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/12.1.0/ucd/UCD.zip";
  27 my $REPORTS = "http://www.unicode.org/reports";
  28 my $RFCS = "http://www.rfc-editor.org/rfc";
  29 my $MSDATA = "https://download.microsoft.com/download/C/F/7/CF713A5E-9FBC-4FD6-9246-275F65C0E498";
  30 my $MSCODEPAGES = "$MSDATA/Windows Supported Code Page Data Files.zip";
  31
  32 # Sort keys file
  33 my $SORTKEYS = "tr10/allkeys.txt";
  34
  35 # RFC3454 (stringprep data)
  36 my $STRINGPREP = "rfc3454.txt";
  37
  38 # Default char for undefined mappings
  39 my $DEF_CHAR = ord '?';
  40
  41 my @allfiles =
  42 (
  43     [ 37,    "CodpageFiles/037.txt",    "IBM EBCDIC US Canada" ],
  44     [ 437,   "CodpageFiles/437.txt",    "OEM United States" ],
  45     [ 500,   "CodpageFiles/500.txt",    "IBM EBCDIC International" ],
  46     [ 737,   "CodpageFiles/737.txt",    "OEM Greek 437G" ],
  47     [ 775,   "CodpageFiles/775.txt",    "OEM Baltic" ],
  48     [ 850,   "CodpageFiles/850.txt",    "OEM Multilingual Latin 1" ],
  49     [ 852,   "CodpageFiles/852.txt",    "OEM Slovak Latin 2" ],
  50     [ 855,   "CodpageFiles/855.txt",    "OEM Cyrillic" ],
  51     [ 857,   "CodpageFiles/857.txt",    "OEM Turkish" ],
  52     [ 860,   "CodpageFiles/860.txt",    "OEM Portuguese" ],
  53     [ 861,   "CodpageFiles/861.txt",    "OEM Icelandic" ],
  54     [ 862,   "CodpageFiles/862.txt",    "OEM Hebrew" ],
  55     [ 863,   "CodpageFiles/863.txt",    "OEM Canadian French" ],
  56     [ 864,   "CodpageFiles/864.txt",    "OEM Arabic" ],
  57     [ 865,   "CodpageFiles/865.txt",    "OEM Nordic" ],
  58     [ 866,   "CodpageFiles/866.txt",    "OEM Russian" ],
  59     [ 869,   "CodpageFiles/869.txt",    "OEM Greek" ],
  60     [ 874,   "CodpageFiles/874.txt",    "ANSI/OEM Thai" ],
  61     [ 875,   "CodpageFiles/875.txt",    "IBM EBCDIC Greek" ],
  62     [ 932,   "CodpageFiles/932.txt",    "ANSI/OEM Japanese Shift-JIS" ],
  63     [ 936,   "CodpageFiles/936.txt",    "ANSI/OEM Simplified Chinese GBK" ],
  64     [ 949,   "CodpageFiles/949.txt",    "ANSI/OEM Korean Unified Hangul" ],
  65     [ 950,   "CodpageFiles/950.txt",    "ANSI/OEM Traditional Chinese Big5" ],
  66     [ 1026,  "CodpageFiles/1026.txt",   "IBM EBCDIC Latin 5 Turkish" ],
  67     [ 1250,  "CodpageFiles/1250.txt",   "ANSI Eastern Europe" ],
  68     [ 1251,  "CodpageFiles/1251.txt",   "ANSI Cyrillic" ],
  69     [ 1252,  "CodpageFiles/1252.txt",   "ANSI Latin 1" ],
  70     [ 1253,  "CodpageFiles/1253.txt",   "ANSI Greek" ],
  71     [ 1254,  "CodpageFiles/1254.txt",   "ANSI Turkish" ],
  72     [ 1255,  "CodpageFiles/1255.txt",   "ANSI Hebrew" ],
  73     [ 1256,  "CodpageFiles/1256.txt",   "ANSI Arabic" ],
  74     [ 1257,  "CodpageFiles/1257.txt",   "ANSI Baltic" ],
  75     [ 1258,  "CodpageFiles/1258.txt",   "ANSI/OEM Viet Nam" ],
  76     [ 1361,  "CodpageFiles/1361.txt",   "Korean Johab" ],
  77     [ 10000, "CodpageFiles/10000.txt",  "Mac Roman" ],
  78     [ 10001, "CodpageFiles/10001.txt",  "Mac Japanese" ],
  79     [ 10002, "CodpageFiles/10002.txt",  "Mac Traditional Chinese" ],
  80     [ 10003, "CodpageFiles/10003.txt",  "Mac Korean" ],
  81     [ 10004, "CodpageFiles/10004.txt",  "Mac Arabic" ],
  82     [ 10005, "CodpageFiles/10005.txt",  "Mac Hebrew" ],
  83     [ 10006, "CodpageFiles/10006.txt",  "Mac Greek" ],
  84     [ 10007, "CodpageFiles/10007.txt",  "Mac Cyrillic" ],
  85     [ 10008, "CodpageFiles/10008.txt",  "Mac Simplified Chinese" ],
  86     [ 10010, "CodpageFiles/10010.txt",  "Mac Romanian" ],
  87     [ 10017, "CodpageFiles/10017.txt",  "Mac Ukrainian" ],
  88     [ 10021, "CodpageFiles/10021.txt",  "Mac Thai" ],
  89     [ 10029, "CodpageFiles/10029.txt",  "Mac Latin 2" ],
  90     [ 10079, "CodpageFiles/10079.txt",  "Mac Icelandic" ],
  91     [ 10081, "CodpageFiles/10081.txt",  "Mac Turkish" ],
  92     [ 10082, "CodpageFiles/10082.txt",  "Mac Croatian" ],
  93     [ 20127, "CodpageFiles/20127.txt",  "US-ASCII (7bit)" ],
  94     [ 20866, "CodpageFiles/20866.txt",  "Russian KOI8" ],
  95     [ 20932, "OBSOLETE/EASTASIA/JIS",   "EUC-JP" ],
  96     [ 21866, "CodpageFiles/21866.txt",  "Ukrainian KOI8" ],
  97     [ 28591, "CodpageFiles/28591.txt",  "ISO 8859-1 Latin 1" ],
  98     [ 28592, "CodpageFiles/28592.txt",  "ISO 8859-2 Latin 2 (East European)" ],
  99     [ 28593, "CodpageFiles/28593.txt",  "ISO 8859-3 Latin 3 (South European)" ],
 100     [ 28594, "CodpageFiles/28594.txt",  "ISO 8859-4 Latin 4 (Baltic old)" ],
 101     [ 28595, "CodpageFiles/28595.txt",  "ISO 8859-5 Cyrillic" ],
 102     [ 28596, "CodpageFiles/28596.txt",  "ISO 8859-6 Arabic" ],
 103     [ 28597, "CodpageFiles/28597.txt",  "ISO 8859-7 Greek" ],
 104     [ 28598, "CodpageFiles/28598.txt",  "ISO 8859-8 Hebrew" ],
 105     [ 28599, "CodpageFiles/28599.txt",  "ISO 8859-9 Latin 5 (Turkish)" ],
 106     [ 28603, "CodpageFiles/28603.txt",  "ISO 8859-13 Latin 7 (Baltic)" ],
 107     [ 28605, "CodpageFiles/28605.txt",  "ISO 8859-15 Latin 9 (Euro)" ],
 108 );
 109
 110
 111 my %ctype =
 112 (
 113     "upper"  => 0x0001,
 114     "lower"  => 0x0002,
 115     "digit"  => 0x0004,
 116     "space"  => 0x0008,
 117     "punct"  => 0x0010,
 118     "cntrl"  => 0x0020,
 119     "blank"  => 0x0040,
 120     "xdigit" => 0x0080,
 121     "alpha"  => 0x0100,
 122     "defin"  => 0x0200
 123 );
 124
 125 my %bracket_types =
 126 (
 127     "o" => 0x0000,
 128     "c" => 0x0001,
 129 );
 130
 131 my %indic_types =
 132 (
 133     "Other"    => 0x0000,
 134     "Bindu"    => 0x0001,
 135     "Visarga"  => 0x0002,
 136     "Avagraha" => 0x0003,
 137     "Nukta"    => 0x0004,
 138     "Virama"   => 0x0005,
 139     "Vowel_Independent"  => 0x0006,
 140     "Vowel_Dependent"  => 0x0007,
 141     "Vowel"  => 0x0008,
 142     "Consonant_Placeholder"  => 0x0009,
 143     "Consonant"  => 0x000a,
 144     "Consonant_Dead"  => 0x000b,
 145     "Consonant_Succeeding_Repha" => 0x000c,
 146     "Consonant_Subjoined"  => 0x000d,
 147     "Consonant_Medial"  => 0x000e,
 148     "Consonant_Final"  => 0x000f,
 149     "Consonant_Head_Letter"  => 0x0010,
 150     "Modifying_Letter"  => 0x0011,
 151     "Tone_Letter"  => 0x0012,
 152     "Tone_Mark"  => 0x0013,
 153     "Register_Shifter"  => 0x0014,
 154     "Consonant_Preceding_Repha" => 0x0015,
 155     "Pure_Killer" => 0x0016,
 156     "Invisible_Stacker" => 0x0017,
 157     "Gemination_Mark" => 0x0018,
 158     "Cantillation_Mark" => 0x0019,
 159     "Non_Joiner" => 0x001a,
 160     "Joiner" => 0x001b,
 161     "Number_Joiner" => 0x001c,
 162     "Number" => 0x001d,
 163     "Brahmi_Joining_Number" => 0x001e,
 164     "Consonant_With_Stacker" => 0x001f,
 165     "Consonant_Prefixed" => 0x0020,
 166     "Syllable_Modifier" => 0x0021,
 167     "Consonant_Killer" => 0x0022,
 168     "Consonant_Initial_Postfixed" => 0x0023,
 169 );
 170
 171 my %matra_types =
 172 (
 173     "Right"    => 0x01,
 174     "Left"  => 0x02,
 175     "Visual_Order_Left" => 0x03,
 176     "Left_And_Right"    => 0x04,
 177     "Top"   => 0x05,
 178     "Bottom"  => 0x06,
 179     "Top_And_Bottom"  => 0x07,
 180     "Top_And_Right"  => 0x08,
 181     "Top_And_Left"  => 0x09,
 182     "Top_And_Left_And_Right"  => 0x0a,
 183     "Bottom_And_Right"  => 0x0b,
 184     "Top_And_Bottom_And_Right"  => 0x0c,
 185     "Overstruck"  => 0x0d,
 186     "Invisible"  => 0x0e,
 187     "Bottom_And_Left"  => 0x0f
 188 );
 189
 190 my %nameprep_flags =
 191 (
 192     "unassigned" => 0x01,
 193     "prohibited" => 0x02,
 194     "bidi_ral"   => 0x04,
 195     "bidi_l"     => 0x08
 196 );
 197
 198 my %break_types =
 199 (
 200     "BK"  => 0x0001,
 201     "CR"  => 0x0002,
 202     "LF"  => 0x0003,
 203     "CM"  => 0x0004,
 204     "SG"  => 0x0005,
 205     "GL"  => 0x0006,
 206     "CB"  => 0x0007,
 207     "SP"  => 0x0008,
 208     "ZW"  => 0x0009,
 209     "NL"  => 0x000a,
 210     "WJ"  => 0x000b,
 211     "JL"  => 0x000c,
 212     "JV"  => 0x000d,
 213     "JT"  => 0x000e,
 214     "H2"  => 0x000f,
 215     "H3"  => 0x0010,
 216     "XX"  => 0x0011,
 217     "OP"  => 0x0012,
 218     "CL"  => 0x0013,
 219     "CP"  => 0x0014,
 220     "QU"  => 0x0015,
 221     "NS"  => 0x0016,
 222     "EX"  => 0x0017,
 223     "SY"  => 0x0018,
 224     "IS"  => 0x0019,
 225     "PR"  => 0x001a,
 226     "PO"  => 0x001b,
 227     "NU"  => 0x001c,
 228     "AL"  => 0x001d,
 229     "ID"  => 0x001e,
 230     "IN"  => 0x001f,
 231     "HY"  => 0x0020,
 232     "BB"  => 0x0021,
 233     "BA"  => 0x0022,
 234     "SA"  => 0x0023,
 235     "AI"  => 0x0024,
 236     "B2"  => 0x0025,
 237     "HL"  => 0x0026,
 238     "CJ"  => 0x0027,
 239     "RI"  => 0x0028,
 240     "EB"  => 0x0029,
 241     "EM"  => 0x002a,
 242     "ZWJ" => 0x002b,
 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 %c2_types =
 332 (
 333     "L"   => 1,    # C2_LEFTTORIGHT
 334     "R"   => 2,    # C2_RIGHTTOLEFT
 335     "AL"  => 2,    # C2_RIGHTTOLEFT
 336     "EN"  => 3,    # C2_EUROPENUMBER
 337     "ES"  => 4,    # C2_EUROPESEPARATOR
 338     "ET"  => 5,    # C2_EUROPETERMINATOR
 339     "AN"  => 6,    # C2_ARABICNUMBER
 340     "CS"  => 7,    # C2_COMMONSEPARATOR
 341     "NSM" => 0,    # C2_NOTAPPLICABLE
 342     "BN"  => 0,    # C2_NOTAPPLICABLE
 343     "B"   => 8,    # C2_BLOCKSEPARATOR
 344     "S"   => 9,    # C2_SEGMENTSEPARATOR
 345     "WS"  => 10,   # C2_WHITESPACE
 346     "ON"  => 11,   # C2_OTHERNEUTRAL
 347     "LRE" => 11,   # C2_OTHERNEUTRAL
 348     "LRO" => 11,   # C2_OTHERNEUTRAL
 349     "RLE" => 11,   # C2_OTHERNEUTRAL
 350     "RLO" => 11,   # C2_OTHERNEUTRAL
 351     "PDF" => 11,   # C2_OTHERNEUTRAL
 352     "LRI" => 11,   # C2_OTHERNEUTRAL
 353     "RLI" => 11,   # C2_OTHERNEUTRAL
 354     "FSI" => 11,   # C2_OTHERNEUTRAL
 355     "PDI" => 11    # C2_OTHERNEUTRAL
 356 );
 357
 358 my %bidi_types =
 359 (
 360     "ON"  => 0,    # Other Neutrals
 361     "L"   => 1,    # Left-to-Right
 362     "R"   => 2,    # Right-to-Left
 363     "AN"  => 3,    # Arabic Number
 364     "EN"  => 4,    # European Number
 365     "AL"  => 5,    # Right-to-Left Arabic
 366     "NSM" => 6,    # Non-Spacing Mark
 367     "CS"  => 7,    # Common Number Separator
 368     "ES"  => 8,    # European Number Separator
 369     "ET"  => 9,    # European Number Terminator
 370     "BN"  => 10,   # Boundary Neutral
 371     "S"   => 11,   # Segment Separator
 372     "WS"  => 12,   # Whitespace
 373     "B"   => 13,   # Paragraph Separator
 374     "RLO" => 14,   # Right-to-Left Override
 375     "RLE" => 15,   # Right-to-Left Embedding
 376     "LRO" => 16,   # Left-to-Right Override
 377     "LRE" => 17,   # Left-to-Right Embedding
 378     "PDF" => 18,   # Pop Directional Format
 379     "LRI" => 19,   # Left-to-Right Isolate
 380     "RLI" => 20,   # Right-to-Left Isolate
 381     "FSI" => 21,   # First Strong Isolate
 382     "PDI" => 22    # Pop Directional Isolate
 383 );
 384
 385 my %joining_types =
 386 (
 387    "U" => 0,    # Non_Joining
 388    "T" => 1,    # Transparent
 389    "R" => 2,    # Right_Joining
 390    "L" => 3,    # Left_Joining
 391    "D" => 4,    # Dual_Joining
 392    "C" => 5,    # Join_Causing
 393 );
 394
 395 my @cp2uni = ();
 396 my @glyph2uni = ();
 397 my @lead_bytes = ();
 398 my @uni2cp = ();
 399 my @tolower_table = ();
 400 my @toupper_table = ();
 401 my @digitmap_table = ();
 402 my @category_table = ();
 403 my @joining_table = ();
 404 my @direction_table = ();
 405 my @decomp_table = ();
 406 my @compose_table = ();
 407 my @combining_class_table = ();
 408 my @decomp_compat_table = ();
 409 my $default_char;
 410 my $default_wchar;
 411
 412 my %joining_forms =
 413 (
 414    "isolated" => [],
 415    "final" => [],
 416    "initial" => [],
 417    "medial" => []
 418 );
 419
 420 sub get_utf16($)
 421 {
 422     my $ch = shift;
 423     return $ch if ($ch < 0x10000);
 424     $ch -= 0x10000;
 425     return ( 0xd800 | ($ch >> 10), 0xdc00 | ($ch & 0x3ff) );
 426 }
 427
 428 ################################################################
 429 # fetch a unicode.org file and open it
 430 sub open_data_file($$)
 431 {
 432     my ($base, $name) = @_;
 433     (my $dir = "data/$name") =~ s/\/[^\/]+$//;
 434     local *FILE;
 435     unless (-f "data/$name")
 436     {
 437         system "mkdir", "-p", $dir;
 438         if ($base =~ /.*\/([^\/]+\.zip)$/)
 439         {
 440             my $zip = $1;
 441             unless (-f "data/$zip")
 442             {
 443                 print "Fetching $base...\n";
 444                 !system "wget", "-q", "-O", "data/$zip", "$base" or die "cannot fetch $base";
 445             }
 446             !system "unzip", "-q", "-d", "data", "data/$zip", $name or die "cannot extract $name from $zip";
 447         }
 448         else
 449         {
 450             print "Fetching $base/$name...\n";
 451             !system "wget", "-q", "-O", "data/$name", "$base/$name" or die "cannot fetch $base/$name";
 452         }
 453     }
 454     open FILE, "<data/$name" or die "cannot open data/$name";
 455     return *FILE;
 456 }
 457
 458 ################################################################
 459 # recursively get the decomposition for a character
 460 sub get_decomposition($$);
 461 sub get_decomposition($$)
 462 {
 463     my ($char, $table) = @_;
 464     my @ret;
 465
 466     return get_utf16($char) unless defined ${$table}[$char];
 467     foreach my $ch (@{${$table}[$char]})
 468     {
 469         push @ret, get_decomposition( $ch, $table );
 470     }
 471     return @ret;
 472 }
 473
 474 ################################################################
 475 # recursively build decompositions
 476 sub build_decompositions(@)
 477 {
 478     my @src = @_;
 479     my @dst;
 480
 481     for (my $i = 0; $i < 65536; $i++)
 482     {
 483         next unless defined $src[$i];
 484         my @decomp = get_decomposition( $i, \@src );
 485         $dst[$i] = \@decomp;
 486     }
 487     return @dst;
 488 }
 489
 490 ################################################################
 491 # read in the Unicode database files
 492 sub load_data()
 493 {
 494     my $start;
 495
 496     # now build mappings from the decomposition field of the Unicode database
 497
 498     my $UNICODE_DATA = open_data_file( $UNIDATA, "UnicodeData.txt" );
 499     while (<$UNICODE_DATA>)
 500     {
 501         # Decode the fields ...
 502         my ($code, $name, $cat, $comb, $bidi,
 503             $decomp, $dec, $dig, $num, $mirror,
 504             $oldname, $comment, $upper, $lower, $title) = split /;/;
 505         my $dst;
 506         my $src = hex $code;
 507
 508         die "unknown category $cat" unless defined $categories{$cat};
 509         die "unknown directionality $bidi" unless defined $directions{$bidi};
 510
 511         $category_table[$src] = $categories{$cat};
 512         $direction_table[$src] = $bidi;
 513         $joining_table[$src] = $joining_types{"T"} if $cat eq "Mn" || $cat eq "Me" || $cat eq "Cf";
 514
 515         if ($lower ne "")
 516         {
 517             $tolower_table[$src] = hex $lower;
 518         }
 519         if ($upper ne "")
 520         {
 521             $toupper_table[$src] = hex $upper;
 522         }
 523         if ($dec ne "")
 524         {
 525             $category_table[$src] |= $ctype{"digit"};
 526         }
 527         if ($dig ne "")
 528         {
 529             $digitmap_table[$src] = ord $dig;
 530         }
 531         if ($comb ne "")
 532         {
 533             $combining_class_table[$src] = $comb;
 534         }
 535
 536         # copy the category and direction for everything between First/Last pairs
 537         if ($name =~ /, First>/) { $start = $src; }
 538         if ($name =~ /, Last>/)
 539         {
 540             while ($start < $src)
 541             {
 542                 $category_table[$start] = $category_table[$src];
 543                 $direction_table[$start] = $direction_table[$src];
 544                 $start++;
 545             }
 546         }
 547
 548         next if $decomp eq "";  # no decomposition, skip it
 549
 550         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)/)
 551         {
 552             my @seq = map { hex $_; } (split /\s+/, (split /\s+/, $decomp, 2)[1]);
 553             $decomp_compat_table[$src] = \@seq;
 554         }
 555
 556         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 557         {
 558             # decomposition of the form "<foo> 1234" -> use char if type is known
 559             if ($1 eq "isolated" || $1 eq "final" || $1 eq "initial" || $1 eq "medial")
 560             {
 561                 ${joining_forms{$1}}[hex $2] = $src;
 562                 next;
 563             }
 564             next unless ($1 eq "font" ||
 565                          $1 eq "noBreak" ||
 566                          $1 eq "circle" ||
 567                          $1 eq "super" ||
 568                          $1 eq "sub" ||
 569                          $1 eq "wide" ||
 570                          $1 eq "narrow" ||
 571                          $1 eq "compat" ||
 572                          $1 eq "small");
 573             $dst = hex $2;
 574         }
 575         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 576         {
 577             # decomposition "<compat> 0020 1234" -> combining accent
 578             $dst = hex $1;
 579         }
 580         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 581         {
 582             # decomposition contains only char values without prefix -> use first char
 583             $dst = hex $1;
 584             $category_table[$src] |= $category_table[$dst] if defined $category_table[$dst];
 585             # store decomposition if it contains two chars
 586             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 587             {
 588                 $decomp_table[$src] = $decomp_compat_table[$src] = [ hex $1, hex $2 ];
 589                 push @compose_table, [ hex $1, hex $2, $src ];
 590             }
 591             elsif ($decomp =~ /^([0-9a-fA-F]+)$/)
 592             {
 593                 # Single char decomposition
 594                 $decomp_table[$src] = $decomp_compat_table[$src] = [ hex $1 ];
 595             }
 596         }
 597         else
 598         {
 599             next;
 600         }
 601     }
 602     close $UNICODE_DATA;
 603
 604     # patch the category of some special characters
 605
 606     foreach my $cat (keys %special_categories)
 607     {
 608         my $flag = $ctype{$cat};
 609         foreach my $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 610     }
 611
 612     # remove case mappings that don't round-trip
 613
 614     for (my $i = 0; $i < @toupper_table; $i++)
 615     {
 616         next unless defined $toupper_table[$i];
 617         my $ch = $toupper_table[$i];
 618         $toupper_table[$i] = undef unless defined $tolower_table[$ch] && $tolower_table[$ch] == $i;
 619     }
 620     for (my $i = 0; $i < @tolower_table; $i++)
 621     {
 622         next unless defined $tolower_table[$i];
 623         my $ch = $tolower_table[$i];
 624         $tolower_table[$i] = undef unless defined $toupper_table[$ch] && $toupper_table[$ch] == $i;
 625     }
 626 }
 627
 628
 629 ################################################################
 630 # define a new lead byte
 631 sub add_lead_byte($)
 632 {
 633     my $ch = shift;
 634     return if defined $cp2uni[$ch];
 635     push @lead_bytes, $ch;
 636     $cp2uni[$ch] = 0;
 637 }
 638
 639 ################################################################
 640 # define a new char mapping
 641 sub add_mapping($$)
 642 {
 643     my ($cp, $uni) = @_;
 644     $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 645     $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 646     if ($cp > 0xff) { add_lead_byte( $cp >> 8 ); }
 647 }
 648
 649 ################################################################
 650 # get a mapping including glyph chars for MB_USEGLYPHCHARS
 651 sub get_glyphs_mapping(@)
 652 {
 653     my @table = @_;
 654
 655     for (my $i = 0; $i < @glyph2uni; $i++)
 656     {
 657         $table[$i] = $glyph2uni[$i] if defined $glyph2uni[$i];
 658     }
 659     return @table;
 660 }
 661
 662 ################################################################
 663 # build EUC-JP table from the JIS 0208/0212 files
 664 sub output_eucjp_codepage($$)
 665 {
 666     my ($dir, $comment) = @_;
 667
 668     $default_char = $DEF_CHAR;
 669     $default_wchar = 0x30fb;
 670
 671     # ASCII chars
 672     foreach my $i (0x00 .. 0x7f) { add_mapping( $i, $i ); }
 673
 674     # lead bytes
 675     foreach my $i (0x8e, 0xa1 .. 0xfe) { add_lead_byte($i); }
 676
 677     # JIS X 0201 right plane
 678     foreach my $i (0xa1 .. 0xdf) { add_mapping( 0x8e00 + $i, 0xfec0 + $i ); }
 679
 680     # undefined chars
 681     foreach my $i (0x80 .. 0x8d, 0x8f .. 0x9f) { $cp2uni[$i] = $i; }
 682     $cp2uni[0xa0] = 0xf8f0;
 683     $cp2uni[0xff] = 0xf8f3;
 684
 685     # Fix backslash conversion
 686     add_mapping( 0xa1c0, 0xff3c );
 687
 688     # Add private mappings for rows undefined in JIS 0208/0212
 689     my $private = 0xe000;
 690     foreach my $hi (0xf5 .. 0xfe)
 691     {
 692         foreach my $lo (0xa1 .. 0xfe)
 693         {
 694             add_mapping( ($hi << 8) + $lo, $private++ );
 695         }
 696     }
 697     foreach my $hi (0xf5 .. 0xfe)
 698     {
 699         foreach my $lo (0x21 .. 0x7e)
 700         {
 701             add_mapping( ($hi << 8) + $lo, $private++ );
 702         }
 703     }
 704
 705     my $name = "$dir/JIS0208.TXT";
 706     my $INPUT = open_data_file( $MAPPINGS, $name );
 707     while (<$INPUT>)
 708     {
 709         next if /^\#/;  # skip comments
 710         next if /^$/;  # skip empty lines
 711         next if /\x1a/;  # skip ^Z
 712         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 713         {
 714             add_mapping( 0x8080 + hex $1, hex $2 );
 715             next;
 716         }
 717         die "$name: Unrecognized line $_\n";
 718     }
 719     close $INPUT;
 720
 721     $name = "$dir/JIS0212.TXT";
 722     $INPUT = open_data_file( $MAPPINGS, $name );
 723     while (<$INPUT>)
 724     {
 725         next if /^\#/;  # skip comments
 726         next if /^$/;  # skip empty lines
 727         next if /\x1a/;  # skip ^Z
 728         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 729         {
 730             add_mapping( 0x8000 + hex $1, hex $2 );
 731             next;
 732         }
 733         die "$name: Unrecognized line $_\n";
 734     }
 735     close $INPUT;
 736
 737     output_codepage_file( 20932, "$MAPPINGS/$dir", $comment );
 738 }
 739
 740
 741 ################################################################
 742 # build the sort keys table
 743 sub dump_sortkeys($)
 744 {
 745     my $filename = shift;
 746     my @sortkeys = ();
 747
 748     my $INPUT = open_data_file( $REPORTS, $SORTKEYS );
 749     while (<$INPUT>)
 750     {
 751         next if /^\#/;  # skip comments
 752         next if /^$/;  # skip empty lines
 753         next if /\x1a/;  # skip ^Z
 754         next if /^\@version/;  # skip @version header
 755         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 756         {
 757             my ($uni,$variable) = (hex $1, $2);
 758             next if $uni > 65535;
 759             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 760             next;
 761         }
 762         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 763         {
 764             # multiple character sequence, ignored for now
 765             next;
 766         }
 767         die "$SORTKEYS: Unrecognized line $_\n";
 768     }
 769     close $INPUT;
 770
 771     # compress the keys to 32 bit:
 772     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 773
 774     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 775                        ${$a}[2] <=> ${$b}[2] or
 776                        ${$a}[3] <=> ${$b}[3] or
 777                        ${$a}[4] <=> ${$b}[4] or
 778                        $a cmp $b; } @sortkeys;
 779
 780     my ($n2, $n3) = (1, 1);
 781     my @keys = (-1, -1, -1, -1, -1 );
 782     my @flatkeys = ();
 783
 784     for (my $i = 0; $i < @sortkeys; $i++)
 785     {
 786         next unless defined $sortkeys[$i];
 787         my @current = @{$sortkeys[$i]};
 788         if ($current[1] == $keys[1])
 789         {
 790             if ($current[2] == $keys[2])
 791             {
 792                 if ($current[3] == $keys[3])
 793                 {
 794                     # nothing
 795                 }
 796                 else
 797                 {
 798                     $keys[3] = $current[3];
 799                     $n3++;
 800                     die if ($n3 >= 16);
 801                 }
 802             }
 803             else
 804             {
 805                 $keys[2] = $current[2];
 806                 $keys[3] = $current[3];
 807                 $n2++;
 808                 $n3 = 1;
 809                 die if ($n2 >= 256);
 810             }
 811         }
 812         else
 813         {
 814             $keys[1] = $current[1];
 815             $keys[2] = $current[2];
 816             $keys[3] = $current[3];
 817             $n2 = 1;
 818             $n3 = 1;
 819         }
 820
 821         if ($current[2]) { $current[2] = $n2; }
 822         if ($current[3]) { $current[3] = $n3; }
 823         if ($current[4]) { $current[4] = 1; }
 824
 825         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 826     }
 827
 828     open OUTPUT,">$filename.new" or die "Cannot create $filename";
 829     printf "Building $filename\n";
 830     printf OUTPUT "/* Unicode collation element table */\n";
 831     printf OUTPUT "/* generated from %s */\n", "$REPORTS/$SORTKEYS";
 832     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 833     print OUTPUT "#include \"windef.h\"\n\n";
 834
 835     dump_two_level_mapping( "collation_table", 0xffffffff, 32, @flatkeys );
 836
 837     close OUTPUT;
 838     save_file($filename);
 839 }
 840
 841
 842 ################################################################
 843 # dump an array of integers
 844 sub dump_array($$@)
 845 {
 846     my ($bit_width, $default, @array) = @_;
 847     my $format = sprintf "0x%%0%ux", $bit_width / 4;
 848     my $i;
 849     my $ret = "    ";
 850     for ($i = 0; $i < $#array; $i++)
 851     {
 852         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 853         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 854     }
 855     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 856     return $ret;
 857 }
 858
 859
 860 ################################################################
 861 # dump a unicode->ascii mapping table
 862 sub dump_uni2cp_table($$)
 863 {
 864     my ($width, $def) = @_;
 865
 866     # count the number of unicode->ascii subtables that contain something
 867
 868     my @lowerbounds;
 869     my @upperbounds;
 870     my $index = 0;
 871     my @filled = (-1) x 256;
 872     for (my $i = 0; $i < 65536; $i++)
 873     {
 874         next unless defined $uni2cp[$i];
 875         if ($filled[$i >> 8] == -1)
 876         {
 877           $lowerbounds[$index] = $i & 0xff;
 878           $upperbounds[$index] = 0xff - $lowerbounds[$index];
 879           $filled[$i >> 8] = $index * 256;
 880           $index++;
 881         }
 882         else
 883         {
 884           $upperbounds[$index-1] = 0xff - ($i & 0xff);
 885         }
 886     }
 887
 888     # add defaults mapping at the end
 889     $filled[256] = $index * 256;
 890     $lowerbounds[$index] = 255;
 891     $upperbounds[$index] = 0;
 892     $index++;
 893
 894     # collapse blocks upwards if possible
 895     my $removed = 0;
 896     $index = 0;
 897     for (my $i = 0; $i < 257; $i++)
 898     {
 899         next if $filled[$i] == -1;
 900         if ($upperbounds[$index - 1] > $lowerbounds[$index])
 901         {
 902            $removed += $lowerbounds[$index];
 903         }
 904         else
 905         {
 906            $removed += $upperbounds[$index - 1];
 907            $lowerbounds[$index] = $upperbounds[$index - 1];
 908         }
 909         $filled[$i] -= $removed;
 910         $index++;
 911     }
 912
 913     # output all the subtables into a single array
 914
 915     printf OUTPUT "static const unsigned %s uni2cp_low[] =\n{\n", $width == 8 ? "char" : "short";
 916
 917     for (my $i = $index = 0; $i < 257; $i++)
 918     {
 919         next if $filled[$i] == -1;
 920         my $start = ($i << 8) + $lowerbounds[$index];
 921         my $end = ($i << 8) + 255;
 922         if ($i == 256)
 923         {
 924             print OUTPUT "    /* defaults */\n";
 925             printf OUTPUT "%s\n};\n\n", dump_array( $width, 0, ($def) x ($end - $start + 1) );
 926         }
 927         else
 928         {
 929             printf OUTPUT "    /* 0x%04x .. 0x%04x */\n", $start, $end;
 930             printf OUTPUT "%s,\n", dump_array( $width, $def, @uni2cp[$start .. $end] );
 931         }
 932         $index++;
 933     }
 934
 935     # output a table of the offsets of the subtables in the previous array
 936
 937     for (my $i = 0; $i < 256; $i++)
 938     {
 939         if ($filled[$i] == -1) { $filled[$i] = $filled[256]; }
 940     }
 941     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 942     printf OUTPUT "{\n%s\n};\n\n", dump_array( 16, 0, @filled[0..255] );
 943 }
 944
 945 ################################################################
 946 # dump an SBCS mapping table
 947 sub dump_sbcs_table($$$$)
 948 {
 949     my ($codepage, $name, $def, $defw) = @_;
 950     my $i;
 951
 952     # output the ascii->unicode table
 953
 954     if (@glyph2uni)
 955     {
 956         printf OUTPUT "static const WCHAR cp2uni[512] =\n";
 957         printf OUTPUT "{\n%s", dump_array( 16, $defw, @cp2uni[0 .. 255] );
 958         printf OUTPUT ",\n    /* glyphs */\n%s\n};\n\n",
 959                       dump_array( 16, $defw, get_glyphs_mapping(@cp2uni[0 .. 255]) );
 960     }
 961     else
 962     {
 963         printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 964         printf OUTPUT "{\n%s\n};\n\n", dump_array( 16, $defw, @cp2uni[0 .. 255] );
 965     }
 966
 967     dump_uni2cp_table( 8, $def );
 968
 969     # output the code page descriptor
 970
 971     printf OUTPUT "const struct sbcs_table DECLSPEC_HIDDEN cptable_%03d =\n{\n", $codepage;
 972     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 973                   $codepage, $def, $defw, $name;
 974     printf OUTPUT "    cp2uni,\n";
 975     if (@glyph2uni) { printf OUTPUT "    cp2uni + 256,\n"; }
 976     else { printf OUTPUT "    cp2uni,\n"; }
 977     printf OUTPUT "    uni2cp_low,\n";
 978     printf OUTPUT "    uni2cp_high\n};\n";
 979 }
 980
 981
 982 ################################################################
 983 # dump a DBCS mapping table
 984 sub dump_dbcs_table($$$$@)
 985 {
 986     my ($codepage, $name, $def, $defw, @lb_ranges) = @_;
 987
 988     # build a list of lead bytes that are actually used
 989
 990     my @lblist = ();
 991     LBLOOP: for (my $y = 0; $y <= $#lead_bytes; $y++)
 992     {
 993         $cp2uni[$lead_bytes[$y]] = $defw;
 994         my $base = $lead_bytes[$y] << 8;
 995         for (my $x = 0; $x < 256; $x++)
 996         {
 997             if (defined $cp2uni[$base+$x])
 998             {
 999                 push @lblist,$lead_bytes[$y];
1000                 next LBLOOP;
1001             }
1002         }
1003     }
1004     my $unused = ($#lead_bytes > $#lblist);
1005
1006     # output the ascii->unicode table for the single byte chars
1007
1008     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
1009     printf OUTPUT "{\n%s,\n", dump_array( 16, $defw, @cp2uni[0 .. 255] );
1010
1011     # output the default table for unused lead bytes
1012
1013     if ($unused)
1014     {
1015         printf OUTPUT "    /* unused lead bytes */\n";
1016         printf OUTPUT "%s,\n", dump_array( 16, 0, ($defw) x 256 );
1017     }
1018
1019     # output the ascii->unicode table for each DBCS lead byte
1020
1021     for (my $y = 0; $y <= $#lblist; $y++)
1022     {
1023         my $base = $lblist[$y] << 8;
1024         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
1025         printf OUTPUT "%s", dump_array( 16, $defw, @cp2uni[$base .. $base+255] );
1026         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
1027     }
1028
1029     # output the lead byte subtables offsets
1030
1031     my @offsets = (0) x 256;
1032     for (my $x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
1033     if ($unused)
1034     {
1035         # increment all lead bytes offset to take into account the unused table
1036         for (my $x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
1037     }
1038     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
1039     printf OUTPUT "{\n%s\n};\n\n", dump_array( 8, 0, @offsets );
1040
1041     dump_uni2cp_table( 16, $def );
1042
1043     # output the code page descriptor
1044
1045     printf OUTPUT "const struct dbcs_table DECLSPEC_HIDDEN cptable_%03d =\n{\n", $codepage;
1046     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
1047                   $codepage, $def, $defw, $name;
1048     printf OUTPUT "    cp2uni,\n";
1049     printf OUTPUT "    cp2uni_leadbytes,\n";
1050     printf OUTPUT "    uni2cp_low,\n";
1051     printf OUTPUT "    uni2cp_high,\n";
1052     printf OUTPUT "    {\n    %s\n    }\n", dump_array( 8, 0, @lb_ranges, 0, 0 );
1053     printf OUTPUT "};\n";
1054 }
1055
1056
1057 ################################################################
1058 # dump an SBCS mapping table in binary format
1059 sub dump_binary_sbcs_table($)
1060 {
1061     my $codepage = shift;
1062
1063     my @header = ( 13, $codepage, 1, $default_char, $default_wchar, $cp2uni[$default_char], $uni2cp[$default_wchar] );
1064     my $wc_offset = 256 + 3 + (@glyph2uni ? 256 : 0);
1065
1066     print OUTPUT pack "S<*", @header;
1067     print OUTPUT pack "C12", (0) x 12;
1068     print OUTPUT pack "S<*", $wc_offset, map { $_ || 0; } @cp2uni[0 .. 255];
1069
1070     if (@glyph2uni)
1071     {
1072         print OUTPUT pack "S<*", 256, get_glyphs_mapping(@cp2uni[0 .. 255]);
1073     }
1074     else
1075     {
1076         print OUTPUT pack "S<*", 0;
1077     }
1078
1079     print OUTPUT pack "S<*", 0, 0;
1080
1081     print OUTPUT pack "C*", map { defined $_ ? $_ : $default_char; } @uni2cp[0 .. 65535];
1082 }
1083
1084
1085 ################################################################
1086 # dump a DBCS mapping table in binary format
1087 sub dump_binary_dbcs_table($)
1088 {
1089     my $codepage = shift;
1090     my @lb_ranges = get_lb_ranges();
1091     my @header = ( 13, $codepage, 2, $default_char, $default_wchar, $cp2uni[$default_char], $uni2cp[$default_wchar] );
1092
1093     my @offsets = (0) x 256;
1094     my $pos = 0;
1095     foreach my $i (@lead_bytes)
1096     {
1097         $offsets[$i] = ($pos += 256);
1098         $cp2uni[$i] = 0;
1099     }
1100
1101     my $wc_offset = 256 + 3 + 256 * (1 + scalar @lead_bytes);
1102
1103     print OUTPUT pack "S<*", @header;
1104     print OUTPUT pack "C12", @lb_ranges, 0 x 12;
1105     print OUTPUT pack "S<*", $wc_offset, map { $_ || 0; } @cp2uni[0 .. 255];
1106     print OUTPUT pack "S<*", 0, scalar @lb_ranges / 2, @offsets;
1107
1108     foreach my $i (@lead_bytes)
1109     {
1110         my $base = $i << 8;
1111         print OUTPUT pack "S<*", map { defined $_ ? $_ : $default_wchar; } @cp2uni[$base .. $base + 255];
1112     }
1113
1114     print OUTPUT pack "S<", 4;
1115     print OUTPUT pack "S<*", map { defined $_ ? $_ : $default_char; } @uni2cp[0 .. 65535];
1116 }
1117
1118
1119 ################################################################
1120 # get the list of defined lead byte ranges
1121 sub get_lb_ranges()
1122 {
1123     my @list = ();
1124     my @ranges = ();
1125
1126     foreach my $i (@lead_bytes) { $list[$i] = 1; }
1127     my $on = 0;
1128     for (my $i = 0; $i < 256; $i++)
1129     {
1130         if ($on)
1131         {
1132             if (!defined $list[$i]) { push @ranges, $i-1; $on = 0; }
1133         }
1134         else
1135         {
1136             if ($list[$i]) { push @ranges, $i; $on = 1; }
1137         }
1138     }
1139     if ($on) { push @ranges, 0xff; }
1140     return @ranges;
1141 }
1142
1143 ################################################################
1144 # dump the Indic Syllabic Category table
1145 sub dump_indic($)
1146 {
1147     my $filename = shift;
1148     my @indic_table;
1149
1150     my $INPUT = open_data_file( $UNIDATA, "IndicSyllabicCategory.txt" );
1151     while (<$INPUT>)
1152     {
1153         next if /^\#/;  # skip comments
1154         next if /^\s*$/;  # skip empty lines
1155         next if /\x1a/;  # skip ^Z
1156         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*#/)
1157         {
1158             my $type = $2;
1159             die "unknown indic $type" unless defined $indic_types{$type};
1160             if (hex $1 < 65536)
1161             {
1162                 $indic_table[hex $1] = $indic_types{$type};
1163             }
1164             next;
1165         }
1166         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1167         {
1168             my $type = $3;
1169             die "unknown indic $type" unless defined $indic_types{$type};
1170             if (hex $1 < 65536 and hex $2 < 65536)
1171             {
1172                 foreach my $i (hex $1 .. hex $2)
1173                 {
1174                     $indic_table[$i] = $indic_types{$type};
1175                 }
1176             }
1177             next;
1178         }
1179         die "malformed line $_";
1180     }
1181     close $INPUT;
1182
1183     $INPUT = open_data_file( $UNIDATA, "IndicPositionalCategory.txt" );
1184     while (<$INPUT>)
1185     {
1186         next if /^\#/;  # skip comments
1187         next if /^\s*$/;  # skip empty lines
1188         next if /\x1a/;  # skip ^Z
1189         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z]+)\s*#/)
1190         {
1191             my $type = $2;
1192             die "unknown matra $type" unless defined $matra_types{$type};
1193             $indic_table[hex $1] += $matra_types{$type} << 8;
1194             next;
1195         }
1196         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1197         {
1198             my $type = $3;
1199             die "unknown matra $type" unless defined $matra_types{$type};
1200             foreach my $i (hex $1 .. hex $2)
1201             {
1202                 $indic_table[$i] += $matra_types{$type} << 8;
1203             }
1204             next;
1205         }
1206         die "malformed line $_";
1207     }
1208     close $INPUT;
1209
1210     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1211     print "Building $filename\n";
1212     print OUTPUT "/* Unicode Indic Syllabic Category */\n";
1213     print OUTPUT "/* generated from $UNIDATA:IndicSyllabicCategory.txt */\n";
1214     print OUTPUT "/*       and from $UNIDATA:IndicPositionalCategory.txt */\n";
1215     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1216     print OUTPUT "#include \"windef.h\"\n\n";
1217
1218     dump_two_level_mapping( "indic_syllabic_table", $indic_types{'Other'}, 16, @indic_table );
1219
1220     close OUTPUT;
1221     save_file($filename);
1222 }
1223
1224 ################################################################
1225 # dump the Line Break Properties table
1226 sub dump_linebreak($)
1227 {
1228     my $filename = shift;
1229     my @break_table;
1230     my $next_group = 0;
1231
1232     my $INPUT = open_data_file( $UNIDATA, "LineBreak.txt" );
1233     while (<$INPUT>)
1234     {
1235         next if /^\#/;  # skip comments
1236         next if /^\s*$/;  # skip empty lines
1237         next if /\x1a/;  # skip ^Z
1238         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z][0-9A-Z])+\s*/)
1239         {
1240             my $type = $2;
1241             die "unknown breaktype $type" unless defined $break_types{$type};
1242             $break_table[hex $1] = $break_types{$type};
1243             next;
1244         }
1245         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z][0-9A-Z])+\s*/)
1246         {
1247             my $type = $3;
1248             die "unknown breaktype $type" unless defined $break_types{$type};
1249             foreach my $i (hex $1 .. hex $2)
1250             {
1251                 $break_table[$i] = $break_types{$type};
1252             }
1253             next;
1254         }
1255         elsif (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1256         {
1257             my $type = $2;
1258             die "unknown breaktype $type" unless defined $break_types{$type};
1259             $break_table[hex $1] = $break_types{$type};
1260             next;
1261         }
1262         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1263         {
1264             my $type = $3;
1265             die "unknown breaktype $type" unless defined $break_types{$type};
1266             foreach my $i (hex $1 .. hex $2)
1267             {
1268                 $break_table[$i] = $break_types{$type};
1269             }
1270             next;
1271         }
1272         die "malformed line $_";
1273     }
1274     close $INPUT;
1275
1276     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1277     print "Building $filename\n";
1278     print OUTPUT "/* Unicode Line Break Properties */\n";
1279     print OUTPUT "/* generated from $UNIDATA:LineBreak.txt */\n";
1280     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1281     print OUTPUT "#include \"windef.h\"\n\n";
1282
1283     dump_two_level_mapping( "wine_linebreak_table", $break_types{'XX'}, 16, @break_table );
1284
1285     close OUTPUT;
1286     save_file($filename);
1287 }
1288
1289 my %scripts =
1290 (
1291     "Unknown"                => 0,
1292     "Common"                 => 1,
1293     "Inherited"              => 2,
1294     "Arabic"                 => 3,
1295     "Armenian"               => 4,
1296     "Avestan"                => 5,
1297     "Balinese"               => 6,
1298     "Bamum"                  => 7,
1299     "Batak"                  => 8,
1300     "Bengali"                => 9,
1301     "Bopomofo"               => 10,
1302     "Brahmi"                 => 11,
1303     "Braille"                => 12,
1304     "Buginese"               => 13,
1305     "Buhid"                  => 14,
1306     "Canadian_Aboriginal"    => 15,
1307     "Carian"                 => 16,
1308     "Cham"                   => 17,
1309     "Cherokee"               => 18,
1310     "Coptic"                 => 19,
1311     "Cuneiform"              => 20,
1312     "Cypriot"                => 21,
1313     "Cyrillic"               => 22,
1314     "Deseret"                => 23,
1315     "Devanagari"             => 24,
1316     "Egyptian_Hieroglyphs"   => 25,
1317     "Ethiopic"               => 26,
1318     "Georgian"               => 27,
1319     "Glagolitic"             => 28,
1320     "Gothic"                 => 29,
1321     "Greek"                  => 30,
1322     "Gujarati"               => 31,
1323     "Gurmukhi"               => 32,
1324     "Han"                    => 33,
1325     "Hangul"                 => 34,
1326     "Hanunoo"                => 35,
1327     "Hebrew"                 => 36,
1328     "Hiragana"               => 37,
1329     "Imperial_Aramaic"       => 38,
1330     "Inscriptional_Pahlavi"  => 39,
1331     "Inscriptional_Parthian" => 40,
1332     "Javanese"               => 41,
1333     "Kaithi"                 => 42,
1334     "Kannada"                => 43,
1335     "Katakana"               => 44,
1336     "Kayah_Li"               => 45,
1337     "Kharoshthi"             => 46,
1338     "Khmer"                  => 47,
1339     "Lao"                    => 48,
1340     "Latin"                  => 49,
1341     "Lepcha"                 => 50,
1342     "Limbu"                  => 51,
1343     "Linear_B"               => 52,
1344     "Lisu"                   => 53,
1345     "Lycian"                 => 54,
1346     "Lydian"                 => 55,
1347     "Malayalam"              => 56,
1348     "Mandaic"                => 57,
1349     "Meetei_Mayek"           => 58,
1350     "Mongolian"              => 59,
1351     "Myanmar"                => 60,
1352     "New_Tai_Lue"            => 61,
1353     "Nko"                    => 62,
1354     "Ogham"                  => 63,
1355     "Ol_Chiki"               => 64,
1356     "Old_Italic"             => 65,
1357     "Old_Persian"            => 66,
1358     "Old_South_Arabian"      => 67,
1359     "Old_Turkic"             => 68,
1360     "Oriya"                  => 69,
1361     "Osmanya"                => 70,
1362     "Phags_Pa"               => 71,
1363     "Phoenician"             => 72,
1364     "Rejang"                 => 73,
1365     "Runic"                  => 74,
1366     "Samaritan"              => 75,
1367     "Saurashtra"             => 76,
1368     "Shavian"                => 77,
1369     "Sinhala"                => 78,
1370     "Sundanese"              => 79,
1371     "Syloti_Nagri"           => 80,
1372     "Syriac"                 => 81,
1373     "Tagalog"                => 82,
1374     "Tagbanwa"               => 83,
1375     "Tai_Le"                 => 84,
1376     "Tai_Tham"               => 85,
1377     "Tai_Viet"               => 86,
1378     "Tamil"                  => 87,
1379     "Telugu"                 => 88,
1380     "Thaana"                 => 89,
1381     "Thai"                   => 90,
1382     "Tibetan"                => 91,
1383     "Tifinagh"               => 92,
1384     "Ugaritic"               => 93,
1385     "Vai"                    => 94,
1386     "Yi"                     => 95,
1387     # Win8/Win8.1
1388     "Chakma"                 => 96,
1389     "Meroitic_Cursive"       => 97,
1390     "Meroitic_Hieroglyphs"   => 98,
1391     "Miao"                   => 99,
1392     "Sharada"                => 100,
1393     "Sora_Sompeng"           => 101,
1394     "Takri"                  => 102,
1395     # Win10
1396     "Bassa_Vah"              => 103,
1397     "Caucasian_Albanian"     => 104,
1398     "Duployan"               => 105,
1399     "Elbasan"                => 106,
1400     "Grantha"                => 107,
1401     "Khojki"                 => 108,
1402     "Khudawadi"              => 109,
1403     "Linear_A"               => 110,
1404     "Mahajani"               => 111,
1405     "Manichaean"             => 112,
1406     "Mende_Kikakui"          => 113,
1407     "Modi"                   => 114,
1408     "Mro"                    => 115,
1409     "Nabataean"              => 116,
1410     "Old_North_Arabian"      => 117,
1411     "Old_Permic"             => 118,
1412     "Pahawh_Hmong"           => 119,
1413     "Palmyrene"              => 120,
1414     "Pau_Cin_Hau"            => 121,
1415     "Psalter_Pahlavi"        => 122,
1416     "Siddham"                => 123,
1417     "Tirhuta"                => 124,
1418     "Warang_Citi"            => 125,
1419     # Win10 RS1
1420     "Adlam"                  => 126,
1421     "Ahom"                   => 127,
1422     "Anatolian_Hieroglyphs"  => 128,
1423     "Bhaiksuki"              => 129,
1424     "Hatran"                 => 130,
1425     "Marchen"                => 131,
1426     "Multani"                => 132,
1427     "Newa"                   => 133,
1428     "Old_Hungarian"          => 134,
1429     "Osage"                  => 135,
1430     "SignWriting"            => 136,
1431     "Tangut"                 => 137,
1432     # Win10 RS4
1433     "Masaram_Gondi"          => 138,
1434     "Nushu"                  => 139,
1435     "Soyombo"                => 140,
1436     "Zanabazar_Square"       => 141,
1437     # Win10 1903
1438     "Dogra"                  => 142,
1439     "Gunjala_Gondi"          => 143,
1440     "Hanifi_Rohingya"        => 144,
1441     "Makasar"                => 145,
1442     "Medefaidrin"            => 146,
1443     "Old_Sogdian"            => 147,
1444     "Sogdian"                => 148,
1445 );
1446
1447 ################################################################
1448 # dump Script IDs table
1449 sub dump_scripts($)
1450 {
1451     my $filename = shift;
1452     my $header = $filename;
1453     my @scripts_table;
1454     my $script_index;
1455     my $i;
1456
1457     my $INPUT = open_data_file( $UNIDATA, "Scripts.txt" );
1458     # Fill the table
1459     # Unknown script id is always 0, so undefined scripts are automatically treated as such
1460     while (<$INPUT>)
1461     {
1462         my $type = "";
1463
1464         next if /^\#/;  # skip comments
1465         next if /^\s*$/;  # skip empty lines
1466         next if /\x1a/;  # skip ^Z
1467         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/)
1468         {
1469             $type = $2;
1470             if (defined $scripts{$type})
1471             {
1472                 $scripts_table[hex $1] = $scripts{$type};
1473             }
1474             next;
1475         }
1476         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/)
1477         {
1478             $type = $3;
1479             if (defined $scripts{$type})
1480             {
1481                 foreach my $i (hex $1 .. hex $2)
1482                 {
1483                     $scripts_table[$i] = $scripts{$type};
1484                 }
1485             }
1486             next;
1487         }
1488     }
1489
1490     close $INPUT;
1491
1492     $header = "$filename.h";
1493     open OUTPUT,">$header.new" or die "Cannot create $header";
1494     print "Building $header\n";
1495     print OUTPUT "/* Unicode Script IDs */\n";
1496     print OUTPUT "/* generated from $UNIDATA:Scripts.txt */\n";
1497     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1498
1499     print OUTPUT "enum unicode_script_id {\n";
1500     foreach my $script (sort { $scripts{$a} <=> $scripts{$b} } keys %scripts)
1501     {
1502         print OUTPUT "    Script_$script = $scripts{$script},\n";
1503     }
1504     print OUTPUT "    Script_LastId = ", (scalar keys %scripts) - 1, "\n";
1505     print OUTPUT "};\n";
1506
1507     close OUTPUT;
1508     save_file($header);
1509
1510     $filename = "$filename.c";
1511     open OUTPUT,">$filename.new" or die "Cannot create $header";
1512     print "Building $filename\n";
1513     print OUTPUT "/* Unicode Script IDs */\n";
1514     print OUTPUT "/* generated from $UNIDATA:Scripts.txt */\n";
1515     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1516     print OUTPUT "#include \"windef.h\"\n\n";
1517
1518     dump_two_level_mapping( "wine_scripts_table", 0, 16, @scripts_table );
1519     close OUTPUT;
1520     save_file($filename);
1521 }
1522
1523 ################################################################
1524 # dump the BiDi mirroring table
1525 sub dump_mirroring($)
1526 {
1527     my $filename = shift;
1528     my @mirror_table = ();
1529
1530     my $INPUT = open_data_file( $UNIDATA, "BidiMirroring.txt" );
1531     while (<$INPUT>)
1532     {
1533         next if /^\#/;  # skip comments
1534         next if /^$/;  # skip empty lines
1535         next if /\x1a/;  # skip ^Z
1536         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+)/)
1537         {
1538             $mirror_table[hex $1] = hex $2;
1539             next;
1540         }
1541         die "malformed line $_";
1542     }
1543     close $INPUT;
1544
1545     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1546     print "Building $filename\n";
1547     print OUTPUT "/* Unicode BiDi mirroring */\n";
1548     print OUTPUT "/* generated from $UNIDATA:BidiMirroring.txt */\n";
1549     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1550     print OUTPUT "#include \"windef.h\"\n\n";
1551     dump_two_level_mapping( "wine_mirror_map", 0, 16, @mirror_table );
1552     close OUTPUT;
1553     save_file($filename);
1554 }
1555
1556 ################################################################
1557 # dump the Bidi Brackets
1558 sub dump_bracket($)
1559 {
1560     my $filename = shift;
1561     my @bracket_table;
1562
1563     my $INPUT = open_data_file( $UNIDATA, "BidiBrackets.txt" );
1564     while (<$INPUT>)
1565     {
1566         next if /^\#/;  # skip comments
1567         next if /^\s*$/;  # skip empty lines
1568         next if /\x1a/;  # skip ^Z
1569         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+);\s*([con])/)
1570         {
1571             my $type = $3;
1572             die "unknown bracket $type" unless defined $bracket_types{$type};
1573             die "characters too distant $1 and $2" if abs(hex($2) - hex($1)) >= 128;
1574             $bracket_table[hex $1] = (hex($2) - hex($1)) % 255;
1575             $bracket_table[hex $1] += $bracket_types{$type} << 8;
1576             next;
1577         }
1578         die "malformed line $_";
1579     }
1580     close $INPUT;
1581
1582     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1583     print "Building $filename\n";
1584     print OUTPUT "/* Unicode Bidirectional Bracket table */\n";
1585     print OUTPUT "/* generated from $UNIDATA:BidiBrackets.txt */\n";
1586     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1587     print OUTPUT "#include \"windef.h\"\n\n";
1588
1589     dump_two_level_mapping( "bidi_bracket_table", 0, 16, @bracket_table );
1590
1591     close OUTPUT;
1592     save_file($filename);
1593 }
1594
1595 ################################################################
1596 # dump the Arabic shaping table
1597 sub dump_shaping($)
1598 {
1599     my $filename = shift;
1600     my %groups;
1601     my $next_group = 0;
1602
1603     $groups{"No_Joining_Group"} = $next_group++;
1604
1605     my $INPUT = open_data_file( $UNIDATA, "ArabicShaping.txt" );
1606     while (<$INPUT>)
1607     {
1608         next if /^\#/;  # skip comments
1609         next if /^\s*$/;  # skip empty lines
1610         next if /\x1a/;  # skip ^Z
1611         if (/^\s*([0-9a-fA-F]+)\s*;.*;\s*([RLDCUT])\s*;\s*(\w+)/)
1612         {
1613             my $type = $2;
1614             my $group = $3;
1615             $groups{$group} = $next_group++ unless defined $groups{$group};
1616             $joining_table[hex $1] = $joining_types{$type} | ($groups{$group} << 8);
1617             next;
1618         }
1619         die "malformed line $_";
1620     }
1621     close $INPUT;
1622
1623     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1624     print "Building $filename\n";
1625     print OUTPUT "/* Unicode Arabic shaping */\n";
1626     print OUTPUT "/* generated from $UNIDATA:ArabicShaping.txt */\n";
1627     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1628     print OUTPUT "#include \"windef.h\"\n\n";
1629
1630     dump_two_level_mapping( "wine_shaping_table", 0, 16, @joining_table );
1631
1632     print OUTPUT "\nconst unsigned short DECLSPEC_HIDDEN wine_shaping_forms[256][4] =\n{\n";
1633     for (my $i = 0x600; $i <= 0x6ff; $i++)
1634     {
1635         printf OUTPUT "    { 0x%04x, 0x%04x, 0x%04x, 0x%04x },\n",
1636             ${joining_forms{"isolated"}}[$i] || $i,
1637             ${joining_forms{"final"}}[$i] || $i,
1638             ${joining_forms{"initial"}}[$i] || $i,
1639             ${joining_forms{"medial"}}[$i] || $i;
1640     }
1641     print OUTPUT "};\n";
1642
1643     close OUTPUT;
1644     save_file($filename);
1645 }
1646
1647 ################################################################
1648 # dump the Vertical Orientation table
1649 sub dump_vertical($)
1650 {
1651     my $filename = shift;
1652     my @vertical_table;
1653
1654     my $INPUT = open_data_file( $UNIDATA, "VerticalOrientation.txt" );
1655     while (<$INPUT>)
1656     {
1657         next if /^\#/;  # skip comments
1658         next if /^\s*$/;  # skip empty lines
1659         next if /\x1a/;  # skip ^Z
1660         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/)
1661         {
1662             my $type = $2;
1663             die "unknown vertical $type" unless defined $vertical_types{$type};
1664             if (hex $1 < 65536)
1665             {
1666                 $vertical_table[hex $1] = $vertical_types{$type};
1667             }
1668             next;
1669         }
1670         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*/)
1671         {
1672             my $type = $3;
1673             die "unknown vertical $type" unless defined $vertical_types{$type};
1674             foreach my $i (hex $1 .. hex $2)
1675             {
1676                 $vertical_table[$i] = $vertical_types{$type};
1677             }
1678             next;
1679         }
1680         die "malformed line $_";
1681     }
1682     close $INPUT;
1683
1684     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1685     print "Building $filename\n";
1686     print OUTPUT "/* Unicode Vertical Orientation */\n";
1687     print OUTPUT "/* generated from $UNIDATA:VerticalOrientation.txt */\n";
1688     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1689     print OUTPUT "#include \"windef.h\"\n\n";
1690
1691     dump_two_level_mapping( "vertical_orientation_table", $vertical_types{'R'}, 16, @vertical_table );
1692
1693     close OUTPUT;
1694     save_file($filename);
1695 }
1696
1697 ################################################################
1698 # dump the digit folding tables
1699 sub dump_digit_folding($)
1700 {
1701     my ($filename) = shift;
1702     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1703     print "Building $filename\n";
1704     print OUTPUT "/* Unicode digit folding mappings */\n";
1705     print OUTPUT "/* generated from $UNIDATA:UnicodeData.txt */\n";
1706     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1707     print OUTPUT "#include \"windef.h\"\n\n";
1708
1709     dump_two_level_mapping( "wine_digitmap", 0, 16, @digitmap_table );
1710     close OUTPUT;
1711     save_file($filename);
1712 }
1713
1714
1715 ################################################################
1716 # dump the case mapping tables
1717 sub dump_case_mappings($)
1718 {
1719     my $filename = shift;
1720     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1721     print "Building $filename\n";
1722     print OUTPUT "/* Unicode case mappings */\n";
1723     print OUTPUT "/* generated from $UNIDATA:UnicodeData.txt */\n";
1724     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1725     print OUTPUT "#include \"windef.h\"\n\n";
1726
1727     dump_case_table( "wine_casemap_lower", @tolower_table );
1728     print OUTPUT "\n";
1729     dump_case_table( "wine_casemap_upper", @toupper_table );
1730     close OUTPUT;
1731     save_file($filename);
1732 }
1733
1734
1735 ################################################################
1736 # dump a case mapping table
1737 sub dump_case_table($@)
1738 {
1739     my ($name,@table) = @_;
1740
1741     # count the number of sub tables that contain something
1742     # also compute the low and upper populated bounds
1743
1744     my @lowerbounds = ( 0, 0 );
1745     my @upperbounds = ( 0, 255 );
1746     my $index = 0;
1747     my @filled = ();
1748     for (my $i = 0; $i < 65536; $i++)
1749     {
1750         next unless defined $table[$i];
1751         if (!defined $filled[$i >> 8])
1752         {
1753           $lowerbounds[$index] = $i & 0xff;
1754           $upperbounds[$index] = 0xff - $lowerbounds[$index];
1755           $filled[$i >> 8] = $index * 256 + 512;
1756           $index++;
1757         }
1758         else
1759         {
1760           $upperbounds[$index-1] = 0xff - ($i & 0xff);
1761         }
1762         $table[$i] = ($table[$i] - $i) & 0xffff;
1763     }
1764
1765     # Collapse blocks upwards if possible
1766     my $removed = 0;
1767     $index = 0;
1768     for (my $i = 0; $i < 256; $i++)
1769     {
1770         next unless defined $filled[$i];
1771         if ($upperbounds[$index - 1] > $lowerbounds[$index])
1772         {
1773            $removed = $removed + $lowerbounds[$index];
1774         }
1775         else
1776         {
1777            $removed = $removed + $upperbounds[$index - 1];
1778            $lowerbounds[$index] = $upperbounds[$index - 1];
1779         }
1780         $filled[$i] = $filled[$i] - $removed;
1781         $index++;
1782     }
1783
1784     # dump the table
1785
1786     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
1787     printf OUTPUT "{\n    /* index */\n";
1788     printf OUTPUT "%s,\n", dump_array( 16, 256, @filled );
1789     printf OUTPUT "    /* defaults */\n";
1790     printf OUTPUT "%s", dump_array( 16, 0, (0) x 256 );
1791     $index = 0;
1792     for (my $i = 0; $i < 256; $i++)
1793     {
1794         next unless $filled[$i];
1795         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
1796         printf OUTPUT "%s", dump_array( 16, 0, @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
1797         $index++;
1798     }
1799     printf OUTPUT "\n};\n";
1800 }
1801
1802 ################################################################
1803 # compress a mapping table by removing identical rows
1804 sub compress_array($$@)
1805 {
1806     my $rows = shift;
1807     my $def = shift;
1808     my @table = @_;
1809     my $len = @table / $rows;
1810     my @array;
1811     my $data = "";
1812
1813     # try to merge table rows
1814     for (my $row = 0; $row < $rows; $row++)
1815     {
1816         my $rowtxt = pack "U*", map { defined($_) ? $_ : $def; } @table[($row * $len)..(($row + 1) * $len - 1)];
1817         my $pos = index $data, $rowtxt;
1818         if ($pos == -1)
1819         {
1820             # check if the tail of the data can match the start of the new row
1821             my $first = substr( $rowtxt, 0, 1 );
1822             for (my $i = length($data) - 1; $i > 0; $i--)
1823             {
1824                 $pos = index( substr( $data, -$i ), $first );
1825                 last if $pos == -1;
1826                 $i -= $pos;
1827                 next unless substr( $data, -$i ) eq substr( $rowtxt, 0, $i );
1828                 substr( $data, -$i ) = "";
1829                 last;
1830             }
1831             $pos = length $data;
1832             $data .= $rowtxt;
1833         }
1834         $array[$row] = $rows + $pos;
1835     }
1836     return @array, unpack "U*", $data;
1837 }
1838
1839 ################################################################
1840 # dump a simple char -> 16-bit value mapping table
1841 sub dump_simple_mapping($$@)
1842 {
1843     my $name = shift;
1844     my $def = shift;
1845     my @array = compress_array( 256, $def, @_[0..65535] );
1846
1847     printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, $#array+1;
1848     printf OUTPUT "    /* offsets */\n%s,\n", dump_array( 16, 0, @array[0..255] );
1849     printf OUTPUT "    /* values */\n%s\n};\n", dump_array( 16, 0, @array[256..$#array] );
1850 }
1851
1852 ################################################################
1853 # dump a char -> 16-bit value mapping table using two-level tables
1854 sub dump_two_level_mapping($$@)
1855 {
1856     my $name = shift;
1857     my $def = shift;
1858     my $size = shift;
1859     my $type = $size == 16 ? "unsigned short" : "unsigned int";
1860     my @row_array = compress_array( 4096, $def, @_[0..65535] );
1861     my @array = compress_array( 256, 0, @row_array[0..4095] );
1862
1863     for (my $i = 256; $i < @array; $i++) { $array[$i] += @array - 4096; }
1864
1865     printf OUTPUT "const %s DECLSPEC_HIDDEN %s[%d] =\n{\n", $type, $name, @array + @row_array - 4096;
1866     printf OUTPUT "    /* level 1 offsets */\n%s,\n", dump_array( $size, 0, @array[0..255] );
1867     printf OUTPUT "    /* level 2 offsets */\n%s,\n", dump_array( $size, 0, @array[256..$#array] );
1868     printf OUTPUT "    /* values */\n%s\n};\n", dump_array( $size, 0, @row_array[4096..$#row_array] );
1869 }
1870
1871 ################################################################
1872 # dump a binary case mapping table in l_intl.nls format
1873 sub dump_binary_case_table(@)
1874 {
1875     my (@table) = @_;
1876     my $max_char = 0x10000;
1877     my $level1 = $max_char / 16;
1878     my $level2 = $level1 / 16;
1879
1880     my @difftable;
1881     for (my $i = 0; $i < @table; $i++)
1882     {
1883         next unless defined $table[$i];
1884         $difftable[$i] = ($table[$i] - $i) & 0xffff;
1885     }
1886
1887     my @row_array = compress_array( $level1, 0, @difftable[0..$max_char-1] );
1888     my @array = compress_array( $level2, 0, @row_array[0..$level1-1] );
1889     for (my $i = $level2; $i < @array; $i++) { $array[$i] += @array - $level1; }
1890     return @array, @row_array[$level1..$#row_array];
1891 }
1892
1893
1894 ################################################################
1895 # dump case mappings for l_intl.nls
1896 sub dump_intl_nls($)
1897 {
1898     my @upper = dump_binary_case_table( @toupper_table );
1899     my @lower = dump_binary_case_table( @tolower_table );
1900
1901     my $filename = shift;
1902     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1903     printf "Building $filename\n";
1904
1905     binmode OUTPUT;
1906     print OUTPUT pack "S<", 1;  # version
1907     print OUTPUT pack "S<*", 1 + scalar @upper, @upper;
1908     print OUTPUT pack "S<*", 1 + scalar @lower, @lower;
1909     close OUTPUT;
1910     save_file($filename);
1911 }
1912
1913
1914 sub load_nameprep_range_table($$$)
1915 {
1916     my ($INPUT, $val, $table_ref) = @_;
1917
1918     while (<$INPUT>)
1919     {
1920         if (/^\s*([0-9a-fA-F]+)-([0-9a-fA-F]+)/)
1921         {
1922             my $last = hex $2;
1923             $last = 65535 if($last >= 65536);
1924             foreach my $i (hex $1 .. $last)
1925             {
1926                 $table_ref->[$i] |= $val;
1927             }
1928             next;
1929         }
1930         elsif (/^\s*([0-9a-fA-F]+)/)
1931         {
1932             if (hex $1 < 65536)
1933             {
1934                 $table_ref->[hex $1] |= $val;
1935             }
1936             next;
1937         }
1938
1939         return if (/End\sTable/);
1940     }
1941 }
1942
1943 sub load_nameprep_map_table($$)
1944 {
1945     my ($INPUT, $table_ref) = @_;
1946
1947     while (<$INPUT>)
1948     {
1949         if (/^\s*([0-9a-fA-F]+);\s;/)
1950         {
1951             # special value for map to nothing
1952             $table_ref->[hex $1] = [0xffff, 0xffff, 0xffff];
1953             next;
1954         }
1955         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+);/)
1956         {
1957             $table_ref->[hex $1] = [hex $2, 0, 0];
1958             next;
1959         }
1960         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1961         {
1962             $table_ref->[hex $1] = [hex $2, hex $3, 0];
1963             next;
1964         }
1965         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1966         {
1967             $table_ref->[hex $1] = [hex $2, hex $3, hex $4];
1968             next;
1969         }
1970
1971         return if (/End\sTable/);
1972     }
1973 }
1974
1975 ################################################################
1976 # dump mapping table, prohibited characters set, unassigned
1977 # characters, bidirectional rules used by nameprep algorithm
1978 sub dump_nameprep($)
1979 {
1980     my $filename = shift;
1981     my @mapping_table = ();
1982     my @flags_table;
1983
1984     my $INPUT = open_data_file( $RFCS, $STRINGPREP );
1985     while (<$INPUT>)
1986     {
1987         next unless /Start\sTable/;
1988
1989         load_nameprep_range_table($INPUT, $nameprep_flags{"unassigned"}, \@flags_table) if (/A.1/);
1990         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.1.2/);
1991         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.2.2/);
1992         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.3/);
1993         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.4/);
1994         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.5/);
1995         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.6/);
1996         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.7/);
1997         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.8/);
1998         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.9/);
1999         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_ral"}, \@flags_table) if (/D.1/);
2000         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_l"}, \@flags_table) if (/D.2/);
2001
2002         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.1/);
2003         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.2/);
2004     }
2005     close $INPUT;
2006
2007     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2008     print "Building $filename\n";
2009     print OUTPUT "/* Nameprep algorithm related data */\n";
2010     print OUTPUT "/* generated from $RFCS/$STRINGPREP */\n";
2011     print OUTPUT "/* DO NOT EDIT!! */\n\n";
2012     print OUTPUT "#include \"windef.h\"\n\n";
2013
2014     dump_two_level_mapping( "nameprep_char_type", 0, 16, @flags_table );
2015
2016     ######### mapping table
2017     # first determine all the 16-char subsets that contain something
2018     my @filled = ();
2019     my $pos = 16*3;  # for the null subset
2020     for (my $i = 0; $i < 65536; $i++)
2021     {
2022         next unless defined $mapping_table[$i];
2023         $filled[$i >> 4] = $pos;
2024         $pos += 16*3;
2025         $i |= 15;
2026     }
2027     my $total = $pos;
2028
2029     # now count the 256-char subsets that contain something
2030     my @filled_idx = (256) x 256;
2031     $pos = 256 + 16;
2032     for (my $i = 0; $i < 4096; $i++)
2033     {
2034         next unless $filled[$i];
2035         $filled_idx[$i >> 4] = $pos;
2036         $pos += 16;
2037         $i |= 15;
2038     }
2039     my $null_offset = $pos;
2040     $total += $pos;
2041
2042     # add the index offsets to the subsets positions
2043     for (my $i = 0; $i < 4096; $i++)
2044     {
2045         next unless $filled[$i];
2046         $filled[$i] += $null_offset;
2047     }
2048
2049     # dump the main index
2050     printf OUTPUT "const unsigned short DECLSPEC_HIDDEN nameprep_mapping[%d] =\n", $total;
2051     printf OUTPUT "{\n    /* index */\n";
2052     printf OUTPUT "%s", dump_array( 16, 0, @filled_idx );
2053     printf OUTPUT ",\n    /* null sub-index */\n%s", dump_array( 16, 0, ($null_offset) x 16 );
2054
2055     # dump the second-level indexes
2056     for (my $i = 0; $i < 256; $i++)
2057     {
2058         next unless ($filled_idx[$i] > 256);
2059         my @table = @filled[($i<<4)..($i<<4)+15];
2060         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
2061         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
2062         printf OUTPUT "%s", dump_array( 16, 0, @table );
2063     }
2064
2065     # dump the 16-char subsets
2066     printf OUTPUT ",\n    /* null mapping */\n";
2067     printf OUTPUT "%s", dump_array( 16, 0, (0) x 48 );
2068
2069     for (my $i = 0; $i < 4096; $i++)
2070     {
2071         next unless $filled[$i];
2072         my @table = (0) x 48;
2073         for (my $j = 0; $j < 16; $j++)
2074         {
2075             if (defined $mapping_table[($i<<4) + $j])
2076             {
2077                 $table[3 * $j] = ${$mapping_table[($i << 4) + $j]}[0];
2078                 $table[3 * $j + 1] = ${$mapping_table[($i << 4) + $j]}[1];
2079                 $table[3 * $j + 2] = ${$mapping_table[($i << 4) + $j]}[2];
2080             }
2081         }
2082         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
2083         printf OUTPUT "%s", dump_array( 16, 0, @table );
2084     }
2085
2086     printf OUTPUT "\n};\n";
2087
2088     close OUTPUT;
2089     save_file($filename);
2090 }
2091
2092
2093 ################################################################
2094 # dump the GetStringTypeW table
2095 sub dump_string_type_table($)
2096 {
2097     my $filename = shift;
2098     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2099     printf "Building $filename\n";
2100     printf OUTPUT "/* Unicode wctype table */\n";
2101     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
2102     printf OUTPUT "#include \"windef.h\"\n\n";
2103
2104     my @table = @category_table;
2105
2106     # add the direction in the high 4 bits of the category
2107     for (my $i = 0; $i < 65536; $i++)
2108     {
2109         $table[$i] |= $c2_types{$direction_table[$i]} << 12 if defined $direction_table[$i];
2110     }
2111
2112     dump_two_level_mapping( "wctype_table", 0, 16, @table );
2113
2114     close OUTPUT;
2115     save_file($filename);
2116 }
2117
2118
2119 ################################################################
2120 # dump the bidi direction table
2121 sub dump_bidi_dir_table($)
2122 {
2123     my $filename = shift;
2124     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2125     printf "Building $filename\n";
2126     printf OUTPUT "/* Unicode BiDi direction table */\n";
2127     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
2128     printf OUTPUT "#include \"windef.h\"\n\n";
2129
2130     my @table;
2131
2132     for (my $i = 0; $i < 65536; $i++)
2133     {
2134         $table[$i] = $bidi_types{$direction_table[$i]} if defined $direction_table[$i];
2135     }
2136
2137     dump_two_level_mapping( "bidi_direction_table", $bidi_types{"L"}, 16, @table );
2138
2139     close OUTPUT;
2140     save_file($filename);
2141 }
2142
2143
2144 ################################################################
2145 # dump the ctype tables
2146 sub dump_ctype_tables($)
2147 {
2148     my $filename = shift;
2149     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2150     printf "Building $filename\n";
2151     printf OUTPUT "/* Unicode ctype tables */\n";
2152     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
2153     printf OUTPUT "#include \"windef.h\"\n\n";
2154
2155     my @table = @category_table;
2156
2157     # add the direction in the high 4 bits of the category
2158     for (my $i = 0; $i < 65536; $i++)
2159     {
2160         $table[$i] |= $directions{$direction_table[$i]} << 12 if defined $direction_table[$i];
2161     }
2162
2163     dump_simple_mapping( "wine_wctype_table", 0, @table );
2164
2165     close OUTPUT;
2166     save_file($filename);
2167 }
2168
2169
2170 ################################################################
2171 # dump the char composition table
2172 sub dump_compose_table($)
2173 {
2174     my $filename = shift;
2175
2176     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2177     print "Building $filename\n";
2178     print OUTPUT "/* Unicode char composition */\n";
2179     print OUTPUT "/* generated from $UNIDATA:UnicodeData.txt */\n";
2180     print OUTPUT "/* DO NOT EDIT!! */\n\n";
2181     print OUTPUT "#include \"windef.h\"\n\n";
2182
2183     my @filled = ();
2184     foreach my $i (@compose_table)
2185     {
2186         my @comp = @$i;
2187         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
2188     }
2189
2190     # count how many different second chars we have
2191
2192     my $count = 0;
2193     for (my $i = 0; $i < 65536; $i++)
2194     {
2195         next unless defined $filled[$i];
2196         $count++;
2197     }
2198
2199     # build the table of second chars and offsets
2200
2201     my $pos = $count + 1;
2202     my @table = ();
2203     for (my $i = 0; $i < 65536; $i++)
2204     {
2205         next unless defined $filled[$i];
2206         push @table, $i, $pos;
2207         $pos += @{$filled[$i]};
2208     }
2209     # terminator with last position
2210     push @table, 0, $pos;
2211     printf OUTPUT "static const WCHAR table[0x%x] =\n{\n", 2*$pos;
2212     printf OUTPUT "    /* second chars + offsets */\n%s", dump_array( 16, 0, @table );
2213
2214     # build the table of first chars and mappings
2215
2216     for (my $i = 0; $i < 65536; $i++)
2217     {
2218         next unless defined $filled[$i];
2219         my @table = ();
2220         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
2221         for (my $j = 0; $j <= $#list; $j++)
2222         {
2223             push @table, $list[$j][0], $list[$j][1];
2224         }
2225         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, dump_array( 16, 0, @table );
2226     }
2227     print OUTPUT "\n};\n\n";
2228     print OUTPUT <<"EOF";
2229 static inline int binary_search( WCHAR ch, int low, int high )
2230 {
2231     while (low <= high)
2232     {
2233         int pos = (low + high) / 2;
2234         if (table[2 * pos] < ch) low = pos + 1;
2235         else if (table[2 * pos] > ch) high = pos - 1;
2236         else return pos;
2237     }
2238     return -1;
2239 }
2240
2241 WCHAR DECLSPEC_HIDDEN wine_compose( const WCHAR *str )
2242 {
2243     int pos, idx = 1, start = 0, count = $count;
2244     for (;;)
2245     {
2246         if ((pos = binary_search( str[idx], start, count - 1 )) == -1) return 0;
2247         if (!idx--) return table[2 * pos + 1];
2248         start = table[2 * pos + 1];
2249         count = table[2 * pos + 3];
2250     }
2251 }
2252 EOF
2253     close OUTPUT;
2254     save_file($filename);
2255 }
2256
2257 ################################################################
2258 # dump a decomposition table
2259 sub dump_decompositions($@)
2260 {
2261     my ($name, @decomp) = @_;
2262
2263     # first determine all the 16-char subsets that contain something
2264
2265     my @filled = (0) x 4096;
2266     my $pos = 16;  # for the null subset
2267     my $data_total = 0;
2268     for (my $i = 0; $i < 65536; $i++)
2269     {
2270         next unless defined $decomp[$i];
2271         if ($filled[$i >> 4] == 0)
2272         {
2273             $filled[$i >> 4] = $pos;
2274             $pos += 16;
2275         }
2276         $data_total += @{$decomp[$i]};
2277     }
2278     my $total = $pos;
2279
2280     # now count the 256-char subsets that contain something
2281
2282     my @filled_idx = (256) x 256;
2283     $pos = 256 + 16;
2284     for (my $i = 0; $i < 4096; $i++)
2285     {
2286         next unless $filled[$i];
2287         $filled_idx[$i >> 4] = $pos;
2288         $pos += 16;
2289         $i |= 15;
2290     }
2291     my $null_offset = $pos;  # null mapping
2292     $total += $pos + 1; # add the offset sentinel
2293
2294     # add the index offsets to the subsets positions
2295
2296     for (my $i = 0; $i < 4096; $i++)
2297     {
2298         next unless $filled[$i];
2299         $filled[$i] += $null_offset;
2300     }
2301
2302     # dump the main index
2303
2304     printf OUTPUT "\nconst WCHAR DECLSPEC_HIDDEN %s[%d] =\n", $name, $total + $data_total;
2305     printf OUTPUT "{\n    /* index */\n";
2306     printf OUTPUT "%s", dump_array( 16, 0, @filled_idx );
2307     printf OUTPUT ",\n    /* null sub-index */\n%s", dump_array( 16, 0, ($null_offset) x 16 );
2308
2309     # dump the second-level indexes
2310
2311     for (my $i = 0; $i < 256; $i++)
2312     {
2313         next unless ($filled_idx[$i] > 256);
2314         my @table = @filled[($i<<4)..($i<<4)+15];
2315         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
2316         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
2317         printf OUTPUT "%s", dump_array( 16, 0, @table );
2318     }
2319
2320     # dump the 16-char offsets
2321
2322     printf OUTPUT ",\n    /* null offsets */\n";
2323     printf OUTPUT "%s", dump_array( 16, 0, ($total) x (16) );
2324
2325     $pos = $total;
2326
2327     my @data;
2328     for (my $i = 0; $i < 4096; $i++)
2329     {
2330         next unless $filled[$i];
2331         my @table = (0) x (16);
2332         for (my $j = 0; $j < 16; $j++)
2333         {
2334             $table[$j] = $pos;
2335             if (defined $decomp[($i<<4) + $j])
2336             {
2337                 $pos += $#{$decomp[($i<<4) + $j]} + 1;
2338                 push @data, @{$decomp[($i<<4) + $j]};
2339             }
2340         }
2341         printf OUTPUT ",\n    /* offsets 0x%03x0 .. 0x%03xf */\n", $i, $i;
2342         printf OUTPUT "%s", dump_array( 16, 0, @table );
2343     }
2344
2345     my @sentinel = $pos;
2346     printf OUTPUT ",\n    /* offset sentinel */\n";
2347     printf OUTPUT "%s", dump_array( 16, 0, @sentinel );
2348
2349     printf OUTPUT ",\n    /* data */\n";
2350     printf OUTPUT "%s", dump_array( 16, 0, @data );
2351
2352     printf OUTPUT "\n};\n";
2353 }
2354
2355 ################################################################
2356 # dump the char decomposition table
2357 sub dump_decompose_table($$)
2358 {
2359     my ($filename, $compat) = @_;
2360
2361     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2362     print "Building $filename\n";
2363     print OUTPUT "/* Unicode char composition */\n";
2364     print OUTPUT "/* generated from $UNIDATA:UnicodeData.txt */\n";
2365     print OUTPUT "/* DO NOT EDIT!! */\n\n";
2366     print OUTPUT "#include \"windef.h\"\n";
2367
2368     dump_decompositions( "nfd_table", build_decompositions( @decomp_table ));
2369     dump_decompositions( "nfkd_table", build_decompositions( @decomp_compat_table )) if $compat;
2370
2371     close OUTPUT;
2372     save_file($filename);
2373 }
2374
2375 ################################################################
2376 # dump the combining class table
2377 sub dump_combining_class($)
2378 {
2379     my $filename = shift;
2380
2381     open OUTPUT,">$filename.new" or die "Cannot create $filename";
2382     print "Building $filename\n";
2383     print OUTPUT "/* Unicode Combining Classes */\n";
2384     print OUTPUT "/* generated from $UNIDATA:UnicodeData.txt */\n";
2385     print OUTPUT "/* DO NOT EDIT!! */\n\n";
2386     print OUTPUT "#include \"windef.h\"\n\n";
2387
2388     dump_two_level_mapping( "combining_class_table", 0, 16, @combining_class_table );
2389     close OUTPUT;
2390     save_file($filename);
2391 }
2392
2393 ################################################################
2394 # output a codepage definition file from the global tables
2395 sub output_codepage_file($$$)
2396 {
2397     my ($codepage, $filename, $comment) = @_;
2398
2399     my $output = sprintf "libs/port/c_%03d.c", $codepage;
2400     open OUTPUT,">$output.new" or die "Cannot create $output";
2401
2402     printf "Building %s (%s)\n", $output, $comment;
2403
2404     # dump all tables
2405
2406     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
2407     print OUTPUT "/* generated from $filename */\n";
2408     print OUTPUT "/* DO NOT EDIT!! */\n\n";
2409     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
2410
2411     if (!@lead_bytes) { dump_sbcs_table( $codepage, $comment, $default_char, $default_wchar ); }
2412     else { dump_dbcs_table( $codepage, $comment, $default_char, $default_wchar, get_lb_ranges() ); }
2413     close OUTPUT;
2414     save_file($output);
2415
2416     # same thing in binary format
2417
2418     $output = sprintf "nls/c_%03d.nls", $codepage;
2419     open OUTPUT,">$output.new" or die "Cannot create $output";
2420
2421     printf "Building %s\n", $output;
2422     if (!@lead_bytes) { dump_binary_sbcs_table( $codepage ); }
2423     else { dump_binary_dbcs_table( $codepage ); }
2424
2425     close OUTPUT;
2426     save_file($output);
2427 }
2428
2429 ################################################################
2430 # output a codepage table from a Microsoft-style mapping file
2431 sub output_msdata_codepage($$)
2432 {
2433     my ($filename, $comment) = @_;
2434     my $state = "";
2435     my ($codepage, $width, $count);
2436     my ($lb_cur, $lb_end);
2437
2438     my $INPUT = open_data_file( $MSCODEPAGES, $filename ) or die "Cannot open $filename";
2439
2440     while (<$INPUT>)
2441     {
2442         next if /^;/;  # skip comments
2443         next if /^\s*$/;  # skip empty lines
2444         next if /\x1a/;  # skip ^Z
2445         last if /^ENDCODEPAGE/;
2446
2447         if (/^CODEPAGE\s+(\d+)/)
2448         {
2449             $codepage = $1;
2450             next;
2451         }
2452         if (/^CPINFO\s+(\d+)\s+0x([0-9a-fA-f]+)\s+0x([0-9a-fA-F]+)/)
2453         {
2454             $width = $1;
2455             $default_char = hex $2;
2456             $default_wchar = hex $3;
2457             next;
2458         }
2459         if (/^(MBTABLE|GLYPHTABLE|WCTABLE|DBCSRANGE|DBCSTABLE)\s+(\d+)/)
2460         {
2461             $state = $1;
2462             $count = $2;
2463             next;
2464         }
2465         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)/)
2466         {
2467             if ($state eq "MBTABLE")
2468             {
2469                 my $cp = hex $1;
2470                 my $uni = hex $2;
2471                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
2472                 next;
2473             }
2474             if ($state eq "GLYPHTABLE")
2475             {
2476                 my $cp = hex $1;
2477                 my $uni = hex $2;
2478                 $glyph2uni[$cp] = $uni unless defined($glyph2uni[$cp]);
2479                 next;
2480             }
2481             if ($state eq "WCTABLE")
2482             {
2483                 my $uni = hex $1;
2484                 my $cp = hex $2;
2485                 $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
2486                 next;
2487             }
2488             if ($state eq "DBCSRANGE")
2489             {
2490                 my $start = hex $1;
2491                 my $end = hex $2;
2492                 for (my $i = $start; $i <= $end; $i++) { add_lead_byte( $i ); }
2493                 $lb_cur = $start;
2494                 $lb_end = $end;
2495                 next;
2496             }
2497             if ($state eq "DBCSTABLE")
2498             {
2499                 my $mb = hex $1;
2500                 my $uni = hex $2;
2501                 my $cp = ($lb_cur << 8) | $mb;
2502                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
2503                 if (!--$count)
2504                 {
2505                     if (++$lb_cur > $lb_end) { $state = "DBCSRANGE"; }
2506                 }
2507                 next;
2508             }
2509         }
2510         die "$filename: Unrecognized line $_\n";
2511     }
2512     close $INPUT;
2513
2514     output_codepage_file( $codepage, "$MSCODEPAGES:$filename", $comment );
2515 }
2516
2517
2518 ################################################################
2519 # read an input file and generate the corresponding .c file
2520 sub handle_codepage_file(@)
2521 {
2522     my ($codepage, $filename, $comment) = @_;
2523
2524     @cp2uni = ();
2525     @glyph2uni = ();
2526     @lead_bytes = ();
2527     @uni2cp = ();
2528     $default_char = $DEF_CHAR;
2529     $default_wchar = $DEF_CHAR;
2530
2531     # some codepage files are special
2532     if ($codepage == 20932)
2533     {
2534         output_eucjp_codepage( $filename, $comment );
2535     }
2536     else
2537     {
2538         output_msdata_codepage( $filename, $comment );
2539     }
2540 }
2541
2542
2543 ################################################################
2544 # save a file if modified
2545 sub save_file($)
2546 {
2547     my $file = shift;
2548     if (-f $file && !system "cmp $file $file.new >/dev/null")
2549     {
2550         unlink "$file.new";
2551     }
2552     else
2553     {
2554         rename "$file.new", "$file";
2555     }
2556 }
2557
2558
2559 ################################################################
2560 # replace the contents of a file between ### cpmap ### marks
2561
2562 sub REPLACE_IN_FILE($@)
2563 {
2564     my $name = shift;
2565     my @data = @_;
2566     my @lines = ();
2567     open(FILE,$name) or die "Can't open $name";
2568     while (<FILE>)
2569     {
2570         push @lines, $_;
2571         last if /\#\#\# cpmap begin \#\#\#/;
2572     }
2573     push @lines, @data;
2574     while (<FILE>)
2575     {
2576         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
2577     }
2578     push @lines, <FILE>;
2579     open(FILE,">$name.new") or die "Can't modify $name";
2580     print FILE @lines;
2581     close(FILE);
2582     save_file($name);
2583 }
2584
2585 ################################################################
2586 # main routine
2587
2588 chdir ".." if -f "./make_unicode";
2589 load_data();
2590 dump_case_mappings( "libs/port/casemap.c" );
2591 dump_sortkeys( "dlls/kernelbase/collation.c" );
2592 dump_compose_table( "libs/port/compose.c" );
2593 dump_compose_table( "dlls/kernelbase/compose.c" );
2594 dump_decompose_table( "libs/port/decompose.c", 1 );
2595 dump_decompose_table( "dlls/kernelbase/decompose.c", 0 );
2596 dump_ctype_tables( "libs/port/wctype.c" );
2597 dump_bidi_dir_table( "dlls/gdi32/direction.c" );
2598 dump_bidi_dir_table( "dlls/usp10/direction.c" );
2599 dump_bidi_dir_table( "dlls/dwrite/direction.c" );
2600 dump_string_type_table( "dlls/kernelbase/wctype.c" );
2601 dump_digit_folding( "dlls/kernelbase/digitmap.c" );
2602 dump_combining_class( "dlls/ntdll/combclass.c" );
2603 dump_mirroring( "dlls/usp10/mirror.c" );
2604 dump_mirroring( "dlls/dwrite/mirror.c" );
2605 dump_bracket( "dlls/usp10/bracket.c" );
2606 dump_bracket( "dlls/dwrite/bracket.c" );
2607 dump_shaping( "dlls/usp10/shaping.c" );
2608 dump_linebreak( "dlls/usp10/linebreak.c" );
2609 dump_linebreak( "dlls/dwrite/linebreak.c" );
2610 dump_scripts( "dlls/dwrite/scripts" );
2611 dump_indic( "dlls/usp10/indicsyllable.c" );
2612 dump_vertical( "dlls/gdi32/vertical.c" );
2613 dump_vertical( "dlls/wineps.drv/vertical.c" );
2614 dump_nameprep( "dlls/kernel32/nameprep.c" );
2615 dump_intl_nls("nls/l_intl.nls");
2616
2617 foreach my $file (@allfiles) { handle_codepage_file( @{$file} ); }
2618
2619 exit 0;
2620
2621 # Local Variables:
2622 # compile-command: "./make_unicode"
2623 # End: