8;
1221             next;
1222         }
1223         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1224         {
1225             my $type = $3;
1226             die "unknown matra $type" unless defined $matra_types{$type};
1227             foreach my $i (hex $1 .. hex $2)
1228             {
1229                 $indic_table[$i] += $matra_types{$type} << 8;
1230             }
1231             next;
1232         }
1233         die "malformed line $_";
1234     }
1235     close $INPUT;
1236
1237     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1238     print "Building $filename\n";
1239     print OUTPUT "/* Unicode Indic Syllabic Category */\n";
1240     print OUTPUT "/* generated from $UNIDATA/IndicSyllabicCategory.txt */\n";
1241     print OUTPUT "/*       and from $UNIDATA/IndicMatraCategory.txt */\n";
1242     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1243
1244     dump_two_level_mapping( "indic_syllabic_table", @indic_table);
1245
1246     close OUTPUT;
1247     save_file($filename);
1248 }
1249
1250 ################################################################
1251 # dump the Line Break Properties table
1252 sub dump_linebreak($)
1253 {
1254     my $filename = shift;
1255     my @break_table = ($break_types{'XX'}) x 65536;
1256     my $next_group = 0;
1257
1258     my $INPUT = open_data_file( $UNIDATA, "LineBreak.txt" );
1259     while (<$INPUT>)
1260     {
1261         next if /^\#/;  # skip comments
1262         next if /^\s*$/;  # skip empty lines
1263         next if /\x1a/;  # skip ^Z
1264         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1265         {
1266             my $type = $2;
1267             die "unknown breaktype $type" unless defined $break_types{$type};
1268             $break_table[hex $1] = $break_types{$type};
1269             next;
1270         }
1271         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1272         {
1273             my $type = $3;
1274             die "unknown breaktype $type" unless defined $break_types{$type};
1275             foreach my $i (hex $1 .. hex $2)
1276             {
1277                 $break_table[$i] = $break_types{$type};
1278             }
1279             next;
1280         }
1281         die "malformed line $_";
1282     }
1283     close $INPUT;
1284
1285     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1286     print "Building $filename\n";
1287     print OUTPUT "/* Unicode Line Break Properties */\n";
1288     print OUTPUT "/* generated from $UNIDATA/LineBreak.txt */\n";
1289     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1290
1291     dump_two_level_mapping( "wine_linebreak_table", @break_table);
1292
1293     close OUTPUT;
1294     save_file($filename);
1295 }
1296
1297 ################################################################
1298 # dump Script IDs table
1299 sub dump_scripts($)
1300 {
1301     my $filename = shift;
1302     my $header = $filename;
1303     my @scripts_table = (0) x 65536; # 0 means unknown script
1304     my $script_index;
1305     my %scripts;
1306     my $i;
1307
1308     my $INPUT = open_data_file( $UNIDATA, "Scripts.txt" );
1309     while (<$INPUT>)
1310     {
1311         my $type = "";
1312
1313         next if /^\#/;  # skip comments
1314         next if /^\s*$/;  # skip empty lines
1315         next if /\x1a/;  # skip ^Z
1316         if (/^\s*([0-9a-fA-F]+)\s*;\s*([_a-zA-Z]+)\s*/)
1317         {
1318             $type = $2;
1319         }
1320         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([_a-zA-Z]+)\s*/)
1321         {
1322             $type = $3;
1323         }
1324         else
1325         {
1326             die "malformed line $_";
1327         }
1328
1329         # ignore some scripts
1330         if ($type eq "" || $type eq "Common" || $type eq "Inherited")
1331         {
1332             next;
1333         }
1334
1335         $scripts{$type} = -1;
1336     }
1337
1338     $script_index = 1;
1339     foreach my $script (sort keys %scripts) {
1340         $scripts{$script} = ++$script_index;
1341     }
1342
1343     # now fill a table
1344     seek $INPUT, 0, 0;
1345     while (<$INPUT>)
1346     {
1347         my $type = "";
1348
1349         next if /^\#/;  # skip comments
1350         next if /^\s*$/;  # skip empty lines
1351         next if /\x1a/;  # skip ^Z
1352         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z]+)\s*/)
1353         {
1354             $type = $2;
1355             if (defined $scripts{$type})
1356             {
1357                 $scripts_table[hex $1] = $scripts{$type};
1358             }
1359             next;
1360         }
1361         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z]+)\s*/)
1362         {
1363             $type = $3;
1364             if (defined $scripts{$type})
1365             {
1366                 foreach my $i (hex $1 .. hex $2)
1367                 {
1368                     $scripts_table[$i] = $scripts{$type};
1369                 }
1370             }
1371             next;
1372         }
1373     }
1374
1375     close $INPUT;
1376
1377     $header = "$filename.h";
1378     open OUTPUT,">$header.new" or die "Cannot create $header";
1379     print "Building $header\n";
1380     print OUTPUT "/* Unicode Script IDs */\n";
1381     print OUTPUT "/* generated from $UNIDATA/Scripts.txt */\n";
1382     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1383
1384     # reserve Unknown and Common ids
1385     print OUTPUT "enum unicode_script_id {\n";
1386     print OUTPUT "    Script_Unknown = 0,\n";
1387     print OUTPUT "    Script_Common = 1,\n";
1388     foreach my $script (sort keys %scripts)
1389     {
1390         print OUTPUT "    Script_$script = $scripts{$script},\n";
1391     }
1392     print OUTPUT "    Script_LastId = $script_index\n";
1393     print OUTPUT "};\n";
1394
1395     close OUTPUT;
1396     save_file($header);
1397
1398     $filename = "$filename.c";
1399     open OUTPUT,">$filename.new" or die "Cannot create $header";
1400     print "Building $filename\n";
1401     print OUTPUT "/* Unicode Script IDs */\n";
1402     print OUTPUT "/* generated from $UNIDATA/Scripts.txt */\n";
1403     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1404
1405     dump_two_level_mapping( "wine_scripts_table", @scripts_table);
1406     close OUTPUT;
1407     save_file($filename);
1408 }
1409
1410 ################################################################
1411 # dump the BiDi mirroring table
1412 sub dump_mirroring($)
1413 {
1414     my $filename = shift;
1415     my @mirror_table = ();
1416
1417     my $INPUT = open_data_file( $UNIDATA, "BidiMirroring.txt" );
1418     while (<$INPUT>)
1419     {
1420         next if /^\#/;  # skip comments
1421         next if /^$/;  # skip empty lines
1422         next if /\x1a/;  # skip ^Z
1423         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+)/)
1424         {
1425             $mirror_table[hex $1] = hex $2;
1426             next;
1427         }
1428         die "malformed line $_";
1429     }
1430     close $INPUT;
1431
1432     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1433     print "Building $filename\n";
1434     print OUTPUT "/* Unicode BiDi mirroring */\n";
1435     print OUTPUT "/* generated from $UNIDATA/BidiMirroring.txt */\n";
1436     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1437     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1438     DUMP_CASE_TABLE( "wine_mirror_map", @mirror_table );
1439     close OUTPUT;
1440     save_file($filename);
1441 }
1442
1443 ################################################################
1444 # dump the Bidi Brackets
1445 sub dump_bracket($)
1446 {
1447     my $filename = shift;
1448     my @bracket_table = (0) x 65536;
1449
1450     my $INPUT = open_data_file( $UNIDATA, "BidiBrackets.txt" );
1451     while (<$INPUT>)
1452     {
1453         next if /^\#/;  # skip comments
1454         next if /^\s*$/;  # skip empty lines
1455         next if /\x1a/;  # skip ^Z
1456         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+);\s*([con])/)
1457         {
1458             my $type = $3;
1459             die "unknown bracket $type" unless defined $bracket_types{$type};
1460             die "characters too distant $1 and $2" if abs(hex($2) - hex($1)) >= 128;
1461             $bracket_table[hex $1] = (hex($2) - hex($1)) % 255;
1462             $bracket_table[hex $1] += $bracket_types{$type} << 8;
1463             next;
1464         }
1465         die "malformed line $_";
1466     }
1467     close $INPUT;
1468
1469     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1470     print "Building $filename\n";
1471     print OUTPUT "/* Unicode Bidirectional Bracket table */\n";
1472     print OUTPUT "/* generated from $UNIDATA/BidiBrackets.txt */\n";
1473     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1474
1475     dump_two_level_mapping( "bidi_bracket_table", @bracket_table);
1476
1477     close OUTPUT;
1478     save_file($filename);
1479 }
1480
1481 ################################################################
1482 # dump the Arabic shaping table
1483 sub dump_shaping($)
1484 {
1485     my $filename = shift;
1486     my %groups;
1487     my $next_group = 0;
1488
1489     $groups{"No_Joining_Group"} = $next_group++;
1490
1491     my $INPUT = open_data_file( $UNIDATA, "ArabicShaping.txt" );
1492     while (<$INPUT>)
1493     {
1494         next if /^\#/;  # skip comments
1495         next if /^\s*$/;  # skip empty lines
1496         next if /\x1a/;  # skip ^Z
1497         if (/^\s*([0-9a-fA-F]+)\s*;.*;\s*([RLDCUT])\s*;\s*(\w+)/)
1498         {
1499             my $type = $2;
1500             my $group = $3;
1501             $groups{$group} = $next_group++ unless defined $groups{$group};
1502             $joining_table[hex $1] = $joining_types{$type} | ($groups{$group} << 8);
1503             next;
1504         }
1505         die "malformed line $_";
1506     }
1507     close $INPUT;
1508
1509     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1510     print "Building $filename\n";
1511     print OUTPUT "/* Unicode Arabic shaping */\n";
1512     print OUTPUT "/* generated from $UNIDATA/ArabicShaping.txt */\n";
1513     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1514
