#!/usr/bin/perl -w # # Generate code page .c files from ftp.unicode.org descriptions # # Copyright 2000 Alexandre Julliard # # This library is free software; you can redistribute it and/or # modify it under the terms of the GNU Lesser General Public # License as published by the Free Software Foundation; either # version 2.1 of the License, or (at your option) any later version. # # This library is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU # Lesser General Public License for more details. # # You should have received a copy of the GNU Lesser General Public # License along with this library; if not, write to the Free Software # Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA # use strict; # base URLs for www.unicode.org files my $MAPPINGS = "http://www.unicode.org/Public/MAPPINGS"; my $UNIDATA = "http://www.unicode.org/Public/12.1.0/ucd"; my $REPORTS = "http://www.unicode.org/reports"; my $RFCS = "http://www.rfc-editor.org/rfc"; # Sort keys file my $SORTKEYS = "tr10/allkeys.txt"; # RFC3454 (stringprep data) my $STRINGPREP = "rfc3454.txt"; # Defaults mapping my $DEFAULTS = "tools/unicode-defaults"; # Default char for undefined mappings my $DEF_CHAR = ord '?'; my @allfiles = ( [ 37, "VENDORS/MICSFT/EBCDIC/CP037.TXT", 0, "IBM EBCDIC US Canada", 0, 0x6f ], [ 424, "VENDORS/MISC/CP424.TXT", 0, "IBM EBCDIC Hebrew" ], [ 437, "VENDORS/MICSFT/PC/CP437.TXT", 1, "OEM United States" ], [ 500, "VENDORS/MICSFT/EBCDIC/CP500.TXT", 0, "IBM EBCDIC International", 0, 0x6f ], [ 737, "VENDORS/MICSFT/PC/CP737.TXT", 1, "OEM Greek 437G" ], [ 775, "VENDORS/MICSFT/PC/CP775.TXT", 1, "OEM Baltic" ], [ 850, "VENDORS/MICSFT/PC/CP850.TXT", 1, "OEM Multilingual Latin 1" ], [ 852, "VENDORS/MICSFT/PC/CP852.TXT", 1, "OEM Slovak Latin 2" ], [ 855, "VENDORS/MICSFT/PC/CP855.TXT", 1, "OEM Cyrillic" ], [ 856, "VENDORS/MISC/CP856.TXT", 0, "Hebrew PC" ], [ 857, "VENDORS/MICSFT/PC/CP857.TXT", 1, "OEM Turkish", 0xf8bb ], [ 860, "VENDORS/MICSFT/PC/CP860.TXT", 1, "OEM Portuguese" ], [ 861, "VENDORS/MICSFT/PC/CP861.TXT", 1, "OEM Icelandic" ], [ 862, "VENDORS/MICSFT/PC/CP862.TXT", 1, "OEM Hebrew" ], [ 863, "VENDORS/MICSFT/PC/CP863.TXT", 1, "OEM Canadian French" ], [ 864, "VENDORS/MICSFT/PC/CP864.TXT", 0, "OEM Arabic", 0xf8bb ], [ 865, "VENDORS/MICSFT/PC/CP865.TXT", 1, "OEM Nordic" ], [ 866, "VENDORS/MICSFT/PC/CP866.TXT", 1, "OEM Russian" ], [ 869, "VENDORS/MICSFT/PC/CP869.TXT", 1, "OEM Greek" ], [ 874, "VENDORS/MICSFT/WindowsBestFit/bestfit874.txt", 1, "ANSI/OEM Thai" ], [ 875, "VENDORS/MICSFT/EBCDIC/CP875.TXT", 0, "IBM EBCDIC Greek", 0, 0x6f ], [ 878, "VENDORS/MISC/KOI8-R.TXT", 0, "Russian KOI8" ], [ 932, "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt", 0, "ANSI/OEM Japanese Shift-JIS" ], [ 936, "VENDORS/MICSFT/WindowsBestFit/bestfit936.txt", 0, "ANSI/OEM Simplified Chinese GBK" ], [ 949, "VENDORS/MICSFT/WindowsBestFit/bestfit949.txt", 0, "ANSI/OEM Korean Unified Hangul" ], [ 950, "VENDORS/MICSFT/WindowsBestFit/bestfit950.txt", 0, "ANSI/OEM Traditional Chinese Big5" ], [ 1006, "VENDORS/MISC/CP1006.TXT", 0, "IBM Arabic" ], [ 1026, "VENDORS/MICSFT/EBCDIC/CP1026.TXT", 0, "IBM EBCDIC Latin 5 Turkish", 0, 0x6f ], [ 1250, "VENDORS/MICSFT/WindowsBestFit/bestfit1250.txt", 0, "ANSI Eastern Europe" ], [ 1251, "VENDORS/MICSFT/WindowsBestFit/bestfit1251.txt", 0, "ANSI Cyrillic" ], [ 1252, "VENDORS/MICSFT/WindowsBestFit/bestfit1252.txt", 0, "ANSI Latin 1" ], [ 1253, "VENDORS/MICSFT/WindowsBestFit/bestfit1253.txt", 0, "ANSI Greek" ], [ 1254, "VENDORS/MICSFT/WindowsBestFit/bestfit1254.txt", 0, "ANSI Turkish" ], [ 1255, "VENDORS/MICSFT/WindowsBestFit/bestfit1255.txt", 0, "ANSI Hebrew" ], [ 1256, "VENDORS/MICSFT/WindowsBestFit/bestfit1256.txt", 0, "ANSI Arabic" ], [ 1257, "VENDORS/MICSFT/WindowsBestFit/bestfit1257.txt", 0, "ANSI Baltic" ], [ 1258, "VENDORS/MICSFT/WindowsBestFit/bestfit1258.txt", 0, "ANSI/OEM Viet Nam" ], [ 1361, "VENDORS/MICSFT/WindowsBestFit/bestfit1361.txt", 0, "Korean Johab" ], [ 10000, "VENDORS/MICSFT/MAC/ROMAN.TXT", 0, "Mac Roman" ], [ 10001, "VENDORS/APPLE/JAPANESE.TXT", 0, "Mac Japanese" ], [ 10002, "VENDORS/APPLE/CHINTRAD.TXT", 0, "Mac Traditional Chinese" ], [ 10003, "VENDORS/APPLE/KOREAN.TXT", 0, "Mac Korean", 0xf8e5 ], [ 10004, "VENDORS/APPLE/ARABIC.TXT", 0, "Mac Arabic", 0xf826 ], [ 10005, "VENDORS/APPLE/HEBREW.TXT", 0, "Mac Hebrew", 0xf7fa ], [ 10006, "VENDORS/MICSFT/MAC/GREEK.TXT", 0, "Mac Greek" ], [ 10007, "VENDORS/MICSFT/MAC/CYRILLIC.TXT", 0, "Mac Cyrillic" ], [ 10008, "VENDORS/APPLE/CHINSIMP.TXT", 0, "Mac Simplified Chinese" ], [ 10010, "VENDORS/APPLE/ROMANIAN.TXT", 0, "Mac Romanian" ], [ 10017, "VENDORS/APPLE/CYRILLIC.TXT", 0, "Mac Ukrainian" ], [ 10021, "VENDORS/APPLE/THAI.TXT", 0, "Mac Thai" ], [ 10029, "VENDORS/MICSFT/MAC/LATIN2.TXT", 0, "Mac Latin 2" ], [ 10079, "VENDORS/MICSFT/MAC/ICELAND.TXT", 0, "Mac Icelandic" ], [ 10081, "VENDORS/MICSFT/MAC/TURKISH.TXT", 0, "Mac Turkish", 0xf89f ], [ 10082, "VENDORS/APPLE/CROATIAN.TXT", 0, "Mac Croatian" ], [ 20127, undef, 0, "US-ASCII (7bit)" ], [ 20866, "VENDORS/MISC/KOI8-R.TXT", 0, "Russian KOI8" ], [ 20932, "OBSOLETE/EASTASIA/JIS/JIS0208.TXT", 0, "EUC-JP", 0, 0x3f, 0x30fb ], [ 21866, "VENDORS/MISC/KOI8-U.TXT", 0, "Ukrainian KOI8" ], [ 28591, "ISO8859/8859-1.TXT", 0, "ISO 8859-1 Latin 1" ], [ 28592, "ISO8859/8859-2.TXT", 0, "ISO 8859-2 Latin 2 (East European)" ], [ 28593, "ISO8859/8859-3.TXT", 0, "ISO 8859-3 Latin 3 (South European)", 0xf7f5 ], [ 28594, "ISO8859/8859-4.TXT", 0, "ISO 8859-4 Latin 4 (Baltic old)" ], [ 28595, "ISO8859/8859-5.TXT", 0, "ISO 8859-5 Cyrillic" ], [ 28596, "ISO8859/8859-6.TXT", 0, "ISO 8859-6 Arabic", 0xf7c8 ], [ 28597, "ISO8859/8859-7.TXT", 0, "ISO 8859-7 Greek", 0xf7c5 ], [ 28598, "ISO8859/8859-8.TXT", 0, "ISO 8859-8 Hebrew", 0xf79c ], [ 28599, "ISO8859/8859-9.TXT", 0, "ISO 8859-9 Latin 5 (Turkish)" ], [ 28600, "ISO8859/8859-10.TXT", 0, "ISO 8859-10 Latin 6 (Nordic)" ], [ 28601, "ISO8859/8859-11.TXT", 0, "ISO 8859-11 Latin (Thai)" ], [ 28603, "ISO8859/8859-13.TXT", 0, "ISO 8859-13 Latin 7 (Baltic)" ], [ 28604, "ISO8859/8859-14.TXT", 0, "ISO 8859-14 Latin 8 (Celtic)" ], [ 28605, "ISO8859/8859-15.TXT", 0, "ISO 8859-15 Latin 9 (Euro)" ], [ 28606, "ISO8859/8859-16.TXT", 0, "ISO 8859-16 Latin 10 (Balkan)" ] ); my %ctype = ( "upper" => 0x0001, "lower" => 0x0002, "digit" => 0x0004, "space" => 0x0008, "punct" => 0x0010, "cntrl" => 0x0020, "blank" => 0x0040, "xdigit" => 0x0080, "alpha" => 0x0100, "defin" => 0x0200 ); my %bracket_types = ( "o" => 0x0000, "c" => 0x0001, ); my %indic_types = ( "Other" => 0x0000, "Bindu" => 0x0001, "Visarga" => 0x0002, "Avagraha" => 0x0003, "Nukta" => 0x0004, "Virama" => 0x0005, "Vowel_Independent" => 0x0006, "Vowel_Dependent" => 0x0007, "Vowel" => 0x0008, "Consonant_Placeholder" => 0x0009, "Consonant" => 0x000a, "Consonant_Dead" => 0x000b, "Consonant_Succeeding_Repha" => 0x000c, "Consonant_Subjoined" => 0x000d, "Consonant_Medial" => 0x000e, "Consonant_Final" => 0x000f, "Consonant_Head_Letter" => 0x0010, "Modifying_Letter" => 0x0011, "Tone_Letter" => 0x0012, "Tone_Mark" => 0x0013, "Register_Shifter" => 0x0014, "Consonant_Preceding_Repha" => 0x0015, "Pure_Killer" => 0x0016, "Invisible_Stacker" => 0x0017, "Gemination_Mark" => 0x0018, "Cantillation_Mark" => 0x0019, "Non_Joiner" => 0x001a, "Joiner" => 0x001b, "Number_Joiner" => 0x001c, "Number" => 0x001d, "Brahmi_Joining_Number" => 0x001e, "Consonant_With_Stacker" => 0x001f, "Consonant_Prefixed" => 0x0020, "Syllable_Modifier" => 0x0021, "Consonant_Killer" => 0x0022, "Consonant_Initial_Postfixed" => 0x0023, ); my %matra_types = ( "Right" => 0x01, "Left" => 0x02, "Visual_Order_Left" => 0x03, "Left_And_Right" => 0x04, "Top" => 0x05, "Bottom" => 0x06, "Top_And_Bottom" => 0x07, "Top_And_Right" => 0x08, "Top_And_Left" => 0x09, "Top_And_Left_And_Right" => 0x0a, "Bottom_And_Right" => 0x0b, "Top_And_Bottom_And_Right" => 0x0c, "Overstruck" => 0x0d, "Invisible" => 0x0e, "Bottom_And_Left" => 0x0f ); my %nameprep_flags = ( "unassigned" => 0x01, "prohibited" => 0x02, "bidi_ral" => 0x04, "bidi_l" => 0x08 ); my %break_types = ( "BK" => 0x0001, "CR" => 0x0002, "LF" => 0x0003, "CM" => 0x0004, "SG" => 0x0005, "GL" => 0x0006, "CB" => 0x0007, "SP" => 0x0008, "ZW" => 0x0009, "NL" => 0x000a, "WJ" => 0x000b, "JL" => 0x000c, "JV" => 0x000d, "JT" => 0x000e, "H2" => 0x000f, "H3" => 0x0010, "XX" => 0x0011, "OP" => 0x0012, "CL" => 0x0013, "CP" => 0x0014, "QU" => 0x0015, "NS" => 0x0016, "EX" => 0x0017, "SY" => 0x0018, "IS" => 0x0019, "PR" => 0x001a, "PO" => 0x001b, "NU" => 0x001c, "AL" => 0x001d, "ID" => 0x001e, "IN" => 0x001f, "HY" => 0x0020, "BB" => 0x0021, "BA" => 0x0022, "SA" => 0x0023, "AI" => 0x0024, "B2" => 0x0025, "HL" => 0x0026, "CJ" => 0x0027, "RI" => 0x0028, "EB" => 0x0029, "EM" => 0x002a, "ZWJ" => 0x002b, ); my %vertical_types = ( "R" => 0x0000, "U" => 0x0001, "Tr" => 0x0002, "Tu" => 0x0003, ); my %categories = ( "Lu" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}, # Letter, Uppercase "Ll" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"lower"}, # Letter, Lowercase "Lt" => $ctype{"defin"}|$ctype{"alpha"}|$ctype{"upper"}|$ctype{"lower"}, # Letter, Titlecase "Mn" => $ctype{"defin"}, # Mark, Non-Spacing "Mc" => $ctype{"defin"}, # Mark, Spacing Combining "Me" => $ctype{"defin"}, # Mark, Enclosing "Nd" => $ctype{"defin"}|$ctype{"digit"}, # Number, Decimal Digit "Nl" => $ctype{"defin"}|$ctype{"alpha"}, # Number, Letter "No" => $ctype{"defin"}, # Number, Other "Zs" => $ctype{"defin"}|$ctype{"space"}, # Separator, Space "Zl" => $ctype{"defin"}|$ctype{"space"}, # Separator, Line "Zp" => $ctype{"defin"}|$ctype{"space"}, # Separator, Paragraph "Cc" => $ctype{"defin"}|$ctype{"cntrl"}, # Other, Control "Cf" => $ctype{"defin"}|$ctype{"cntrl"}, # Other, Format "Cs" => $ctype{"defin"}, # Other, Surrogate "Co" => $ctype{"defin"}, # Other, Private Use "Cn" => $ctype{"defin"}, # Other, Not Assigned "Lm" => $ctype{"defin"}|$ctype{"alpha"}, # Letter, Modifier "Lo" => $ctype{"defin"}|$ctype{"alpha"}, # Letter, Other "Pc" => $ctype{"defin"}|$ctype{"punct"}, # Punctuation, Connector "Pd" => $ctype{"defin"}|$ctype{"punct"}, # Punctuation, Dash "Ps" => $ctype{"defin"}|$ctype{"punct"}, # Punctuation, Open "Pe" => $ctype{"defin"}|$ctype{"punct"}, # Punctuation, Close "Pi" => $ctype{"defin"}|$ctype{"punct"}, # Punctuation, Initial quote "Pf" => $ctype{"defin"}|$ctype{"punct"}, # Punctuation, Final quote "Po" => $ctype{"defin"}|$ctype{"punct"}, # Punctuation, Other "Sm" => $ctype{"defin"}, # Symbol, Math "Sc" => $ctype{"defin"}, # Symbol, Currency "Sk" => $ctype{"defin"}, # Symbol, Modifier "So" => $ctype{"defin"} # Symbol, Other ); # a few characters need additional categories that cannot be determined automatically my %special_categories = ( "xdigit" => [ ord('0')..ord('9'),ord('A')..ord('F'),ord('a')..ord('f'), 0xff10..0xff19, 0xff21..0xff26, 0xff41..0xff46 ], "space" => [ 0x09..0x0d, 0x85 ], "blank" => [ 0x09, 0x20, 0xa0, 0x3000, 0xfeff ], "cntrl" => [ 0x070f, 0x200c, 0x200d, 0x200e, 0x200f, 0x202a, 0x202b, 0x202c, 0x202d, 0x202e, 0x206a, 0x206b, 0x206c, 0x206d, 0x206e, 0x206f, 0xfeff, 0xfff9, 0xfffa, 0xfffb ], "punct" => [ 0x24, 0x2b, 0x3c..0x3e, 0x5e, 0x60, 0x7c, 0x7e, 0xa2..0xbe, 0xd7, 0xf7 ], "digit" => [ 0xb2, 0xb3, 0xb9 ], "lower" => [ 0x2071, 0x207f ] ); my %directions = ( "L" => 1, # Left-to-Right "R" => 2, # Right-to-Left "AL" => 12, # Right-to-Left Arabic "EN" => 3, # European Number "ES" => 4, # European Number Separator "ET" => 5, # European Number Terminator "AN" => 6, # Arabic Number "CS" => 7, # Common Number Separator "NSM" => 13, # Non-Spacing Mark "BN" => 14, # Boundary Neutral "B" => 8, # Paragraph Separator "S" => 9, # Segment Separator "WS" => 10, # Whitespace "ON" => 11, # Other Neutrals "LRE" => 15, # Left-to-Right Embedding "LRO" => 15, # Left-to-Right Override "RLE" => 15, # Right-to-Left Embedding "RLO" => 15, # Right-to-Left Override "PDF" => 15, # Pop Directional Format "LRI" => 15, # Left-to-Right Isolate "RLI" => 15, # Right-to-Left Isolate "FSI" => 15, # First Strong Isolate "PDI" => 15 # Pop Directional Isolate ); my %c2_types = ( "L" => 1, # C2_LEFTTORIGHT "R" => 2, # C2_RIGHTTOLEFT "AL" => 2, # C2_RIGHTTOLEFT "EN" => 3, # C2_EUROPENUMBER "ES" => 4, # C2_EUROPESEPARATOR "ET" => 5, # C2_EUROPETERMINATOR "AN" => 6, # C2_ARABICNUMBER "CS" => 7, # C2_COMMONSEPARATOR "NSM" => 0, # C2_NOTAPPLICABLE "BN" => 0, # C2_NOTAPPLICABLE "B" => 8, # C2_BLOCKSEPARATOR "S" => 9, # C2_SEGMENTSEPARATOR "WS" => 10, # C2_WHITESPACE "ON" => 11, # C2_OTHERNEUTRAL "LRE" => 11, # C2_OTHERNEUTRAL "LRO" => 11, # C2_OTHERNEUTRAL "RLE" => 11, # C2_OTHERNEUTRAL "RLO" => 11, # C2_OTHERNEUTRAL "PDF" => 11, # C2_OTHERNEUTRAL "LRI" => 11, # C2_OTHERNEUTRAL "RLI" => 11, # C2_OTHERNEUTRAL "FSI" => 11, # C2_OTHERNEUTRAL "PDI" => 11 # C2_OTHERNEUTRAL ); my %bidi_types = ( "ON" => 0, # Other Neutrals "L" => 1, # Left-to-Right "R" => 2, # Right-to-Left "AN" => 3, # Arabic Number "EN" => 4, # European Number "AL" => 5, # Right-to-Left Arabic "NSM" => 6, # Non-Spacing Mark "CS" => 7, # Common Number Separator "ES" => 8, # European Number Separator "ET" => 9, # European Number Terminator "BN" => 10, # Boundary Neutral "S" => 11, # Segment Separator "WS" => 12, # Whitespace "B" => 13, # Paragraph Separator "RLO" => 14, # Right-to-Left Override "RLE" => 15, # Right-to-Left Embedding "LRO" => 16, # Left-to-Right Override "LRE" => 17, # Left-to-Right Embedding "PDF" => 18, # Pop Directional Format "LRI" => 19, # Left-to-Right Isolate "RLI" => 20, # Right-to-Left Isolate "FSI" => 21, # First Strong Isolate "PDI" => 22 # Pop Directional Isolate ); my %joining_types = ( "U" => 0, # Non_Joining "T" => 1, # Transparent "R" => 2, # Right_Joining "L" => 3, # Left_Joining "D" => 4, # Dual_Joining "C" => 5, # Join_Causing ); my @cp2uni = (); my @lead_bytes = (); my @uni2cp = (); my @unicode_defaults = (); my @unicode_aliases = (); my @tolower_table = (); my @toupper_table = (); my @digitmap_table = (); my @compatmap_table = (); my @category_table = (); my @joining_table = (); my @direction_table = (); my @decomp_table = (); my @compose_table = (); my @combining_class_table = (); my @full_compatmap_table = (); my $default_char; my $default_wchar; my %joining_forms = ( "isolated" => [], "final" => [], "initial" => [], "medial" => [] ); ################################################################ # fetch a unicode.org file and open it sub open_data_file($$) { my ($base, $name) = @_; (my $dir = "data/$name") =~ s/\/[^\/]+$//; local *FILE; unless (-f "data/$name") { print "Fetching $base/$name...\n"; system "mkdir", "-p", $dir; !system "wget", "-q", "-O", "data/$name", "$base/$name" or die "cannot fetch $base/$name"; } open FILE, ") { next if /^\#/; # skip comments next if /^$/; # skip empty lines if (/^(([0-9a-fA-F]+)(,[0-9a-fA-F]+)*)\s+([0-9a-fA-F]+|'.'|none)\s+(\#.*)?/) { my @src = map hex, split /,/,$1; my $dst = $4; my $comment = $5; if ($#src > 0) { push @unicode_aliases, \@src; } next if ($dst eq "none"); $dst = ($dst =~ /\'.\'/) ? ord substr($dst,1,1) : hex $dst; foreach my $src (@src) { die "Duplicate value" if defined($unicode_defaults[$src]); $unicode_defaults[$src] = $dst; } next; } die "Unrecognized line $_\n"; } close DEFAULTS; # now build mappings from the decomposition field of the Unicode database my $UNICODE_DATA = open_data_file( $UNIDATA, "UnicodeData.txt" ); while (<$UNICODE_DATA>) { # Decode the fields ... my ($code, $name, $cat, $comb, $bidi, $decomp, $dec, $dig, $num, $mirror, $oldname, $comment, $upper, $lower, $title) = split /;/; my $dst; my $src = hex $code; die "unknown category $cat" unless defined $categories{$cat}; die "unknown directionality $bidi" unless defined $directions{$bidi}; $category_table[$src] = $categories{$cat}; $direction_table[$src] = $bidi; $joining_table[$src] = $joining_types{"T"} if $cat eq "Mn" || $cat eq "Me" || $cat eq "Cf"; if ($lower ne "") { $tolower_table[$src] = hex $lower; } if ($upper ne "") { $toupper_table[$src] = hex $upper; } if ($dec ne "") { $category_table[$src] |= $ctype{"digit"}; } if ($dig ne "") { $digitmap_table[$src] = ord $dig; } if ($comb ne "") { $combining_class_table[$src] = $comb; } # copy the category and direction for everything between First/Last pairs if ($name =~ /, First>/) { $start = $src; } if ($name =~ /, Last>/) { while ($start < $src) { $category_table[$start] = $category_table[$src]; $direction_table[$start] = $direction_table[$src]; $start++; } } next if $decomp eq ""; # no decomposition, skip it if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)/) { my @seq = (); for my $ch (split /\s+/, (split /\s+/, $decomp, 2)[1]) { push @seq, (hex $ch); } $full_compatmap_table[$src] = \@seq; } if ($decomp =~ /^<([a-zA-Z]+)>\s+([0-9a-fA-F]+)$/) { # decomposition of the form " 1234" -> use char if type is known if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)) { # Single char decomposition in the compatibility range $compatmap_table[$src] = hex $2; } if ($1 eq "isolated" || $1 eq "final" || $1 eq "initial" || $1 eq "medial") { ${joining_forms{$1}}[hex $2] = $src; next; } next unless ($1 eq "font" || $1 eq "noBreak" || $1 eq "circle" || $1 eq "super" || $1 eq "sub" || $1 eq "wide" || $1 eq "narrow" || $1 eq "compat" || $1 eq "small"); $dst = hex $2; } elsif ($decomp =~ /^\s+0020\s+([0-9a-fA-F]+)/) { # decomposition " 0020 1234" -> combining accent $dst = hex $1; } elsif ($decomp =~ /^([0-9a-fA-F]+)/) { # decomposition contains only char values without prefix -> use first char $dst = hex $1; $category_table[$src] |= $category_table[$dst] if defined $category_table[$dst]; # store decomposition if it contains two chars if ($decomp =~ /^([0-9a-fA-F]+)\s+([0-9a-fA-F]+)$/) { $decomp_table[$src] = [ hex $1, hex $2 ]; push @compose_table, [ hex $1, hex $2, $src ]; } elsif ($decomp =~ /^([0-9a-fA-F]+)$/) { # Single char decomposition if (hex $1 < 65536) { $decomp_table[$src] = [ hex $1, 0 ]; } if (($src >= 0xf900 && $src < 0xfb00) || ($src >= 0xfe30 && $src < 0xfffd)) { # Single char decomposition in the compatibility range $compatmap_table[$src] = hex $1; } } } else { next; } next if defined($unicode_defaults[$src]); # may have been set in the defaults file # check for loops for (my $i = $dst; ; $i = $unicode_defaults[$i]) { die sprintf("loop detected for %04x -> %04x",$src,$dst) if $i == $src; last unless defined($unicode_defaults[$i]); } $unicode_defaults[$src] = $dst; } close $UNICODE_DATA; # patch the category of some special characters foreach my $cat (keys %special_categories) { my $flag = $ctype{$cat}; foreach my $i (@{$special_categories{$cat}}) { $category_table[$i] |= $flag; } } # remove case mappings that don't round-trip for (my $i = 0; $i < @toupper_table; $i++) { next unless defined $toupper_table[$i]; my $ch = $toupper_table[$i]; $toupper_table[$i] = undef unless defined $tolower_table[$ch] && $tolower_table[$ch] == $i; } for (my $i = 0; $i < @tolower_table; $i++) { next unless defined $tolower_table[$i]; my $ch = $tolower_table[$i]; $tolower_table[$i] = undef unless defined $toupper_table[$ch] && $toupper_table[$ch] == $i; } } ################################################################ # define a new lead byte sub add_lead_byte($) { my $ch = shift; return if defined $cp2uni[$ch]; push @lead_bytes, $ch; $cp2uni[$ch] = 0; } ################################################################ # define a new char mapping sub add_mapping($$) { my ($cp, $uni) = @_; $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]); $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]); if ($cp > 0xff) { add_lead_byte( $cp >> 8 ); } } ################################################################ # parse a standard codepage file sub read_codepage_file($) { my $name = shift; my $INPUT = open_data_file( $MAPPINGS, $name ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^$/; # skip empty lines next if /\x1a/; # skip ^Z next if (/^0x([0-9a-fA-F]+)\s+\#UNDEFINED/); # undefined char if (/^0x([0-9a-fA-F]+)\s+\#DBCS LEAD BYTE/) { add_lead_byte( hex $1 ); next; } # 0x12 0x3456 if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/) { add_mapping( hex $1, hex $2 ); next; } # 0x12 0x3456+0x7890 if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\+.*\s+(\#.*)?/) { add_mapping( hex $1, hex $2 ); next; } # 0x12 +0x3456 if (/^0x([0-9a-fA-F]+)\s+\+0x([0-9a-fA-F]+)\s+(\#.*)?/) { add_mapping( hex $1, hex $2 ); next; } # 0x12 +0x3456 if (/^0x([0-9a-fA-F]+)\s+\+0x([0-9a-fA-F]+)\s+(\#.*)?/) { my $uni = hex $2; add_mapping( hex $1, $uni ) if $uni >= 0x80; next; } die "$name: Unrecognized line $_\n"; } close $INPUT; } ################################################################ # fill input data for the 20127 (us-ascii) codepage sub fill_20127_codepage() { for (my $i = 0; $i < 128; $i++) { $cp2uni[$i] = $uni2cp[$i] = $i; } for (my $i = 128; $i < 256; $i++) { $cp2uni[$i] = $i & 0x7f; } } ################################################################ # get a mapping including glyph chars for MB_USEGLYPHCHARS sub get_glyphs_mapping(@) { $_[0x01] = 0x263a; # (WHITE SMILING FACE) $_[0x02] = 0x263b; # (BLACK SMILING FACE) $_[0x03] = 0x2665; # (BLACK HEART SUIT) $_[0x04] = 0x2666; # (BLACK DIAMOND SUIT) $_[0x05] = 0x2663; # (BLACK CLUB SUIT) $_[0x06] = 0x2660; # (BLACK SPADE SUIT) $_[0x07] = 0x2022; # (BULLET) $_[0x08] = 0x25d8; # (INVERSE BULLET) $_[0x09] = 0x25cb; # (WHITE CIRCLE) $_[0x0a] = 0x25d9; # (INVERSE WHITE CIRCLE) $_[0x0b] = 0x2642; # (MALE SIGN) $_[0x0c] = 0x2640; # (FEMALE SIGN) $_[0x0d] = 0x266a; # (EIGHTH NOTE) $_[0x0e] = 0x266b; # (BEAMED EIGHTH NOTES) $_[0x0f] = 0x263c; # (WHITE SUN WITH RAYS) $_[0x10] = 0x25ba; # (BLACK RIGHT-POINTING POINTER) $_[0x11] = 0x25c4; # (BLACK LEFT-POINTING POINTER) $_[0x12] = 0x2195; # (UP DOWN ARROW) $_[0x13] = 0x203c; # (DOUBLE EXCLAMATION MARK) $_[0x14] = 0x00b6; # (PILCROW SIGN) $_[0x15] = 0x00a7; # (SECTION SIGN) $_[0x16] = 0x25ac; # (BLACK RECTANGLE) $_[0x17] = 0x21a8; # (UP DOWN ARROW WITH BASE) $_[0x18] = 0x2191; # (UPWARDS ARROW) $_[0x19] = 0x2193; # (DOWNWARDS ARROW) $_[0x1a] = 0x2192; # (RIGHTWARDS ARROW) $_[0x1b] = 0x2190; # (LEFTWARDS ARROW) $_[0x1c] = 0x221f; # (RIGHT ANGLE) $_[0x1d] = 0x2194; # (LEFT RIGHT ARROW) $_[0x1e] = 0x25b2; # (BLACK UP-POINTING TRIANGLE) $_[0x1f] = 0x25bc; # (BLACK DOWN-POINTING TRIANGLE) $_[0x7f] = 0x2302; # (HOUSE) return @_; } ################################################################ # build EUC-JP table from the JIS 0208 file # FIXME: for proper EUC-JP we should probably read JIS 0212 too # but this would require 3-byte DBCS characters sub READ_JIS0208_FILE($) { my $name = shift; # ASCII chars for (my $i = 0x00; $i <= 0x7f; $i++) { $cp2uni[$i] = $i; $uni2cp[$i] = $i; } # JIS X 0201 right plane for (my $i = 0xa1; $i <= 0xdf; $i++) { $cp2uni[0x8e00 + $i] = 0xfec0 + $i; $uni2cp[0xfec0 + $i] = 0x8e00 + $i; } # lead bytes foreach my $i (0x8e, 0x8f, 0xa1 .. 0xfe) { add_lead_byte($i); } # undefined chars foreach my $i (0x80 .. 0x8d, 0x90 .. 0xa0, 0xff) { $cp2uni[$i] = $DEF_CHAR; } # Shift-JIS compatibility $uni2cp[0x00a5] = 0x5c; $uni2cp[0x203e] = 0x7e; # Fix backslash conversion $cp2uni[0xa1c0] = 0xff3c; $uni2cp[0xff3c] = 0xa1c0; my $INPUT = open_data_file( $MAPPINGS, $name ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^0x[0-9a-fA-F]+\s+0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)\s+(\#.*)?/) { my $cp = 0x8080 + hex $1; my $uni = hex $2; $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]); $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]); next; } die "$name: Unrecognized line $_\n"; } close $INPUT; } ################################################################ # build the sort keys table sub dump_sortkeys($) { my $filename = shift; my @sortkeys = (); my $INPUT = open_data_file( $REPORTS, $SORTKEYS ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^$/; # skip empty lines next if /\x1a/; # skip ^Z next if /^\@version/; # skip @version header if (/^([0-9a-fA-F]+)\s+;\s+\[([*.])([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/) { my ($uni,$variable) = (hex $1, $2); next if $uni > 65535; $sortkeys[$uni] = [ $uni, hex $3, hex $4, hex $5, hex $6 ]; next; } if (/^([0-9a-fA-F]+\s+)+;\s+\[[*.]([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]{4})\.([0-9a-fA-F]+)\]/) { # multiple character sequence, ignored for now next; } die "$SORTKEYS: Unrecognized line $_\n"; } close $INPUT; # compress the keys to 32 bit: # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit @sortkeys = sort { ${$a}[1] <=> ${$b}[1] or ${$a}[2] <=> ${$b}[2] or ${$a}[3] <=> ${$b}[3] or ${$a}[4] <=> ${$b}[4] or $a cmp $b; } @sortkeys; my ($n2, $n3) = (1, 1); my @keys = (-1, -1, -1, -1, -1 ); my @flatkeys = (); for (my $i = 0; $i < @sortkeys; $i++) { next unless defined $sortkeys[$i]; my @current = @{$sortkeys[$i]}; if ($current[1] == $keys[1]) { if ($current[2] == $keys[2]) { if ($current[3] == $keys[3]) { # nothing } else { $keys[3] = $current[3]; $n3++; die if ($n3 >= 16); } } else { $keys[2] = $current[2]; $keys[3] = $current[3]; $n2++; $n3 = 1; die if ($n2 >= 256); } } else { $keys[1] = $current[1]; $keys[2] = $current[2]; $keys[3] = $current[3]; $n2 = 1; $n3 = 1; } if ($current[2]) { $current[2] = $n2; } if ($current[3]) { $current[3] = $n3; } if ($current[4]) { $current[4] = 1; } $flatkeys[$current[0]] = ($current[1] << 16) | ($current[2] << 8) | ($current[3] << 4) | $current[4]; } open OUTPUT,">$filename.new" or die "Cannot create $filename"; printf "Building $filename\n"; printf OUTPUT "/* Unicode collation element table */\n"; printf OUTPUT "/* generated from %s */\n", "$REPORTS/$SORTKEYS"; printf OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "collation_table", 0xffffffff, 32, @flatkeys ); close OUTPUT; save_file($filename); } ################################################################ # add default mappings once the file had been read sub add_default_mappings($) { my $first_private = shift; # Apply aliases foreach my $alias (@unicode_aliases) { my $target = undef; foreach my $src (@$alias) { if (defined($uni2cp[$src])) { $target = $uni2cp[$src]; last; } } next unless defined($target); # At least one char of the alias set is defined, set the others to the same value foreach my $src (@$alias) { $uni2cp[$src] = $target unless defined($uni2cp[$src]); } } # For every src -> target mapping in the defaults table, # make uni2cp[src] = uni2cp[target] if uni2cp[target] is defined for (my $src = 0; $src < 65536; $src++) { next if defined($uni2cp[$src]); # source has a definition already next unless defined($unicode_defaults[$src]); # no default for this char my $target = $unicode_defaults[$src]; # do a recursive mapping until we find a target char that is defined while (!defined($uni2cp[$target]) && defined($unicode_defaults[$target])) { $target = $unicode_defaults[$target]; } if (defined($uni2cp[$target])) { $uni2cp[$src] = $uni2cp[$target]; } } # Add a private char mapping for undefined chars if ($first_private) { for (my $i = 32; $i < 256; $i++) { next if defined($cp2uni[$i]); $cp2uni[$i] = $first_private; $uni2cp[$first_private] = $i; $first_private++; } } # Add an identity mapping for all undefined chars for (my $i = 0; $i < 256; $i++) { next if defined($cp2uni[$i]); next if defined($uni2cp[$i]); $cp2uni[$i] = $uni2cp[$i] = $i; } } ################################################################ # dump an array of integers sub dump_array($$@) { my ($bit_width, $default, @array) = @_; my $format = sprintf "0x%%0%ux", $bit_width / 4; my $i; my $ret = " "; for ($i = 0; $i < $#array; $i++) { $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default); $ret .= (($i % 8) != 7) ? ", " : ",\n "; } $ret .= sprintf($format, defined $array[$i] ? $array[$i] : $default); return $ret; } ################################################################ # dump a unicode->ascii mapping table sub dump_uni2cp_table($$) { my ($width, $def) = @_; # count the number of unicode->ascii subtables that contain something my @lowerbounds; my @upperbounds; my $index = 0; my @filled = (-1) x 256; for (my $i = 0; $i < 65536; $i++) { next unless defined $uni2cp[$i]; if ($filled[$i >> 8] == -1) { $lowerbounds[$index] = $i & 0xff; $upperbounds[$index] = 0xff - $lowerbounds[$index]; $filled[$i >> 8] = $index * 256; $index++; } else { $upperbounds[$index-1] = 0xff - ($i & 0xff); } } # add defaults mapping at the end $filled[256] = $index * 256; $lowerbounds[$index] = 255; $upperbounds[$index] = 0; $index++; # collapse blocks upwards if possible my $removed = 0; $index = 0; for (my $i = 0; $i < 257; $i++) { next if $filled[$i] == -1; if ($upperbounds[$index - 1] > $lowerbounds[$index]) { $removed += $lowerbounds[$index]; } else { $removed += $upperbounds[$index - 1]; $lowerbounds[$index] = $upperbounds[$index - 1]; } $filled[$i] -= $removed; $index++; } # output all the subtables into a single array printf OUTPUT "static const unsigned %s uni2cp_low[] =\n{\n", $width == 8 ? "char" : "short"; for (my $i = $index = 0; $i < 257; $i++) { next if $filled[$i] == -1; my $start = ($i << 8) + $lowerbounds[$index]; my $end = ($i << 8) + 255; if ($i == 256) { print OUTPUT " /* defaults */\n"; printf OUTPUT "%s\n};\n\n", dump_array( $width, 0, ($def) x ($end - $start + 1) ); } else { printf OUTPUT " /* 0x%04x .. 0x%04x */\n", $start, $end; printf OUTPUT "%s,\n", dump_array( $width, $def, @uni2cp[$start .. $end] ); } $index++; } # output a table of the offsets of the subtables in the previous array for (my $i = 0; $i < 256; $i++) { if ($filled[$i] == -1) { $filled[$i] = $filled[256]; } } printf OUTPUT "static const unsigned short uni2cp_high[256] =\n"; printf OUTPUT "{\n%s\n};\n\n", dump_array( 16, 0, @filled[0..255] ); } ################################################################ # dump an SBCS mapping table sub dump_sbcs_table($$$$$) { my ($codepage, $has_glyphs, $name, $def, $defw) = @_; my $i; # output the ascii->unicode table if ($has_glyphs) { printf OUTPUT "static const WCHAR cp2uni[512] =\n"; printf OUTPUT "{\n%s", dump_array( 16, $defw, @cp2uni[0 .. 255] ); printf OUTPUT ",\n /* glyphs */\n%s\n};\n\n", dump_array( 16, $defw, get_glyphs_mapping(@cp2uni[0 .. 255]) ); } else { printf OUTPUT "static const WCHAR cp2uni[256] =\n"; printf OUTPUT "{\n%s\n};\n\n", dump_array( 16, $defw, @cp2uni[0 .. 255] ); } dump_uni2cp_table( 8, $def ); # output the code page descriptor printf OUTPUT "const struct sbcs_table DECLSPEC_HIDDEN cptable_%03d =\n{\n", $codepage; printf OUTPUT " { %d, 1, 0x%04x, 0x%04x, \"%s\" },\n", $codepage, $def, $defw, $name; printf OUTPUT " cp2uni,\n"; if ($has_glyphs) { printf OUTPUT " cp2uni + 256,\n"; } else { printf OUTPUT " cp2uni,\n"; } printf OUTPUT " uni2cp_low,\n"; printf OUTPUT " uni2cp_high\n};\n"; } ################################################################ # dump a DBCS mapping table sub dump_dbcs_table($$$$@) { my ($codepage, $name, $def, $defw, @lb_ranges) = @_; # build a list of lead bytes that are actually used my @lblist = (); LBLOOP: for (my $y = 0; $y <= $#lead_bytes; $y++) { $cp2uni[$lead_bytes[$y]] = $defw; my $base = $lead_bytes[$y] << 8; for (my $x = 0; $x < 256; $x++) { if (defined $cp2uni[$base+$x]) { push @lblist,$lead_bytes[$y]; next LBLOOP; } } } my $unused = ($#lead_bytes > $#lblist); # output the ascii->unicode table for the single byte chars printf OUTPUT "static const WCHAR cp2uni[%d] =\n", 256 * ($#lblist + 2 + $unused); printf OUTPUT "{\n%s,\n", dump_array( 16, $defw, @cp2uni[0 .. 255] ); # output the default table for unused lead bytes if ($unused) { printf OUTPUT " /* unused lead bytes */\n"; printf OUTPUT "%s,\n", dump_array( 16, 0, ($defw) x 256 ); } # output the ascii->unicode table for each DBCS lead byte for (my $y = 0; $y <= $#lblist; $y++) { my $base = $lblist[$y] << 8; printf OUTPUT " /* lead byte %02x */\n", $lblist[$y]; printf OUTPUT "%s", dump_array( 16, $defw, @cp2uni[$base .. $base+255] ); printf OUTPUT ($y < $#lblist) ? ",\n" : "\n};\n\n"; } # output the lead byte subtables offsets my @offsets = (0) x 256; for (my $x = 0; $x <= $#lblist; $x++) { $offsets[$lblist[$x]] = $x + 1; } if ($unused) { # increment all lead bytes offset to take into account the unused table for (my $x = 0; $x <= $#lead_bytes; $x++) { $offsets[$lead_bytes[$x]]++; } } printf OUTPUT "static const unsigned char cp2uni_leadbytes[256] =\n"; printf OUTPUT "{\n%s\n};\n\n", dump_array( 8, 0, @offsets ); dump_uni2cp_table( 16, $def ); # output the code page descriptor printf OUTPUT "const struct dbcs_table DECLSPEC_HIDDEN cptable_%03d =\n{\n", $codepage; printf OUTPUT " { %d, 2, 0x%04x, 0x%04x, \"%s\" },\n", $codepage, $def, $defw, $name; printf OUTPUT " cp2uni,\n"; printf OUTPUT " cp2uni_leadbytes,\n"; printf OUTPUT " uni2cp_low,\n"; printf OUTPUT " uni2cp_high,\n"; printf OUTPUT " {\n %s\n }\n", dump_array( 8, 0, @lb_ranges, 0, 0 ); printf OUTPUT "};\n"; } ################################################################ # get the list of defined lead byte ranges sub get_lb_ranges() { my @list = (); my @ranges = (); my $i = 0; foreach $i (@lead_bytes) { $list[$i] = 1; } my $on = 0; for (my $i = 0; $i < 256; $i++) { if ($on) { if (!defined $list[$i]) { push @ranges, $i-1; $on = 0; } } else { if ($list[$i]) { push @ranges, $i; $on = 1; } } } if ($on) { push @ranges, 0xff; } return @ranges; } ################################################################ # dump the Indic Syllabic Category table sub dump_indic($) { my $filename = shift; my @indic_table; my $INPUT = open_data_file( $UNIDATA, "IndicSyllabicCategory.txt" ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*#/) { my $type = $2; die "unknown indic $type" unless defined $indic_types{$type}; if (hex $1 < 65536) { $indic_table[hex $1] = $indic_types{$type}; } next; } elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/) { my $type = $3; die "unknown indic $type" unless defined $indic_types{$type}; if (hex $1 < 65536 and hex $2 < 65536) { foreach my $i (hex $1 .. hex $2) { $indic_table[$i] = $indic_types{$type}; } } next; } die "malformed line $_"; } close $INPUT; $INPUT = open_data_file( $UNIDATA, "IndicPositionalCategory.txt" ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z]+)\s*#/) { my $type = $2; die "unknown matra $type" unless defined $matra_types{$type}; $indic_table[hex $1] += $matra_types{$type} << 8; next; } elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*#/) { my $type = $3; die "unknown matra $type" unless defined $matra_types{$type}; foreach my $i (hex $1 .. hex $2) { $indic_table[$i] += $matra_types{$type} << 8; } next; } die "malformed line $_"; } close $INPUT; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode Indic Syllabic Category */\n"; print OUTPUT "/* generated from $UNIDATA/IndicSyllabicCategory.txt */\n"; print OUTPUT "/* and from $UNIDATA/IndicPositionalCategory.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "indic_syllabic_table", $indic_types{'Other'}, 16, @indic_table ); close OUTPUT; save_file($filename); } ################################################################ # dump the Line Break Properties table sub dump_linebreak($) { my $filename = shift; my @break_table; my $next_group = 0; my $INPUT = open_data_file( $UNIDATA, "LineBreak.txt" ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z][0-9A-Z])+\s*/) { my $type = $2; die "unknown breaktype $type" unless defined $break_types{$type}; $break_table[hex $1] = $break_types{$type}; next; } elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z][0-9A-Z])+\s*/) { my $type = $3; die "unknown breaktype $type" unless defined $break_types{$type}; foreach my $i (hex $1 .. hex $2) { $break_table[$i] = $break_types{$type}; } next; } elsif (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/) { my $type = $2; die "unknown breaktype $type" unless defined $break_types{$type}; $break_table[hex $1] = $break_types{$type}; next; } elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([0-9A-Z][0-9A-Z])+\s*/) { my $type = $3; die "unknown breaktype $type" unless defined $break_types{$type}; foreach my $i (hex $1 .. hex $2) { $break_table[$i] = $break_types{$type}; } next; } die "malformed line $_"; } close $INPUT; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode Line Break Properties */\n"; print OUTPUT "/* generated from $UNIDATA/LineBreak.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "wine_linebreak_table", $break_types{'XX'}, 16, @break_table ); close OUTPUT; save_file($filename); } my %scripts = ( "Unknown" => 0, "Common" => 1, "Inherited" => 2, "Arabic" => 3, "Armenian" => 4, "Avestan" => 5, "Balinese" => 6, "Bamum" => 7, "Batak" => 8, "Bengali" => 9, "Bopomofo" => 10, "Brahmi" => 11, "Braille" => 12, "Buginese" => 13, "Buhid" => 14, "Canadian_Aboriginal" => 15, "Carian" => 16, "Cham" => 17, "Cherokee" => 18, "Coptic" => 19, "Cuneiform" => 20, "Cypriot" => 21, "Cyrillic" => 22, "Deseret" => 23, "Devanagari" => 24, "Egyptian_Hieroglyphs" => 25, "Ethiopic" => 26, "Georgian" => 27, "Glagolitic" => 28, "Gothic" => 29, "Greek" => 30, "Gujarati" => 31, "Gurmukhi" => 32, "Han" => 33, "Hangul" => 34, "Hanunoo" => 35, "Hebrew" => 36, "Hiragana" => 37, "Imperial_Aramaic" => 38, "Inscriptional_Pahlavi" => 39, "Inscriptional_Parthian" => 40, "Javanese" => 41, "Kaithi" => 42, "Kannada" => 43, "Katakana" => 44, "Kayah_Li" => 45, "Kharoshthi" => 46, "Khmer" => 47, "Lao" => 48, "Latin" => 49, "Lepcha" => 50, "Limbu" => 51, "Linear_B" => 52, "Lisu" => 53, "Lycian" => 54, "Lydian" => 55, "Malayalam" => 56, "Mandaic" => 57, "Meetei_Mayek" => 58, "Mongolian" => 59, "Myanmar" => 60, "New_Tai_Lue" => 61, "Nko" => 62, "Ogham" => 63, "Ol_Chiki" => 64, "Old_Italic" => 65, "Old_Persian" => 66, "Old_South_Arabian" => 67, "Old_Turkic" => 68, "Oriya" => 69, "Osmanya" => 70, "Phags_Pa" => 71, "Phoenician" => 72, "Rejang" => 73, "Runic" => 74, "Samaritan" => 75, "Saurashtra" => 76, "Shavian" => 77, "Sinhala" => 78, "Sundanese" => 79, "Syloti_Nagri" => 80, "Syriac" => 81, "Tagalog" => 82, "Tagbanwa" => 83, "Tai_Le" => 84, "Tai_Tham" => 85, "Tai_Viet" => 86, "Tamil" => 87, "Telugu" => 88, "Thaana" => 89, "Thai" => 90, "Tibetan" => 91, "Tifinagh" => 92, "Ugaritic" => 93, "Vai" => 94, "Yi" => 95, # Win8/Win8.1 "Chakma" => 96, "Meroitic_Cursive" => 97, "Meroitic_Hieroglyphs" => 98, "Miao" => 99, "Sharada" => 100, "Sora_Sompeng" => 101, "Takri" => 102, # Win10 "Bassa_Vah" => 103, "Caucasian_Albanian" => 104, "Duployan" => 105, "Elbasan" => 106, "Grantha" => 107, "Khojki" => 108, "Khudawadi" => 109, "Linear_A" => 110, "Mahajani" => 111, "Manichaean" => 112, "Mende_Kikakui" => 113, "Modi" => 114, "Mro" => 115, "Nabataean" => 116, "Old_North_Arabian" => 117, "Old_Permic" => 118, "Pahawh_Hmong" => 119, "Palmyrene" => 120, "Pau_Cin_Hau" => 121, "Psalter_Pahlavi" => 122, "Siddham" => 123, "Tirhuta" => 124, "Warang_Citi" => 125, # Win10 RS1 "Adlam" => 126, "Ahom" => 127, "Anatolian_Hieroglyphs" => 128, "Bhaiksuki" => 129, "Hatran" => 130, "Marchen" => 131, "Multani" => 132, "Newa" => 133, "Old_Hungarian" => 134, "Osage" => 135, "SignWriting" => 136, "Tangut" => 137, # Win10 RS4 "Masaram_Gondi" => 138, "Nushu" => 139, "Soyombo" => 140, "Zanabazar_Square" => 141, # Win10 1903 "Dogra" => 142, "Gunjala_Gondi" => 143, "Hanifi_Rohingya" => 144, "Makasar" => 145, "Medefaidrin" => 146, "Old_Sogdian" => 147, "Sogdian" => 148, ); ################################################################ # dump Script IDs table sub dump_scripts($) { my $filename = shift; my $header = $filename; my @scripts_table; my $script_index; my $i; my $INPUT = open_data_file( $UNIDATA, "Scripts.txt" ); # Fill the table # Unknown script id is always 0, so undefined scripts are automatically treated as such while (<$INPUT>) { my $type = ""; next if /^\#/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/) { $type = $2; if (defined $scripts{$type}) { $scripts_table[hex $1] = $scripts{$type}; } next; } elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/) { $type = $3; if (defined $scripts{$type}) { foreach my $i (hex $1 .. hex $2) { $scripts_table[$i] = $scripts{$type}; } } next; } } close $INPUT; $header = "$filename.h"; open OUTPUT,">$header.new" or die "Cannot create $header"; print "Building $header\n"; print OUTPUT "/* Unicode Script IDs */\n"; print OUTPUT "/* generated from $UNIDATA/Scripts.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "enum unicode_script_id {\n"; foreach my $script (sort { $scripts{$a} <=> $scripts{$b} } keys %scripts) { print OUTPUT " Script_$script = $scripts{$script},\n"; } print OUTPUT " Script_LastId = ", (scalar keys %scripts) - 1, "\n"; print OUTPUT "};\n"; close OUTPUT; save_file($header); $filename = "$filename.c"; open OUTPUT,">$filename.new" or die "Cannot create $header"; print "Building $filename\n"; print OUTPUT "/* Unicode Script IDs */\n"; print OUTPUT "/* generated from $UNIDATA/Scripts.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "wine_scripts_table", 0, 16, @scripts_table ); close OUTPUT; save_file($filename); } ################################################################ # dump the BiDi mirroring table sub dump_mirroring($) { my $filename = shift; my @mirror_table = (); my $INPUT = open_data_file( $UNIDATA, "BidiMirroring.txt" ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+)/) { $mirror_table[hex $1] = hex $2; next; } die "malformed line $_"; } close $INPUT; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode BiDi mirroring */\n"; print OUTPUT "/* generated from $UNIDATA/BidiMirroring.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "wine_mirror_map", 0, 16, @mirror_table ); close OUTPUT; save_file($filename); } ################################################################ # dump the Bidi Brackets sub dump_bracket($) { my $filename = shift; my @bracket_table; my $INPUT = open_data_file( $UNIDATA, "BidiBrackets.txt" ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;\s*([0-9a-fA-F]+);\s*([con])/) { my $type = $3; die "unknown bracket $type" unless defined $bracket_types{$type}; die "characters too distant $1 and $2" if abs(hex($2) - hex($1)) >= 128; $bracket_table[hex $1] = (hex($2) - hex($1)) % 255; $bracket_table[hex $1] += $bracket_types{$type} << 8; next; } die "malformed line $_"; } close $INPUT; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode Bidirectional Bracket table */\n"; print OUTPUT "/* generated from $UNIDATA/BidiBrackets.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "bidi_bracket_table", 0, 16, @bracket_table ); close OUTPUT; save_file($filename); } ################################################################ # dump the Arabic shaping table sub dump_shaping($) { my $filename = shift; my %groups; my $next_group = 0; $groups{"No_Joining_Group"} = $next_group++; my $INPUT = open_data_file( $UNIDATA, "ArabicShaping.txt" ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;.*;\s*([RLDCUT])\s*;\s*(\w+)/) { my $type = $2; my $group = $3; $groups{$group} = $next_group++ unless defined $groups{$group}; $joining_table[hex $1] = $joining_types{$type} | ($groups{$group} << 8); next; } die "malformed line $_"; } close $INPUT; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode Arabic shaping */\n"; print OUTPUT "/* generated from $UNIDATA/ArabicShaping.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "wine_shaping_table", 0, 16, @joining_table ); print OUTPUT "\nconst unsigned short DECLSPEC_HIDDEN wine_shaping_forms[256][4] =\n{\n"; for (my $i = 0x600; $i <= 0x6ff; $i++) { printf OUTPUT " { 0x%04x, 0x%04x, 0x%04x, 0x%04x },\n", ${joining_forms{"isolated"}}[$i] || $i, ${joining_forms{"final"}}[$i] || $i, ${joining_forms{"initial"}}[$i] || $i, ${joining_forms{"medial"}}[$i] || $i; } print OUTPUT "};\n"; close OUTPUT; save_file($filename); } ################################################################ # dump the Vertical Orientation table sub dump_vertical($) { my $filename = shift; my @vertical_table; my $INPUT = open_data_file( $UNIDATA, "VerticalOrientation.txt" ); while (<$INPUT>) { next if /^\#/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z if (/^\s*([0-9a-fA-F]+)\s*;\s*([a-zA-Z_]+)\s*/) { my $type = $2; die "unknown vertical $type" unless defined $vertical_types{$type}; if (hex $1 < 65536) { $vertical_table[hex $1] = $vertical_types{$type}; } next; } elsif (/^\s*([0-9a-fA-F]+)..\s*([0-9a-fA-F]+)\s*;\s*([A-Za-z_]+)\s*/) { my $type = $3; die "unknown vertical $type" unless defined $vertical_types{$type}; foreach my $i (hex $1 .. hex $2) { $vertical_table[$i] = $vertical_types{$type}; } next; } die "malformed line $_"; } close $INPUT; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode Vertical Orientation */\n"; print OUTPUT "/* generated from $UNIDATA/VerticalOrientation.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "vertical_orientation_table", $vertical_types{'R'}, 16, @vertical_table ); close OUTPUT; save_file($filename); } ################################################################ # dump the digit folding tables sub dump_digit_folding($$) { my ($filename, $compat) = @_; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode digit folding mappings */\n"; print OUTPUT "/* generated from $UNIDATA/UnicodeData.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "wine_digitmap", 0, 16, @digitmap_table ); if ($compat) { print OUTPUT "\n"; dump_case_table( "DECLSPEC_HIDDEN wine_compatmap", @compatmap_table ); } close OUTPUT; save_file($filename); } ################################################################ # dump the case mapping tables sub dump_case_mappings($) { my $filename = shift; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode case mappings */\n"; print OUTPUT "/* generated from $UNIDATA/UnicodeData.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_case_table( "wine_casemap_lower", @tolower_table ); print OUTPUT "\n"; dump_case_table( "wine_casemap_upper", @toupper_table ); close OUTPUT; save_file($filename); } ################################################################ # dump a case mapping table sub dump_case_table($@) { my ($name,@table) = @_; # count the number of sub tables that contain something # also compute the low and upper populated bounds my @lowerbounds = ( 0, 0 ); my @upperbounds = ( 0, 255 ); my $index = 0; my @filled = (); for (my $i = 0; $i < 65536; $i++) { next unless defined $table[$i]; if (!defined $filled[$i >> 8]) { $lowerbounds[$index] = $i & 0xff; $upperbounds[$index] = 0xff - $lowerbounds[$index]; $filled[$i >> 8] = $index * 256 + 512; $index++; } else { $upperbounds[$index-1] = 0xff - ($i & 0xff); } $table[$i] = ($table[$i] - $i) & 0xffff; } # Collapse blocks upwards if possible my $removed = 0; $index = 0; for (my $i = 0; $i < 256; $i++) { next unless defined $filled[$i]; if ($upperbounds[$index - 1] > $lowerbounds[$index]) { $removed = $removed + $lowerbounds[$index]; } else { $removed = $removed + $upperbounds[$index - 1]; $lowerbounds[$index] = $upperbounds[$index - 1]; } $filled[$i] = $filled[$i] - $removed; $index++; } # dump the table printf OUTPUT "const WCHAR %s[%d] =\n", $name, $index * 256 + 512 - $removed; printf OUTPUT "{\n /* index */\n"; printf OUTPUT "%s,\n", dump_array( 16, 256, @filled ); printf OUTPUT " /* defaults */\n"; printf OUTPUT "%s", dump_array( 16, 0, (0) x 256 ); $index = 0; for (my $i = 0; $i < 256; $i++) { next unless $filled[$i]; printf OUTPUT ",\n /* 0x%02x%02x .. 0x%02xff */\n", $i, $lowerbounds[$index], $i; printf OUTPUT "%s", dump_array( 16, 0, @table[($i<<8) + $lowerbounds[$index] .. ($i<<8)+255] ); $index++; } printf OUTPUT "\n};\n"; } ################################################################ # compress a mapping table by removing identical rows sub compress_array($$@) { my $rows = shift; my $def = shift; my @table = @_; my $len = @table / $rows; my @array; my $data = ""; # try to merge table rows for (my $row = 0; $row < $rows; $row++) { my $rowtxt = pack "U*", map { defined($_) ? $_ : $def; } @table[($row * $len)..(($row + 1) * $len - 1)]; my $pos = index $data, $rowtxt; if ($pos == -1) { # check if the tail of the data can match the start of the new row my $first = substr( $rowtxt, 0, 1 ); for (my $i = length($data) - 1; $i > 0; $i--) { $pos = index( substr( $data, -$i ), $first ); last if $pos == -1; $i -= $pos; next unless substr( $data, -$i ) eq substr( $rowtxt, 0, $i ); substr( $data, -$i ) = ""; last; } $pos = length $data; $data .= $rowtxt; } $array[$row] = $rows + $pos; } return @array, unpack "U*", $data; } ################################################################ # dump a simple char -> 16-bit value mapping table sub dump_simple_mapping($$@) { my $name = shift; my $def = shift; my @array = compress_array( 256, $def, @_[0..65535] ); printf OUTPUT "const unsigned short %s[%d] =\n{\n", $name, $#array+1; printf OUTPUT " /* offsets */\n%s,\n", dump_array( 16, 0, @array[0..255] ); printf OUTPUT " /* values */\n%s\n};\n", dump_array( 16, 0, @array[256..$#array] ); } ################################################################ # dump a char -> 16-bit value mapping table using two-level tables sub dump_two_level_mapping($$@) { my $name = shift; my $def = shift; my $size = shift; my $type = $size == 16 ? "unsigned short" : "unsigned int"; my @row_array = compress_array( 4096, $def, @_[0..65535] ); my @array = compress_array( 256, 0, @row_array[0..4095] ); for (my $i = 256; $i < @array; $i++) { $array[$i] += @array - 4096; } printf OUTPUT "const %s DECLSPEC_HIDDEN %s[%d] =\n{\n", $type, $name, @array + @row_array - 4096; printf OUTPUT " /* level 1 offsets */\n%s,\n", dump_array( $size, 0, @array[0..255] ); printf OUTPUT " /* level 2 offsets */\n%s,\n", dump_array( $size, 0, @array[256..$#array] ); printf OUTPUT " /* values */\n%s\n};\n", dump_array( $size, 0, @row_array[4096..$#row_array] ); } ################################################################ # dump a binary case mapping table in l_intl.nls format sub dump_binary_case_table(@) { my (@table) = @_; my $max_char = 0x10000; my $level1 = $max_char / 16; my $level2 = $level1 / 16; my @difftable; for (my $i = 0; $i < @table; $i++) { next unless defined $table[$i]; $difftable[$i] = ($table[$i] - $i) & 0xffff; } my @row_array = compress_array( $level1, 0, @difftable[0..$max_char-1] ); my @array = compress_array( $level2, 0, @row_array[0..$level1-1] ); for (my $i = $level2; $i < @array; $i++) { $array[$i] += @array - $level1; } return @array, @row_array[$level1..$#row_array]; } ################################################################ # dump case mappings for l_intl.nls sub dump_intl_nls($) { my @upper = dump_binary_case_table( @toupper_table ); my @lower = dump_binary_case_table( @tolower_table ); my $filename = shift; open OUTPUT,">$filename.new" or die "Cannot create $filename"; printf "Building $filename\n"; binmode OUTPUT; print OUTPUT pack "S<", 1; # version print OUTPUT pack "S<*", 1 + scalar @upper, @upper; print OUTPUT pack "S<*", 1 + scalar @lower, @lower; close OUTPUT; save_file($filename); } sub load_nameprep_range_table($$$) { my ($INPUT, $val, $table_ref) = @_; while (<$INPUT>) { if (/^\s*([0-9a-fA-F]+)-([0-9a-fA-F]+)/) { my $last = hex $2; $last = 65535 if($last >= 65536); foreach my $i (hex $1 .. $last) { $table_ref->[$i] |= $val; } next; } elsif (/^\s*([0-9a-fA-F]+)/) { if (hex $1 < 65536) { $table_ref->[hex $1] |= $val; } next; } return if (/End\sTable/); } } sub load_nameprep_map_table($$) { my ($INPUT, $table_ref) = @_; while (<$INPUT>) { if (/^\s*([0-9a-fA-F]+);\s;/) { # special value for map to nothing $table_ref->[hex $1] = [0xffff, 0xffff, 0xffff]; next; } elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+);/) { $table_ref->[hex $1] = [hex $2, 0, 0]; next; } elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/) { $table_ref->[hex $1] = [hex $2, hex $3, 0]; next; } elsif (/^\s*([0-9a-fA-F]+);\s([0-9a-fA-F]+)\s([0-9a-fA-F]+)\s([0-9a-fA-F]+);/) { $table_ref->[hex $1] = [hex $2, hex $3, hex $4]; next; } return if (/End\sTable/); } } ################################################################ # dump mapping table, prohibited characters set, unassigned # characters, bidirectional rules used by nameprep algorithm sub dump_nameprep($) { my $filename = shift; my @mapping_table = (); my @flags_table; my $INPUT = open_data_file( $RFCS, $STRINGPREP ); while (<$INPUT>) { next unless /Start\sTable/; load_nameprep_range_table($INPUT, $nameprep_flags{"unassigned"}, \@flags_table) if (/A.1/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.1.2/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.2.2/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.3/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.4/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.5/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.6/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.7/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.8/); load_nameprep_range_table($INPUT, $nameprep_flags{"prohibited"}, \@flags_table) if (/C.9/); load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_ral"}, \@flags_table) if (/D.1/); load_nameprep_range_table($INPUT, $nameprep_flags{"bidi_l"}, \@flags_table) if (/D.2/); load_nameprep_map_table($INPUT, \@mapping_table) if (/B.1/); load_nameprep_map_table($INPUT, \@mapping_table) if (/B.2/); } close $INPUT; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Nameprep algorithm related data */\n"; print OUTPUT "/* generated from $RFCS/$STRINGPREP */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "nameprep_char_type", 0, 16, @flags_table ); ######### mapping table # first determine all the 16-char subsets that contain something my @filled = (); my $pos = 16*3; # for the null subset for (my $i = 0; $i < 65536; $i++) { next unless defined $mapping_table[$i]; $filled[$i >> 4] = $pos; $pos += 16*3; $i |= 15; } my $total = $pos; # now count the 256-char subsets that contain something my @filled_idx = (256) x 256; $pos = 256 + 16; for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; $filled_idx[$i >> 4] = $pos; $pos += 16; $i |= 15; } my $null_offset = $pos; $total += $pos; # add the index offsets to the subsets positions for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; $filled[$i] += $null_offset; } # dump the main index printf OUTPUT "const unsigned short DECLSPEC_HIDDEN nameprep_mapping[%d] =\n", $total; printf OUTPUT "{\n /* index */\n"; printf OUTPUT "%s", dump_array( 16, 0, @filled_idx ); printf OUTPUT ",\n /* null sub-index */\n%s", dump_array( 16, 0, ($null_offset) x 16 ); # dump the second-level indexes for (my $i = 0; $i < 256; $i++) { next unless ($filled_idx[$i] > 256); my @table = @filled[($i<<4)..($i<<4)+15]; for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; } printf OUTPUT ",\n /* sub-index %02x */\n", $i; printf OUTPUT "%s", dump_array( 16, 0, @table ); } # dump the 16-char subsets printf OUTPUT ",\n /* null mapping */\n"; printf OUTPUT "%s", dump_array( 16, 0, (0) x 48 ); for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; my @table = (0) x 48; for (my $j = 0; $j < 16; $j++) { if (defined $mapping_table[($i<<4) + $j]) { $table[3 * $j] = ${$mapping_table[($i << 4) + $j]}[0]; $table[3 * $j + 1] = ${$mapping_table[($i << 4) + $j]}[1]; $table[3 * $j + 2] = ${$mapping_table[($i << 4) + $j]}[2]; } } printf OUTPUT ",\n /* 0x%03x0 .. 0x%03xf */\n", $i, $i; printf OUTPUT "%s", dump_array( 16, 0, @table ); } printf OUTPUT "\n};\n"; close OUTPUT; save_file($filename); } ################################################################ # dump the GetStringTypeW table sub dump_string_type_table($) { my $filename = shift; open OUTPUT,">$filename.new" or die "Cannot create $filename"; printf "Building $filename\n"; printf OUTPUT "/* Unicode wctype table */\n"; printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n"; printf OUTPUT "#include \"windef.h\"\n\n"; my @table = @category_table; # add the direction in the high 4 bits of the category for (my $i = 0; $i < 65536; $i++) { $table[$i] |= $c2_types{$direction_table[$i]} << 12 if defined $direction_table[$i]; } dump_two_level_mapping( "wctype_table", 0, 16, @table ); close OUTPUT; save_file($filename); } ################################################################ # dump the bidi direction table sub dump_bidi_dir_table($) { my $filename = shift; open OUTPUT,">$filename.new" or die "Cannot create $filename"; printf "Building $filename\n"; printf OUTPUT "/* Unicode BiDi direction table */\n"; printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n"; printf OUTPUT "#include \"windef.h\"\n\n"; my @table; for (my $i = 0; $i < 65536; $i++) { $table[$i] = $bidi_types{$direction_table[$i]} if defined $direction_table[$i]; } dump_two_level_mapping( "bidi_direction_table", $bidi_types{"L"}, 16, @table ); close OUTPUT; save_file($filename); } ################################################################ # dump the ctype tables sub dump_ctype_tables($) { my $filename = shift; open OUTPUT,">$filename.new" or die "Cannot create $filename"; printf "Building $filename\n"; printf OUTPUT "/* Unicode ctype tables */\n"; printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n"; printf OUTPUT "#include \"windef.h\"\n\n"; my @table = @category_table; # add the direction in the high 4 bits of the category for (my $i = 0; $i < 65536; $i++) { $table[$i] |= $directions{$direction_table[$i]} << 12 if defined $direction_table[$i]; } dump_simple_mapping( "wine_wctype_table", 0, @table ); close OUTPUT; save_file($filename); } ################################################################ # dump the char composition table sub dump_compose_table($) { my $filename = shift; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode char composition */\n"; print OUTPUT "/* generated from $UNIDATA/UnicodeData.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; my @filled = (); foreach my $i (@compose_table) { my @comp = @$i; push @{$filled[$comp[1]]}, [ $comp[0], $comp[2] ]; } # count how many different second chars we have my $count = 0; for (my $i = 0; $i < 65536; $i++) { next unless defined $filled[$i]; $count++; } # build the table of second chars and offsets my $pos = $count + 1; my @table = (); for (my $i = 0; $i < 65536; $i++) { next unless defined $filled[$i]; push @table, $i, $pos; $pos += @{$filled[$i]}; } # terminator with last position push @table, 0, $pos; printf OUTPUT "static const WCHAR table[0x%x] =\n{\n", 2*$pos; printf OUTPUT " /* second chars + offsets */\n%s", dump_array( 16, 0, @table ); # build the table of first chars and mappings for (my $i = 0; $i < 65536; $i++) { next unless defined $filled[$i]; my @table = (); my @list = sort { $a->[0] <=> $b->[0] } @{$filled[$i]}; for (my $j = 0; $j <= $#list; $j++) { push @table, $list[$j][0], $list[$j][1]; } printf OUTPUT ",\n /* 0x%04x */\n%s", $i, dump_array( 16, 0, @table ); } print OUTPUT "\n};\n\n"; print OUTPUT <<"EOF"; static inline int binary_search( WCHAR ch, int low, int high ) { while (low <= high) { int pos = (low + high) / 2; if (table[2 * pos] < ch) low = pos + 1; else if (table[2 * pos] > ch) high = pos - 1; else return pos; } return -1; } WCHAR DECLSPEC_HIDDEN wine_compose( const WCHAR *str ) { int pos, idx = 1, start = 0, count = $count; for (;;) { if ((pos = binary_search( str[idx], start, count - 1 )) == -1) return 0; if (!idx--) return table[2 * pos + 1]; start = table[2 * pos + 1]; count = table[2 * pos + 3]; } } EOF close OUTPUT; save_file($filename); } sub dump_full_compat_table() { # first determine all the 16-char subsets that contain something my @filled = (0) x 4096; my $pos = 16; # for the null subset my $data_total = 0; for (my $i = 0; $i < 65536; $i++) { next unless defined $full_compatmap_table[$i]; if ($filled[$i >> 4] == 0) { $filled[$i >> 4] = $pos; $pos += 16; } $data_total += $#{$full_compatmap_table[$i]} + 1; } my $total = $pos; # now count the 256-char subsets that contain something my @filled_idx = (256) x 256; $pos = 256 + 16; for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; $filled_idx[$i >> 4] = $pos; $pos += 16; $i |= 15; } my $null_offset = $pos; # null mapping $total += $pos + 1; # add the offset sentinel # add the index offsets to the subsets positions for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; $filled[$i] += $null_offset; } # dump the main index printf OUTPUT "static const WCHAR compatmap_table[%d] =\n", $total + $data_total; printf OUTPUT "{\n /* index */\n"; printf OUTPUT "%s", dump_array( 16, 0, @filled_idx ); printf OUTPUT ",\n /* null sub-index */\n%s", dump_array( 16, 0, ($null_offset) x 16 ); # dump the second-level indexes for (my $i = 0; $i < 256; $i++) { next unless ($filled_idx[$i] > 256); my @table = @filled[($i<<4)..($i<<4)+15]; for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; } printf OUTPUT ",\n /* sub-index %02x */\n", $i; printf OUTPUT "%s", dump_array( 16, 0, @table ); } # dump the 16-char offsets printf OUTPUT ",\n /* null offsets */\n"; printf OUTPUT "%s", dump_array( 16, 0, ($total) x (16) ); $pos = $total; my @data; for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; my @table = (0) x (16); for (my $j = 0; $j < 16; $j++) { $table[$j] = $pos; if (defined $full_compatmap_table[($i<<4) + $j]) { $pos += $#{$full_compatmap_table[($i<<4) + $j]} + 1; push @data, @{$full_compatmap_table[($i<<4) + $j]}; } } printf OUTPUT ",\n /* offsets 0x%03x0 .. 0x%03xf */\n", $i, $i; printf OUTPUT "%s", dump_array( 16, 0, @table ); } my @sentinel = $pos; printf OUTPUT ",\n /* offset sentinel */\n"; printf OUTPUT "%s", dump_array( 16, 0, @sentinel ); printf OUTPUT ",\n /* data */\n"; printf OUTPUT "%s", dump_array( 16, 0, @data ); printf OUTPUT "\n};\n\n"; } ################################################################ # dump the char decomposition table sub dump_decompose_table($$) { my ($filename, $compat) = @_; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode char composition */\n"; print OUTPUT "/* generated from $UNIDATA/UnicodeData.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; # first determine all the 16-char subsets that contain something my @filled = (0) x 4096; my $pos = 16*2; # for the null subset for (my $i = 0; $i < 65536; $i++) { next unless defined $decomp_table[$i]; $filled[$i >> 4] = $pos; $pos += 16*2; $i |= 15; } my $total = $pos; # now count the 256-char subsets that contain something my @filled_idx = (256) x 256; $pos = 256 + 16; for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; $filled_idx[$i >> 4] = $pos; $pos += 16; $i |= 15; } my $null_offset = $pos; # null mapping $total += $pos; # add the index offsets to the subsets positions for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; $filled[$i] += $null_offset; } # dump the main index printf OUTPUT "static const WCHAR table[%d] =\n", $total; printf OUTPUT "{\n /* index */\n"; printf OUTPUT "%s", dump_array( 16, 0, @filled_idx ); printf OUTPUT ",\n /* null sub-index */\n%s", dump_array( 16, 0, ($null_offset) x 16 ); # dump the second-level indexes for (my $i = 0; $i < 256; $i++) { next unless ($filled_idx[$i] > 256); my @table = @filled[($i<<4)..($i<<4)+15]; for (my $j = 0; $j < 16; $j++) { $table[$j] ||= $null_offset; } printf OUTPUT ",\n /* sub-index %02x */\n", $i; printf OUTPUT "%s", dump_array( 16, 0, @table ); } # dump the 16-char subsets printf OUTPUT ",\n /* null mapping */\n"; printf OUTPUT "%s", dump_array( 16, 0, (0) x 32 ); for (my $i = 0; $i < 4096; $i++) { next unless $filled[$i]; my @table = (0) x 32; for (my $j = 0; $j < 16; $j++) { if (defined $decomp_table[($i<<4) + $j]) { $table[2 * $j] = ${$decomp_table[($i << 4) + $j]}[0]; $table[2 * $j + 1] = ${$decomp_table[($i << 4) + $j]}[1]; } } printf OUTPUT ",\n /* 0x%03x0 .. 0x%03xf */\n", $i, $i; printf OUTPUT "%s", dump_array( 16, 0, @table ); } print OUTPUT "\n};\n\n"; if ($compat) { dump_full_compat_table(); print OUTPUT <<"EOF"; #include "wine/unicode.h" static const WCHAR *get_compatmap_entry( WCHAR ch, unsigned int *len ) { unsigned short offset_offs = compatmap_table[compatmap_table[ch >> 8] + ((ch >> 4) & 0xf)] + (ch & 0xf); unsigned short start = compatmap_table[offset_offs]; unsigned short end = compatmap_table[offset_offs + 1]; if (end > start) { *len = end - start; return compatmap_table + start; } return NULL; } unsigned int DECLSPEC_HIDDEN wine_decompose( int flags, WCHAR ch, WCHAR *dst, unsigned int dstlen ) { const WCHAR *ptr = NULL; unsigned int res, len, dst_pos = 0; *dst = ch; if (flags & WINE_DECOMPOSE_COMPAT) ptr = get_compatmap_entry( ch, &len ); if (!ptr) { ptr = table + table[table[ch >> 8] + ((ch >> 4) & 0x0f)] + 2 * (ch & 0xf); len = ptr[1] ? 2 : 1; } if (!*ptr) return 1; if (dstlen < len) return 0; /* apply the decomposition recursively */ while (len--) { if (!(res = wine_decompose( flags, *ptr++, dst + dst_pos, dstlen - dst_pos - len ))) return 0; dst_pos += res; } return dst_pos; } EOF } else { print OUTPUT <<"EOF"; unsigned int DECLSPEC_HIDDEN wine_decompose( int flags, WCHAR ch, WCHAR *dst, unsigned int dstlen ) { const WCHAR *ptr = table + table[table[ch >> 8] + ((ch >> 4) & 0x0f)] + 2 * (ch & 0xf); unsigned int res, len, dst_pos = 0; *dst = ch; if (!*ptr) return 1; len = ptr[1] ? 2 : 1; if (dstlen < len) return 0; /* apply the decomposition recursively */ while (len--) { if (!(res = wine_decompose( flags, *ptr++, dst + dst_pos, dstlen - dst_pos - len ))) return 0; dst_pos += res; } return dst_pos; } EOF } close OUTPUT; save_file($filename); } ################################################################ # dump the combining class table sub dump_combining_class($) { my $filename = shift; open OUTPUT,">$filename.new" or die "Cannot create $filename"; print "Building $filename\n"; print OUTPUT "/* Unicode Combining Classes */\n"; print OUTPUT "/* generated from $UNIDATA/UnicodeData.txt */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; print OUTPUT "#include \"windef.h\"\n\n"; dump_two_level_mapping( "combining_class_table", 0, 16, @combining_class_table ); close OUTPUT; save_file($filename); } ################################################################ # output a codepage definition file from the global tables sub output_codepage_file($$$$) { my ($codepage, $filename, $comment, $has_glyphs) = @_; my $output = sprintf "libs/port/c_%03d.c", $codepage; open OUTPUT,">$output.new" or die "Cannot create $output"; printf "Building %s from %s (%s)\n", $output, $filename || "hardcoded data", $comment; # dump all tables printf OUTPUT "/* code page %03d (%s) */\n", $codepage, $comment; if ($filename) { print OUTPUT "/* generated from $MAPPINGS/$filename */\n"; print OUTPUT "/* DO NOT EDIT!! */\n\n"; } else { printf OUTPUT "/* Automatically generated; DO NOT EDIT!! */\n\n"; } printf OUTPUT "#include \"wine/unicode.h\"\n\n"; if (!@lead_bytes) { dump_sbcs_table( $codepage, $has_glyphs, $comment, $default_char, $default_wchar ); } else { dump_dbcs_table( $codepage, $comment, $default_char, $default_wchar, get_lb_ranges() ); } close OUTPUT; save_file($output); } ################################################################ # read a "bestfit" Windows mapping file sub read_bestfit_file($) { my ($filename) = @_; my $state = ""; my ($codepage, $width, $count); my ($lb_cur, $lb_end); my $INPUT = open_data_file( $MAPPINGS, $filename ) or die "Cannot open $filename"; while (<$INPUT>) { next if /^;/; # skip comments next if /^\s*$/; # skip empty lines next if /\x1a/; # skip ^Z last if /^ENDCODEPAGE/; if (/^CODEPAGE\s+(\d+)/) { $codepage = $1; next; } if (/^CPINFO\s+(\d+)\s+0x([0-9a-fA-f]+)\s+0x([0-9a-fA-F]+)/) { $width = $1; $default_char = hex $2; $default_wchar = hex $3; next; } if (/^(MBTABLE|WCTABLE|DBCSRANGE|DBCSTABLE)\s+(\d+)/) { $state = $1; $count = $2; next; } if (/^0x([0-9a-fA-F]+)\s+0x([0-9a-fA-F]+)/) { if ($state eq "MBTABLE") { my $cp = hex $1; my $uni = hex $2; $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]); next; } if ($state eq "WCTABLE") { my $uni = hex $1; my $cp = hex $2; $uni2cp[$uni] = $cp unless defined($uni2cp[$uni]); next; } if ($state eq "DBCSRANGE") { my $start = hex $1; my $end = hex $2; for (my $i = $start; $i <= $end; $i++) { add_lead_byte( $i ); } $lb_cur = $start; $lb_end = $end; next; } if ($state eq "DBCSTABLE") { my $mb = hex $1; my $uni = hex $2; my $cp = ($lb_cur << 8) | $mb; $cp2uni[$cp] = $uni unless defined($cp2uni[$cp]); if (!--$count) { if (++$lb_cur > $lb_end) { $state = "DBCSRANGE"; } } next; } } die "$filename: Unrecognized line $_\n"; } close $INPUT; } ################################################################ # read an input file and generate the corresponding .c file sub HANDLE_FILE(@) { my ($codepage,$filename,$has_glyphs,$comment,$first_private,$def,$defw) = @_; @cp2uni = (); @lead_bytes = (); @uni2cp = (); $default_char = $def || $DEF_CHAR; $default_wchar = $defw || $DEF_CHAR; # some codepage files are special if ($codepage == 20932) { READ_JIS0208_FILE( $filename ); add_default_mappings( $first_private ); } elsif ($codepage == 20127) { fill_20127_codepage(); add_default_mappings( $first_private ); } elsif ($filename =~ /\/bestfit/) { read_bestfit_file( $filename ); } else { read_codepage_file( $filename ); if ($codepage == 10001) { # add Shift-JIS mappings read_bestfit_file( "VENDORS/MICSFT/WindowsBestFit/bestfit932.txt" ); } else { add_default_mappings( $first_private ); } } output_codepage_file( $codepage, $filename, $comment, $has_glyphs ); } ################################################################ # save a file if modified sub save_file($) { my $file = shift; if (-f $file && !system "cmp $file $file.new >/dev/null") { unlink "$file.new"; } else { rename "$file.new", "$file"; } } ################################################################ # output the list of codepage tables into the cptable.c file sub output_cptable($) { my $output = shift; my @tables_decl = (); printf "Building %s\n", $output; foreach my $file (@allfiles) { my ($codepage,$filename) = @$file; push @tables_decl, sprintf("extern union cptable cptable_%03d;\n",$codepage); } push @tables_decl, sprintf("\nstatic const union cptable * const cptables[%d] =\n{\n",$#allfiles+1); foreach my $file (@allfiles) { my ($codepage,$filename) = @$file; push @tables_decl, sprintf(" &cptable_%03d,\n", $codepage); } push @tables_decl, "};"; REPLACE_IN_FILE( $output, @tables_decl ); } ################################################################ # replace the contents of a file between ### cpmap ### marks sub REPLACE_IN_FILE($@) { my $name = shift; my @data = @_; my @lines = (); open(FILE,$name) or die "Can't open $name"; while () { push @lines, $_; last if /\#\#\# cpmap begin \#\#\#/; } push @lines, @data; while () { if (/\#\#\# cpmap end \#\#\#/) { push @lines, "\n", $_; last; } } push @lines, ; open(FILE,">$name.new") or die "Can't modify $name"; print FILE @lines; close(FILE); save_file($name); } ################################################################ # main routine chdir ".." if -f "./make_unicode"; READ_DEFAULTS( $DEFAULTS ); dump_case_mappings( "libs/port/casemap.c" ); dump_sortkeys( "libs/port/collation.c" ); dump_sortkeys( "dlls/kernelbase/collation.c" ); dump_compose_table( "libs/port/compose.c" ); dump_compose_table( "dlls/kernelbase/compose.c" ); dump_decompose_table( "libs/port/decompose.c", 1 ); dump_decompose_table( "dlls/kernelbase/decompose.c", 0 ); dump_ctype_tables( "libs/port/wctype.c" ); dump_bidi_dir_table( "dlls/gdi32/direction.c" ); dump_bidi_dir_table( "dlls/usp10/direction.c" ); dump_bidi_dir_table( "dlls/dwrite/direction.c" ); dump_string_type_table( "dlls/kernelbase/wctype.c" ); dump_digit_folding( "libs/port/digitmap.c", 1 ); dump_digit_folding( "dlls/kernelbase/digitmap.c", 0 ); dump_combining_class( "libs/port/combclass.c" ); dump_mirroring( "dlls/usp10/mirror.c" ); dump_mirroring( "dlls/dwrite/mirror.c" ); dump_bracket( "dlls/usp10/bracket.c" ); dump_bracket( "dlls/dwrite/bracket.c" ); dump_shaping( "dlls/usp10/shaping.c" ); dump_linebreak( "dlls/usp10/linebreak.c" ); dump_linebreak( "dlls/dwrite/linebreak.c" ); dump_scripts( "dlls/dwrite/scripts" ); dump_indic( "dlls/usp10/indicsyllable.c" ); dump_intl_nls("loader/l_intl.nls"); dump_vertical( "dlls/gdi32/vertical.c" ); dump_vertical( "dlls/wineps.drv/vertical.c" ); dump_nameprep( "dlls/kernel32/nameprep.c" ); foreach my $file (@allfiles) { HANDLE_FILE( @{$file} ); } output_cptable("libs/port/cptable.c"); exit 0; # Local Variables: # compile-command: "./make_unicode" # End: