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