1515     dump_two_level_mapping( "wine_shaping_table", @joining_table );
1516
1517     print OUTPUT "\nconst unsigned short wine_shaping_forms[256][4] =\n{\n";
1518     for (my $i = 0x600; $i <= 0x6ff; $i++)
1519     {
1520         printf OUTPUT "    { 0x%04x, 0x%04x, 0x%04x, 0x%04x },\n",
1521             ${joining_forms{"isolated"}}[$i] || $i,
1522             ${joining_forms{"final"}}[$i] || $i,
1523             ${joining_forms{"initial"}}[$i] || $i,
1524             ${joining_forms{"medial"}}[$i] || $i;
1525     }
1526     print OUTPUT "};\n";
1527
1528     close OUTPUT;
1529     save_file($filename);
1530 }
1531
1532 ################################################################
1533 # dump the Vertical Orientation table
1534 sub dump_vertical($)
1535 {
1536     my $filename = shift;
1537     my @vertical_table = ($vertical_types{'R'}) x 65536;
1538
1539     my $INPUT = open_data_file( $VERTICALDATA, "VerticalOrientation-11.txt" );
1540     while (<$INPUT>)
1541     {
1542         next if /^\#/;  # skip comments
1543         next if /^\s*$/;  # skip empty lines
1544         next if /\x1a/;  # skip ^Z
1545         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/)
1546         {
1547             my $type = $2;
1548             die "unknown vertical $type" unless defined $vertical_types{$type};
1549             if (hex $1 < 65536)
1550             {
1551                 $vertical_table[hex $1] = $vertical_types{$type};
1552             }
1553             next;
1554         }
1555         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*/)
1556         {
1557             my $type = $3;
1558             die "unknown vertical $type" unless defined $vertical_types{$type};
1559             foreach my $i (hex $1 .. hex $2)
1560             {
1561                 $vertical_table[$i] = $vertical_types{$type};
1562             }
1563             next;
1564         }
1565         die "malformed line $_";
1566     }
1567     close $INPUT;
1568
1569     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1570     print "Building $filename\n";
1571     print OUTPUT "/* Unicode Vertical Orientation */\n";
1572     print OUTPUT "/* generated from $VERTICALDATA/VerticalOrientation-11.txt */\n";
1573     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1574
1575     dump_two_level_mapping( "vertical_orientation_table", @vertical_table);
1576
1577     close OUTPUT;
1578     save_file($filename);
1579 }
1580
1581 ################################################################
1582 # dump the case mapping tables
1583 sub DUMP_CASE_MAPPINGS($)
1584 {
1585     my $filename = shift;
1586     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1587     printf "Building $filename\n";
1588     printf OUTPUT "/* Unicode case mappings */\n";
1589     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1590     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1591
1592     DUMP_CASE_TABLE( "wine_casemap_lower", @tolower_table );
1593     DUMP_CASE_TABLE( "wine_casemap_upper", @toupper_table );
1594     DUMP_CASE_TABLE( "wine_digitmap",  @digitmap_table );
1595     DUMP_CASE_TABLE( "wine_compatmap", @compatmap_table );
1596     close OUTPUT;
1597     save_file($filename);
1598 }
1599
1600
1601 ################################################################
1602 # dump a case mapping table
1603 sub DUMP_CASE_TABLE($@)
1604 {
1605     my ($name,@table) = @_;
1606
1607     # count the number of sub tables that contain something
1608     # also compute the low and upper populated bounds
1609
1610     my @lowerbounds = ( 0, 0 );
1611     my @upperbounds = ( 0, 255 );
1612     my $index = 0;
1613     my @filled = ();
1614     for (my $i = 0; $i < 65536; $i++)
1615     {
1616         next unless defined $table[$i];
1617         if (!defined $filled[$i >> 8])
1618         {
1619           $lowerbounds[$index] = $i & 0xff;
1620           $upperbounds[$index] = 0xff - $lowerbounds[$index];
1621           $filled[$i >> 8] = $index * 256 + 512;
1622           $index++;
1623         }
1624         else
1625         {
1626           $upperbounds[$index-1] = 0xff - ($i & 0xff);
1627         }
1628         $table[$i] = ($table[$i] - $i) & 0xffff;
1629     }
1630
1631     # Collapse blocks upwards if possible
1632     my $removed = 0;
1633     $index = 0;
1634     for (my $i = 0; $i < 256; $i++)
1635     {
1636         next unless defined $filled[$i];
1637         if ($upperbounds[$index - 1] > $lowerbounds[$index])
1638         {
1639            $removed = $removed + $lowerbounds[$index];
1640         }
1641         else
1642         {
1643            $removed = $removed + $upperbounds[$index - 1];
1644            $lowerbounds[$index] = $upperbounds[$index - 1];
1645         }
1646         $filled[$i] = $filled[$i] - $removed;
1647         $index++;
1648     }
1649
1650     # dump the table
1651
1652     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
1653     printf OUTPUT "{\n    /* index */\n";
1654     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
1655     printf OUTPUT "    /* defaults */\n";
1656     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
