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/6.3.0/ucd";
  27 my $REPORTS = "http://www.unicode.org/reports";
  28 my $RFCS = "http://www.rfc-editor.org/rfc";
  29 my $VERTICALDATA = "http://www.unicode.org/Public/vertical/revision-11";
  30
  31 # Sort keys file
  32 my $SORTKEYS = "tr10/allkeys.txt";
  33
  34 # RFC3454 (stringprep data)
  35 my $STRINGPREP = "rfc3454.txt";
  36
  37 # Defaults mapping
  38 my $DEFAULTS = "tools/unicode-defaults";
  39
  40 # Default char for undefined mappings
  41 my $DEF_CHAR = ord '?';
  42
  43 my @allfiles =
  44 (
  45     [ 37,    "VENDORS/MICSFT/EBCDIC/CP037.TXT",   0, "IBM EBCDIC US Canada", 0, 0x6f ],
  46     [ 424,   "VENDORS/MISC/CP424.TXT",            0, "IBM EBCDIC Hebrew" ],
  47     [ 437,   "VENDORS/MICSFT/PC/CP437.TXT",       1, "OEM United States" ],
  48     [ 500,   "VENDORS/MICSFT/EBCDIC/CP500.TXT",   0, "IBM EBCDIC International", 0, 0x6f ],
  49     [ 737,   "VENDORS/MICSFT/PC/CP737.TXT",       1, "OEM Greek 437G" ],
  50     [ 775,   "VENDORS/MICSFT/PC/CP775.TXT",       1, "OEM Baltic" ],
  51     [ 850,   "VENDORS/MICSFT/PC/CP850.TXT",       1, "OEM Multilingual Latin 1" ],
  52     [ 852,   "VENDORS/MICSFT/PC/CP852.TXT",       1, "OEM Slovak Latin 2" ],
  53     [ 855,   "VENDORS/MICSFT/PC/CP855.TXT",       1, "OEM Cyrillic" ],
  54     [ 856,   "VENDORS/MISC/CP856.TXT",            0, "Hebrew PC" ],
  55     [ 857,   "VENDORS/MICSFT/PC/CP857.TXT",       1, "OEM Turkish", 0xf8bb ],
  56     [ 860,   "VENDORS/MICSFT/PC/CP860.TXT",       1, "OEM Portuguese" ],
  57     [ 861,   "VENDORS/MICSFT/PC/CP861.TXT",       1, "OEM Icelandic" ],
  58     [ 862,   "VENDORS/MICSFT/PC/CP862.TXT",       1, "OEM Hebrew" ],
  59     [ 863,   "VENDORS/MICSFT/PC/CP863.TXT",       1, "OEM Canadian French" ],
  60     [ 864,   "VENDORS/MICSFT/PC/CP864.TXT",       0, "OEM Arabic", 0xf8bb ],
  61     [ 865,   "VENDORS/MICSFT/PC/CP865.TXT",       1, "OEM Nordic" ],
  62     [ 866,   "VENDORS/MICSFT/PC/CP866.TXT",       1, "OEM Russian" ],
  63     [ 869,   "VENDORS/MICSFT/PC/CP869.TXT",       1, "OEM Greek" ],
  64     [ 874,   "VENDORS/MICSFT/WindowsBestFit/bestfit874.txt",  1, "ANSI/OEM Thai" ],
  65     [ 875,   "VENDORS/MICSFT/EBCDIC/CP875.TXT",               0, "IBM EBCDIC Greek", 0, 0x6f ],
  66     [ 878,   "VENDORS/MISC/KOI8-R.TXT",                       0, "Russian KOI8" ],
  67     [ 932,   "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt",  0, "ANSI/OEM Japanese Shift-JIS" ],
  68     [ 936,   "VENDORS/MICSFT/WindowsBestFit/bestfit936.txt",  0, "ANSI/OEM Simplified Chinese GBK" ],
  69     [ 949,   "VENDORS/MICSFT/WindowsBestFit/bestfit949.txt",  0, "ANSI/OEM Korean Unified Hangul" ],
  70     [ 950,   "VENDORS/MICSFT/WindowsBestFit/bestfit950.txt",  0, "ANSI/OEM Traditional Chinese Big5" ],
  71     [ 1006,  "VENDORS/MISC/CP1006.TXT",                       0, "IBM Arabic" ],
  72     [ 1026,  "VENDORS/MICSFT/EBCDIC/CP1026.TXT",              0, "IBM EBCDIC Latin 5 Turkish", 0, 0x6f ],
  73     [ 1250,  "VENDORS/MICSFT/WindowsBestFit/bestfit1250.txt", 0, "ANSI Eastern Europe" ],
  74     [ 1251,  "VENDORS/MICSFT/WindowsBestFit/bestfit1251.txt", 0, "ANSI Cyrillic" ],
  75     [ 1252,  "VENDORS/MICSFT/WindowsBestFit/bestfit1252.txt", 0, "ANSI Latin 1" ],
  76     [ 1253,  "VENDORS/MICSFT/WindowsBestFit/bestfit1253.txt", 0, "ANSI Greek" ],
  77     [ 1254,  "VENDORS/MICSFT/WindowsBestFit/bestfit1254.txt", 0, "ANSI Turkish" ],
  78     [ 1255,  "VENDORS/MICSFT/WindowsBestFit/bestfit1255.txt", 0, "ANSI Hebrew" ],
  79     [ 1256,  "VENDORS/MICSFT/WindowsBestFit/bestfit1256.txt", 0, "ANSI Arabic" ],
  80     [ 1257,  "VENDORS/MICSFT/WindowsBestFit/bestfit1257.txt", 0, "ANSI Baltic" ],
  81     [ 1258,  "VENDORS/MICSFT/WindowsBestFit/bestfit1258.txt", 0, "ANSI/OEM Viet Nam" ],
  82     [ 1361,  "VENDORS/MICSFT/WindowsBestFit/bestfit1361.txt", 0, "Korean Johab" ],
  83     [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT",      0, "Mac Roman" ],
  84     [ 10001, "VENDORS/APPLE/JAPANESE.TXT",        0, "Mac Japanese" ],
  85     [ 10002, "VENDORS/APPLE/CHINTRAD.TXT",        0, "Mac Traditional Chinese" ],
  86     [ 10003, "VENDORS/APPLE/KOREAN.TXT",          0, "Mac Korean", 0xf8e5 ],
  87     [ 10004, "VENDORS/APPLE/ARABIC.TXT",          0, "Mac Arabic", 0xf826 ],
  88     [ 10005, "VENDORS/APPLE/HEBREW.TXT",          0, "Mac Hebrew", 0xf7fa ],
  89     [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT",      0, "Mac Greek" ],
  90     [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT",   0, "Mac Cyrillic" ],
  91     [ 10008, "VENDORS/APPLE/CHINSIMP.TXT",        0, "Mac Simplified Chinese" ],
  92     [ 10010, "VENDORS/APPLE/ROMANIAN.TXT",        0, "Mac Romanian" ],
  93     [ 10017, "VENDORS/APPLE/CYRILLIC.TXT",        0, "Mac Ukrainian" ],
  94     [ 10021, "VENDORS/APPLE/THAI.TXT",            0, "Mac Thai" ],
  95     [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT",     0, "Mac Latin 2" ],
  96     [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT",    0, "Mac Icelandic" ],
  97     [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT",    0, "Mac Turkish", 0xf89f ],
  98     [ 10082, "VENDORS/APPLE/CROATIAN.TXT",        0, "Mac Croatian" ],
  99     [ 20127, undef,                               0, "US-ASCII (7bit)" ],
 100     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           0, "Russian KOI8" ],
 101     [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", 0, "EUC-JP", 0, 0x3f, 0x30fb  ],
 102     [ 21866, "VENDORS/MISC/KOI8-U.TXT",           0, "Ukrainian KOI8" ],
 103     [ 28591, "ISO8859/8859-1.TXT",                0, "ISO 8859-1 Latin 1" ],
 104     [ 28592, "ISO8859/8859-2.TXT",                0, "ISO 8859-2 Latin 2 (East European)" ],
 105     [ 28593, "ISO8859/8859-3.TXT",                0, "ISO 8859-3 Latin 3 (South European)", 0xf7f5 ],
 106     [ 28594, "ISO8859/8859-4.TXT",                0, "ISO 8859-4 Latin 4 (Baltic old)" ],
 107     [ 28595, "ISO8859/8859-5.TXT",                0, "ISO 8859-5 Cyrillic" ],
 108     [ 28596, "ISO8859/8859-6.TXT",                0, "ISO 8859-6 Arabic", 0xf7c8 ],
 109     [ 28597, "ISO8859/8859-7.TXT",                0, "ISO 8859-7 Greek", 0xf7c5 ],
 110     [ 28598, "ISO8859/8859-8.TXT",                0, "ISO 8859-8 Hebrew", 0xf79c ],
 111     [ 28599, "ISO8859/8859-9.TXT",                0, "ISO 8859-9 Latin 5 (Turkish)" ],
 112     [ 28600, "ISO8859/8859-10.TXT",               0, "ISO 8859-10 Latin 6 (Nordic)" ],
 113     [ 28603, "ISO8859/8859-13.TXT",               0, "ISO 8859-13 Latin 7 (Baltic)" ],
 114     [ 28604, "ISO8859/8859-14.TXT",               0, "ISO 8859-14 Latin 8 (Celtic)" ],
 115     [ 28605, "ISO8859/8859-15.TXT",               0, "ISO 8859-15 Latin 9 (Euro)" ],
 116     [ 28606, "ISO8859/8859-16.TXT",               0, "ISO 8859-16 Latin 10 (Balkan)" ]
 117 );
 118
 119
 120 my %ctype =
 121 (
 122     "upper"  => 0x0001,
 123     "lower"  => 0x0002,
 124     "digit"  => 0x0004,
 125     "space"  => 0x0008,
 126     "punct"  => 0x0010,
 127     "cntrl"  => 0x0020,
 128     "blank"  => 0x0040,
 129     "xdigit" => 0x0080,
 130     "alpha"  => 0x0100,
 131     "defin"  => 0x0200
 132 );
 133
 134 my %indic_types =
 135 (
 136     "Other"    => 0x0000,
 137     "Bindu"    => 0x0001,
 138     "Visarga"  => 0x0002,
 139     "Avagraha" => 0x0003,
 140     "Nukta"    => 0x0004,
 141     "Virama"   => 0x0005,
 142     "Vowel_Independent"  => 0x0006,
 143     "Vowel_Dependent"  => 0x0007,
 144     "Vowel"  => 0x0008,
 145     "Consonant_Placeholder"  => 0x0009,
 146     "Consonant"  => 0x000a,
 147     "Consonant_Dead"  => 0x000b,
 148     "Consonant_Repha"  => 0x000c,
 149     "Consonant_Subjoined"  => 0x000d,
 150     "Consonant_Medial"  => 0x000e,
 151     "Consonant_Final"  => 0x000f,
 152     "Consonant_Head_Letter"  => 0x0010,
 153     "Modifying_Letter"  => 0x0011,
 154     "Tone_Letter"  => 0x0012,
 155     "Tone_Mark"  => 0x0013,
 156     "Register_Shifter"  => 0x0014
 157 );
 158
 159 my %matra_types =
 160 (
 161     "Right"    => 0x01,
 162     "Left"  => 0x02,
 163     "Visual_Order_Left" => 0x03,
 164     "Left_And_Right"    => 0x04,
 165     "Top"   => 0x05,
 166     "Bottom"  => 0x06,
 167     "Top_And_Bottom"  => 0x07,
 168     "Top_And_Right"  => 0x08,
 169     "Top_And_Left"  => 0x09,
 170     "Top_And_Left_And_Right"  => 0x0a,
 171     "Bottom_And_Right"  => 0x0b,
 172     "Top_And_Bottom_And_Right"  => 0x0c,
 173     "Overstruck"  => 0x0d,
 174     "Invisible"  => 0x0e
 175 );
 176
 177 my %nameprep_flags =
 178 (
 179     "unassigned" => 0x01,
 180     "prohibited" => 0x02,
 181     "bidi_ral"   => 0x04,
 182     "bidi_l"     => 0x08
 183 );
 184
 185 my %break_types =
 186 (
 187     "BK"  => 0x0001,
 188     "CR"  => 0x0002,
 189     "LF"  => 0x0003,
 190     "CM"  => 0x0004,
 191     "SG"  => 0x0005,
 192     "GL"  => 0x0006,
 193     "CB"  => 0x0007,
 194     "SP"  => 0x0008,
 195     "ZW"  => 0x0009,
 196     "NL"  => 0x000a,
 197     "WJ"  => 0x000b,
 198     "JL"  => 0x000c,
 199     "JV"  => 0x000d,
 200     "JT"  => 0x000e,
 201     "H2"  => 0x000f,
 202     "H3"  => 0x0010,
 203     "XX"  => 0x0011,
 204     "OP"  => 0x0012,
 205     "CL"  => 0x0013,
 206     "CP"  => 0x0014,
 207     "QU"  => 0x0015,
 208     "NS"  => 0x0016,
 209     "EX"  => 0x0017,
 210     "SY"  => 0x0018,
 211     "IS"  => 0x0019,
 212     "PR"  => 0x001a,
 213     "PO"  => 0x001b,
 214     "NU"  => 0x001c,
 215     "AL"  => 0x001d,
 216     "ID"  => 0x001e,
 217     "IN"  => 0x001f,
 218     "HY"  => 0x0020,
 219     "BB"  => 0x0021,
 220     "BA"  => 0x0022,
 221     "SA"  => 0x0023,
 222     "AI"  => 0x0024,
 223     "B2"  => 0x0025,
 224     "HL"  => 0x0026,
 225     "CJ"  => 0x0027,
 226     "RI"  => 0x0028,
 227 );
 228
 229 my %vertical_types =
 230 (
 231     "R"  => 0x0000,
 232     "U"  => 0x0001,
 233     "Tr" => 0x0002,
 234     "Tu" => 0x0003,
 235 );
 236
 237 my %categories =
 238 (
 239     "Lu" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
 240     "Ll" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
 241     "Lt" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}|$ctype{"lower"},    # Letter, Titlecase
 242     "Mn" => $ctype{"defin"},                    # Mark, Non-Spacing
 243     "Mc" => $ctype{"defin"},                    # Mark, Spacing Combining
 244     "Me" => $ctype{"defin"},                    # Mark, Enclosing
 245     "Nd" => $ctype{"defin"}|$ctype{"digit"},    # Number, Decimal Digit
 246     "Nl" => $ctype{"defin"}|$ctype{"alpha"},    # Number, Letter
 247     "No" => $ctype{"defin"},                    # Number, Other
 248     "Zs" => $ctype{"defin"}|$ctype{"space"},    # Separator, Space
 249     "Zl" => $ctype{"defin"}|$ctype{"space"},    # Separator, Line
 250     "Zp" => $ctype{"defin"}|$ctype{"space"},    # Separator, Paragraph
 251     "Cc" => $ctype{"defin"}|$ctype{"cntrl"},    # Other, Control
 252     "Cf" => $ctype{"defin"}|$ctype{"cntrl"},    # Other, Format
 253     "Cs" => $ctype{"defin"},                    # Other, Surrogate
 254     "Co" => $ctype{"defin"},                    # Other, Private Use
 255     "Cn" => $ctype{"defin"},                    # Other, Not Assigned
 256     "Lm" => $ctype{"defin"}|$ctype{"alpha"},    # Letter, Modifier
 257     "Lo" => $ctype{"defin"}|$ctype{"alpha"},    # Letter, Other
 258     "Pc" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Connector
 259     "Pd" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Dash
 260     "Ps" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Open
 261     "Pe" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Close
 262     "Pi" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Initial quote
 263     "Pf" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Final quote
 264     "Po" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Other
 265     "Sm" => $ctype{"defin"},                    # Symbol, Math
 266     "Sc" => $ctype{"defin"},                    # Symbol, Currency
 267     "Sk" => $ctype{"defin"},                    # Symbol, Modifier
 268     "So" => $ctype{"defin"}                     # Symbol, Other
 269 );
 270
 271 # a few characters need additional categories that cannot be determined automatically
 272 my %special_categories =
 273 (
 274     "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
 275                   0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ],
 276     "space"  => [ 0x09..0x0d, 0x85 ],
 277     "blank"  => [ 0x09, 0x20, 0xa0, 0x3000, 0xfeff ],
 278     "cntrl"  => [ 0x070f, 0x200c, 0x200d,
 279                   0x200e, 0x200f, 0x202a, 0x202b, 0x202c, 0x202d, 0x202e,
 280                   0x206a, 0x206b, 0x206c, 0x206d, 0x206e, 0x206f, 0xfeff,
 281                   0xfff9, 0xfffa, 0xfffb ],
 282     "punct"  => [ 0x24, 0x2b, 0x3c..0x3e, 0x5e, 0x60, 0x7c, 0x7e, 0xa2..0xbe,
 283                   0xd7, 0xf7 ],
 284     "digit"  => [ 0xb2, 0xb3, 0xb9 ],
 285     "lower"  => [ 0x2071, 0x207f ]
 286 );
 287
 288 my %directions =
 289 (
 290     "L"   => 1,    # Left-to-Right
 291     "R"   => 2,    # Right-to-Left
 292     "AL"  => 12,   # Right-to-Left Arabic
 293     "EN"  => 3,    # European Number
 294     "ES"  => 4,    # European Number Separator
 295     "ET"  => 5,    # European Number Terminator
 296     "AN"  => 6,    # Arabic Number
 297     "CS"  => 7,    # Common Number Separator
 298     "NSM" => 13,   # Non-Spacing Mark
 299     "BN"  => 14,   # Boundary Neutral
 300     "B"   => 8,    # Paragraph Separator
 301     "S"   => 9,    # Segment Separator
 302     "WS"  => 10,   # Whitespace
 303     "ON"  => 11,   # Other Neutrals
 304     "LRE" => 15,   # Left-to-Right Embedding
 305     "LRO" => 15,   # Left-to-Right Override
 306     "RLE" => 15,   # Right-to-Left Embedding
 307     "RLO" => 15,   # Right-to-Left Override
 308     "PDF" => 15,   # Pop Directional Format
 309     "LRI" => 15,   # Left-to-Right Isolate
 310     "RLI" => 15,   # Right-to-Left Isolate
 311     "FSI" => 15,   # First Strong Isolate
 312     "PDI" => 15    # Pop Directional Isolate
 313 );
 314
 315 my %joining_types =
 316 (
 317    "U" => 0,    # Non_Joining
 318    "T" => 1,    # Transparent
 319    "R" => 2,    # Right_Joining
 320    "L" => 3,    # Left_Joining
 321    "D" => 4,    # Dual_Joining
 322    "C" => 5,    # Join_Causing
 323 );
 324
 325 my @cp2uni = ();
 326 my @lead_bytes = ();
 327 my @uni2cp = ();
 328 my @unicode_defaults = ();
 329 my @unicode_aliases = ();
 330 my @tolower_table = ();
 331 my @toupper_table = ();
 332 my @digitmap_table = ();
 333 my @compatmap_table = ();
 334 my @category_table = (0) x 65536;
 335 my @joining_table = (0) x 65536;
 336 my @direction_table = ();
 337 my @decomp_table = ();
 338 my @compose_table = ();
 339 my $default_char;
 340 my $default_wchar;
 341
 342 my %joining_forms =
 343 (
 344    "isolated" => [],
 345    "final" => [],
 346    "initial" => [],
 347    "medial" => []
 348 );
 349
 350 ################################################################
 351 # fetch a unicode.org file and open it
 352 sub open_data_file($$)
 353 {
 354     my ($base, $name) = @_;
 355     (my $dir = "data/$name") =~ s/\/[^\/]+$//;
 356     local *FILE;
 357     unless (-f "data/$name")
 358     {
 359         print "Fetching $base/$name...\n";
 360         system "mkdir", "-p", $dir;
 361         !system "wget", "-q", "-O", "data/$name", "$base/$name" or die "cannot fetch $base/$name";
 362     }
 363     open FILE, "<data/$name" or die "cannot open data/$name";
 364     return *FILE;
 365 }
 366
 367 ################################################################
 368 # read in the defaults file
 369 sub READ_DEFAULTS($)
 370 {
 371     my $filename = shift;
 372     my $start;
 373
 374     # first setup a few default mappings
 375
 376     open DEFAULTS, "$filename" or die "Cannot open $filename";
 377     print "Loading $filename\n";
 378     while (<DEFAULTS>)
 379     {
 380         next if /^\#/;  # skip comments
 381         next if /^$/;  # skip empty lines
 382         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 383         {
 384             my @src = map hex, split /,/,$1;
 385             my $dst = $4;
 386             my $comment = $5;
 387             if ($#src > 0) { push @unicode_aliases, \@src; }
 388             next if ($dst eq "none");
 389             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 390             foreach my $src (@src)
 391             {
 392                 die "Duplicate value" if defined($unicode_defaults[$src]);
 393                 $unicode_defaults[$src] = $dst;
 394             }
 395             next;
 396         }
 397         die "Unrecognized line $_\n";
 398     }
 399     close DEFAULTS;
 400
 401     # now build mappings from the decomposition field of the Unicode database
 402
 403     my $UNICODE_DATA = open_data_file( $UNIDATA, "UnicodeData.txt" );
 404     while (<$UNICODE_DATA>)
 405     {
 406         # Decode the fields ...
 407         my ($code, $name, $cat, $comb, $bidi,
 408             $decomp, $dec, $dig, $num, $mirror,
 409             $oldname, $comment, $upper, $lower, $title) = split /;/;
 410         my $dst;
 411         my $src = hex $code;
 412
 413         die "unknown category $cat" unless defined $categories{$cat};
 414         die "unknown directionality $bidi" unless defined $directions{$bidi};
 415
 416         $category_table[$src] = $categories{$cat};
 417         $direction_table[$src] = $directions{$bidi};
 418         $joining_table[$src] = $joining_types{"T"} if $cat eq "Mn" || $cat eq "Me" || $cat eq "Cf";
 419
 420         if ($lower ne "")
 421         {
 422             $tolower_table[$src] = hex $lower;
 423         }
 424         if ($upper ne "")
 425         {
 426             $toupper_table[$src] = hex $upper;
 427         }
 428         if ($dec ne "")
 429         {
 430             $category_table[$src] |= $ctype{"digit"};
 431         }
 432         if ($dig ne "")
 433         {
 434             $digitmap_table[$src] = ord $dig;
 435         }
 436
 437         # copy the category and direction for everything between First/Last pairs
 438         if ($name =~ /, First>/) { $start = $src; }
 439         if ($name =~ /, Last>/)
 440         {
 441             while ($start < $src)
 442             {
 443                 $category_table[$start] = $category_table[$src];
 444                 $direction_table[$start] = $direction_table[$src];
 445                 $start++;
 446             }
 447         }
 448
 449         next if $decomp eq "";  # no decomposition, skip it
 450
 451         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 452         {
 453             # decomposition of the form "<foo> 1234" -> use char if type is known
 454             if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd))
 455             {
 456                 # Single char decomposition in the compatibility range
 457                 $compatmap_table[$src] = hex $2;
 458             }
 459             if ($1 eq "isolated" || $1 eq "final" || $1 eq "initial" || $1 eq "medial")
 460             {
 461                 ${joining_forms{$1}}[hex $2] = $src;
 462                 next;
 463             }
 464             next unless ($1 eq "font" ||
 465                          $1 eq "noBreak" ||
 466                          $1 eq "circle" ||
 467                          $1 eq "super" ||
 468                          $1 eq "sub" ||
 469                          $1 eq "wide" ||
 470                          $1 eq "narrow" ||
 471                          $1 eq "compat" ||
 472                          $1 eq "small");
 473             $dst = hex $2;
 474         }
 475         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 476         {
 477             # decomposition "<compat> 0020 1234" -> combining accent
 478             $dst = hex $1;
 479         }
 480         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 481         {
 482             # decomposition contains only char values without prefix -> use first char
 483             $dst = hex $1;
 484             $category_table[$src] |= $category_table[$dst] if defined $category_table[$dst];
 485             # store decomposition if it contains two chars
 486             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 487             {
 488                 $decomp_table[$src] = [ hex $1, hex $2 ];
 489                 push @compose_table, [ hex $1, hex $2, $src ];
 490             }
 491             elsif ($decomp =~ /^(<[a-z]+>\s)*([0-9a-fA-F]+)$/ &&
 492                    (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)))
 493             {
 494                 # Single char decomposition in the compatibility range
 495                 $compatmap_table[$src] = hex $2;
 496             }
 497         }
 498         else
 499         {
 500             next;
 501         }
 502
 503         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 504
 505         # check for loops
 506         for (my $i = $dst; ; $i = $unicode_defaults[$i])
 507         {
 508             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 509             last unless defined($unicode_defaults[$i]);
 510         }
 511         $unicode_defaults[$src] = $dst;
 512     }
 513     close $UNICODE_DATA;
 514
 515     # patch the category of some special characters
 516
 517     foreach my $cat (keys %special_categories)
 518     {
 519         my $flag = $ctype{$cat};
 520         foreach my $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 521     }
 522 }
 523
 524
 525 ################################################################
 526 # define a new lead byte
 527 sub add_lead_byte($)
 528 {
 529     my $ch = shift;
 530     push @lead_bytes, $ch;
 531     $cp2uni[$ch] = 0;
 532 }
 533
 534 ################################################################
 535 # define a new char mapping
 536 sub add_mapping($$)
 537 {
 538     my ($cp, $uni) = @_;
 539     $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 540     $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 541     if ($cp > 0xff && !defined($cp2uni[$cp >> 8]))
 542     {
 543         add_lead_byte( $cp >> 8 );
 544     }
 545 }
 546
 547 ################################################################
 548 # parse a standard codepage file
 549 sub read_codepage_file($)
 550 {
 551     my $name = shift;
 552     my $INPUT = open_data_file( $MAPPINGS, $name );
 553
 554     while (<$INPUT>)
 555     {
 556         next if /^\#/;  # skip comments
 557         next if /^$/;  # skip empty lines
 558         next if /\x1a/;  # skip ^Z
 559         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 560
 561         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 562         {
 563             add_lead_byte( hex $1 );
 564             next;
 565         }
 566         # 0x12 0x3456
 567         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 568         {
 569             add_mapping( hex $1, hex $2 );
 570             next;
 571         }
 572         # 0x12 0x3456+0x7890
 573         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\+.*\s+(\#.*)?/)
 574         {
 575             add_mapping( hex $1, hex $2 );
 576             next;
 577         }
 578         # 0x12 <LR>+0x3456
 579         if (/^0x([0-9a-fA-F]+)\s+<LR>\+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 580         {
 581             add_mapping( hex $1, hex $2 );
 582             next;
 583         }
 584         # 0x12 <RL>+0x3456
 585         if (/^0x([0-9a-fA-F]+)\s+<RL>\+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 586         {
 587             my $uni = hex $2;
 588             add_mapping( hex $1, $uni ) if $uni >= 0x80;
 589             next;
 590         }
 591         die "$name: Unrecognized line $_\n";
 592     }
 593     close $INPUT;
 594 }
 595
 596
 597 ################################################################
 598 # fill input data for the 20127 (us-ascii) codepage
 599 sub fill_20127_codepage()
 600 {
 601     for (my $i = 0; $i < 128; $i++) { $cp2uni[$i] = $uni2cp[$i] = $i; }
 602     for (my $i = 128; $i < 256; $i++) { $cp2uni[$i] = $i & 0x7f; }
 603 }
 604
 605 ################################################################
 606 # get a mapping including glyph chars for MB_USEGLYPHCHARS
 607
 608 sub get_glyphs_mapping(@)
 609 {
 610     $_[0x01] = 0x263a;  # (WHITE SMILING FACE)
 611     $_[0x02] = 0x263b;  # (BLACK SMILING FACE)
 612     $_[0x03] = 0x2665;  # (BLACK HEART SUIT)
 613     $_[0x04] = 0x2666;  # (BLACK DIAMOND SUIT)
 614     $_[0x05] = 0x2663;  # (BLACK CLUB SUIT)
 615     $_[0x06] = 0x2660;  # (BLACK SPADE SUIT)
 616     $_[0x07] = 0x2022;  # (BULLET)
 617     $_[0x08] = 0x25d8;  # (INVERSE BULLET)
 618     $_[0x09] = 0x25cb;  # (WHITE CIRCLE)
 619     $_[0x0a] = 0x25d9;  # (INVERSE WHITE CIRCLE)
 620     $_[0x0b] = 0x2642;  # (MALE SIGN)
 621     $_[0x0c] = 0x2640;  # (FEMALE SIGN)
 622     $_[0x0d] = 0x266a;  # (EIGHTH NOTE)
 623     $_[0x0e] = 0x266b;  # (BEAMED EIGHTH NOTES)
 624     $_[0x0f] = 0x263c;  # (WHITE SUN WITH RAYS)
 625     $_[0x10] = 0x25ba;  # (BLACK RIGHT-POINTING POINTER)
 626     $_[0x11] = 0x25c4;  # (BLACK LEFT-POINTING POINTER)
 627     $_[0x12] = 0x2195;  # (UP DOWN ARROW)
 628     $_[0x13] = 0x203c;  # (DOUBLE EXCLAMATION MARK)
 629     $_[0x14] = 0x00b6;  # (PILCROW SIGN)
 630     $_[0x15] = 0x00a7;  # (SECTION SIGN)
 631     $_[0x16] = 0x25ac;  # (BLACK RECTANGLE)
 632     $_[0x17] = 0x21a8;  # (UP DOWN ARROW WITH BASE)
 633     $_[0x18] = 0x2191;  # (UPWARDS ARROW)
 634     $_[0x19] = 0x2193;  # (DOWNWARDS ARROW)
 635     $_[0x1a] = 0x2192;  # (RIGHTWARDS ARROW)
 636     $_[0x1b] = 0x2190;  # (LEFTWARDS ARROW)
 637     $_[0x1c] = 0x221f;  # (RIGHT ANGLE)
 638     $_[0x1d] = 0x2194;  # (LEFT RIGHT ARROW)
 639     $_[0x1e] = 0x25b2;  # (BLACK UP-POINTING TRIANGLE)
 640     $_[0x1f] = 0x25bc;  # (BLACK DOWN-POINTING TRIANGLE)
 641     $_[0x7f] = 0x2302;  # (HOUSE)
 642     return @_;
 643 }
 644
 645 ################################################################
 646 # build EUC-JP table from the JIS 0208 file
 647 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 648 # but this would require 3-byte DBCS characters
 649 sub READ_JIS0208_FILE($)
 650 {
 651     my $name = shift;
 652
 653     # ASCII chars
 654     for (my $i = 0x00; $i <= 0x7f; $i++)
 655     {
 656         $cp2uni[$i] = $i;
 657         $uni2cp[$i] = $i;
 658     }
 659
 660     # JIS X 0201 right plane
 661     for (my $i = 0xa1; $i <= 0xdf; $i++)
 662     {
 663         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 664         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 665     }
 666
 667     # lead bytes
 668     foreach my $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 669     {
 670         push @lead_bytes,$i;
 671         $cp2uni[$i] = 0;
 672     }
 673
 674     # undefined chars
 675     foreach my $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 676     {
 677         $cp2uni[$i] = $DEF_CHAR;
 678     }
 679
 680     # Shift-JIS compatibility
 681     $uni2cp[0x00a5] = 0x5c;
 682     $uni2cp[0x203e] = 0x7e;
 683
 684     # Fix backslash conversion
 685     $cp2uni[0xa1c0] = 0xff3c;
 686     $uni2cp[0xff3c] = 0xa1c0;
 687
 688     my $INPUT = open_data_file( $MAPPINGS, $name );
 689     while (<$INPUT>)
 690     {
 691         next if /^\#/;  # skip comments
 692         next if /^$/;  # skip empty lines
 693         next if /\x1a/;  # skip ^Z
 694         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 695         {
 696             my $cp = 0x8080 + hex $1;
 697             my $uni = hex $2;
 698             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 699             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 700             next;
 701         }
 702         die "$name: Unrecognized line $_\n";
 703     }
 704     close $INPUT;
 705 }
 706
 707
 708 ################################################################
 709 # build the sort keys table
 710 sub READ_SORTKEYS_FILE()
 711 {
 712     my @sortkeys = ();
 713     for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
 714
 715     my $INPUT = open_data_file( $REPORTS, $SORTKEYS );
 716     while (<$INPUT>)
 717     {
 718         next if /^\#/;  # skip comments
 719         next if /^$/;  # skip empty lines
 720         next if /\x1a/;  # skip ^Z
 721         next if /^\@version/;  # skip @version header
 722         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 723         {
 724             my ($uni,$variable) = (hex $1, $2);
 725             next if $uni > 65535;
 726             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 727             next;
 728         }
 729         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 730         {
 731             # multiple character sequence, ignored for now
 732             next;
 733         }
 734         die "$SORTKEYS: Unrecognized line $_\n";
 735     }
 736     close $INPUT;
 737
 738     # compress the keys to 32 bit:
 739     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 740
 741     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 742                        ${$a}[2] <=> ${$b}[2] or
 743                        ${$a}[3] <=> ${$b}[3] or
 744                        ${$a}[4] <=> ${$b}[4] or
 745                        $a cmp $b; } @sortkeys;
 746
 747     my ($n2, $n3) = (1, 1);
 748     my @keys = (-1, -1, -1, -1, -1 );
 749     my @flatkeys = ();
 750
 751     for (my $i = 0; $i < 65536; $i++)
 752     {
 753         my @current = @{$sortkeys[$i]};
 754         next if $current[0] == -1;
 755         if ($current[1] == $keys[1])
 756         {
 757             if ($current[2] == $keys[2])
 758             {
 759                 if ($current[3] == $keys[3])
 760                 {
 761                     # nothing
 762                 }
 763                 else
 764                 {
 765                     $keys[3] = $current[3];
 766                     $n3++;
 767                     die if ($n3 >= 16);
 768                 }
 769             }
 770             else
 771             {
 772                 $keys[2] = $current[2];
 773                 $keys[3] = $current[3];
 774                 $n2++;
 775                 $n3 = 1;
 776                 die if ($n2 >= 256);
 777             }
 778         }
 779         else
 780         {
 781             $keys[1] = $current[1];
 782             $keys[2] = $current[2];
 783             $keys[3] = $current[3];
 784             $n2 = 1;
 785             $n3 = 1;
 786         }
 787
 788         if ($current[2]) { $current[2] = $n2; }
 789         if ($current[3]) { $current[3] = $n3; }
 790         if ($current[4]) { $current[4] = 1; }
 791
 792         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 793     }
 794     return @flatkeys;
 795 }
 796
 797
 798 ################################################################
 799 # build the sort keys table
 800 sub DUMP_SORTKEYS($@)
 801 {
 802     my ($filename, @keys) = @_;
 803
 804     # count the number of 256-key ranges that contain something
 805
 806     my @offsets = ();
 807     my $ranges = 2;
 808     for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
 809     for (my $i = 0; $i < 65536; $i++)
 810     {
 811         next unless defined $keys[$i];
 812         $offsets[$i >> 8] = $ranges * 256;
 813         $ranges++;
 814         $i |= 255;
 815     }
 816
 817     # output the range offsets
 818
 819     open OUTPUT,">$filename.new" or die "Cannot create $filename";
 820     printf "Building $filename\n";
 821     printf OUTPUT "/* Unicode collation element table */\n";
 822     printf OUTPUT "/* generated from %s */\n", "$REPORTS/$SORTKEYS";
 823     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 824
 825     printf OUTPUT "const unsigned int collation_table[%d] =\n{\n", $ranges*256;
 826     printf OUTPUT "    /* index */\n";
 827     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%08x", 0, @offsets );
 828
 829     # output the default values
 830
 831     printf OUTPUT "    /* defaults */\n";
 832     printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0, (0xffffffff) x 256 );
 833
 834     # output all the key ranges
 835
 836     for (my $i = 0; $i < 256; $i++)
 837     {
 838         next if $offsets[$i] == 256;
 839         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 840         printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0xffffffff, @keys[($i<<8) .. ($i<<8)+255] );
 841     }
 842     printf OUTPUT "\n};\n";
 843     close OUTPUT;
 844     save_file($filename);
 845 }
 846
 847
 848 ################################################################
 849 # add default mappings once the file had been read
 850 sub add_default_mappings($)
 851 {
 852     my $first_private = shift;
 853
 854     # Apply aliases
 855
 856     foreach my $alias (@unicode_aliases)
 857     {
 858         my $target = undef;
 859         foreach my $src (@$alias)
 860         {
 861             if (defined($uni2cp[$src]))
 862             {
 863                 $target = $uni2cp[$src];
 864                 last;
 865             }
 866         }
 867         next unless defined($target);
 868
 869         # At least one char of the alias set is defined, set the others to the same value
 870         foreach my $src (@$alias)
 871         {
 872             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 873         }
 874     }
 875
 876     # For every src -> target mapping in the defaults table,
 877     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 878
 879     for (my $src = 0; $src < 65536; $src++)
 880     {
 881         next if defined($uni2cp[$src]);  # source has a definition already
 882         next unless defined($unicode_defaults[$src]);  # no default for this char
 883         my $target = $unicode_defaults[$src];
 884
 885         # do a recursive mapping until we find a target char that is defined
 886         while (!defined($uni2cp[$target]) &&
 887                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 888
 889         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 890     }
 891
 892     # Add a private char mapping for undefined chars
 893
 894     if ($first_private)
 895     {
 896         for (my $i = 32; $i < 256; $i++)
 897         {
 898             next if defined($cp2uni[$i]);
 899             $cp2uni[$i] = $first_private;
 900             $uni2cp[$first_private] = $i;
 901             $first_private++;
 902         }
 903     }
 904
 905     # Add an identity mapping for all undefined chars
 906
 907     for (my $i = 0; $i < 256; $i++)
 908     {
 909         next if defined($cp2uni[$i]);
 910         next if defined($uni2cp[$i]);
 911         $cp2uni[$i] = $uni2cp[$i] = $i;
 912     }
 913 }
 914
 915 ################################################################
 916 # dump an array of integers
 917 sub DUMP_ARRAY($$@)
 918 {
 919     my ($format,$default,@array) = @_;
 920     my $i;
 921     my $ret = "    ";
 922     for ($i = 0; $i < $#array; $i++)
 923     {
 924         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 925         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 926     }
 927     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 928     return $ret;
 929 }
 930
 931
 932 ################################################################
 933 # dump a unicode->ascii mapping table
 934 sub dump_uni2cp_table($$)
 935 {
 936     my ($width, $def) = @_;
 937
 938     # count the number of unicode->ascii subtables that contain something
 939
 940     my @lowerbounds;
 941     my @upperbounds;
 942     my $index = 0;
 943     my @filled = (-1) x 256;
 944     for (my $i = 0; $i < 65536; $i++)
 945     {
 946         next unless defined $uni2cp[$i];
 947         if ($filled[$i >> 8] == -1)
 948         {
 949           $lowerbounds[$index] = $i & 0xff;
 950           $upperbounds[$index] = 0xff - $lowerbounds[$index];
 951           $filled[$i >> 8] = $index * 256;
 952           $index++;
 953         }
 954         else
 955         {
 956           $upperbounds[$index-1] = 0xff - ($i & 0xff);
 957         }
 958     }
 959
 960     # add defaults mapping at the end
 961     $filled[256] = $index * 256;
 962     $lowerbounds[$index] = 255;
 963     $upperbounds[$index] = 0;
 964     $index++;
 965
 966     # collapse blocks upwards if possible
 967     my $removed = 0;
 968     $index = 0;
 969     for (my $i = 0; $i < 257; $i++)
 970     {
 971         next if $filled[$i] == -1;
 972         if ($upperbounds[$index - 1] > $lowerbounds[$index])
 973         {
 974            $removed += $lowerbounds[$index];
 975         }
 976         else
 977         {
 978            $removed += $upperbounds[$index - 1];
 979            $lowerbounds[$index] = $upperbounds[$index - 1];
 980         }
 981         $filled[$i] -= $removed;
 982         $index++;
 983     }
 984
 985     # output all the subtables into a single array
 986
 987     printf OUTPUT "static const unsigned %s uni2cp_low[%d] =\n{\n",
 988                   $width == 8 ? "char" : "short", ($index + 1) * 256 - $removed;
 989
 990     my $format = $width == 8 ? "0x%02x" : "0x%04x";
 991     for (my $i = $index = 0; $i < 257; $i++)
 992     {
 993         next if $filled[$i] == -1;
 994         my $start = ($i << 8) + $lowerbounds[$index];
 995         my $end = ($i << 8) + 255;
 996         if ($i == 256)
 997         {
 998             print OUTPUT "    /* defaults */\n";
 999             printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( $format, 0, ($def) x ($end - $start + 1) );
1000         }
1001         else
1002         {
1003             printf OUTPUT "    /* 0x%04x .. 0x%04x */\n", $start, $end;
1004             printf OUTPUT "%s,\n", DUMP_ARRAY( $format, $def, @uni2cp[$start .. $end] );
1005         }
1006         $index++;
1007     }
1008
1009     # output a table of the offsets of the subtables in the previous array
1010
1011     for (my $i = 0; $i < 256; $i++)
1012     {
1013         if ($filled[$i] == -1) { $filled[$i] = $filled[256]; }
1014     }
1015     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
1016     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @filled[0..255] );
1017 }
1018
1019 ################################################################
1020 # dump an SBCS mapping table
1021 sub dump_sbcs_table($$$$$)
1022 {
1023     my ($codepage, $has_glyphs, $name, $def, $defw) = @_;
1024     my $i;
1025
1026     # output the ascii->unicode table
1027
1028     if ($has_glyphs)
1029     {
1030         printf OUTPUT "static const WCHAR cp2uni[512] =\n";
1031         printf OUTPUT "{\n%s", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
1032         printf OUTPUT ",\n    /* glyphs */\n%s\n};\n\n",
1033                       DUMP_ARRAY( "0x%04x", $defw, get_glyphs_mapping(@cp2uni[0 .. 255]) );
1034     }
1035     else
1036     {
1037         printf OUTPUT "static const WCHAR cp2uni[256] =\n";
1038         printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
1039     }
1040
1041     dump_uni2cp_table( 8, $def );
1042
1043     # output the code page descriptor
1044
1045     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
1046     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
1047                   $codepage, $def, $defw, $name;
1048     printf OUTPUT "    cp2uni,\n";
1049     if ($has_glyphs) { printf OUTPUT "    cp2uni + 256,\n"; }
1050     else { printf OUTPUT "    cp2uni,\n"; }
1051     printf OUTPUT "    uni2cp_low,\n";
1052     printf OUTPUT "    uni2cp_high\n};\n";
1053 }
1054
1055
1056 ################################################################
1057 # dump a DBCS mapping table
1058 sub dump_dbcs_table($$$$@)
1059 {
1060     my ($codepage, $name, $def, $defw, @lb_ranges) = @_;
1061
1062     # build a list of lead bytes that are actually used
1063
1064     my @lblist = ();
1065     LBLOOP: for (my $y = 0; $y <= $#lead_bytes; $y++)
1066     {
1067         my $base = $lead_bytes[$y] << 8;
1068         for (my $x = 0; $x < 256; $x++)
1069         {
1070             if (defined $cp2uni[$base+$x])
1071             {
1072                 push @lblist,$lead_bytes[$y];
1073                 next LBLOOP;
1074             }
1075         }
1076     }
1077     my $unused = ($#lead_bytes > $#lblist);
1078
1079     # output the ascii->unicode table for the single byte chars
1080
1081     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
1082     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
1083
1084     # output the default table for unused lead bytes
1085
1086     if ($unused)
1087     {
1088         printf OUTPUT "    /* unused lead bytes */\n";
1089         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($defw) x 256 );
1090     }
1091
1092     # output the ascii->unicode table for each DBCS lead byte
1093
1094     for (my $y = 0; $y <= $#lblist; $y++)
1095     {
1096         my $base = $lblist[$y] << 8;
1097         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
1098         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[$base .. $base+255] );
1099         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
1100     }
1101
1102     # output the lead byte subtables offsets
1103
1104     my @offsets = (0) x 256;
1105     for (my $x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
1106     if ($unused)
1107     {
1108         # increment all lead bytes offset to take into account the unused table
1109         for (my $x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
1110     }
1111     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
1112     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
1113
1114     dump_uni2cp_table( 16, $def );
1115
1116     # output the code page descriptor
1117
1118     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
1119     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
1120                   $codepage, $def, $defw, $name;
1121     printf OUTPUT "    cp2uni,\n";
1122     printf OUTPUT "    cp2uni_leadbytes,\n";
1123     printf OUTPUT "    uni2cp_low,\n";
1124     printf OUTPUT "    uni2cp_high,\n";
1125     printf OUTPUT "    {\n    %s\n    }\n", DUMP_ARRAY( "0x%02x", 0, @lb_ranges, 0, 0 );
1126     printf OUTPUT "};\n";
1127 }
1128
1129
1130 ################################################################
1131 # get the list of defined lead byte ranges
1132 sub get_lb_ranges()
1133 {
1134     my @list = ();
1135     my @ranges = ();
1136     my $i = 0;
1137     foreach $i (@lead_bytes) { $list[$i] = 1; }
1138     my $on = 0;
1139     for (my $i = 0; $i < 256; $i++)
1140     {
1141         if ($on)
1142         {
1143             if (!defined $list[$i]) { push @ranges, $i-1; $on = 0; }
1144         }
1145         else
1146         {
1147             if ($list[$i]) { push @ranges, $i; $on = 1; }
1148         }
1149     }
1150     if ($on) { push @ranges, 0xff; }
1151     return @ranges;
1152 }
1153
1154 ################################################################
1155 # dump the Indic Syllabic Category table
1156 sub dump_indic($)
1157 {
1158     my $filename = shift;
1159     my @indic_table = ($indic_types{'Other'}) x 65536;;
1160
1161     my $INPUT = open_data_file( $UNIDATA, "IndicSyllabicCategory.txt" );
1162     while (<$INPUT>)
1163     {
1164         next if /^\#/;  # skip comments
1165         next if /^\s*$/;  # skip empty lines
1166         next if /\x1a/;  # skip ^Z
1167         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*#/)
1168         {
1169             my $type = $2;
1170             die "unknown indic $type" unless defined $indic_types{$type};
1171             if (hex $1 < 65536)
1172             {
1173                 $indic_table[hex $1] = $indic_types{$type};
1174             }
1175             next;
1176         }
1177         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1178         {
1179             my $type = $3;
1180             die "unknown indic $type" unless defined $indic_types{$type};
1181             if (hex $1 < 65536 and hex $2 < 6536)
1182             {
1183                 foreach my $i (hex $1 .. hex $2)
1184                 {
1185                     $indic_table[$i] = $indic_types{$type};
1186                 }
1187             }
1188             next;
1189         }
1190         die "malformed line $_";
1191     }
1192     close $INPUT;
1193
1194     $INPUT = open_data_file( $UNIDATA, "IndicMatraCategory.txt" );
1195     while (<$INPUT>)
1196     {
1197         next if /^\#/;  # skip comments
1198         next if /^\s*$/;  # skip empty lines
1199         next if /\x1a/;  # skip ^Z
1200         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z]+)\s*#/)
1201         {
1202             my $type = $2;
1203             die "unknown matra $type" unless defined $matra_types{$type};
1204             $indic_table[hex $1] += $matra_types{$type} << 8;
1205             next;
1206         }
1207         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1208         {
1209             my $type = $3;
1210             die "unknown matra $type" unless defined $matra_types{$type};
1211             foreach my $i (hex $1 .. hex $2)
1212             {
1213                 $indic_table[$i] += $matra_types{$type} << 8;
1214             }
1215             next;
1216         }
1217         die "malformed line $_";
1218     }
1219     close $INPUT;
1220
1221     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1222     print "Building $filename\n";
1223     print OUTPUT "/* Unicode Indic Syllabic Category */\n";
1224     print OUTPUT "/* generated from $UNIDATA/IndicSyllabicCategory.txt */\n";
1225     print OUTPUT "/*       and from $UNIDATA/IndicMatraCategory.txt */\n";
1226     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1227     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1228
1229     dump_two_level_mapping( "indic_syllabic_table", @indic_table);
1230
1231     close OUTPUT;
1232     save_file($filename);
1233 }
1234
1235 ################################################################
1236 # dump the Line Break Properties table
1237 sub dump_linebreak($)
1238 {
1239     my $filename = shift;
1240     my @break_table = ($break_types{'XX'}) x 65536;;
1241     my $next_group = 0;
1242
1243     my $INPUT = open_data_file( $UNIDATA, "LineBreak.txt" );
1244     while (<$INPUT>)
1245     {
1246         next if /^\#/;  # skip comments
1247         next if /^\s*$/;  # skip empty lines
1248         next if /\x1a/;  # skip ^Z
1249         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1250         {
1251             my $type = $2;
1252             die "unknown breaktype $type" unless defined $break_types{$type};
1253             $break_table[hex $1] = $break_types{$type};
1254             next;
1255         }
1256         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1257         {
1258             my $type = $3;
1259             die "unknown breaktype $type" unless defined $break_types{$type};
1260             foreach my $i (hex $1 .. hex $2)
1261             {
1262                 $break_table[$i] = $break_types{$type};
1263             }
1264             next;
1265         }
1266         die "malformed line $_";
1267     }
1268     close $INPUT;
1269
1270     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1271     print "Building $filename\n";
1272     print OUTPUT "/* Unicode Line Break Properties */\n";
1273     print OUTPUT "/* generated from $UNIDATA/LineBreak.txt */\n";
1274     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1275     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1276
1277     dump_two_level_mapping( "wine_linebreak_table", @break_table);
1278
1279     close OUTPUT;
1280     save_file($filename);
1281 }
1282
1283
1284 ################################################################
1285 # dump the BiDi mirroring table
1286 sub dump_mirroring($)
1287 {
1288     my $filename = shift;
1289     my @mirror_table = ();
1290
1291     my $INPUT = open_data_file( $UNIDATA, "BidiMirroring.txt" );
1292     while (<$INPUT>)
1293     {
1294         next if /^\#/;  # skip comments
1295         next if /^$/;  # skip empty lines
1296         next if /\x1a/;  # skip ^Z
1297         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+)/)
1298         {
1299             $mirror_table[hex $1] = hex $2;
1300             next;
1301         }
1302         die "malformed line $_";
1303     }
1304     close $INPUT;
1305
1306     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1307     print "Building $filename\n";
1308     print OUTPUT "/* Unicode BiDi mirroring */\n";
1309     print OUTPUT "/* generated from $UNIDATA/BidiMirroring.txt */\n";
1310     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1311     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1312     DUMP_CASE_TABLE( "wine_mirror_map", @mirror_table );
1313     close OUTPUT;
1314     save_file($filename);
1315 }
1316
1317
1318 ################################################################
1319 # dump the Arabic shaping table
1320 sub dump_shaping($)
1321 {
1322     my $filename = shift;
1323     my %groups;
1324     my $next_group = 0;
1325
1326     $groups{"No_Joining_Group"} = $next_group++;
1327
1328     my $INPUT = open_data_file( $UNIDATA, "ArabicShaping.txt" );
1329     while (<$INPUT>)
1330     {
1331         next if /^\#/;  # skip comments
1332         next if /^\s*$/;  # skip empty lines
1333         next if /\x1a/;  # skip ^Z
1334         if (/^\s*([0-9a-fA-F]+)\s*;.*;\s*([RLDCUT])\s*;\s*(\w+)/)
1335         {
1336             my $type = $2;
1337             my $group = $3;
1338             $groups{$group} = $next_group++ unless defined $groups{$group};
1339             $joining_table[hex $1] = $joining_types{$type} | ($groups{$group} << 8);
1340             next;
1341         }
1342         die "malformed line $_";
1343     }
1344     close $INPUT;
1345
1346     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1347     print "Building $filename\n";
1348     print OUTPUT "/* Unicode Arabic shaping */\n";
1349     print OUTPUT "/* generated from $UNIDATA/ArabicShaping.txt */\n";
1350     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1351     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1352
1353     dump_two_level_mapping( "wine_shaping_table", @joining_table );
1354
1355     print OUTPUT "\nconst unsigned short wine_shaping_forms[256][4] =\n{\n";
1356     for (my $i = 0x600; $i <= 0x6ff; $i++)
1357     {
1358         printf OUTPUT "    { 0x%04x, 0x%04x, 0x%04x, 0x%04x },\n",
1359             ${joining_forms{"isolated"}}[$i] || $i,
1360             ${joining_forms{"final"}}[$i] || $i,
1361             ${joining_forms{"initial"}}[$i] || $i,
1362             ${joining_forms{"medial"}}[$i] || $i;
1363     }
1364     print OUTPUT "};\n";
1365
1366     close OUTPUT;
1367     save_file($filename);
1368 }
1369
1370 ################################################################
1371 # dump the Vertical Orientation table
1372 sub dump_vertical($)
1373 {
1374     my $filename = shift;
1375     my @vertical_table = ($vertical_types{'R'}) x 65536;
1376
1377     my $INPUT = open_data_file( $VERTICALDATA, "VerticalOrientation-11.txt" );
1378     while (<$INPUT>)
1379     {
1380         next if /^\#/;  # skip comments
1381         next if /^\s*$/;  # skip empty lines
1382         next if /\x1a/;  # skip ^Z
1383         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/)
1384         {
1385             my $type = $2;
1386             die "unknown vertical $type" unless defined $vertical_types{$type};
1387             if (hex $1 < 65536)
1388             {
1389                 $vertical_table[hex $1] = $vertical_types{$type};
1390             }
1391             next;
1392         }
1393         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*/)
1394         {
1395             my $type = $3;
1396             die "unknown vertical $type" unless defined $vertical_types{$type};
1397             foreach my $i (hex $1 .. hex $2)
1398             {
1399                 $vertical_table[$i] = $vertical_types{$type};
1400             }
1401             next;
1402         }
1403         die "malformed line $_";
1404     }
1405     close $INPUT;
1406
1407     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1408     print "Building $filename\n";
1409     print OUTPUT "/* Unicode Vertical Orientation */\n";
1410     print OUTPUT "/* generated from $VERTICALDATA/VerticalOrientation-11.txt */\n";
1411     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1412     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1413
1414     dump_two_level_mapping( "vertical_orientation_table", @vertical_table);
1415
1416     close OUTPUT;
1417     save_file($filename);
1418 }
1419
1420 ################################################################
1421 # dump the case mapping tables
1422 sub DUMP_CASE_MAPPINGS($)
1423 {
1424     my $filename = shift;
1425     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1426     printf "Building $filename\n";
1427     printf OUTPUT "/* Unicode case mappings */\n";
1428     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1429     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1430
1431     DUMP_CASE_TABLE( "wine_casemap_lower", @tolower_table );
1432     DUMP_CASE_TABLE( "wine_casemap_upper", @toupper_table );
1433     DUMP_CASE_TABLE( "wine_digitmap",  @digitmap_table );
1434     DUMP_CASE_TABLE( "wine_compatmap", @compatmap_table );
1435     close OUTPUT;
1436     save_file($filename);
1437 }
1438
1439
1440 ################################################################
1441 # dump a case mapping table
1442 sub DUMP_CASE_TABLE($@)
1443 {
1444     my ($name,@table) = @_;
1445
1446     # count the number of sub tables that contain something
1447     # also compute the low and upper populated bounds
1448
1449     my @lowerbounds = ( 0, 0 );
1450     my @upperbounds = ( 0, 255 );
1451     my $index = 0;
1452     my @filled = ();
1453     for (my $i = 0; $i < 65536; $i++)
1454     {
1455         next unless defined $table[$i];
1456         if (!defined $filled[$i >> 8])
1457         {
1458           $lowerbounds[$index] = $i & 0xff;
1459           $upperbounds[$index] = 0xff - $lowerbounds[$index];
1460           $filled[$i >> 8] = $index * 256 + 512;
1461           $index++;
1462         }
1463         else
1464         {
1465           $upperbounds[$index-1] = 0xff - ($i & 0xff);
1466         }
1467         $table[$i] = ($table[$i] - $i) & 0xffff;
1468     }
1469
1470     # Collapse blocks upwards if possible
1471     my $removed = 0;
1472     $index = 0;
1473     for (my $i = 0; $i < 256; $i++)
1474     {
1475         next unless defined $filled[$i];
1476         if ($upperbounds[$index - 1] > $lowerbounds[$index])
1477         {
1478            $removed = $removed + $lowerbounds[$index];
1479         }
1480         else
1481         {
1482            $removed = $removed + $upperbounds[$index - 1];
1483            $lowerbounds[$index] = $upperbounds[$index - 1];
1484         }
1485         $filled[$i] = $filled[$i] - $removed;
1486         $index++;
1487     }
1488
1489     # dump the table
1490
1491     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
1492     printf OUTPUT "{\n    /* index */\n";
1493     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
1494     printf OUTPUT "    /* defaults */\n";
1495     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
1496     $index = 0;
1497     for (my $i = 0; $i < 256; $i++)
1498     {
1499         next unless $filled[$i];
1500         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
1501         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0,
1502                       @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
1503         $index++;
1504     }
1505     printf OUTPUT "\n};\n";
1506 }
1507
1508 ################################################################
1509 # compress a mapping table by removing identical rows
1510 sub compress_array($@)
1511 {
1512     my $rows = shift;
1513     my @table = @_;
1514     my $len = @table / $rows;
1515     my @array = (0) x $rows;
1516     my %sequences;
1517
1518     # try to merge table rows
1519     for (my $row = 0; $row < $rows; $row++)
1520     {
1521         my $rowtxt = pack "S*", @table[($row * $len)..($row * $len + $len - 1)];
1522         if (defined($sequences{$rowtxt}))
1523         {
1524             # reuse an existing row
1525             $array[$row] = $sequences{$rowtxt};
1526         }
1527         else
1528         {
1529             # create a new row
1530             $sequences{$rowtxt} = $array[$row] = $#array + 1;
1531             push @array, @table[$row * $len..$row * $len + $len - 1];
1532         }
1533     }
1534     return @array;
1535 }
1536
1537 ################################################################
1538 # dump a simple char -> 16-bit value mapping table
1539 sub dump_simple_mapping($@)
1540 {
1541     my $name = shift;
1542     my @array = compress_array( 256, @_[0..65535] );
1543
1544     printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, $#array+1;
1545     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1546     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1547 }
1548
1549 ################################################################
1550 # dump a char -> 16-bit value mapping table using two-level tables
1551 sub dump_two_level_mapping($@)
1552 {
1553     my $name = shift;
1554     my @row_array = compress_array( 4096, @_[0..65535] );
1555     my @array = compress_array( 256, @row_array[0..4095] );
1556
1557     for (my $i = 256; $i < @array; $i++) { $array[$i] += @array - 4096; }
1558
1559     printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, @array + @row_array - 4096;
1560     printf OUTPUT "    /* level 1 offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1561     printf OUTPUT "    /* level 2 offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1562     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @row_array[4096..$#row_array] );
1563 }
1564
1565 ################################################################
1566 # dump a binary case mapping table in l_intl.nls format
1567 sub dump_binary_case_table(@)
1568 {
1569     my (@table) = @_;
1570
1571     my %difftables_hash = ();
1572     my @difftables;
1573     my %offtables2_hash = ();
1574     my @offtables2 = ();
1575
1576     my @offtable = ();
1577     for (my $i = 0; $i < 256; $i++)
1578     {
1579         my @offtable2 = ();
1580         for(my $j = 0; $j < 16; $j++) # offset table for xx00-xxFF characters
1581         {
1582             my @difftable;
1583             for (my $k = 0; $k < 16; $k++) # case map table for xxx0-xxxF characters
1584             {
1585                 my $char = ($i<<8) + ($j<<4) + $k;
1586                 $difftable[$k] = (defined $table[$char]) ? (($table[$char]-$char) & 0xffff) : 0;
1587             }
1588
1589             my $diff_key = pack "S*", @difftable;
1590             my $offset3 = $difftables_hash{$diff_key};
1591             if (!defined $offset3)
1592             {
1593                 $offset3 = scalar @difftables;
1594                 $difftables_hash{$diff_key} = $offset3;
1595                 push @difftables, @difftable;
1596             }
1597             $offtable2[$j] = $offset3;
1598         }
1599
1600         my $offtable2_key = pack "S*", @offtable2;
1601         my $offset2 = $offtables2_hash{$offtable2_key};
1602         if (!defined $offset2)
1603         {
1604             $offset2 = scalar @offtables2;
1605             $offtables2_hash{$offtable2_key} = $offset2;
1606             push @offtables2, \@offtable2;
1607         }
1608         $offtable[$i] = $offset2;
1609     }
1610
1611     my @output;
1612     my $offset = 0x100; # offset of first subtable in words
1613     foreach (@offtable)
1614     {
1615         push @output, 0x10 * $_ + $offset; # offset of subtable in words
1616     }
1617
1618     $offset = 0x100 + 0x10 * scalar @offtables2; # offset of first difftable in words
1619     foreach(@offtables2)
1620     {
1621         my $table = $_;
1622         foreach(@$table)
1623         {
1624             push @output, $_ + $offset; # offset of difftable in words
1625         }
1626     }
1627
1628     my $len = 1 + scalar @output + scalar @difftables;
1629     return pack "S<*", $len, @output, @difftables;
1630 }
1631
1632
1633 ################################################################
1634 # dump case mappings for l_intl.nls
1635 sub dump_intl_nls($)
1636 {
1637     my $filename = shift;
1638     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1639     printf "Building $filename\n";
1640
1641     binmode OUTPUT;
1642     print OUTPUT pack "S<", 1;  # version
1643     print OUTPUT dump_binary_case_table( @toupper_table );
1644     print OUTPUT dump_binary_case_table( @tolower_table );
1645     close OUTPUT;
1646     save_file($filename);
1647 }
1648
1649
1650 sub load_nameprep_range_table($$$)
1651 {
1652     my ($INPUT, $val, $table_ref) = @_;
1653
1654     while (<$INPUT>)
1655     {
1656         if (/^\s*([0-9a-fA-F]+)-([0-9a-fA-F]+)/)
1657         {
1658             my $last = hex $2;
1659             $last = 65535 if($last >= 65536);
1660             foreach my $i (hex $1 .. $last)
1661             {
1662                 $table_ref->[$i] |= $val;
1663             }
1664             next;
1665         }
1666         elsif (/^\s*([0-9a-fA-F]+)/)
1667         {
1668             if (hex $1 < 65536)
1669             {
1670                 $table_ref->[hex $1] |= $val;
1671             }
1672             next;
1673         }
1674
1675         return if (/End\sTable/);
1676     }
1677 }
1678
1679 sub load_nameprep_map_table($$)
1680 {
1681     my ($INPUT, $table_ref) = @_;
1682
1683     while (<$INPUT>)
1684     {
1685         if (/^\s*([0-9a-fA-F]+);\s;/)
1686         {
1687             # special value for map to nothing
1688             $table_ref->[hex $1] = [0xffff, 0xffff, 0xffff];
1689             next;
1690         }
1691         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+);/)
1692         {
1693             $table_ref->[hex $1] = [hex $2, 0, 0];
1694             next;
1695         }
1696         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1697         {
1698             $table_ref->[hex $1] = [hex $2, hex $3, 0];
1699             next;
1700         }
1701         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1702         {
1703             $table_ref->[hex $1] = [hex $2, hex $3, hex $4];
1704             next;
1705         }
1706
1707         return if (/End\sTable/);
1708     }
1709 }
1710
1711 ################################################################
1712 # dump mapping table, prohibited characters set, unassigned
1713 # characters, bidirectional rules used by nameprep algorithm
1714 sub dump_nameprep($)
1715 {
1716     my $filename = shift;
1717     my @mapping_table = ();
1718     my @flags_table = (0) x 65536;
1719
1720     my $INPUT = open_data_file( $RFCS, $STRINGPREP );
1721     while (<$INPUT>)
1722     {
1723         next unless /Start\sTable/;
1724
1725         load_nameprep_range_table($INPUT, $nameprep_flags{"unassigned"}, \@flags_table) if (/A.1/);
1726         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.1.2/);
1727         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.2.2/);
1728         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.3/);
1729         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.4/);
1730         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.5/);
1731         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.6/);
1732         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.7/);
1733         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.8/);
1734         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.9/);
1735         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_ral"}, \@flags_table) if (/D.1/);
1736         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_l"}, \@flags_table) if (/D.2/);
1737
1738         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.1/);
1739         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.2/);
1740     }
1741     close $INPUT;
1742
1743     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1744     print "Building $filename\n";
1745     print OUTPUT "/* Nameprep algorithm related data */\n";
1746     print OUTPUT "/* generated from $RFCS/$STRINGPREP */\n";
1747     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1748     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1749
1750     dump_two_level_mapping( "nameprep_char_type", @flags_table );
1751
1752     ######### mapping table
1753     # first determine all the 16-char subsets that contain something
1754     my @filled = ();
1755     my $pos = 16*3;  # for the null subset
1756     for (my $i = 0; $i < 65536; $i++)
1757     {
1758         next unless defined $mapping_table[$i];
1759         $filled[$i >> 4] = $pos;
1760         $pos += 16*3;
1761         $i |= 15;
1762     }
1763     my $total = $pos;
1764
1765     # now count the 256-char subsets that contain something
1766     my @filled_idx = (256) x 256;
1767     $pos = 256 + 16;
1768     for (my $i = 0; $i < 4096; $i++)
1769     {
1770         next unless $filled[$i];
1771         $filled_idx[$i >> 4] = $pos;
1772         $pos += 16;
1773         $i |= 15;
1774     }
1775     my $null_offset = $pos;
1776     $total += $pos;
1777
1778     # add the index offsets to the subsets positions
1779     for (my $i = 0; $i < 4096; $i++)
1780     {
1781         next unless $filled[$i];
1782         $filled[$i] += $null_offset;
1783     }
1784
1785     # dump the main index
1786     printf OUTPUT "const WCHAR nameprep_mapping[%d] =\n", $total;
1787     printf OUTPUT "{\n    /* index */\n";
1788     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1789     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1790
1791     # dump the second-level indexes
1792     for (my $i = 0; $i < 256; $i++)
1793     {
1794         next unless ($filled_idx[$i] > 256);
1795         my @table = @filled[($i<<4)..($i<<4)+15];
1796         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1797         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1798         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1799     }
1800
1801     # dump the 16-char subsets
1802     printf OUTPUT ",\n    /* null mapping */\n";
1803         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 48 );
1804
1805     for (my $i = 0; $i < 4096; $i++)
1806     {
1807         next unless $filled[$i];
1808         my @table = (0) x 48;
1809         for (my $j = 0; $j < 16; $j++)
1810         {
1811             if (defined $mapping_table[($i<<4) + $j])
1812             {
1813                 $table[3 * $j] = ${$mapping_table[($i << 4) + $j]}[0];
1814                 $table[3 * $j + 1] = ${$mapping_table[($i << 4) + $j]}[1];
1815                 $table[3 * $j + 2] = ${$mapping_table[($i << 4) + $j]}[2];
1816             }
1817         }
1818         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1819         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1820     }
1821
1822     printf OUTPUT "\n};\n";
1823
1824     close OUTPUT;
1825     save_file($filename);
1826 }
1827
1828 ################################################################
1829 # dump the ctype tables
1830 sub DUMP_CTYPE_TABLES($)
1831 {
1832     my $filename = shift;
1833     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1834     printf "Building $filename\n";
1835     printf OUTPUT "/* Unicode ctype tables */\n";
1836     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1837     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1838
1839     # add the direction in the high 4 bits of the category
1840     for (my $i = 0; $i < 65536; $i++)
1841     {
1842         $category_table[$i] |= $direction_table[$i] << 12 if defined $direction_table[$i];
1843     }
1844
1845     dump_simple_mapping( "wine_wctype_table", @category_table );
1846
1847     close OUTPUT;
1848     save_file($filename);
1849 }
1850
1851
1852 ################################################################
1853 # dump the char composition tables
1854 sub DUMP_COMPOSE_TABLES($)
1855 {
1856     my $filename = shift;
1857
1858     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1859     printf "Building $filename\n";
1860     printf OUTPUT "/* Unicode char composition */\n";
1861     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1862     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1863
1864     ######### composition table
1865
1866     my @filled = ();
1867     foreach my $i (@compose_table)
1868     {
1869         my @comp = @$i;
1870         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
1871     }
1872
1873     # count how many different second chars we have
1874
1875     my $count = 0;
1876     for (my $i = 0; $i < 65536; $i++)
1877     {
1878         next unless defined $filled[$i];
1879         $count++;
1880     }
1881
1882     # build the table of second chars and offsets
1883
1884     my $pos = $count + 1;
1885     my @table = ();
1886     for (my $i = 0; $i < 65536; $i++)
1887     {
1888         next unless defined $filled[$i];
1889         push @table, $i, $pos;
1890         $pos += @{$filled[$i]};
1891     }
1892     # terminator with last position
1893     push @table, 0, $pos;
1894     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
1895     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
1896
1897     # build the table of first chars and mappings
1898
1899     for (my $i = 0; $i < 65536; $i++)
1900     {
1901         next unless defined $filled[$i];
1902         my @table = ();
1903         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
1904         for (my $j = 0; $j <= $#list; $j++)
1905         {
1906             push @table, $list[$j][0], $list[$j][1];
1907         }
1908         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
1909     }
1910     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
1911
1912     ######### decomposition table
1913
1914     # first determine all the 16-char subsets that contain something
1915
1916     @filled = (0) x 4096;
1917     $pos = 16*2;  # for the null subset
1918     for (my $i = 0; $i < 65536; $i++)
1919     {
1920         next unless defined $decomp_table[$i];
1921         $filled[$i >> 4] = $pos;
1922         $pos += 16*2;
1923         $i |= 15;
1924     }
1925     my $total = $pos;
1926
1927     # now count the 256-char subsets that contain something
1928
1929     my @filled_idx = (256) x 256;
1930     $pos = 256 + 16;
1931     for (my $i = 0; $i < 4096; $i++)
1932     {
1933         next unless $filled[$i];
1934         $filled_idx[$i >> 4] = $pos;
1935         $pos += 16;
1936         $i |= 15;
1937     }
1938     my $null_offset = $pos;  # null mapping
1939     $total += $pos;
1940
1941     # add the index offsets to the subsets positions
1942
1943     for (my $i = 0; $i < 4096; $i++)
1944     {
1945         next unless $filled[$i];
1946         $filled[$i] += $null_offset;
1947     }
1948
1949     # dump the main index
1950
1951     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
1952     printf OUTPUT "{\n    /* index */\n";
1953     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1954     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1955
1956     # dump the second-level indexes
1957
1958     for (my $i = 0; $i < 256; $i++)
1959     {
1960         next unless ($filled_idx[$i] > 256);
1961         my @table = @filled[($i<<4)..($i<<4)+15];
1962         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1963         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1964         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1965     }
1966
1967     # dump the 16-char subsets
1968
1969     printf OUTPUT ",\n    /* null mapping */\n";
1970     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
1971
1972     for (my $i = 0; $i < 4096; $i++)
1973     {
1974         next unless $filled[$i];
1975         my @table = (0) x 32;
1976         for (my $j = 0; $j < 16; $j++)
1977         {
1978             if (defined $decomp_table[($i<<4) + $j])
1979             {
1980                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
1981                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
1982             }
1983         }
1984         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1985         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1986     }
1987
1988     printf OUTPUT "\n};\n";
1989     close OUTPUT;
1990     save_file($filename);
1991 }
1992
1993 ################################################################
1994 # output a codepage definition file from the global tables
1995 sub output_codepage_file($$$$)
1996 {
1997     my ($codepage, $filename, $comment, $has_glyphs) = @_;
1998
1999     my $output = sprintf "libs/wine/c_%03d.c", $codepage;
2000     open OUTPUT,">$output.new" or die "Cannot create $output";
2001
2002     printf "Building %s from %s (%s)\n", $output, $filename || "hardcoded data", $comment;
2003
2004     # dump all tables
2005
2006     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
2007     if ($filename)
2008     {
2009         print OUTPUT "/* generated from $MAPPINGS/$filename */\n";
2010         print OUTPUT "/* DO NOT EDIT!! */\n\n";
2011     }
2012     else
2013     {
2014         printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
2015     }
2016     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
2017
2018     if (!@lead_bytes) { dump_sbcs_table( $codepage, $has_glyphs, $comment, $default_char, $default_wchar ); }
2019     else { dump_dbcs_table( $codepage, $comment, $default_char, $default_wchar, get_lb_ranges() ); }
2020     close OUTPUT;
2021     save_file($output);
2022 }
2023
2024 ################################################################
2025 # read a "bestfit" Windows mapping file
2026 sub read_bestfit_file($)
2027 {
2028     my ($filename) = @_;
2029     my $state = "";
2030     my ($codepage, $width, $count);
2031     my ($lb_cur, $lb_end);
2032
2033     my $INPUT = open_data_file( $MAPPINGS, $filename ) or die "Cannot open $filename";
2034
2035     while (<$INPUT>)
2036     {
2037         next if /^;/;  # skip comments
2038         next if /^\s*$/;  # skip empty lines
2039         next if /\x1a/;  # skip ^Z
2040         last if /^ENDCODEPAGE/;
2041
2042         if (/^CODEPAGE\s+(\d+)/)
2043         {
2044             $codepage = $1;
2045             next;
2046         }
2047         if (/^CPINFO\s+(\d+)\s+0x([0-9a-fA-f]+)\s+0x([0-9a-fA-F]+)/)
2048         {
2049             $width = $1;
2050             $default_char = hex $2;
2051             $default_wchar = hex $3;
2052             next;
2053         }
2054         if (/^(MBTABLE|WCTABLE|DBCSRANGE|DBCSTABLE)\s+(\d+)/)
2055         {
2056             $state = $1;
2057             $count = $2;
2058             next;
2059         }
2060         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)/)
2061         {
2062             if ($state eq "MBTABLE")
2063             {
2064                 my $cp = hex $1;
2065                 my $uni = hex $2;
2066                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
2067                 next;
2068             }
2069             if ($state eq "WCTABLE")
2070             {
2071                 my $uni = hex $1;
2072                 my $cp = hex $2;
2073                 $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
2074                 next;
2075             }
2076             if ($state eq "DBCSRANGE")
2077             {
2078                 my $start = hex $1;
2079                 my $end = hex $2;
2080                 for (my $i = $start; $i <= $end; $i++) { add_lead_byte( $i ); }
2081                 $lb_cur = $start;
2082                 $lb_end = $end;
2083                 next;
2084             }
2085             if ($state eq "DBCSTABLE")
2086             {
2087                 my $mb = hex $1;
2088                 my $uni = hex $2;
2089                 my $cp = ($lb_cur << 8) | $mb;
2090                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
2091                 if (!--$count)
2092                 {
2093                     if (++$lb_cur > $lb_end) { $state = "DBCSRANGE"; }
2094                 }
2095                 next;
2096             }
2097         }
2098         die "$filename: Unrecognized line $_\n";
2099     }
2100     close $INPUT;
2101 }
2102
2103
2104 ################################################################
2105 # read an input file and generate the corresponding .c file
2106 sub HANDLE_FILE(@)
2107 {
2108     my ($codepage,$filename,$has_glyphs,$comment,$first_private,$def,$defw) = @_;
2109
2110     @cp2uni = ();
2111     @lead_bytes = ();
2112     @uni2cp = ();
2113     $default_char = $def || $DEF_CHAR;
2114     $default_wchar = $defw || $DEF_CHAR;
2115
2116     # some codepage files are special
2117     if ($codepage == 20932)
2118     {
2119         READ_JIS0208_FILE( $filename );
2120         add_default_mappings( $first_private );
2121     }
2122     elsif ($codepage == 20127)
2123     {
2124         fill_20127_codepage();
2125         add_default_mappings( $first_private );
2126     }
2127     elsif ($filename =~ /\/bestfit/)
2128     {
2129         read_bestfit_file( $filename );
2130     }
2131     else
2132     {
2133         read_codepage_file( $filename );
2134         if ($codepage == 10001)
2135         {
2136             # add Shift-JIS mappings
2137             read_bestfit_file( "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt" );
2138         }
2139         else
2140         {
2141             add_default_mappings( $first_private );
2142         }
2143     }
2144
2145     output_codepage_file( $codepage, $filename, $comment, $has_glyphs );
2146 }
2147
2148
2149 ################################################################
2150 # save a file if modified
2151 sub save_file($)
2152 {
2153     my $file = shift;
2154     if (-f $file && !system "cmp $file $file.new >/dev/null")
2155     {
2156         unlink "$file.new";
2157     }
2158     else
2159     {
2160         rename "$file.new", "$file";
2161     }
2162 }
2163
2164
2165 ################################################################
2166 # output the list of codepage tables into the cptable.c file
2167 sub output_cptable($)
2168 {
2169     my $output = shift;
2170     my @tables_decl = ();
2171
2172     printf "Building %s\n", $output;
2173
2174     foreach my $file (@allfiles)
2175     {
2176         my ($codepage,$filename) = @$file;
2177         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
2178     }
2179
2180     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
2181     foreach my $file (@allfiles)
2182     {
2183         my ($codepage,$filename) = @$file;
2184         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
2185     }
2186     push @tables_decl, "};";
2187     REPLACE_IN_FILE( $output, @tables_decl );
2188 }
2189
2190 ################################################################
2191 # replace the contents of a file between ### cpmap ### marks
2192
2193 sub REPLACE_IN_FILE($@)
2194 {
2195     my $name = shift;
2196     my @data = @_;
2197     my @lines = ();
2198     open(FILE,$name) or die "Can't open $name";
2199     while (<FILE>)
2200     {
2201         push @lines, $_;
2202         last if /\#\#\# cpmap begin \#\#\#/;
2203     }
2204     push @lines, @data;
2205     while (<FILE>)
2206     {
2207         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
2208     }
2209     push @lines, <FILE>;
2210     open(FILE,">$name.new") or die "Can't modify $name";
2211     print FILE @lines;
2212     close(FILE);
2213     save_file($name);
2214 }
2215
2216 ################################################################
2217 # main routine
2218
2219 chdir ".." if -f "./make_unicode";
2220 READ_DEFAULTS( $DEFAULTS );
2221 DUMP_CASE_MAPPINGS( "libs/wine/casemap.c" );
2222 DUMP_SORTKEYS( "libs/wine/collation.c", READ_SORTKEYS_FILE() );
2223 DUMP_COMPOSE_TABLES( "libs/wine/compose.c" );
2224 DUMP_CTYPE_TABLES( "libs/wine/wctype.c" );
2225 dump_mirroring( "dlls/usp10/mirror.c" );
2226 dump_shaping( "dlls/usp10/shaping.c" );
2227 dump_linebreak( "dlls/usp10/linebreak.c" );
2228 dump_indic( "dlls/usp10/indicsyllable.c" );
2229 dump_intl_nls("loader/l_intl.nls");
2230 dump_vertical( "dlls/gdi32/vertical.c" );
2231 dump_vertical( "dlls/wineps.drv/vertical.c" );
2232 dump_nameprep( "dlls/kernel32/nameprep.c" );
2233
2234 foreach my $file (@allfiles) { HANDLE_FILE( @{$file} ); }
2235
2236 output_cptable("libs/wine/cptable.c");
2237
2238 exit 0;
2239
2240 # Local Variables:
2241 # compile-command: "./make_unicode"
2242 # End: