Commit | Line | Data |
a0ed51b3 |
1 | #!../../miniperl |
2 | |
11695a73 |
3 | $UnicodeData = "UnicodeData-Latest.txt"; |
4 | |
a0ed51b3 |
5 | # Note: we try to keep filenames unique within first 8 chars. Using |
6 | # subdirectories for the following helps. |
7 | mkdir "In", 0777; |
8 | mkdir "Is", 0777; |
9 | mkdir "To", 0777; |
11695a73 |
10 | mkdir "Eq", 0777; |
a0ed51b3 |
11 | |
12 | @todo = ( |
13 | # typical |
14 | |
b8c5462f |
15 | ['IsWord', '$cat =~ /^L[ulo]|^Nd/ or $code eq "005F"', ''], |
16 | ['IsAlnum', '$cat =~ /^L[ulo]|^Nd/', ''], |
17 | ['IsAlpha', '$cat =~ /^L[ulo]/', ''], |
18 | ['IsSpace', '$cat =~ /^Z/ or $code lt "0020" and chr(hex $code) =~ /^\s/', ''], |
19 | ['IsDigit', '$cat =~ /^Nd$/', ''], |
20 | ['IsUpper', '$cat =~ /^Lu$/', ''], |
21 | ['IsLower', '$cat =~ /^Ll$/', ''], |
22 | ['IsASCII', 'hex $code <= 127', ''], |
23 | ['IsCntrl', '$cat =~ /^C/', ''], |
24 | ['IsGraph', '$cat =~ /^[^C]/ and $code ne "0020"', ''], |
25 | ['IsPrint', '$cat =~ /^[^C]/', ''], |
26 | ['IsPunct', '$cat =~ /^P/', ''], |
27 | ['IsXDigit', '$code =~ /^00(3[0-9]|[46][1-6])$/', ''], |
28 | ['ToUpper', '$up', '$up'], |
29 | ['ToLower', '$down', '$down'], |
30 | ['ToTitle', '$title', '$title'], |
31 | ['ToDigit', '$dec ne ""', '$dec'], |
a0ed51b3 |
32 | |
33 | # Name |
34 | |
35 | ['Name', '$name', '$name'], |
36 | |
37 | # Category |
38 | |
39 | ['Category', '$cat', '$cat'], |
40 | |
41 | # Normative |
42 | |
43 | ['IsM', '$cat =~ /^M/', ''], # Mark |
44 | ['IsMn', '$cat eq "Mn"', ''], # Mark, Non-Spacing |
45 | ['IsMc', '$cat eq "Mc"', ''], # Mark, Combining |
46 | |
47 | ['IsN', '$cat =~ /^N/', ''], # Number |
48 | ['IsNd', '$cat eq "Nd"', ''], # Number, Decimal Digit |
49 | ['IsNo', '$cat eq "No"', ''], # Number, Other |
50 | |
51 | ['IsZ', '$cat =~ /^Z/', ''], # Zeparator |
52 | ['IsZs', '$cat eq "Zs"', ''], # Separator, Space |
53 | ['IsZl', '$cat eq "Zl"', ''], # Separator, Line |
54 | ['IsZp', '$cat eq "Zp"', ''], # Separator, Paragraph |
55 | |
56 | ['IsC', '$cat =~ /^C/', ''], # Crazy |
57 | ['IsCc', '$cat eq "Cc"', ''], # Other, Control or Format |
58 | ['IsCo', '$cat eq "Co"', ''], # Other, Private Use |
59 | ['IsCn', '$cat eq "Cn"', ''], # Other, Not Assigned |
60 | |
61 | # Informative |
62 | |
63 | ['IsL', '$cat =~ /^L/', ''], # Letter |
64 | ['IsLu', '$cat eq "Lu"', ''], # Letter, Uppercase |
65 | ['IsLl', '$cat eq "Ll"', ''], # Letter, Lowercase |
66 | ['IsLt', '$cat eq "Lt"', ''], # Letter, Titlecase |
67 | ['IsLm', '$cat eq "Lm"', ''], # Letter, Modifier |
68 | ['IsLo', '$cat eq "Lo"', ''], # Letter, Other |
69 | |
70 | ['IsP', '$cat =~ /^P/', ''], # Punctuation |
71 | ['IsPd', '$cat eq "Pd"', ''], # Punctuation, Dash |
72 | ['IsPs', '$cat eq "Ps"', ''], # Punctuation, Open |
73 | ['IsPe', '$cat eq "Pe"', ''], # Punctuation, Close |
74 | ['IsPo', '$cat eq "Po"', ''], # Punctuation, Other |
75 | |
76 | ['IsS', '$cat =~ /^S/', ''], # Symbol |
77 | ['IsSm', '$cat eq "Sm"', ''], # Symbol, Math |
78 | ['IsSc', '$cat eq "Sc"', ''], # Symbol, Currency |
79 | ['IsSo', '$cat eq "So"', ''], # Symbol, Other |
80 | |
81 | # Combining class |
82 | ['CombiningClass', '$comb', '$comb'], |
83 | |
84 | # BIDIRECTIONAL PROPERTIES |
85 | |
86 | ['Bidirectional', '$bid', '$bid'], |
87 | |
88 | # Strong types: |
89 | |
90 | ['IsBidiL', '$bid eq "L"', ''], # Left-Right; Most alphabetic, |
91 | # syllabic, and logographic |
92 | # characters (e.g., CJK |
93 | # ideographs) |
94 | ['IsBidiR', '$bid eq "R"', ''], # Right-Left; Arabic, Hebrew, |
95 | # and punctuation specific to |
96 | # those scripts |
97 | |
98 | # Weak types: |
99 | |
100 | ['IsBidiEN','$bid eq "EN"', ''], # European Number |
101 | ['IsBidiES','$bid eq "ES"', ''], # European Number Separator |
102 | ['IsBidiET','$bid eq "ET"', ''], # European Number Terminator |
103 | ['IsBidiAN','$bid eq "AN"', ''], # Arabic Number |
104 | ['IsBidiCS','$bid eq "CS"', ''], # Common Number Separator |
105 | |
106 | # Separators: |
107 | |
108 | ['IsBidiB', '$bid eq "B"', ''], # Block Separator |
109 | ['IsBidiS', '$bid eq "S"', ''], # Segment Separator |
110 | |
111 | # Neutrals: |
112 | |
113 | ['IsBidiWS','$bid eq "WS"', ''], # Whitespace |
114 | ['IsBidiON','$bid eq "ON"', ''], # Other Neutrals ; All other |
115 | # characters: punctuation, |
116 | # symbols |
117 | |
118 | # Decomposition |
119 | |
120 | ['Decomposition', '$decomp', '$decomp'], |
121 | ['IsDecoCanon', '$decomp && $decomp !~ /^</', ''], |
122 | ['IsDecoCompat', '$decomp =~ /^</', ''], |
123 | ['IsDCfont', '$decomp =~ /^<font>/', ''], |
124 | ['IsDCnoBreak', '$decomp =~ /^<noBreak>/', ''], |
125 | ['IsDCinitial', '$decomp =~ /^<initial>/', ''], |
126 | ['IsDCinital', '$decomp =~ /^<medial>/', ''], |
127 | ['IsDCfinal', '$decomp =~ /^<final>/', ''], |
128 | ['IsDCisolated', '$decomp =~ /^<isolated>/', ''], |
129 | ['IsDCcircle', '$decomp =~ /^<circle>/', ''], |
130 | ['IsDCsuper', '$decomp =~ /^<super>/', ''], |
131 | ['IsDCsub', '$decomp =~ /^<sub>/', ''], |
132 | ['IsDCvertical', '$decomp =~ /^<vertical>/', ''], |
133 | ['IsDCwide', '$decomp =~ /^<wide>/', ''], |
134 | ['IsDCnarrow', '$decomp =~ /^<narrow>/', ''], |
135 | ['IsDCsmall', '$decomp =~ /^<small>/', ''], |
136 | ['IsDCsquare', '$decomp =~ /^<square>/', ''], |
137 | ['IsDCcompat', '$decomp =~ /^<compat>/', ''], |
138 | |
139 | # Number |
140 | |
141 | ['Number', '$num', '$num'], |
142 | |
143 | # Mirrored |
144 | |
145 | ['IsMirrored', '$mir eq "Y"', ''], |
146 | |
147 | # Arabic |
148 | |
149 | ['ArabLink', '1', '$link'], |
150 | ['ArabLnkGrp', '1', '$linkgroup'], |
151 | |
152 | # Jamo |
153 | |
154 | ['JamoShort', '1', '$short'], |
155 | ); |
156 | |
157 | # This is not written for speed... |
158 | |
159 | foreach $file (@todo) { |
160 | my ($table, $wanted, $val) = @$file; |
161 | next if @ARGV and not grep { $_ eq $table } @ARGV; |
162 | print $table,"\n"; |
163 | if ($table =~ /^(Is|In|To)(.*)/) { |
164 | open(OUT, ">$1/$2.pl") or die "Can't create $1/$2.pl: $!\n"; |
165 | } |
166 | else { |
167 | open(OUT, ">$table.pl") or die "Can't create $table.pl: $!\n"; |
168 | } |
169 | print OUT <<"END"; |
170 | return <<'END'; |
171 | END |
172 | print OUT proplist($table, $wanted, $val); |
173 | print OUT "END\n"; |
174 | close OUT; |
175 | } |
176 | |
177 | # Must treat blocks specially. |
178 | |
179 | exit if @ARGV and not grep { $_ eq Block } @ARGV; |
180 | print "Block\n"; |
181 | open(UD, 'blocks.txt') or die "Can't open blocks.txt: $!\n"; |
182 | open(OUT, ">Block.pl") or die "Can't create $table.pl: $!\n"; |
183 | print OUT <<"END"; |
184 | return <<'END'; |
185 | END |
186 | |
187 | while (<UD>) { |
188 | next if /^#/; |
189 | next if /^$/; |
190 | chomp; |
191 | ($code, $last, $name) = split(/; */); |
192 | if ($name) { |
193 | print OUT "$code $last $name\n"; |
194 | $name =~ s/\s+//g; |
195 | open(BLOCK, ">In/$name.pl"); |
196 | print BLOCK <<"END2"; |
197 | return <<'END'; |
198 | $code $last |
199 | END |
200 | END2 |
201 | close BLOCK; |
202 | } |
203 | } |
204 | |
205 | print OUT "END\n"; |
206 | close OUT; |
207 | |
208 | ################################################## |
209 | |
210 | sub proplist { |
211 | my ($table, $wanted, $val) = @_; |
212 | my @wanted; |
213 | my $out; |
214 | my $split; |
215 | |
216 | if ($table =~ /^Arab/) { |
217 | open(UD, "arabshp.txt") or warn "Can't open $table: $!"; |
218 | |
219 | $split = '($code, $name, $link, $linkgroup) = split(/; */);'; |
220 | } |
221 | elsif ($table =~ /^Jamo/) { |
222 | open(UD, "jamo2.txt") or warn "Can't open $table: $!"; |
223 | |
224 | $split = '($code, $short, $name) = split(/; */); $code =~ s/^U\+//;'; |
225 | } |
226 | else { |
11695a73 |
227 | open(UD, $UnicodeData) or warn "Can't open $UnicodeData: $!"; |
a0ed51b3 |
228 | |
229 | $split = '($code, $name, $cat, $comb, $bid, $decomp, $dec, $dig, $num, $mir, $uni1, |
230 | $comment, $up, $down, $title) = split(/;/);'; |
231 | } |
232 | |
233 | if ($table =~ /^(?:To|Is)[A-Z]/) { |
234 | eval <<"END"; |
235 | while (<UD>) { |
236 | next if /^#/; |
237 | next if /^\s/; |
238 | chop; |
239 | $split |
240 | if ($wanted) { |
241 | push(\@wanted, [hex \$code, hex $val, \$name =~ /, First>\$/]); |
242 | } |
243 | } |
244 | END |
245 | die $@ if $@; |
246 | |
247 | while (@wanted) { |
248 | $beg = shift @wanted; |
249 | $last = $beg; |
250 | while (@wanted and $wanted[0]->[0] == $last->[0] + 1 and |
251 | (not $val or $wanted[0]->[1] == $last->[1] + 1)) { |
252 | $last = shift @wanted; |
253 | } |
254 | $out .= sprintf "%04x", $beg->[0]; |
255 | if ($beg->[2]) { |
256 | $last = shift @wanted; |
257 | } |
258 | if ($beg == $last) { |
259 | $out .= "\t"; |
260 | } |
261 | else { |
262 | $out .= sprintf "\t%04x", $last->[0]; |
263 | } |
264 | $out .= sprintf "\t%04x", $beg->[1] if $val; |
265 | $out .= "\n"; |
266 | } |
267 | } |
268 | else { |
269 | eval <<"END"; |
270 | while (<UD>) { |
271 | next if /^#/; |
272 | next if /^\s*\$/; |
273 | chop; |
274 | $split |
275 | if ($wanted) { |
276 | push(\@wanted, [hex \$code, $val, \$name =~ /, First>\$/]); |
277 | } |
278 | } |
279 | END |
280 | die $@ if $@; |
281 | |
282 | while (@wanted) { |
283 | $beg = shift @wanted; |
284 | $last = $beg; |
285 | while (@wanted and $wanted[0]->[0] == $last->[0] + 1 and |
286 | ($wanted[0]->[1] eq $last->[1])) { |
287 | $last = shift @wanted; |
288 | } |
289 | $out .= sprintf "%04x", $beg->[0]; |
290 | if ($beg->[2]) { |
291 | $last = shift @wanted; |
292 | } |
293 | if ($beg == $last) { |
294 | $out .= "\t"; |
295 | } |
296 | else { |
297 | $out .= sprintf "\t%04x", $last->[0]; |
298 | } |
299 | $out .= sprintf "\t%s\n", $beg->[1]; |
300 | } |
301 | } |
302 | $out; |
303 | } |
11695a73 |
304 | |
11eeea96 |
305 | # Create the equivalence mappings. |
306 | |
11695a73 |
307 | open(UNICODEDATA, $UnicodeData) || die "$0: $UnicodeData: $!\n"; |
308 | |
309 | while (<UNICODEDATA>) { |
310 | ($code, $name) = split /;/; |
311 | |
312 | $code{$name} = $code; |
313 | $name{$code} = $name; |
314 | |
315 | if ($name =~ /^((?:LATIN|GREEK|CYRILLIC|HEBREW|BENGALI) .+? LETTER .+?) WITH /) { |
316 | push @base, [ $code, $1 ]; |
317 | } elsif ($name =~ /^(ARABIC LETTER \w+?) WITH .+ (\w+ FORM)$/) { |
318 | push @base, [ $code, "$1 $2" ]; |
319 | } elsif ($name =~ /^(ARABIC LETTER \w+?) WITH /) { |
320 | push @base, [ $code, $1 ]; |
321 | # Is the concept of turning ligatures into character classes sound? |
322 | } elsif ($name =~ /^(ARABIC) LIGATURE (.+?) (WITH .+ )+(\w+ FORM)$/) { |
323 | my $script = $1; |
324 | my $base = $2; |
325 | my $with = $3; |
326 | my $form = $4; |
327 | push @base, [ $code, "$script LETTER $base" ]; |
328 | push @base, [ $code, "$script LETTER $base $form" ]; |
329 | my @with = split(/\bWITH\s+/, $with); |
330 | shift @with; |
331 | @with = grep { ! /^ (?:ABOVE|BELOW)/ } @with; |
332 | foreach my $base (@with) { |
333 | push @base, [ $code, "$script LETTER $base" ]; |
334 | push @base, [ $code, "$script LETTER $base $form" ]; |
335 | } |
336 | } elsif ($name =~ /^((?:ARMENIAN|CYRILLIC) .+) LIGATURE (\w+) (\w+)$/) { |
337 | push @base, [ $code, "$1 LETTER $2" ]; |
338 | push @base, [ $code, "$1 LETTER $3" ]; |
339 | # Latin ligatures (ae, oe, ij, ff, fi, fl, ffi, ffl, long st, st) ignored. |
340 | # Hebrew Yiddish ligatures (double vav, vav yod, double yod, yod yod patah, |
341 | # alef lamed) ignored. |
342 | } else { |
343 | next; |
344 | } |
345 | |
346 | } |
347 | |
348 | foreach my $b (@base) { |
349 | ($code, $base) = @$b; |
350 | next unless exists $code{$base}; |
351 | push @{$unicode{$code{$base}}}, $code; |
352 | print "$code: $name{$code} -> $base\n", |
353 | } |
354 | |
355 | @unicode = sort keys %unicode; |
356 | |
357 | if (open(EQ_UNICODE, ">Eq/Unicode")) { |
358 | foreach my $c (@unicode) { |
359 | print EQ_UNICODE "$c @{$unicode{$c}}\n"; |
360 | } |
361 | close EQ_UNICODE; |
362 | } else { |
363 | die "$0: failed to open Eq/Unicode for writing: $!\n"; |
364 | } |
365 | |
366 | if (open(EQ_LATIN1, ">Eq/Latin1")) { |
367 | foreach my $c (@unicode) { |
368 | last if hex($c) > 255; |
369 | my @c = grep { hex($_) <= 255 } @{$unicode{$c}}; |
370 | next unless @c; |
371 | print EQ_LATIN1 "$c @c\n"; |
372 | } |
373 | close EQ_LATIN1; |
374 | } else { |
375 | die "$0: failed to open Eq/Latin1 for writing: $!\n"; |
376 | } |
377 | |