1657     $index = 0;
1658     for (my $i = 0; $i < 256; $i++)
1659     {
1660         next unless $filled[$i];
1661         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
1662         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0,
1663                       @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
1664         $index++;
1665     }
1666     printf OUTPUT "\n};\n";
1667 }
1668
1669 ################################################################
1670 # compress a mapping table by removing identical rows
1671 sub compress_array($@)
1672 {
1673     my $rows = shift;
1674     my @table = @_;
1675     my $len = @table / $rows;
1676     my @array = (0) x $rows;
1677     my %sequences;
1678
1679     # try to merge table rows
1680     for (my $row = 0; $row < $rows; $row++)
1681     {
1682         my $rowtxt = pack "S*", @table[($row * $len)..($row * $len + $len - 1)];
1683         if (defined($sequences{$rowtxt}))
1684         {
1685             # reuse an existing row
1686             $array[$row] = $sequences{$rowtxt};
1687         }
1688         else
1689         {
1690             # create a new row
1691             $sequences{$rowtxt} = $array[$row] = $#array + 1;
1692             push @array, @table[$row * $len..$row * $len + $len - 1];
1693         }
1694     }
1695     return @array;
1696 }
1697
1698 ################################################################
1699 # dump a simple char -> 16-bit value mapping table
1700 sub dump_simple_mapping($@)
1701 {
1702     my $name = shift;
1703     my @array = compress_array( 256, @_[0..65535] );
1704
1705     printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, $#array+1;
1706     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1707     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1708 }
1709
1710 ################################################################
1711 # dump a char -> 16-bit value mapping table using two-level tables
1712 sub dump_two_level_mapping($@)
1713 {
1714     my $name = shift;
1715     my @row_array = compress_array( 4096, @_[0..65535] );
1716     my @array = compress_array( 256, @row_array[0..4095] );
1717
1718     for (my $i = 256; $i < @array; $i++) { $array[$i] += @array - 4096; }
1719
1720     printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, @array + @row_array - 4096;
1721     printf OUTPUT "    /* level 1 offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1722     printf OUTPUT "    /* level 2 offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1723     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @row_array[4096..$#row_array] );
1724 }
1725
1726 ################################################################
1727 # dump a binary case mapping table in l_intl.nls format
1728 sub dump_binary_case_table(@)
1729 {
1730     my (@table) = @_;
1731
1732     my %difftables_hash = ();
1733     my @difftables;
1734     my %offtables2_hash = ();
1735     my @offtables2 = ();
1736
1737     my @offtable = ();
1738     for (my $i = 0; $i < 256; $i++)
1739     {
1740         my @offtable2 = ();
1741         for(my $j = 0; $j < 16; $j++) # offset table for xx00-xxFF characters
1742         {
1743             my @difftable;
1744             for (my $k = 0; $k < 16; $k++) # case map table for xxx0-xxxF characters
1745             {
1746                 my $char = ($i<<8) + ($j<<4) + $k;
1747                 $difftable[$k] = (defined $table[$char]) ? (($table[$char]-$char) & 0xffff) : 0;
1748             }
1749
1750             my $diff_key = pack "S*", @difftable;
1751             my $offset3 = $difftables_hash{$diff_key};
1752             if (!defined $offset3)
1753             {
1754                 $offset3 = scalar @difftables;
1755                 $difftables_hash{$diff_key} = $offset3;
1756                 push @difftables, @difftable;
1757             }
1758             $offtable2[$j] = $offset3;
1759         }
1760
1761         my $offtable2_key = pack "S*", @offtable2;
1762         my $offset2 = $offtables2_hash{$offtable2_key};
1763         if (!defined $offset2)
1764         {
1765             $offset2 = scalar @offtables2;
1766             $offtables2_hash{$offtable2_key} = $offset2;
1767             push @offtables2, \@offtable2;
1768         }
1769         $offtable[$i] = $offset2;
1770     }
1771
1772     my @output;
1773     my $offset = 0x100; # offset of first subtable in words
1774     foreach (@offtable)
1775     {
1776         push @output, 0x10 * $_ + $offset; # offset of subtable in words
1777     }
1778
1779     $offset = 0x100 + 0x10 * scalar @offtables2; # offset of first difftable in words
1780     foreach(@offtables2)
1781     {
1782         my $table = $_;
1783         foreach(@$table)
1784         {
1785             push @output, $_ + $offset; # offset of difftable in words
1786         }
1787     }
1788
1789     my $len = 1 + scalar @output + scalar @difftables;
1790     return pack "S<*", $len, @output, @difftables;
1791 }
1792
1793
1794 ################################################################
1795 # dump case mappings for l_intl.nls
1796 sub dump_intl_nls($)
1797 {
1798     my $filename = shift;
1799     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1800     printf "Building $filename\n";
1801
1802     binmode OUTPUT;
1803     print OUTPUT pack "S<", 1;  # version
1804     print OUTPUT dump_binary_case_table( @toupper_table );
1805     print OUTPUT dump_binary_case_table( @tolower_table );
1806     close OUTPUT;
1807     save_file($filename);
1808 }
1809
1810
1811 sub load_nameprep_range_table($$$)
1812 {
1813     my ($INPUT, $val, $table_ref) = @_;
1814
1815     while (<$INPUT>)
1816     {
1817         if (/^\s*([0-9a-fA-F]+)-([0-9a-fA-F]+)/)
1818         {
1819             my $last = hex $2;
1820             $last = 65535 if($last >= 65536);
1821             foreach my $i (hex $1 .. $last)
1822             {
1823                 $table_ref->[$i] |= $val;
1824             }
1825             next;
1826         }
1827         elsif (/^\s*([0-9a-fA-F]+)/)
1828         {
1829             if (hex $1 < 65536)
1830             {
1831                 $table_ref->[hex $1] |= $val;
1832             }
1833             next;
1834         }
1835
1836         return if (/End\sTable/);
1837     }
1838 }
1839
1840 sub load_nameprep_map_table($$)
1841 {
1842     my ($INPUT, $table_ref) = @_;
1843
1844     while (<$INPUT>)
1845     {
1846         if (/^\s*([0-9a-fA-F]+);\s;/)
1847         {
1848             # special value for map to nothing
1849             $table_ref->[hex $1] = [0xffff, 0xffff, 0xffff];
1850             next;
1851         }
1852         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+);/)
1853         {
1854             $table_ref->[hex $1] = [hex $2, 0, 0];
1855             next;
1856         }
1857         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1858         {
1859             $table_ref->[hex $1] = [hex $2, hex $3, 0];
1860             next;
1861         }
1862         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1863         {
1864             $table_ref->[hex $1] = [hex $2, hex $3, hex $4];
1865             next;
1866         }
1867
1868         return if (/End\sTable/);
1869     }
1870 }
1871
1872 ################################################################
1873 # dump mapping table, prohibited characters set, unassigned
1874 # characters, bidirectional rules used by nameprep algorithm
1875 sub dump_nameprep($)
1876 {
1877     my $filename = shift;
1878     my @mapping_table = ();
1879     my @flags_table = (0) x 65536;
1880
1881     my $INPUT = open_data_file( $RFCS, $STRINGPREP );
1882     while (<$INPUT>)
1883     {
1884         next unless /Start\sTable/;
1885
1886         load_nameprep_range_table($INPUT, $nameprep_flags{"unassigned"}, \@flags_table) if (/A.1/);
1887         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.1.2/);
1888         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.2.2/);
1889         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.3/);
1890         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.4/);
1891         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.5/);
1892         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.6/);
1893         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.7/);
1894         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.8/);
1895         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.9/);
1896         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_ral"}, \@flags_table) if (/D.1/);
1897         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_l"}, \@flags_table) if (/D.2/);
1898
1899         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.1/);
1900         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.2/);
1901     }
1902     close $INPUT;
1903
1904     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1905     print "Building $filename\n";
1906     print OUTPUT "/* Nameprep algorithm related data */\n";
1907     print OUTPUT "/* generated from $RFCS/$STRINGPREP */\n";
1908     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1909
1910     dump_two_level_mapping( "nameprep_char_type", @flags_table );
1911
1912     ######### mapping table
1913     # first determine all the 16-char subsets that contain something
1914     my @filled = ();
1915     my $pos = 16*3;  # for the null subset
1916     for (my $i = 0; $i < 65536; $i++)
1917     {
1918         next unless defined $mapping_table[$i];
1919         $filled[$i >> 4] = $pos;
1920         $pos += 16*3;
1921         $i |= 15;
1922     }
1923     my $total = $pos;
1924
1925     # now count the 256-char subsets that contain something
1926     my @filled_idx = (256) x 256;
1927     $pos = 256 + 16;
1928     for (my $i = 0; $i < 4096; $i++)
1929     {
1930         next unless $filled[$i];
1931         $filled_idx[$i >> 4] = $pos;
1932         $pos += 16;
1933         $i |= 15;
1934     }
1935     my $null_offset = $pos;
1936     $total += $pos;
1937
1938     # add the index offsets to the subsets positions
1939     for (my $i = 0; $i < 4096; $i++)
1940     {
1941         next unless $filled[$i];
1942         $filled[$i] += $null_offset;
1943     }
1944
1945     # dump the main index
1946     printf OUTPUT "const unsigned short nameprep_mapping[%d] =\n", $total;
1947     printf OUTPUT "{\n    /* index */\n";
1948     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1949     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1950
1951     # dump the second-level indexes
1952     for (my $i = 0; $i < 256; $i++)
1953     {
1954         next unless ($filled_idx[$i] > 256);
1955         my @table = @filled[($i<<4)..($i<<4)+15];
1956         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1957         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1958         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1959     }
1960
1961     # dump the 16-char subsets
1962     printf OUTPUT ",\n    /* null mapping */\n";
1963         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 48 );
1964
1965     for (my $i = 0; $i < 4096; $i++)
1966     {
1967         next unless $filled[$i];
1968         my @table = (0) x 48;
1969         for (my $j = 0; $j < 16; $j++)
1970         {
1971             if (defined $mapping_table[($i<<4) + $j])
1972             {
1973                 $table[3 * $j] = ${$mapping_table[($i << 4) + $j]}[0];
1974                 $table[3 * $j + 1] = ${$mapping_table[($i << 4) + $j]}[1];
1975                 $table[3 * $j + 2] = ${$mapping_table[($i << 4) + $j]}[2];
1976             }
1977         }
1978         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1979         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1980     }
1981
1982     printf OUTPUT "\n};\n";
1983
1984     close OUTPUT;
1985     save_file($filename);
1986 }
1987
1988 ################################################################
1989 # dump the ctype tables
1990 sub DUMP_CTYPE_TABLES($)
1991 {
1992     my $filename = shift;
1993     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1994     printf "Building $filename\n";
1995     printf OUTPUT "/* Unicode ctype tables */\n";
1996     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1997     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1998
1999     # add the direction in the high 4 bits of the category
2000     for (my $i = 0; $i < 65536; $i++)
2001     {
2002         $category_table[$i] |= $direction_table[$i] << 12 if defined $direction_table[$i];
2003     }
2004
2005     dump_simple_mapping( "wine_wctype_table", @category_table );
2006
2007     close OUTPUT;
2008     save_file($filename);
2009 }
2010
2011
2012 ################################################################
2013 # dump the char composition tables
2014 sub DUMP_COMPOSE_TABLES($)
2015 {
2016     my $filename = shift;
2017
2018     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2019     printf "Building $filename\n";
2020     printf OUTPUT "/* Unicode char composition */\n";
2021     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
2022     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
2023
2024     ######### composition table
2025
2026     my @filled = ();
2027     foreach my $i (@compose_table)
2028     {
2029         my @comp = @$i;
2030         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
2031     }
2032
2033     # count how many different second chars we have
2034
2035     my $count = 0;
2036     for (my $i = 0; $i < 65536; $i++)
2037     {
2038         next unless defined $filled[$i];
2039         $count++;
2040     }
2041
2042     # build the table of second chars and offsets
2043
2044     my $pos = $count + 1;
2045     my @table = ();
2046     for (my $i = 0; $i < 65536; $i++)
2047     {
2048         next unless defined $filled[$i];
2049         push @table, $i, $pos;
2050         $pos += @{$filled[$i]};
2051     }
2052     # terminator with last position
2053     push @table, 0, $pos;
2054     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
2055     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
2056
2057     # build the table of first chars and mappings
2058
2059     for (my $i = 0; $i < 65536; $i++)
2060     {
2061         next unless defined $filled[$i];
2062         my @table = ();
2063         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
2064         for (my $j = 0; $j <= $#list; $j++)
2065         {
2066             push @table, $list[$j][0], $list[$j][1];
2067         }
2068         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
2069     }
2070     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
2071
2072     ######### decomposition table
2073
2074     # first determine all the 16-char subsets that contain something
2075
2076     @filled = (0) x 4096;
2077     $pos = 16*2;  # for the null subset
2078     for (my $i = 0; $i < 65536; $i++)
2079     {
2080         next unless defined $decomp_table[$i];
2081         $filled[$i >> 4] = $pos;
2082         $pos += 16*2;
2083         $i |= 15;
2084     }
2085     my $total = $pos;
2086
2087     # now count the 256-char subsets that contain something
2088
2089     my @filled_idx = (256) x 256;
2090     $pos = 256 + 16;
2091     for (my $i = 0; $i < 4096; $i++)
2092     {
2093         next unless $filled[$i];
2094         $filled_idx[$i >> 4] = $pos;
2095         $pos += 16;
2096         $i |= 15;
2097     }
2098     my $null_offset = $pos;  # null mapping
2099     $total += $pos;
2100
2101     # add the index offsets to the subsets positions
2102
2103     for (my $i = 0; $i < 4096; $i++)
2104     {
2105         next unless $filled[$i];
2106         $filled[$i] += $null_offset;
2107     }
2108
2109     # dump the main index
2110
2111     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
2112     printf OUTPUT "{\n    /* index */\n";
2113     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
2114     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
2115
2116     # dump the second-level indexes
2117
2118     for (my $i = 0; $i < 256; $i++)
2119     {
2120         next unless ($filled_idx[$i] > 256);
2121         my @table = @filled[($i<<4)..($i<<4)+15];
2122         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
2123         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
2124         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
2125     }
2126
2127     # dump the 16-char subsets
2128
2129     printf OUTPUT ",\n    /* null mapping */\n";
2130     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
2131
2132     for (my $i = 0; $i < 4096; $i++)
2133     {
2134         next unless $filled[$i];
2135         my @table = (0) x 32;
2136         for (my $j = 0; $j < 16; $j++)
2137         {
2138             if (defined $decomp_table[($i<<4) + $j])
2139             {
2140                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
2141                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
2142             }
2143         }
2144         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
2145         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
2146     }
2147
2148     printf OUTPUT "\n};\n";
2149     close OUTPUT;
2150     save_file($filename);
2151 }
2152
2153 ################################################################
2154 # output a codepage definition file from the global tables
2155 sub output_codepage_file($$$$)
2156 {
2157     my ($codepage, $filename, $comment, $has_glyphs) = @_;
2158
2159     my $output = sprintf "libs/wine/c_%03d.c", $codepage;
2160     open OUTPUT,">$output.new" or die "Cannot create $output";
2161
2162     printf "Building %s from %s (%s)\n", $output, $filename || "hardcoded data", $comment;
2163
2164     # dump all tables
2165
2166     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
2167     if ($filename)
2168     {
2169         print OUTPUT "/* generated from $MAPPINGS/$filename */\n";
2170         print OUTPUT "/* DO NOT EDIT!! */\n\n";
2171     }
2172     else
2173     {
2174         printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
2175     }
2176     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
2177
2178     if (!@lead_bytes) { dump_sbcs_table( $codepage, $has_glyphs, $comment, $default_char, $default_wchar ); }
2179     else { dump_dbcs_table( $codepage, $comment, $default_char, $default_wchar, get_lb_ranges() ); }
2180     close OUTPUT;
2181     save_file($output);
2182 }
2183
2184 ################################################################
2185 # read a "bestfit" Windows mapping file
2186 sub read_bestfit_file($)
2187 {
2188     my ($filename) = @_;
2189     my $state = "";
2190     my ($codepage, $width, $count);
2191     my ($lb_cur, $lb_end);
2192
2193     my $INPUT = open_data_file( $MAPPINGS, $filename ) or die "Cannot open $filename";
2194
2195     while (<$INPUT>)
2196     {
2197         next if /^;/;  # skip comments
2198         next if /^\s*$/;  # skip empty lines
2199         next if /\x1a/;  # skip ^Z
2200         last if /^ENDCODEPAGE/;
2201
2202         if (/^CODEPAGE\s+(\d+)/)
2203         {
2204             $codepage = $1;
2205             next;
2206         }
2207         if (/^CPINFO\s+(\d+)\s+0x([0-9a-fA-f]+)\s+0x([0-9a-fA-F]+)/)
2208         {
2209             $width = $1;
2210             $default_char = hex $2;
2211             $default_wchar = hex $3;
2212             next;
2213         }
2214         if (/^(MBTABLE|WCTABLE|DBCSRANGE|DBCSTABLE)\s+(\d+)/)
2215         {
2216             $state = $1;
2217             $count = $2;
2218             next;
2219         }
2220         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)/)
2221         {
2222             if ($state eq "MBTABLE")
2223             {
2224                 my $cp = hex $1;
2225                 my $uni = hex $2;
2226                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
2227                 next;
2228             }
2229             if ($state eq "WCTABLE")
2230             {
2231                 my $uni = hex $1;
2232                 my $cp = hex $2;
2233                 $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
2234                 next;
2235             }
2236             if ($state eq "DBCSRANGE")
2237             {
2238                 my $start = hex $1;
2239                 my $end = hex $2;
2240                 for (my $i = $start; $i <= $end; $i++) { add_lead_byte( $i ); }
2241                 $lb_cur = $start;
2242                 $lb_end = $end;
2243                 next;
2244             }
2245             if ($state eq "DBCSTABLE")
2246             {
2247                 my $mb = hex $1;
2248                 my $uni = hex $2;
2249                 my $cp = ($lb_cur << 8) | $mb;
2250                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
2251                 if (!--$count)
2252                 {
2253                     if (++$lb_cur > $lb_end) { $state = "DBCSRANGE"; }
2254                 }
2255                 next;
2256             }
2257         }
2258         die "$filename: Unrecognized line $_\n";
2259     }
2260     close $INPUT;
2261 }
2262
2263
2264 ################################################################
2265 # read an input file and generate the corresponding .c file
2266 sub HANDLE_FILE(@)
2267 {
2268     my ($codepage,$filename,$has_glyphs,$comment,$first_private,$def,$defw) = @_;
2269
2270     @cp2uni = ();
2271     @lead_bytes = ();
2272     @uni2cp = ();
2273     $default_char = $def || $DEF_CHAR;
2274     $default_wchar = $defw || $DEF_CHAR;
2275
2276     # some codepage files are special
2277     if ($codepage == 20932)
2278     {
2279         READ_JIS0208_FILE( $filename );
2280         add_default_mappings( $first_private );
2281     }
2282     elsif ($codepage == 20127)
2283     {
2284         fill_20127_codepage();
2285         add_default_mappings( $first_private );
2286     }
2287     elsif ($filename =~ /\/bestfit/)
2288     {
2289         read_bestfit_file( $filename );
2290     }
2291     else
2292     {
2293         read_codepage_file( $filename );
2294         if ($codepage == 10001)
2295         {
2296             # add Shift-JIS mappings
2297             read_bestfit_file( "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt" );
2298         }
2299         else
2300         {
2301             add_default_mappings( $first_private );
2302         }
2303     }
2304
2305     output_codepage_file( $codepage, $filename, $comment, $has_glyphs );
2306 }
2307
2308
2309 ################################################################
2310 # save a file if modified
2311 sub save_file($)
2312 {
2313     my $file = shift;
2314     if (-f $file && !system "cmp $file $file.new >/dev/null")
2315     {
2316         unlink "$file.new";
2317     }
2318     else
2319     {
2320         rename "$file.new", "$file";
2321     }
2322 }
2323
2324
2325 ################################################################
2326 # output the list of codepage tables into the cptable.c file
2327 sub output_cptable($)
2328 {
2329     my $output = shift;
2330     my @tables_decl = ();
2331
2332     printf "Building %s\n", $output;
2333
2334     foreach my $file (@allfiles)
2335     {
2336         my ($codepage,$filename) = @$file;
2337         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
2338     }
2339
2340     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
2341     foreach my $file (@allfiles)
2342     {
2343         my ($codepage,$filename) = @$file;
2344         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
2345     }
2346     push @tables_decl, "};";
2347     REPLACE_IN_FILE( $output, @tables_decl );
2348 }
2349
2350 ################################################################
2351 # replace the contents of a file between ### cpmap ### marks
2352
2353 sub REPLACE_IN_FILE($@)
2354 {
2355     my $name = shift;
2356     my @data = @_;
2357     my @lines = ();
2358     open(FILE,$name) or die "Can't open $name";
2359     while (<FILE>)
2360     {
2361         push @lines, $_;
2362         last if /\#\#\# cpmap begin \#\#\#/;
2363     }
2364     push @lines, @data;
2365     while (<FILE>)
2366     {
2367         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
2368     }
2369     push @lines, <FILE>;
2370     open(FILE,">$name.new") or die "Can't modify $name";
2371     print FILE @lines;
2372     close(FILE);
2373     save_file($name);
2374 }
2375
2376 ################################################################
2377 # main routine
2378
2379 chdir ".." if -f "./make_unicode";
2380 READ_DEFAULTS( $DEFAULTS );
2381 DUMP_CASE_MAPPINGS( "libs/wine/casemap.c" );
2382 DUMP_SORTKEYS( "libs/wine/collation.c", READ_SORTKEYS_FILE() );
2383 DUMP_COMPOSE_TABLES( "libs/wine/compose.c" );
2384 DUMP_CTYPE_TABLES( "libs/wine/wctype.c" );
2385 dump_mirroring( "dlls/usp10/mirror.c" );
2386 dump_mirroring( "dlls/dwrite/mirror.c" );
2387 dump_bracket( "dlls/usp10/bracket.c" );
2388 dump_bracket( "dlls/dwrite/bracket.c" );
2389 dump_shaping( "dlls/usp10/shaping.c" );
2390 dump_linebreak( "dlls/usp10/linebreak.c" );
2391 dump_linebreak( "dlls/dwrite/linebreak.c" );
2392 dump_scripts( "dlls/dwrite/scripts" );
2393 dump_indic( "dlls/usp10/indicsyllable.c" );
2394 dump_intl_nls("loader/l_intl.nls");
2395 dump_vertical( "dlls/gdi32/vertical.c" );
2396 dump_vertical( "dlls/wineps.drv/vertical.c" );
2397 dump_nameprep( "dlls/kernel32/nameprep.c" );
2398
2399 foreach my $file (@allfiles) { HANDLE_FILE( @{$file} ); }
2400
2401 output_cptable("libs/wine/cptable.c");
2402
2403 exit 0;
2404
2405 # Local Variables:
2406 # compile-command: "./make_unicode"
2407 # End: