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