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/6.2.0/ucd";
  27 my $REPORTS = "http://www.unicode.org/reports";
  28 my $RFCS = "http://www.rfc-editor.org/rfc";
  29
  30 # Sort keys file
  31 my $SORTKEYS = "tr10/allkeys.txt";
  32
  33 # RFC3454 (stringprep data)
  34 my $STRINGPREP = "rfc3454.txt";
  35
  36 # Defaults mapping
  37 my $DEFAULTS = "tools/unicode-defaults";
  38
  39 # Default char for undefined mappings
  40 my $DEF_CHAR = ord '?';
  41
  42 my @allfiles =
  43 (
  44     [ 37,    "VENDORS/MICSFT/EBCDIC/CP037.TXT",   0, "IBM EBCDIC US Canada", 0x6f ],
  45     [ 424,   "VENDORS/MISC/CP424.TXT",            0, "IBM EBCDIC Hebrew" ],
  46     [ 437,   "VENDORS/MICSFT/PC/CP437.TXT",       1, "OEM United States" ],
  47     [ 500,   "VENDORS/MICSFT/EBCDIC/CP500.TXT",   0, "IBM EBCDIC International", 0x6f ],
  48     [ 737,   "VENDORS/MICSFT/PC/CP737.TXT",       1, "OEM Greek 437G" ],
  49     [ 775,   "VENDORS/MICSFT/PC/CP775.TXT",       1, "OEM Baltic" ],
  50     [ 850,   "VENDORS/MICSFT/PC/CP850.TXT",       1, "OEM Multilingual Latin 1" ],
  51     [ 852,   "VENDORS/MICSFT/PC/CP852.TXT",       1, "OEM Slovak Latin 2" ],
  52     [ 855,   "VENDORS/MICSFT/PC/CP855.TXT",       1, "OEM Cyrillic" ],
  53     [ 856,   "VENDORS/MISC/CP856.TXT",            0, "Hebrew PC" ],
  54     [ 857,   "VENDORS/MICSFT/PC/CP857.TXT",       1, "OEM Turkish" ],
  55     [ 860,   "VENDORS/MICSFT/PC/CP860.TXT",       1, "OEM Portuguese" ],
  56     [ 861,   "VENDORS/MICSFT/PC/CP861.TXT",       1, "OEM Icelandic" ],
  57     [ 862,   "VENDORS/MICSFT/PC/CP862.TXT",       1, "OEM Hebrew" ],
  58     [ 863,   "VENDORS/MICSFT/PC/CP863.TXT",       1, "OEM Canadian French" ],
  59     [ 864,   "VENDORS/MICSFT/PC/CP864.TXT",       0, "OEM Arabic" ],
  60     [ 865,   "VENDORS/MICSFT/PC/CP865.TXT",       1, "OEM Nordic" ],
  61     [ 866,   "VENDORS/MICSFT/PC/CP866.TXT",       1, "OEM Russian" ],
  62     [ 869,   "VENDORS/MICSFT/PC/CP869.TXT",       1, "OEM Greek" ],
  63     [ 874,   "VENDORS/MICSFT/WindowsBestFit/bestfit874.txt",  1, "ANSI/OEM Thai" ],
  64     [ 875,   "VENDORS/MICSFT/EBCDIC/CP875.TXT",               0, "IBM EBCDIC Greek", 0x6f ],
  65     [ 878,   "VENDORS/MISC/KOI8-R.TXT",                       0, "Russian KOI8" ],
  66     [ 932,   "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt",  0, "ANSI/OEM Japanese Shift-JIS" ],
  67     [ 936,   "VENDORS/MICSFT/WindowsBestFit/bestfit936.txt",  0, "ANSI/OEM Simplified Chinese GBK" ],
  68     [ 949,   "VENDORS/MICSFT/WindowsBestFit/bestfit949.txt",  0, "ANSI/OEM Korean Unified Hangul" ],
  69     [ 950,   "VENDORS/MICSFT/WindowsBestFit/bestfit950.txt",  0, "ANSI/OEM Traditional Chinese Big5" ],
  70     [ 1006,  "VENDORS/MISC/CP1006.TXT",                       0, "IBM Arabic" ],
  71     [ 1026,  "VENDORS/MICSFT/EBCDIC/CP1026.TXT",              0, "IBM EBCDIC Latin 5 Turkish", 0x6f ],
  72     [ 1250,  "VENDORS/MICSFT/WindowsBestFit/bestfit1250.txt", 0, "ANSI Eastern Europe" ],
  73     [ 1251,  "VENDORS/MICSFT/WindowsBestFit/bestfit1251.txt", 0, "ANSI Cyrillic" ],
  74     [ 1252,  "VENDORS/MICSFT/WindowsBestFit/bestfit1252.txt", 0, "ANSI Latin 1" ],
  75     [ 1253,  "VENDORS/MICSFT/WindowsBestFit/bestfit1253.txt", 0, "ANSI Greek" ],
  76     [ 1254,  "VENDORS/MICSFT/WindowsBestFit/bestfit1254.txt", 0, "ANSI Turkish" ],
  77     [ 1255,  "VENDORS/MICSFT/WindowsBestFit/bestfit1255.txt", 0, "ANSI Hebrew" ],
  78     [ 1256,  "VENDORS/MICSFT/WindowsBestFit/bestfit1256.txt", 0, "ANSI Arabic" ],
  79     [ 1257,  "VENDORS/MICSFT/WindowsBestFit/bestfit1257.txt", 0, "ANSI Baltic" ],
  80     [ 1258,  "VENDORS/MICSFT/WindowsBestFit/bestfit1258.txt", 0, "ANSI/OEM Viet Nam" ],
  81     [ 1361,  "VENDORS/MICSFT/WindowsBestFit/bestfit1361.txt", 0, "Korean Johab" ],
  82     [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT",      0, "Mac Roman" ],
  83     [ 10001, "VENDORS/APPLE/JAPANESE.TXT",        0, "Mac Japanese" ],
  84     [ 10002, "VENDORS/APPLE/CHINTRAD.TXT",        0, "Mac Traditional Chinese" ],
  85     [ 10003, "VENDORS/APPLE/KOREAN.TXT",          0, "Mac Korean" ],
  86     [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT",      0, "Mac Greek" ],
  87     [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT",   0, "Mac Cyrillic" ],
  88     [ 10008, "VENDORS/APPLE/CHINSIMP.TXT",        0, "Mac Simplified Chinese" ],
  89     [ 10010, "VENDORS/APPLE/ROMANIAN.TXT",        0, "Mac Romanian" ],
  90     [ 10017, "VENDORS/APPLE/CYRILLIC.TXT",        0, "Mac Ukrainian" ],
  91     [ 10021, "VENDORS/APPLE/THAI.TXT",            0, "Mac Thai" ],
  92     [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT",     0, "Mac Latin 2" ],
  93     [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT",    0, "Mac Icelandic" ],
  94     [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT",    0, "Mac Turkish" ],
  95     [ 10082, "VENDORS/APPLE/CROATIAN.TXT",        0, "Mac Croatian" ],
  96     [ 20127, undef,                               0, "US-ASCII (7bit)" ],
  97     [ 20866, "VENDORS/MISC/KOI8-R.TXT",           0, "Russian KOI8" ],
  98     [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", 0, "EUC-JP", 0x3f, 0x30fb  ],
  99     [ 21866, "VENDORS/MISC/KOI8-U.TXT",           0, "Ukrainian KOI8" ],
 100     [ 28591, "ISO8859/8859-1.TXT",                0, "ISO 8859-1 Latin 1" ],
 101     [ 28592, "ISO8859/8859-2.TXT",                0, "ISO 8859-2 Latin 2 (East European)" ],
 102     [ 28593, "ISO8859/8859-3.TXT",                0, "ISO 8859-3 Latin 3 (South European)" ],
 103     [ 28594, "ISO8859/8859-4.TXT",                0, "ISO 8859-4 Latin 4 (Baltic old)" ],
 104     [ 28595, "ISO8859/8859-5.TXT",                0, "ISO 8859-5 Cyrillic" ],
 105     [ 28596, "ISO8859/8859-6.TXT",                0, "ISO 8859-6 Arabic" ],
 106     [ 28597, "ISO8859/8859-7.TXT",                0, "ISO 8859-7 Greek" ],
 107     [ 28598, "ISO8859/8859-8.TXT",                0, "ISO 8859-8 Hebrew" ],
 108     [ 28599, "ISO8859/8859-9.TXT",                0, "ISO 8859-9 Latin 5 (Turkish)" ],
 109     [ 28600, "ISO8859/8859-10.TXT",               0, "ISO 8859-10 Latin 6 (Nordic)" ],
 110     [ 28603, "ISO8859/8859-13.TXT",               0, "ISO 8859-13 Latin 7 (Baltic)" ],
 111     [ 28604, "ISO8859/8859-14.TXT",               0, "ISO 8859-14 Latin 8 (Celtic)" ],
 112     [ 28605, "ISO8859/8859-15.TXT",               0, "ISO 8859-15 Latin 9 (Euro)" ],
 113     [ 28606, "ISO8859/8859-16.TXT",               0, "ISO 8859-16 Latin 10 (Balkan)" ]
 114 );
 115
 116
 117 my %ctype =
 118 (
 119     "upper"  => 0x0001,
 120     "lower"  => 0x0002,
 121     "digit"  => 0x0004,
 122     "space"  => 0x0008,
 123     "punct"  => 0x0010,
 124     "cntrl"  => 0x0020,
 125     "blank"  => 0x0040,
 126     "xdigit" => 0x0080,
 127     "alpha"  => 0x0100,
 128     "defin"  => 0x0200
 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_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 );
 155
 156 my %matra_types =
 157 (
 158     "Right"    => 0x01,
 159     "Left"  => 0x02,
 160     "Visual_Order_Left" => 0x03,
 161     "Left_And_Right"    => 0x04,
 162     "Top"   => 0x05,
 163     "Bottom"  => 0x06,
 164     "Top_And_Bottom"  => 0x07,
 165     "Top_And_Right"  => 0x08,
 166     "Top_And_Left"  => 0x09,
 167     "Top_And_Left_And_Right"  => 0x0a,
 168     "Bottom_And_Right"  => 0x0b,
 169     "Top_And_Bottom_And_Right"  => 0x0c,
 170     "Overstruck"  => 0x0d,
 171     "Invisible"  => 0x0e
 172 );
 173
 174 my %nameprep_flags =
 175 (
 176     "unassigned" => 0x01,
 177     "prohibited" => 0x02,
 178     "bidi_ral"   => 0x04,
 179     "bidi_l"     => 0x08
 180 );
 181
 182 my %break_types =
 183 (
 184     "BK"  => 0x0001,
 185     "CR"  => 0x0002,
 186     "LF"  => 0x0003,
 187     "CM"  => 0x0004,
 188     "SG"  => 0x0005,
 189     "GL"  => 0x0006,
 190     "CB"  => 0x0007,
 191     "SP"  => 0x0008,
 192     "ZW"  => 0x0009,
 193     "NL"  => 0x000a,
 194     "WJ"  => 0x000b,
 195     "JL"  => 0x000c,
 196     "JV"  => 0x000d,
 197     "JT"  => 0x000e,
 198     "H2"  => 0x000f,
 199     "H3"  => 0x0010,
 200     "XX"  => 0x0011,
 201     "OP"  => 0x0012,
 202     "CL"  => 0x0013,
 203     "CP"  => 0x0014,
 204     "QU"  => 0x0015,
 205     "NS"  => 0x0016,
 206     "EX"  => 0x0017,
 207     "SY"  => 0x0018,
 208     "IS"  => 0x0019,
 209     "PR"  => 0x001a,
 210     "PO"  => 0x001b,
 211     "NU"  => 0x001c,
 212     "AL"  => 0x001d,
 213     "ID"  => 0x001e,
 214     "IN"  => 0x001f,
 215     "HY"  => 0x0020,
 216     "BB"  => 0x0021,
 217     "BA"  => 0x0022,
 218     "SA"  => 0x0023,
 219     "AI"  => 0x0024,
 220     "B2"  => 0x0025,
 221     "HL"  => 0x0026,
 222     "CJ"  => 0x0027,
 223     "RI"  => 0x0028,
 224 );
 225
 226 my %categories =
 227 (
 228     "Lu" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase
 229     "Ll" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase
 230     "Lt" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}|$ctype{"lower"},    # Letter, Titlecase
 231     "Mn" => $ctype{"defin"},                    # Mark, Non-Spacing
 232     "Mc" => $ctype{"defin"},                    # Mark, Spacing Combining
 233     "Me" => $ctype{"defin"},                    # Mark, Enclosing
 234     "Nd" => $ctype{"defin"}|$ctype{"digit"},    # Number, Decimal Digit
 235     "Nl" => $ctype{"defin"}|$ctype{"alpha"},    # Number, Letter
 236     "No" => $ctype{"defin"},                    # Number, Other
 237     "Zs" => $ctype{"defin"}|$ctype{"space"},    # Separator, Space
 238     "Zl" => $ctype{"defin"}|$ctype{"space"},    # Separator, Line
 239     "Zp" => $ctype{"defin"}|$ctype{"space"},    # Separator, Paragraph
 240     "Cc" => $ctype{"defin"}|$ctype{"cntrl"},    # Other, Control
 241     "Cf" => $ctype{"defin"}|$ctype{"cntrl"},    # Other, Format
 242     "Cs" => $ctype{"defin"},                    # Other, Surrogate
 243     "Co" => $ctype{"defin"},                    # Other, Private Use
 244     "Cn" => $ctype{"defin"},                    # Other, Not Assigned
 245     "Lm" => $ctype{"defin"}|$ctype{"alpha"},    # Letter, Modifier
 246     "Lo" => $ctype{"defin"}|$ctype{"alpha"},    # Letter, Other
 247     "Pc" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Connector
 248     "Pd" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Dash
 249     "Ps" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Open
 250     "Pe" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Close
 251     "Pi" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Initial quote
 252     "Pf" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Final quote
 253     "Po" => $ctype{"defin"}|$ctype{"punct"},    # Punctuation, Other
 254     "Sm" => $ctype{"defin"},                    # Symbol, Math
 255     "Sc" => $ctype{"defin"},                    # Symbol, Currency
 256     "Sk" => $ctype{"defin"},                    # Symbol, Modifier
 257     "So" => $ctype{"defin"}                     # Symbol, Other
 258 );
 259
 260 # a few characters need additional categories that cannot be determined automatically
 261 my %special_categories =
 262 (
 263     "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'),
 264                   0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ],
 265     "space"  => [ 0x09..0x0d, 0x85 ],
 266     "blank"  => [ 0x09, 0x20, 0xa0, 0x3000, 0xfeff ],
 267     "cntrl"  => [ 0x070f, 0x200c, 0x200d,
 268                   0x200e, 0x200f, 0x202a, 0x202b, 0x202c, 0x202d, 0x202e,
 269                   0x206a, 0x206b, 0x206c, 0x206d, 0x206e, 0x206f, 0xfeff,
 270                   0xfff9, 0xfffa, 0xfffb ],
 271     "punct"  => [ 0x24, 0x2b, 0x3c..0x3e, 0x5e, 0x60, 0x7c, 0x7e, 0xa2..0xbe,
 272                   0xd7, 0xf7 ],
 273     "digit"  => [ 0xb2, 0xb3, 0xb9 ],
 274     "lower"  => [ 0x2071, 0x207f ]
 275 );
 276
 277 my %directions =
 278 (
 279     "L"   => 1,    # Left-to-Right
 280     "LRE" => 15,   # Left-to-Right Embedding
 281     "LRO" => 15,   # Left-to-Right Override
 282     "R"   => 2,    # Right-to-Left
 283     "AL"  => 12,   # Right-to-Left Arabic
 284     "RLE" => 15,   # Right-to-Left Embedding
 285     "RLO" => 15,   # Right-to-Left Override
 286     "PDF" => 15,   # Pop Directional Format
 287     "EN"  => 3,    # European Number
 288     "ES"  => 4,    # European Number Separator
 289     "ET"  => 5,    # European Number Terminator
 290     "AN"  => 6,    # Arabic Number
 291     "CS"  => 7,    # Common Number Separator
 292     "NSM" => 13,   # Non-Spacing Mark
 293     "BN"  => 14,   # Boundary Neutral
 294     "B"   => 8,    # Paragraph Separator
 295     "S"   => 9,    # Segment Separator
 296     "WS"  => 10,   # Whitespace
 297     "ON"  => 11    # Other Neutrals
 298 );
 299
 300 my %joining_types =
 301 (
 302    "U" => 0,    # Non_Joining
 303    "T" => 1,    # Transparent
 304    "R" => 2,    # Right_Joining
 305    "L" => 3,    # Left_Joining
 306    "D" => 4,    # Dual_Joining
 307    "C" => 5,    # Join_Causing
 308 );
 309
 310 my @cp2uni = ();
 311 my @lead_bytes = ();
 312 my @uni2cp = ();
 313 my @unicode_defaults = ();
 314 my @unicode_aliases = ();
 315 my @tolower_table = ();
 316 my @toupper_table = ();
 317 my @digitmap_table = ();
 318 my @compatmap_table = ();
 319 my @category_table = (0) x 65536;
 320 my @joining_table = (0) x 65536;
 321 my @direction_table = ();
 322 my @decomp_table = ();
 323 my @compose_table = ();
 324 my $default_char;
 325 my $default_wchar;
 326
 327 my %joining_forms =
 328 (
 329    "isolated" => [],
 330    "final" => [],
 331    "initial" => [],
 332    "medial" => []
 333 );
 334
 335 ################################################################
 336 # fetch a unicode.org file and open it
 337 sub open_data_file($$)
 338 {
 339     my ($base, $name) = @_;
 340     (my $dir = "data/$name") =~ s/\/[^\/]+$//;
 341     local *FILE;
 342     unless (-f "data/$name")
 343     {
 344         print "Fetching $base/$name...\n";
 345         system "mkdir", "-p", $dir;
 346         !system "wget", "-q", "-O", "data/$name", "$base/$name" or die "cannot fetch $base/$name";
 347     }
 348     open FILE, "<data/$name" or die "cannot open data/$name";
 349     return *FILE;
 350 }
 351
 352 ################################################################
 353 # read in the defaults file
 354 sub READ_DEFAULTS($)
 355 {
 356     my $filename = shift;
 357     my $start;
 358
 359     # first setup a few default mappings
 360
 361     open DEFAULTS, "$filename" or die "Cannot open $filename";
 362     print "Loading $filename\n";
 363     while (<DEFAULTS>)
 364     {
 365         next if /^\#/;  # skip comments
 366         next if /^$/;  # skip empty lines
 367         if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/)
 368         {
 369             my @src = map hex, split /,/,$1;
 370             my $dst = $4;
 371             my $comment = $5;
 372             if ($#src > 0) { push @unicode_aliases, \@src; }
 373             next if ($dst eq "none");
 374             $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst;
 375             foreach my $src (@src)
 376             {
 377                 die "Duplicate value" if defined($unicode_defaults[$src]);
 378                 $unicode_defaults[$src] = $dst;
 379             }
 380             next;
 381         }
 382         die "Unrecognized line $_\n";
 383     }
 384     close DEFAULTS;
 385
 386     # now build mappings from the decomposition field of the Unicode database
 387
 388     my $UNICODE_DATA = open_data_file( $UNIDATA, "UnicodeData.txt" );
 389     while (<$UNICODE_DATA>)
 390     {
 391         # Decode the fields ...
 392         my ($code, $name, $cat, $comb, $bidi,
 393             $decomp, $dec, $dig, $num, $mirror,
 394             $oldname, $comment, $upper, $lower, $title) = split /;/;
 395         my $dst;
 396         my $src = hex $code;
 397
 398         die "unknown category $cat" unless defined $categories{$cat};
 399         die "unknown directionality $bidi" unless defined $directions{$bidi};
 400
 401         $category_table[$src] = $categories{$cat};
 402         $direction_table[$src] = $directions{$bidi};
 403         $joining_table[$src] = $joining_types{"T"} if $cat eq "Mn" || $cat eq "Me" || $cat eq "Cf";
 404
 405         if ($lower ne "")
 406         {
 407             $tolower_table[$src] = hex $lower;
 408         }
 409         if ($upper ne "")
 410         {
 411             $toupper_table[$src] = hex $upper;
 412         }
 413         if ($dec ne "")
 414         {
 415             $category_table[$src] |= $ctype{"digit"};
 416         }
 417         if ($dig ne "")
 418         {
 419             $digitmap_table[$src] = ord $dig;
 420         }
 421
 422         # copy the category and direction for everything between First/Last pairs
 423         if ($name =~ /, First>/) { $start = $src; }
 424         if ($name =~ /, Last>/)
 425         {
 426             while ($start < $src)
 427             {
 428                 $category_table[$start] = $category_table[$src];
 429                 $direction_table[$start] = $direction_table[$src];
 430                 $start++;
 431             }
 432         }
 433
 434         next if $decomp eq "";  # no decomposition, skip it
 435
 436         if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/)
 437         {
 438             # decomposition of the form "<foo> 1234" -> use char if type is known
 439             if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd))
 440             {
 441                 # Single char decomposition in the compatibility range
 442                 $compatmap_table[$src] = hex $2;
 443             }
 444             if ($1 eq "isolated" || $1 eq "final" || $1 eq "initial" || $1 eq "medial")
 445             {
 446                 ${joining_forms{$1}}[hex $2] = $src;
 447                 next;
 448             }
 449             next unless ($1 eq "font" ||
 450                          $1 eq "noBreak" ||
 451                          $1 eq "circle" ||
 452                          $1 eq "super" ||
 453                          $1 eq "sub" ||
 454                          $1 eq "wide" ||
 455                          $1 eq "narrow" ||
 456                          $1 eq "compat" ||
 457                          $1 eq "small");
 458             $dst = hex $2;
 459         }
 460         elsif ($decomp =~ /^<compat>\s+0020\s+([0-9a-fA-F]+)/)
 461         {
 462             # decomposition "<compat> 0020 1234" -> combining accent
 463             $dst = hex $1;
 464         }
 465         elsif ($decomp =~ /^([0-9a-fA-F]+)/)
 466         {
 467             # decomposition contains only char values without prefix -> use first char
 468             $dst = hex $1;
 469             $category_table[$src] |= $category_table[$dst] if defined $category_table[$dst];
 470             # store decomposition if it contains two chars
 471             if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/)
 472             {
 473                 $decomp_table[$src] = [ hex $1, hex $2 ];
 474                 push @compose_table, [ hex $1, hex $2, $src ];
 475             }
 476             elsif ($decomp =~ /^(<[a-z]+>\s)*([0-9a-fA-F]+)$/ &&
 477                    (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)))
 478             {
 479                 # Single char decomposition in the compatibility range
 480                 $compatmap_table[$src] = hex $2;
 481             }
 482         }
 483         else
 484         {
 485             next;
 486         }
 487
 488         next if defined($unicode_defaults[$src]);  # may have been set in the defaults file
 489
 490         # check for loops
 491         for (my $i = $dst; ; $i = $unicode_defaults[$i])
 492         {
 493             die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src;
 494             last unless defined($unicode_defaults[$i]);
 495         }
 496         $unicode_defaults[$src] = $dst;
 497     }
 498     close $UNICODE_DATA;
 499
 500     # patch the category of some special characters
 501
 502     foreach my $cat (keys %special_categories)
 503     {
 504         my $flag = $ctype{$cat};
 505         foreach my $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; }
 506     }
 507 }
 508
 509
 510 ################################################################
 511 # define a new lead byte
 512 sub add_lead_byte($)
 513 {
 514     my $ch = shift;
 515     push @lead_bytes, $ch;
 516     $cp2uni[$ch] = 0;
 517 }
 518
 519 ################################################################
 520 # define a new char mapping
 521 sub add_mapping($$)
 522 {
 523     my ($cp, $uni) = @_;
 524     $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 525     $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 526     if ($cp > 0xff && !defined($cp2uni[$cp >> 8]))
 527     {
 528         add_lead_byte( $cp >> 8 );
 529     }
 530 }
 531
 532 ################################################################
 533 # parse a standard codepage file
 534 sub read_codepage_file($)
 535 {
 536     my $name = shift;
 537     my $INPUT = open_data_file( $MAPPINGS, $name );
 538
 539     while (<$INPUT>)
 540     {
 541         next if /^\#/;  # skip comments
 542         next if /^$/;  # skip empty lines
 543         next if /\x1a/;  # skip ^Z
 544         next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/);  # undefined char
 545
 546         if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/)
 547         {
 548             add_lead_byte( hex $1 );
 549             next;
 550         }
 551         # 0x12 0x3456
 552         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 553         {
 554             add_mapping( hex $1, hex $2 );
 555             next;
 556         }
 557         # 0x12 0x3456+0x7890
 558         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\+.*\s+(\#.*)?/)
 559         {
 560             add_mapping( hex $1, hex $2 );
 561             next;
 562         }
 563         # 0x12 <LR>+0x3456
 564         if (/^0x([0-9a-fA-F]+)\s+<[LR]+>\+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 565         {
 566             add_mapping( hex $1, hex $2 );
 567             next;
 568         }
 569         die "$name: Unrecognized line $_\n";
 570     }
 571     close $INPUT;
 572 }
 573
 574
 575 ################################################################
 576 # fill input data for the 20127 (us-ascii) codepage
 577 sub fill_20127_codepage()
 578 {
 579     for (my $i = 0; $i < 128; $i++) { $cp2uni[$i] = $uni2cp[$i] = $i; }
 580     for (my $i = 128; $i < 256; $i++) { $cp2uni[$i] = $i & 0x7f; }
 581 }
 582
 583 ################################################################
 584 # get a mapping including glyph chars for MB_USEGLYPHCHARS
 585
 586 sub get_glyphs_mapping(@)
 587 {
 588     $_[0x01] = 0x263a;  # (WHITE SMILING FACE)
 589     $_[0x02] = 0x263b;  # (BLACK SMILING FACE)
 590     $_[0x03] = 0x2665;  # (BLACK HEART SUIT)
 591     $_[0x04] = 0x2666;  # (BLACK DIAMOND SUIT)
 592     $_[0x05] = 0x2663;  # (BLACK CLUB SUIT)
 593     $_[0x06] = 0x2660;  # (BLACK SPADE SUIT)
 594     $_[0x07] = 0x2022;  # (BULLET)
 595     $_[0x08] = 0x25d8;  # (INVERSE BULLET)
 596     $_[0x09] = 0x25cb;  # (WHITE CIRCLE)
 597     $_[0x0a] = 0x25d9;  # (INVERSE WHITE CIRCLE)
 598     $_[0x0b] = 0x2642;  # (MALE SIGN)
 599     $_[0x0c] = 0x2640;  # (FEMALE SIGN)
 600     $_[0x0d] = 0x266a;  # (EIGHTH NOTE)
 601     $_[0x0e] = 0x266b;  # (BEAMED EIGHTH NOTES)
 602     $_[0x0f] = 0x263c;  # (WHITE SUN WITH RAYS)
 603     $_[0x10] = 0x25ba;  # (BLACK RIGHT-POINTING POINTER)
 604     $_[0x11] = 0x25c4;  # (BLACK LEFT-POINTING POINTER)
 605     $_[0x12] = 0x2195;  # (UP DOWN ARROW)
 606     $_[0x13] = 0x203c;  # (DOUBLE EXCLAMATION MARK)
 607     $_[0x14] = 0x00b6;  # (PILCROW SIGN)
 608     $_[0x15] = 0x00a7;  # (SECTION SIGN)
 609     $_[0x16] = 0x25ac;  # (BLACK RECTANGLE)
 610     $_[0x17] = 0x21a8;  # (UP DOWN ARROW WITH BASE)
 611     $_[0x18] = 0x2191;  # (UPWARDS ARROW)
 612     $_[0x19] = 0x2193;  # (DOWNWARDS ARROW)
 613     $_[0x1a] = 0x2192;  # (RIGHTWARDS ARROW)
 614     $_[0x1b] = 0x2190;  # (LEFTWARDS ARROW)
 615     $_[0x1c] = 0x221f;  # (RIGHT ANGLE)
 616     $_[0x1d] = 0x2194;  # (LEFT RIGHT ARROW)
 617     $_[0x1e] = 0x25b2;  # (BLACK UP-POINTING TRIANGLE)
 618     $_[0x1f] = 0x25bc;  # (BLACK DOWN-POINTING TRIANGLE)
 619     $_[0x7f] = 0x2302;  # (HOUSE)
 620     return @_;
 621 }
 622
 623 ################################################################
 624 # build EUC-JP table from the JIS 0208 file
 625 # FIXME: for proper EUC-JP we should probably read JIS 0212 too
 626 # but this would require 3-byte DBCS characters
 627 sub READ_JIS0208_FILE($)
 628 {
 629     my $name = shift;
 630
 631     # ASCII chars
 632     for (my $i = 0x00; $i <= 0x7f; $i++)
 633     {
 634         $cp2uni[$i] = $i;
 635         $uni2cp[$i] = $i;
 636     }
 637
 638     # JIS X 0201 right plane
 639     for (my $i = 0xa1; $i <= 0xdf; $i++)
 640     {
 641         $cp2uni[0x8e00 + $i] = 0xfec0 + $i;
 642         $uni2cp[0xfec0 + $i] = 0x8e00 + $i;
 643     }
 644
 645     # lead bytes
 646     foreach my $i (0x8e, 0x8f, 0xa1 .. 0xfe)
 647     {
 648         push @lead_bytes,$i;
 649         $cp2uni[$i] = 0;
 650     }
 651
 652     # undefined chars
 653     foreach my $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff)
 654     {
 655         $cp2uni[$i] = $DEF_CHAR;
 656     }
 657
 658     # Shift-JIS compatibility
 659     $uni2cp[0x00a5] = 0x5c;
 660     $uni2cp[0x203e] = 0x7e;
 661
 662     # Fix backslash conversion
 663     $cp2uni[0xa1c0] = 0xff3c;
 664     $uni2cp[0xff3c] = 0xa1c0;
 665
 666     my $INPUT = open_data_file( $MAPPINGS, $name );
 667     while (<$INPUT>)
 668     {
 669         next if /^\#/;  # skip comments
 670         next if /^$/;  # skip empty lines
 671         next if /\x1a/;  # skip ^Z
 672         if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/)
 673         {
 674             my $cp = 0x8080 + hex $1;
 675             my $uni = hex $2;
 676             $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
 677             $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
 678             next;
 679         }
 680         die "$name: Unrecognized line $_\n";
 681     }
 682     close $INPUT;
 683 }
 684
 685
 686 ################################################################
 687 # build the sort keys table
 688 sub READ_SORTKEYS_FILE()
 689 {
 690     my @sortkeys = ();
 691     for (my $i = 0; $i < 65536; $i++) { $sortkeys[$i] = [ -1, 0, 0, 0, 0 ] };
 692
 693     my $INPUT = open_data_file( $REPORTS, $SORTKEYS );
 694     while (<$INPUT>)
 695     {
 696         next if /^\#/;  # skip comments
 697         next if /^$/;  # skip empty lines
 698         next if /\x1a/;  # skip ^Z
 699         next if /^\@version/;  # skip @version header
 700         if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 701         {
 702             my ($uni,$variable) = (hex $1, $2);
 703             next if $uni > 65535;
 704             $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ];
 705             next;
 706         }
 707         if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/)
 708         {
 709             # multiple character sequence, ignored for now
 710             next;
 711         }
 712         die "$SORTKEYS: Unrecognized line $_\n";
 713     }
 714     close $INPUT;
 715
 716     # compress the keys to 32 bit:
 717     # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
 718
 719     @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or
 720                        ${$a}[2] <=> ${$b}[2] or
 721                        ${$a}[3] <=> ${$b}[3] or
 722                        ${$a}[4] <=> ${$b}[4] or
 723                        $a cmp $b; } @sortkeys;
 724
 725     my ($n2, $n3) = (1, 1);
 726     my @keys = (-1, -1, -1, -1, -1 );
 727     my @flatkeys = ();
 728
 729     for (my $i = 0; $i < 65536; $i++)
 730     {
 731         my @current = @{$sortkeys[$i]};
 732         next if $current[0] == -1;
 733         if ($current[1] == $keys[1])
 734         {
 735             if ($current[2] == $keys[2])
 736             {
 737                 if ($current[3] == $keys[3])
 738                 {
 739                     # nothing
 740                 }
 741                 else
 742                 {
 743                     $keys[3] = $current[3];
 744                     $n3++;
 745                     die if ($n3 >= 16);
 746                 }
 747             }
 748             else
 749             {
 750                 $keys[2] = $current[2];
 751                 $keys[3] = $current[3];
 752                 $n2++;
 753                 $n3 = 1;
 754                 die if ($n2 >= 256);
 755             }
 756         }
 757         else
 758         {
 759             $keys[1] = $current[1];
 760             $keys[2] = $current[2];
 761             $keys[3] = $current[3];
 762             $n2 = 1;
 763             $n3 = 1;
 764         }
 765
 766         if ($current[2]) { $current[2] = $n2; }
 767         if ($current[3]) { $current[3] = $n3; }
 768         if ($current[4]) { $current[4] = 1; }
 769
 770         $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4];
 771     }
 772     return @flatkeys;
 773 }
 774
 775
 776 ################################################################
 777 # build the sort keys table
 778 sub DUMP_SORTKEYS($@)
 779 {
 780     my ($filename, @keys) = @_;
 781
 782     # count the number of 256-key ranges that contain something
 783
 784     my @offsets = ();
 785     my $ranges = 2;
 786     for (my $i = 0; $i < 256; $i++) { $offsets[$i] = 256; }
 787     for (my $i = 0; $i < 65536; $i++)
 788     {
 789         next unless defined $keys[$i];
 790         $offsets[$i >> 8] = $ranges * 256;
 791         $ranges++;
 792         $i |= 255;
 793     }
 794
 795     # output the range offsets
 796
 797     open OUTPUT,">$filename.new" or die "Cannot create $filename";
 798     printf "Building $filename\n";
 799     printf OUTPUT "/* Unicode collation element table */\n";
 800     printf OUTPUT "/* generated from %s */\n", "$REPORTS/$SORTKEYS";
 801     printf OUTPUT "/* DO NOT EDIT!! */\n\n";
 802
 803     printf OUTPUT "const unsigned int collation_table[%d] =\n{\n", $ranges*256;
 804     printf OUTPUT "    /* index */\n";
 805     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%08x", 0, @offsets );
 806
 807     # output the default values
 808
 809     printf OUTPUT "    /* defaults */\n";
 810     printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0, (0xffffffff) x 256 );
 811
 812     # output all the key ranges
 813
 814     for (my $i = 0; $i < 256; $i++)
 815     {
 816         next if $offsets[$i] == 256;
 817         printf OUTPUT ",\n    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 818         printf OUTPUT "%s", DUMP_ARRAY( "0x%08x", 0xffffffff, @keys[($i<<8) .. ($i<<8)+255] );
 819     }
 820     printf OUTPUT "\n};\n";
 821     close OUTPUT;
 822     save_file($filename);
 823 }
 824
 825
 826 ################################################################
 827 # add default mappings once the file had been read
 828 sub ADD_DEFAULT_MAPPINGS()
 829 {
 830     # Apply aliases
 831
 832     foreach my $alias (@unicode_aliases)
 833     {
 834         my $target = undef;
 835         foreach my $src (@$alias)
 836         {
 837             if (defined($uni2cp[$src]))
 838             {
 839                 $target = $uni2cp[$src];
 840                 last;
 841             }
 842         }
 843         next unless defined($target);
 844
 845         # At least one char of the alias set is defined, set the others to the same value
 846         foreach my $src (@$alias)
 847         {
 848             $uni2cp[$src] = $target unless defined($uni2cp[$src]);
 849         }
 850     }
 851
 852     # For every src -> target mapping in the defaults table,
 853     # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined
 854
 855     for (my $src = 0; $src < 65536; $src++)
 856     {
 857         next if defined($uni2cp[$src]);  # source has a definition already
 858         next unless defined($unicode_defaults[$src]);  # no default for this char
 859         my $target = $unicode_defaults[$src];
 860
 861         # do a recursive mapping until we find a target char that is defined
 862         while (!defined($uni2cp[$target]) &&
 863                defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; }
 864
 865         if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; }
 866     }
 867
 868     # Add an identity mapping for all undefined chars
 869
 870     for (my $i = 0; $i < 256; $i++)
 871     {
 872         next if defined($cp2uni[$i]);
 873         next if defined($uni2cp[$i]);
 874         $cp2uni[$i] = $uni2cp[$i] = $i;
 875     }
 876 }
 877
 878 ################################################################
 879 # dump an array of integers
 880 sub DUMP_ARRAY($$@)
 881 {
 882     my ($format,$default,@array) = @_;
 883     my $i;
 884     my $ret = "    ";
 885     for ($i = 0; $i < $#array; $i++)
 886     {
 887         $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 888         $ret .= (($i % 8) != 7) ? ", " : ",\n    ";
 889     }
 890     $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default);
 891     return $ret;
 892 }
 893
 894 ################################################################
 895 # dump an SBCS mapping table
 896 sub dump_sbcs_table($$$$$)
 897 {
 898     my ($codepage, $has_glyphs, $name, $def, $defw) = @_;
 899     my $i;
 900
 901     # output the ascii->unicode table
 902
 903     if ($has_glyphs)
 904     {
 905         printf OUTPUT "static const WCHAR cp2uni[512] =\n";
 906         printf OUTPUT "{\n%s", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
 907         printf OUTPUT ",\n    /* glyphs */\n%s\n};\n\n",
 908                       DUMP_ARRAY( "0x%04x", $defw, get_glyphs_mapping(@cp2uni[0 .. 255]) );
 909     }
 910     else
 911     {
 912         printf OUTPUT "static const WCHAR cp2uni[256] =\n";
 913         printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
 914     }
 915
 916     # count the number of unicode->ascii subtables that contain something
 917
 918     my @filled = ();
 919     my $subtables = 1;
 920     for (my $i = 0; $i < 65536; $i++)
 921     {
 922         next unless defined $uni2cp[$i];
 923         $filled[$i >> 8] = 1;
 924         $subtables++;
 925         $i |= 255;
 926     }
 927
 928     # output all the subtables into a single array
 929
 930     printf OUTPUT "static const unsigned char uni2cp_low[%d] =\n{\n", $subtables*256;
 931     for (my $i = 0; $i < 256; $i++)
 932     {
 933         next unless $filled[$i];
 934         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $i, $i;
 935         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%02x", $def, @uni2cp[($i<<8) .. ($i<<8)+255] );
 936     }
 937     printf OUTPUT "    /* defaults */\n";
 938     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, ($def) x 256 );
 939
 940     # output a table of the offsets of the subtables in the previous array
 941
 942     my $pos = 0;
 943     my @offsets = ();
 944     for (my $i = 0; $i < 256; $i++)
 945     {
 946         if ($filled[$i]) { push @offsets, $pos; $pos += 256; }
 947         else { push @offsets, ($subtables-1) * 256; }
 948     }
 949     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
 950     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
 951
 952     # output the code page descriptor
 953
 954     printf OUTPUT "const struct sbcs_table cptable_%03d =\n{\n", $codepage;
 955     printf OUTPUT "    { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n",
 956                   $codepage, $def, $defw, $name;
 957     printf OUTPUT "    cp2uni,\n";
 958     if ($has_glyphs) { printf OUTPUT "    cp2uni + 256,\n"; }
 959     else { printf OUTPUT "    cp2uni,\n"; }
 960     printf OUTPUT "    uni2cp_low,\n";
 961     printf OUTPUT "    uni2cp_high\n};\n";
 962 }
 963
 964
 965 ################################################################
 966 # dump a DBCS mapping table
 967 sub dump_dbcs_table($$$$@)
 968 {
 969     my ($codepage, $name, $def, $defw, @lb_ranges) = @_;
 970
 971     # build a list of lead bytes that are actually used
 972
 973     my @lblist = ();
 974     LBLOOP: for (my $y = 0; $y <= $#lead_bytes; $y++)
 975     {
 976         my $base = $lead_bytes[$y] << 8;
 977         for (my $x = 0; $x < 256; $x++)
 978         {
 979             if (defined $cp2uni[$base+$x])
 980             {
 981                 push @lblist,$lead_bytes[$y];
 982                 next LBLOOP;
 983             }
 984         }
 985     }
 986     my $unused = ($#lead_bytes > $#lblist);
 987
 988     # output the ascii->unicode table for the single byte chars
 989
 990     printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused);
 991     printf OUTPUT "{\n%s,\n", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[0 .. 255] );
 992
 993     # output the default table for unused lead bytes
 994
 995     if ($unused)
 996     {
 997         printf OUTPUT "    /* unused lead bytes */\n";
 998         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 0, ($defw) x 256 );
 999     }
1000
1001     # output the ascii->unicode table for each DBCS lead byte
1002
1003     for (my $y = 0; $y <= $#lblist; $y++)
1004     {
1005         my $base = $lblist[$y] << 8;
1006         printf OUTPUT "    /* lead byte %02x */\n", $lblist[$y];
1007         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", $defw, @cp2uni[$base .. $base+255] );
1008         printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n";
1009     }
1010
1011     # output the lead byte subtables offsets
1012
1013     my @offsets = ();
1014     for (my $x = 0; $x < 256; $x++) { $offsets[$x] = 0; }
1015     for (my $x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; }
1016     if ($unused)
1017     {
1018         # increment all lead bytes offset to take into account the unused table
1019         for (my $x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; }
1020     }
1021     printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n";
1022     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%02x", 0, @offsets );
1023
1024     # count the number of unicode->ascii subtables that contain something
1025
1026     my @filled = ();
1027     my $subtables = 1;
1028     for (my $i = 0; $i < 65536; $i++)
1029     {
1030         next unless defined $uni2cp[$i];
1031         $filled[$i >> 8] = 1;
1032         $subtables++;
1033         $i |= 255;
1034     }
1035
1036     # output all the subtables into a single array
1037
1038     printf OUTPUT "static const unsigned short uni2cp_low[%d] =\n{\n", $subtables*256;
1039     for (my $y = 0; $y < 256; $y++)
1040     {
1041         next unless $filled[$y];
1042         printf OUTPUT "    /* 0x%02x00 .. 0x%02xff */\n", $y, $y;
1043         printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", $def, @uni2cp[($y<<8) .. ($y<<8)+255] );
1044     }
1045     printf OUTPUT "    /* defaults */\n";
1046     printf OUTPUT "%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, ($def) x 256 );
1047
1048     # output a table of the offsets of the subtables in the previous array
1049
1050     my $pos = 0;
1051     @offsets = ();
1052     for (my $y = 0; $y < 256; $y++)
1053     {
1054         if ($filled[$y]) { push @offsets, $pos; $pos += 256; }
1055         else { push @offsets, ($subtables-1) * 256; }
1056     }
1057     printf OUTPUT "static const unsigned short uni2cp_high[256] =\n";
1058     printf OUTPUT "{\n%s\n};\n\n", DUMP_ARRAY( "0x%04x", 0, @offsets );
1059
1060     # output the code page descriptor
1061
1062     printf OUTPUT "const struct dbcs_table cptable_%03d =\n{\n", $codepage;
1063     printf OUTPUT "    { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n",
1064                   $codepage, $def, $defw, $name;
1065     printf OUTPUT "    cp2uni,\n";
1066     printf OUTPUT "    cp2uni_leadbytes,\n";
1067     printf OUTPUT "    uni2cp_low,\n";
1068     printf OUTPUT "    uni2cp_high,\n";
1069     printf OUTPUT "    {\n    %s\n    }\n", DUMP_ARRAY( "0x%02x", 0, @lb_ranges, 0, 0 );
1070     printf OUTPUT "};\n";
1071 }
1072
1073
1074 ################################################################
1075 # get the list of defined lead byte ranges
1076 sub get_lb_ranges()
1077 {
1078     my @list = ();
1079     my @ranges = ();
1080     my $i = 0;
1081     foreach $i (@lead_bytes) { $list[$i] = 1; }
1082     my $on = 0;
1083     for (my $i = 0; $i < 256; $i++)
1084     {
1085         if ($on)
1086         {
1087             if (!defined $list[$i]) { push @ranges, $i-1; $on = 0; }
1088         }
1089         else
1090         {
1091             if ($list[$i]) { push @ranges, $i; $on = 1; }
1092         }
1093     }
1094     if ($on) { push @ranges, 0xff; }
1095     return @ranges;
1096 }
1097
1098 ################################################################
1099 # dump the Indic Syllabic Category table
1100 sub dump_indic($)
1101 {
1102     my $filename = shift;
1103     my @indic_table = ($indic_types{'Other'}) x 65536;;
1104
1105     my $INPUT = open_data_file( $UNIDATA, "IndicSyllabicCategory.txt" );
1106     while (<$INPUT>)
1107     {
1108         next if /^\#/;  # skip comments
1109         next if /^\s*$/;  # skip empty lines
1110         next if /\x1a/;  # skip ^Z
1111         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*#/)
1112         {
1113             my $type = $2;
1114             die "unknown indic $type" unless defined $indic_types{$type};
1115             if (hex $1 < 65536)
1116             {
1117                 $indic_table[hex $1] = $indic_types{$type};
1118             }
1119             next;
1120         }
1121         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1122         {
1123             my $type = $3;
1124             die "unknown indic $type" unless defined $indic_types{$type};
1125             if (hex $1 < 65536 and hex $2 < 6536)
1126             {
1127                 foreach my $i (hex $1 .. hex $2)
1128                 {
1129                     $indic_table[$i] = $indic_types{$type};
1130                 }
1131             }
1132             next;
1133         }
1134         die "malformed line $_";
1135     }
1136     close $INPUT;
1137
1138     $INPUT = open_data_file( $UNIDATA, "IndicMatraCategory.txt" );
1139     while (<$INPUT>)
1140     {
1141         next if /^\#/;  # skip comments
1142         next if /^\s*$/;  # skip empty lines
1143         next if /\x1a/;  # skip ^Z
1144         if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z]+)\s*#/)
1145         {
1146             my $type = $2;
1147             die "unknown matra $type" unless defined $matra_types{$type};
1148             $indic_table[hex $1] += $matra_types{$type} << 8;
1149             next;
1150         }
1151         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/)
1152         {
1153             my $type = $3;
1154             die "unknown matra $type" unless defined $matra_types{$type};
1155             foreach my $i (hex $1 .. hex $2)
1156             {
1157                 $indic_table[$i] += $matra_types{$type} << 8;
1158             }
1159             next;
1160         }
1161         die "malformed line $_";
1162     }
1163     close $INPUT;
1164
1165     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1166     print "Building $filename\n";
1167     print OUTPUT "/* Unicode Indic Syllabic Category */\n";
1168     print OUTPUT "/* generated from $UNIDATA/IndicSyllabicCategory.txt */\n";
1169     print OUTPUT "/*       and from $UNIDATA/IndicMatraCategory.txt */\n";
1170     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1171     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1172
1173     dump_two_level_mapping( "indic_syllabic_table", @indic_table);
1174
1175     close OUTPUT;
1176     save_file($filename);
1177 }
1178
1179 ################################################################
1180 # dump the Line Break Properties table
1181 sub dump_linebreak($)
1182 {
1183     my $filename = shift;
1184     my @break_table = ($break_types{'XX'}) x 65536;;
1185     my $next_group = 0;
1186
1187     my $INPUT = open_data_file( $UNIDATA, "LineBreak.txt" );
1188     while (<$INPUT>)
1189     {
1190         next if /^\#/;  # skip comments
1191         next if /^\s*$/;  # skip empty lines
1192         next if /\x1a/;  # skip ^Z
1193         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1194         {
1195             my $type = $2;
1196             die "unknown breaktype $type" unless defined $break_types{$type};
1197             $break_table[hex $1] = $break_types{$type};
1198             next;
1199         }
1200         elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/)
1201         {
1202             my $type = $3;
1203             die "unknown breaktype $type" unless defined $break_types{$type};
1204             foreach my $i (hex $1 .. hex $2)
1205             {
1206                 $break_table[$i] = $break_types{$type};
1207             }
1208             next;
1209         }
1210         die "malformed line $_";
1211     }
1212     close $INPUT;
1213
1214     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1215     print "Building $filename\n";
1216     print OUTPUT "/* Unicode Line Break Properties */\n";
1217     print OUTPUT "/* generated from $UNIDATA/LineBreak.txt */\n";
1218     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1219     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1220
1221     dump_two_level_mapping( "wine_linebreak_table", @break_table);
1222
1223     close OUTPUT;
1224     save_file($filename);
1225 }
1226
1227
1228 ################################################################
1229 # dump the BiDi mirroring table
1230 sub dump_mirroring($)
1231 {
1232     my $filename = shift;
1233     my @mirror_table = ();
1234
1235     my $INPUT = open_data_file( $UNIDATA, "BidiMirroring.txt" );
1236     while (<$INPUT>)
1237     {
1238         next if /^\#/;  # skip comments
1239         next if /^$/;  # skip empty lines
1240         next if /\x1a/;  # skip ^Z
1241         if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+)/)
1242         {
1243             $mirror_table[hex $1] = hex $2;
1244             next;
1245         }
1246         die "malformed line $_";
1247     }
1248     close $INPUT;
1249
1250     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1251     print "Building $filename\n";
1252     print OUTPUT "/* Unicode BiDi mirroring */\n";
1253     print OUTPUT "/* generated from $UNIDATA/BidiMirroring.txt */\n";
1254     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1255     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1256     DUMP_CASE_TABLE( "wine_mirror_map", @mirror_table );
1257     close OUTPUT;
1258     save_file($filename);
1259 }
1260
1261
1262 ################################################################
1263 # dump the Arabic shaping table
1264 sub dump_shaping($)
1265 {
1266     my $filename = shift;
1267     my %groups;
1268     my $next_group = 0;
1269
1270     $groups{"No_Joining_Group"} = $next_group++;
1271
1272     my $INPUT = open_data_file( $UNIDATA, "ArabicShaping.txt" );
1273     while (<$INPUT>)
1274     {
1275         next if /^\#/;  # skip comments
1276         next if /^\s*$/;  # skip empty lines
1277         next if /\x1a/;  # skip ^Z
1278         if (/^\s*([0-9a-fA-F]+)\s*;.*;\s*([RLDCUT])\s*;\s*(\w+)/)
1279         {
1280             my $type = $2;
1281             my $group = $3;
1282             $groups{$group} = $next_group++ unless defined $groups{$group};
1283             $joining_table[hex $1] = $joining_types{$type} | ($groups{$group} << 8);
1284             next;
1285         }
1286         die "malformed line $_";
1287     }
1288     close $INPUT;
1289
1290     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1291     print "Building $filename\n";
1292     print OUTPUT "/* Unicode Arabic shaping */\n";
1293     print OUTPUT "/* generated from $UNIDATA/ArabicShaping.txt */\n";
1294     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1295     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1296
1297     dump_two_level_mapping( "wine_shaping_table", @joining_table );
1298
1299     print OUTPUT "\nconst unsigned short wine_shaping_forms[256][4] =\n{\n";
1300     for (my $i = 0x600; $i <= 0x6ff; $i++)
1301     {
1302         printf OUTPUT "    { 0x%04x, 0x%04x, 0x%04x, 0x%04x },\n",
1303             ${joining_forms{"isolated"}}[$i] || $i,
1304             ${joining_forms{"final"}}[$i] || $i,
1305             ${joining_forms{"initial"}}[$i] || $i,
1306             ${joining_forms{"medial"}}[$i] || $i;
1307     }
1308     print OUTPUT "};\n";
1309
1310     close OUTPUT;
1311     save_file($filename);
1312 }
1313
1314
1315 ################################################################
1316 # dump the case mapping tables
1317 sub DUMP_CASE_MAPPINGS($)
1318 {
1319     my $filename = shift;
1320     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1321     printf "Building $filename\n";
1322     printf OUTPUT "/* Unicode case mappings */\n";
1323     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1324     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1325
1326     DUMP_CASE_TABLE( "wine_casemap_lower", @tolower_table );
1327     DUMP_CASE_TABLE( "wine_casemap_upper", @toupper_table );
1328     DUMP_CASE_TABLE( "wine_digitmap",  @digitmap_table );
1329     DUMP_CASE_TABLE( "wine_compatmap", @compatmap_table );
1330     close OUTPUT;
1331     save_file($filename);
1332 }
1333
1334
1335 ################################################################
1336 # dump a case mapping table
1337 sub DUMP_CASE_TABLE($@)
1338 {
1339     my ($name,@table) = @_;
1340
1341     # count the number of sub tables that contain something
1342     # also compute the low and upper populated bounds
1343
1344     my @lowerbounds = ( 0, 0 );
1345     my @upperbounds = ( 0, 255 );
1346     my $index = 0;
1347     my @filled = ();
1348     for (my $i = 0; $i < 65536; $i++)
1349     {
1350         next unless defined $table[$i];
1351         if (!defined $filled[$i >> 8])
1352         {
1353           $lowerbounds[$index] = $i & 0xff;
1354           $upperbounds[$index] = 0xff - $lowerbounds[$index];
1355           $filled[$i >> 8] = $index * 256 + 512;
1356           $index++;
1357         }
1358         else
1359         {
1360           $upperbounds[$index-1] = 0xff - ($i & 0xff);
1361         }
1362         $table[$i] = ($table[$i] - $i) & 0xffff;
1363     }
1364
1365     # Collapse blocks upwards if possible
1366     my $removed = 0;
1367     $index = 0;
1368     for (my $i = 0; $i < 256; $i++)
1369     {
1370         next unless defined $filled[$i];
1371         if ($upperbounds[$index - 1] > $lowerbounds[$index])
1372         {
1373            $removed = $removed + $lowerbounds[$index];
1374         }
1375         else
1376         {
1377            $removed = $removed + $upperbounds[$index - 1];
1378            $lowerbounds[$index] = $upperbounds[$index - 1];
1379         }
1380         $filled[$i] = $filled[$i] - $removed;
1381         $index++;
1382     }
1383
1384     # dump the table
1385
1386     printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed;
1387     printf OUTPUT "{\n    /* index */\n";
1388     printf OUTPUT "%s,\n", DUMP_ARRAY( "0x%04x", 256, @filled );
1389     printf OUTPUT "    /* defaults */\n";
1390     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 256 );
1391     $index = 0;
1392     for (my $i = 0; $i < 256; $i++)
1393     {
1394         next unless $filled[$i];
1395         printf OUTPUT ",\n    /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i;
1396         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0,
1397                       @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] );
1398         $index++;
1399     }
1400     printf OUTPUT "\n};\n";
1401 }
1402
1403 ################################################################
1404 # compress a mapping table by removing identical rows
1405 sub compress_array($@)
1406 {
1407     my $rows = shift;
1408     my @table = @_;
1409     my $len = @table / $rows;
1410     my @array = (0) x $rows;
1411     my %sequences;
1412
1413     # try to merge table rows
1414     for (my $row = 0; $row < $rows; $row++)
1415     {
1416         my $rowtxt = pack "S*", @table[($row * $len)..($row * $len + $len - 1)];
1417         if (defined($sequences{$rowtxt}))
1418         {
1419             # reuse an existing row
1420             $array[$row] = $sequences{$rowtxt};
1421         }
1422         else
1423         {
1424             # create a new row
1425             $sequences{$rowtxt} = $array[$row] = $#array + 1;
1426             push @array, @table[$row * $len..$row * $len + $len - 1];
1427         }
1428     }
1429     return @array;
1430 }
1431
1432 ################################################################
1433 # dump a simple char -> 16-bit value mapping table
1434 sub dump_simple_mapping($@)
1435 {
1436     my $name = shift;
1437     my @array = compress_array( 256, @_[0..65535] );
1438
1439     printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, $#array+1;
1440     printf OUTPUT "    /* offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1441     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1442 }
1443
1444 ################################################################
1445 # dump a char -> 16-bit value mapping table using two-level tables
1446 sub dump_two_level_mapping($@)
1447 {
1448     my $name = shift;
1449     my @row_array = compress_array( 4096, @_[0..65535] );
1450     my @array = compress_array( 256, @row_array[0..4095] );
1451
1452     for (my $i = 256; $i < @array; $i++) { $array[$i] += @array - 4096; }
1453
1454     printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, @array + @row_array - 4096;
1455     printf OUTPUT "    /* level 1 offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[0..255] );
1456     printf OUTPUT "    /* level 2 offsets */\n%s,\n", DUMP_ARRAY( "0x%04x", 0, @array[256..$#array] );
1457     printf OUTPUT "    /* values */\n%s\n};\n", DUMP_ARRAY( "0x%04x", 0, @row_array[4096..$#row_array] );
1458 }
1459
1460 ################################################################
1461 # dump a binary case mapping table in l_intl.nls format
1462 sub dump_binary_case_table(@)
1463 {
1464     my (@table) = @_;
1465
1466     my %difftables_hash = ();
1467     my @difftables;
1468     my %offtables2_hash = ();
1469     my @offtables2 = ();
1470
1471     my @offtable = ();
1472     for (my $i = 0; $i < 256; $i++)
1473     {
1474         my @offtable2 = ();
1475         for(my $j = 0; $j < 16; $j++) # offset table for xx00-xxFF characters
1476         {
1477             my @difftable;
1478             for (my $k = 0; $k < 16; $k++) # case map table for xxx0-xxxF characters
1479             {
1480                 my $char = ($i<<8) + ($j<<4) + $k;
1481                 $difftable[$k] = (defined $table[$char]) ? (($table[$char]-$char) & 0xffff) : 0;
1482             }
1483
1484             my $diff_key = pack "S*", @difftable;
1485             my $offset3 = $difftables_hash{$diff_key};
1486             if (!defined $offset3)
1487             {
1488                 $offset3 = scalar @difftables;
1489                 $difftables_hash{$diff_key} = $offset3;
1490                 push @difftables, @difftable;
1491             }
1492             $offtable2[$j] = $offset3;
1493         }
1494
1495         my $offtable2_key = pack "S*", @offtable2;
1496         my $offset2 = $offtables2_hash{$offtable2_key};
1497         if (!defined $offset2)
1498         {
1499             $offset2 = scalar @offtables2;
1500             $offtables2_hash{$offtable2_key} = $offset2;
1501             push @offtables2, \@offtable2;
1502         }
1503         $offtable[$i] = $offset2;
1504     }
1505
1506     my @output;
1507     my $offset = 0x100; # offset of first subtable in words
1508     foreach (@offtable)
1509     {
1510         push @output, 0x10 * $_ + $offset; # offset of subtable in words
1511     }
1512
1513     $offset = 0x100 + 0x10 * scalar @offtables2; # offset of first difftable in words
1514     foreach(@offtables2)
1515     {
1516         my $table = $_;
1517         foreach(@$table)
1518         {
1519             push @output, $_ + $offset; # offset of difftable in words
1520         }
1521     }
1522
1523     my $len = 1 + scalar @output + scalar @difftables;
1524     return pack "S<*", $len, @output, @difftables;
1525 }
1526
1527
1528 ################################################################
1529 # dump case mappings for l_intl.nls
1530 sub dump_intl_nls($)
1531 {
1532     my $filename = shift;
1533     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1534     printf "Building $filename\n";
1535
1536     binmode OUTPUT;
1537     print OUTPUT pack "S<", 1;  # version
1538     print OUTPUT dump_binary_case_table( @toupper_table );
1539     print OUTPUT dump_binary_case_table( @tolower_table );
1540     close OUTPUT;
1541     save_file($filename);
1542 }
1543
1544
1545 sub load_nameprep_range_table($$$)
1546 {
1547     my ($INPUT, $val, $table_ref) = @_;
1548
1549     while (<$INPUT>)
1550     {
1551         if (/^\s*([0-9a-fA-F]+)-([0-9a-fA-F]+)/)
1552         {
1553             my $last = hex $2;
1554             $last = 65535 if($last >= 65536);
1555             foreach my $i (hex $1 .. $last)
1556             {
1557                 $table_ref->[$i] |= $val;
1558             }
1559             next;
1560         }
1561         elsif (/^\s*([0-9a-fA-F]+)/)
1562         {
1563             if (hex $1 < 65536)
1564             {
1565                 $table_ref->[hex $1] |= $val;
1566             }
1567             next;
1568         }
1569
1570         return if (/End\sTable/);
1571     }
1572 }
1573
1574 sub load_nameprep_map_table($$)
1575 {
1576     my ($INPUT, $table_ref) = @_;
1577
1578     while (<$INPUT>)
1579     {
1580         if (/^\s*([0-9a-fA-F]+);\s;/)
1581         {
1582             # special value for map to nothing
1583             $table_ref->[hex $1] = [0xffff, 0xffff, 0xffff];
1584             next;
1585         }
1586         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+);/)
1587         {
1588             $table_ref->[hex $1] = [hex $2, 0, 0];
1589             next;
1590         }
1591         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1592         {
1593             $table_ref->[hex $1] = [hex $2, hex $3, 0];
1594             next;
1595         }
1596         elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/)
1597         {
1598             $table_ref->[hex $1] = [hex $2, hex $3, hex $4];
1599             next;
1600         }
1601
1602         return if (/End\sTable/);
1603     }
1604 }
1605
1606 ################################################################
1607 # dump mapping table, prohibited characters set, unassigned
1608 # characters, bidirectional rules used by nameprep algorithm
1609 sub dump_nameprep($)
1610 {
1611     my $filename = shift;
1612     my @mapping_table = ();
1613     my @flags_table = (0) x 65536;
1614
1615     my $INPUT = open_data_file( $RFCS, $STRINGPREP );
1616     while (<$INPUT>)
1617     {
1618         next unless /Start\sTable/;
1619
1620         load_nameprep_range_table($INPUT, $nameprep_flags{"unassigned"}, \@flags_table) if (/A.1/);
1621         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.1.2/);
1622         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.2.2/);
1623         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.3/);
1624         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.4/);
1625         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.5/);
1626         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.6/);
1627         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.7/);
1628         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.8/);
1629         load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.9/);
1630         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_ral"}, \@flags_table) if (/D.1/);
1631         load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_l"}, \@flags_table) if (/D.2/);
1632
1633         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.1/);
1634         load_nameprep_map_table($INPUT, \@mapping_table) if (/B.2/);
1635     }
1636     close $INPUT;
1637
1638     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1639     print "Building $filename\n";
1640     print OUTPUT "/* Nameprep algorithm related data */\n";
1641     print OUTPUT "/* generated from $RFCS/$STRINGPREP */\n";
1642     print OUTPUT "/* DO NOT EDIT!! */\n\n";
1643     print OUTPUT "#include \"wine/unicode.h\"\n\n";
1644
1645     dump_two_level_mapping( "nameprep_char_type", @flags_table );
1646
1647     ######### mapping table
1648     # first determine all the 16-char subsets that contain something
1649     my @filled = ();
1650     my $pos = 16*3;  # for the null subset
1651     for (my $i = 0; $i < 65536; $i++)
1652     {
1653         next unless defined $mapping_table[$i];
1654         $filled[$i >> 4] = $pos;
1655         $pos += 16*3;
1656         $i |= 15;
1657     }
1658     my $total = $pos;
1659
1660     # now count the 256-char subsets that contain something
1661     my @filled_idx = (256) x 256;
1662     $pos = 256 + 16;
1663     for (my $i = 0; $i < 4096; $i++)
1664     {
1665         next unless $filled[$i];
1666         $filled_idx[$i >> 4] = $pos;
1667         $pos += 16;
1668         $i |= 15;
1669     }
1670     my $null_offset = $pos;
1671     $total += $pos;
1672
1673     # add the index offsets to the subsets positions
1674     for (my $i = 0; $i < 4096; $i++)
1675     {
1676         next unless $filled[$i];
1677         $filled[$i] += $null_offset;
1678     }
1679
1680     # dump the main index
1681     printf OUTPUT "const WCHAR nameprep_mapping[%d] =\n", $total;
1682     printf OUTPUT "{\n    /* index */\n";
1683     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1684     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1685
1686     # dump the second-level indexes
1687     for (my $i = 0; $i < 256; $i++)
1688     {
1689         next unless ($filled_idx[$i] > 256);
1690         my @table = @filled[($i<<4)..($i<<4)+15];
1691         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1692         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1693         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1694     }
1695
1696     # dump the 16-char subsets
1697     printf OUTPUT ",\n    /* null mapping */\n";
1698         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 48 );
1699
1700     for (my $i = 0; $i < 4096; $i++)
1701     {
1702         next unless $filled[$i];
1703         my @table = (0) x 48;
1704         for (my $j = 0; $j < 16; $j++)
1705         {
1706             if (defined $mapping_table[($i<<4) + $j])
1707             {
1708                 $table[3 * $j] = ${$mapping_table[($i << 4) + $j]}[0];
1709                 $table[3 * $j + 1] = ${$mapping_table[($i << 4) + $j]}[1];
1710                 $table[3 * $j + 2] = ${$mapping_table[($i << 4) + $j]}[2];
1711             }
1712         }
1713         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1714         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1715     }
1716
1717     printf OUTPUT "\n};\n";
1718
1719     close OUTPUT;
1720     save_file($filename);
1721 }
1722
1723 ################################################################
1724 # dump the ctype tables
1725 sub DUMP_CTYPE_TABLES($)
1726 {
1727     my $filename = shift;
1728     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1729     printf "Building $filename\n";
1730     printf OUTPUT "/* Unicode ctype tables */\n";
1731     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1732     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1733
1734     # add the direction in the high 4 bits of the category
1735     for (my $i = 0; $i < 65536; $i++)
1736     {
1737         $category_table[$i] |= $direction_table[$i] << 12 if defined $direction_table[$i];
1738     }
1739
1740     dump_simple_mapping( "wine_wctype_table", @category_table );
1741
1742     close OUTPUT;
1743     save_file($filename);
1744 }
1745
1746
1747 ################################################################
1748 # dump the char composition tables
1749 sub DUMP_COMPOSE_TABLES($)
1750 {
1751     my $filename = shift;
1752
1753     open OUTPUT,">$filename.new" or die "Cannot create $filename";
1754     printf "Building $filename\n";
1755     printf OUTPUT "/* Unicode char composition */\n";
1756     printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1757     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1758
1759     ######### composition table
1760
1761     my @filled = ();
1762     foreach my $i (@compose_table)
1763     {
1764         my @comp = @$i;
1765         push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ];
1766     }
1767
1768     # count how many different second chars we have
1769
1770     my $count = 0;
1771     for (my $i = 0; $i < 65536; $i++)
1772     {
1773         next unless defined $filled[$i];
1774         $count++;
1775     }
1776
1777     # build the table of second chars and offsets
1778
1779     my $pos = $count + 1;
1780     my @table = ();
1781     for (my $i = 0; $i < 65536; $i++)
1782     {
1783         next unless defined $filled[$i];
1784         push @table, $i, $pos;
1785         $pos += @{$filled[$i]};
1786     }
1787     # terminator with last position
1788     push @table, 0, $pos;
1789     printf OUTPUT "const WCHAR unicode_compose_table[0x%x] =\n{\n", 2*$pos;
1790     printf OUTPUT "    /* second chars + offsets */\n%s", DUMP_ARRAY( "0x%04x", 0, @table );
1791
1792     # build the table of first chars and mappings
1793
1794     for (my $i = 0; $i < 65536; $i++)
1795     {
1796         next unless defined $filled[$i];
1797         my @table = ();
1798         my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]};
1799         for (my $j = 0; $j <= $#list; $j++)
1800         {
1801             push @table, $list[$j][0], $list[$j][1];
1802         }
1803         printf OUTPUT ",\n    /* 0x%04x */\n%s", $i, DUMP_ARRAY( "0x%04x", 0, @table );
1804     }
1805     printf OUTPUT "\n};\n\nconst unsigned int unicode_compose_table_size = %d;\n\n", $count;
1806
1807     ######### decomposition table
1808
1809     # first determine all the 16-char subsets that contain something
1810
1811     @filled = (0) x 4096;
1812     $pos = 16*2;  # for the null subset
1813     for (my $i = 0; $i < 65536; $i++)
1814     {
1815         next unless defined $decomp_table[$i];
1816         $filled[$i >> 4] = $pos;
1817         $pos += 16*2;
1818         $i |= 15;
1819     }
1820     my $total = $pos;
1821
1822     # now count the 256-char subsets that contain something
1823
1824     my @filled_idx = (256) x 256;
1825     $pos = 256 + 16;
1826     for (my $i = 0; $i < 4096; $i++)
1827     {
1828         next unless $filled[$i];
1829         $filled_idx[$i >> 4] = $pos;
1830         $pos += 16;
1831         $i |= 15;
1832     }
1833     my $null_offset = $pos;  # null mapping
1834     $total += $pos;
1835
1836     # add the index offsets to the subsets positions
1837
1838     for (my $i = 0; $i < 4096; $i++)
1839     {
1840         next unless $filled[$i];
1841         $filled[$i] += $null_offset;
1842     }
1843
1844     # dump the main index
1845
1846     printf OUTPUT "const WCHAR unicode_decompose_table[%d] =\n", $total;
1847     printf OUTPUT "{\n    /* index */\n";
1848     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @filled_idx );
1849     printf OUTPUT ",\n    /* null sub-index */\n%s", DUMP_ARRAY( "0x%04x", 0, ($null_offset) x 16 );
1850
1851     # dump the second-level indexes
1852
1853     for (my $i = 0; $i < 256; $i++)
1854     {
1855         next unless ($filled_idx[$i] > 256);
1856         my @table = @filled[($i<<4)..($i<<4)+15];
1857         for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; }
1858         printf OUTPUT ",\n    /* sub-index %02x */\n", $i;
1859         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1860     }
1861
1862     # dump the 16-char subsets
1863
1864     printf OUTPUT ",\n    /* null mapping */\n";
1865     printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, (0) x 32 );
1866
1867     for (my $i = 0; $i < 4096; $i++)
1868     {
1869         next unless $filled[$i];
1870         my @table = (0) x 32;
1871         for (my $j = 0; $j < 16; $j++)
1872         {
1873             if (defined $decomp_table[($i<<4) + $j])
1874             {
1875                 $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0];
1876                 $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1];
1877             }
1878         }
1879         printf OUTPUT ",\n    /* 0x%03x0 .. 0x%03xf */\n", $i, $i;
1880         printf OUTPUT "%s", DUMP_ARRAY( "0x%04x", 0, @table );
1881     }
1882
1883     printf OUTPUT "\n};\n";
1884     close OUTPUT;
1885     save_file($filename);
1886 }
1887
1888 ################################################################
1889 # output a codepage definition file from the global tables
1890 sub output_codepage_file($$$$)
1891 {
1892     my ($codepage, $filename, $comment, $has_glyphs) = @_;
1893
1894     my $output = sprintf "libs/wine/c_%03d.c", $codepage;
1895     open OUTPUT,">$output.new" or die "Cannot create $output";
1896
1897     printf "Building %s from %s (%s)\n", $output, $filename || "hardcoded data", $comment;
1898
1899     # dump all tables
1900
1901     printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment;
1902     if ($filename)
1903     {
1904         print OUTPUT "/* generated from $MAPPINGS/$filename */\n";
1905         print OUTPUT "/* DO NOT EDIT!! */\n\n";
1906     }
1907     else
1908     {
1909         printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n";
1910     }
1911     printf OUTPUT "#include \"wine/unicode.h\"\n\n";
1912
1913     if (!@lead_bytes) { dump_sbcs_table( $codepage, $has_glyphs, $comment, $default_char, $default_wchar ); }
1914     else { dump_dbcs_table( $codepage, $comment, $default_char, $default_wchar, get_lb_ranges() ); }
1915     close OUTPUT;
1916     save_file($output);
1917 }
1918
1919 ################################################################
1920 # read a "bestfit" Windows mapping file
1921 sub read_bestfit_file($)
1922 {
1923     my ($filename) = @_;
1924     my $state = "";
1925     my ($codepage, $width, $count);
1926     my ($lb_cur, $lb_end);
1927
1928     my $INPUT = open_data_file( $MAPPINGS, $filename ) or die "Cannot open $filename";
1929
1930     while (<$INPUT>)
1931     {
1932         next if /^;/;  # skip comments
1933         next if /^\s*$/;  # skip empty lines
1934         next if /\x1a/;  # skip ^Z
1935         last if /^ENDCODEPAGE/;
1936
1937         if (/^CODEPAGE\s+(\d+)/)
1938         {
1939             $codepage = $1;
1940             next;
1941         }
1942         if (/^CPINFO\s+(\d+)\s+0x([0-9a-fA-f]+)\s+0x([0-9a-fA-F]+)/)
1943         {
1944             $width = $1;
1945             $default_char = hex $2;
1946             $default_wchar = hex $3;
1947             next;
1948         }
1949         if (/^(MBTABLE|WCTABLE|DBCSRANGE|DBCSTABLE)\s+(\d+)/)
1950         {
1951             $state = $1;
1952             $count = $2;
1953             next;
1954         }
1955         if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)/)
1956         {
1957             if ($state eq "MBTABLE")
1958             {
1959                 my $cp = hex $1;
1960                 my $uni = hex $2;
1961                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
1962                 next;
1963             }
1964             if ($state eq "WCTABLE")
1965             {
1966                 my $uni = hex $1;
1967                 my $cp = hex $2;
1968                 $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]);
1969                 next;
1970             }
1971             if ($state eq "DBCSRANGE")
1972             {
1973                 my $start = hex $1;
1974                 my $end = hex $2;
1975                 for (my $i = $start; $i <= $end; $i++) { add_lead_byte( $i ); }
1976                 $lb_cur = $start;
1977                 $lb_end = $end;
1978                 next;
1979             }
1980             if ($state eq "DBCSTABLE")
1981             {
1982                 my $mb = hex $1;
1983                 my $uni = hex $2;
1984                 my $cp = ($lb_cur << 8) | $mb;
1985                 $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]);
1986                 if (!--$count)
1987                 {
1988                     if (++$lb_cur > $lb_end) { $state = "DBCSRANGE"; }
1989                 }
1990                 next;
1991             }
1992         }
1993         die "$filename: Unrecognized line $_\n";
1994     }
1995     close $INPUT;
1996 }
1997
1998
1999 ################################################################
2000 # read an input file and generate the corresponding .c file
2001 sub HANDLE_FILE(@)
2002 {
2003     my ($codepage,$filename,$has_glyphs,$comment,$def,$defw) = @_;
2004
2005     @cp2uni = ();
2006     @lead_bytes = ();
2007     @uni2cp = ();
2008     $default_char = $def || $DEF_CHAR;
2009     $default_wchar = $defw || $DEF_CHAR;
2010
2011     # some codepage files are special
2012     if ($codepage == 20932)
2013     {
2014         READ_JIS0208_FILE( $filename );
2015         ADD_DEFAULT_MAPPINGS();
2016     }
2017     elsif ($codepage == 20127)
2018     {
2019         fill_20127_codepage();
2020         ADD_DEFAULT_MAPPINGS();
2021     }
2022     elsif ($filename =~ /\/bestfit/)
2023     {
2024         read_bestfit_file( $filename );
2025     }
2026     elsif ($codepage == 10001)
2027     {
2028         read_codepage_file( $filename );
2029         # add Shift-JIS mappings
2030         read_bestfit_file( "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt" );
2031     }
2032     else
2033     {
2034         read_codepage_file( $filename );
2035         ADD_DEFAULT_MAPPINGS();
2036     }
2037
2038     output_codepage_file( $codepage, $filename, $comment, $has_glyphs );
2039 }
2040
2041
2042 ################################################################
2043 # save a file if modified
2044 sub save_file($)
2045 {
2046     my $file = shift;
2047     if (-f $file && !system "cmp $file $file.new >/dev/null")
2048     {
2049         unlink "$file.new";
2050     }
2051     else
2052     {
2053         rename "$file.new", "$file";
2054     }
2055 }
2056
2057
2058 ################################################################
2059 # output the list of codepage tables into the cptable.c file
2060 sub output_cptable($)
2061 {
2062     my $output = shift;
2063     my @tables_decl = ();
2064
2065     printf "Building %s\n", $output;
2066
2067     foreach my $file (@allfiles)
2068     {
2069         my ($codepage,$filename,$comment) = @$file;
2070         push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage);
2071     }
2072
2073     push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1);
2074     foreach my $file (@allfiles)
2075     {
2076         my ($codepage,$filename,$comment) = @$file;
2077         push @tables_decl, sprintf("    &cptable_%03d,\n", $codepage);
2078     }
2079     push @tables_decl, "};";
2080     REPLACE_IN_FILE( $output, @tables_decl );
2081 }
2082
2083 ################################################################
2084 # replace the contents of a file between ### cpmap ### marks
2085
2086 sub REPLACE_IN_FILE($@)
2087 {
2088     my $name = shift;
2089     my @data = @_;
2090     my @lines = ();
2091     open(FILE,$name) or die "Can't open $name";
2092     while (<FILE>)
2093     {
2094         push @lines, $_;
2095         last if /\#\#\# cpmap begin \#\#\#/;
2096     }
2097     push @lines, @data;
2098     while (<FILE>)
2099     {
2100         if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; }
2101     }
2102     push @lines, <FILE>;
2103     open(FILE,">$name.new") or die "Can't modify $name";
2104     print FILE @lines;
2105     close(FILE);
2106     save_file($name);
2107 }
2108
2109 ################################################################
2110 # main routine
2111
2112 chdir ".." if -f "./make_unicode";
2113 READ_DEFAULTS( $DEFAULTS );
2114 DUMP_CASE_MAPPINGS( "libs/wine/casemap.c" );
2115 DUMP_SORTKEYS( "libs/wine/collation.c", READ_SORTKEYS_FILE() );
2116 DUMP_COMPOSE_TABLES( "libs/wine/compose.c" );
2117 DUMP_CTYPE_TABLES( "libs/wine/wctype.c" );
2118 dump_mirroring( "dlls/usp10/mirror.c" );
2119 dump_shaping( "dlls/usp10/shaping.c" );
2120 dump_linebreak( "dlls/usp10/linebreak.c" );
2121 dump_indic( "dlls/usp10/indicsyllable.c" );
2122 dump_intl_nls("tools/l_intl.nls");
2123 dump_nameprep( "dlls/kernel32/nameprep.c" );
2124
2125 foreach my $file (@allfiles) { HANDLE_FILE( @{$file} ); }
2126
2127 output_cptable("libs/wine/cptable.c");
2128
2129 exit 0;
2130
2131 # Local Variables:
2132 # compile-command: "./make_unicode"
2133 # End: