Commit | Line | Data |
d73e5302 |
1 | #!/usr/bin/perl -w |
d73e5302 |
2 | use strict; |
cf25bb62 |
3 | use Carp; |
4 | ## |
5 | ## mktables -- create the runtime Perl Unicode files (lib/unicore/**/*.pl) |
6 | ## from the Unicode database files (lib/unicore/*.txt). |
7 | ## |
d2d499f5 |
8 | |
d73e5302 |
9 | mkdir("In", 0755); |
10 | mkdir("Is", 0755); |
11 | mkdir("To", 0755); |
12 | |
cf25bb62 |
13 | ## |
14 | ## Process any args. |
15 | ## |
16 | my $Verbose = 0; |
17 | |
18 | while (@ARGV) |
19 | { |
20 | my $arg = shift @ARGV; |
21 | if ($arg eq '-v') { |
22 | $Verbose = 1; |
23 | } elsif ($arg eq '-q') { |
24 | $Verbose = 0; |
d73e5302 |
25 | } else { |
cf25bb62 |
26 | die "usage: $0 [-v|-q]"; |
d73e5302 |
27 | } |
28 | } |
29 | |
cf25bb62 |
30 | my $LastUnicodeCodepoint = 0x10FFFF; # As of Unicode 3.1.1. |
d73e5302 |
31 | |
cf25bb62 |
32 | my $now = localtime; |
33 | my $HEADER=<<"EOF"; |
d73e5302 |
34 | # !!!!!!! DO NOT EDIT THIS FILE !!!!!!! |
35 | # This file is built by $0 from e.g. Unicode.txt. |
36 | # Any changes made here will be lost! |
cf25bb62 |
37 | # Built $now. |
38 | |
39 | EOF |
40 | |
41 | ## |
42 | ## The main datastructure (a "Table") represents a set of code points that |
43 | ## are part of a particular quality (that are part of \pL, \p{InGreek}, |
44 | ## etc.). They are kept as ranges of code points (starting and ending of |
45 | ## each range). |
46 | ## |
47 | ## For example, a range ASCII LETTERS would be represented as: |
48 | ## [ [ 0x41 => 0x5A, 'UPPER' ], |
49 | ## [ 0x61 => 0x7A, 'LOWER, ] ] |
50 | ## |
51 | sub RANGE_START() { 0 } ## index into range element |
52 | sub RANGE_END() { 1 } ## index into range element |
53 | sub RANGE_NAME() { 2 } ## index into range element |
54 | |
55 | my %TableInfo; |
56 | my %FuzzyNames; |
57 | my %AliasInfo; |
58 | |
59 | ## |
60 | ## Associates a property ("Greek", "Lu", "Assigned",...) with a Table. |
61 | ## |
62 | ## Called like: |
63 | ## New_Prop(In => 'Greek', $Table, AllowFuzzy => 1); |
64 | ## |
65 | ## Normally, these parameters are set when the Table is created (when the |
66 | ## Table->New constructor is called), but there are times when it needs to |
67 | ## be done after-the-fact...) |
68 | ## |
69 | sub New_Prop($$$@) |
70 | { |
71 | my $Type = shift; ## "Is" or "In"; |
72 | my $Name = shift; |
73 | my $Table = shift; |
74 | |
75 | ## remaining args are optional key/val |
76 | my %Args = @_; |
77 | |
78 | my $AllowFuzzy = delete $Args{AllowFuzzy}; |
79 | |
80 | ## sanity check a few args |
81 | if (%Args or ($Type ne 'Is' and $Type ne 'In') or not ref $Table) { |
82 | confess "$0: bad args to New_Prop" |
83 | } |
84 | |
85 | if (not $TableInfo{$Type}->{$Name}) |
86 | { |
87 | $TableInfo{$Type}->{$Name} = $Table; |
88 | if ($AllowFuzzy) { |
89 | $FuzzyNames{$Type}->{$Name} = $Name; |
90 | } |
91 | } |
d73e5302 |
92 | } |
93 | |
d73e5302 |
94 | |
cf25bb62 |
95 | ## |
96 | ## Creates a new Table object. |
97 | ## |
98 | ## Args are key/value pairs: |
99 | ## In => Name -- Name of "In" property to be associated with |
100 | ## Is => Name -- Name of "Is" property to be associated with |
101 | ## AllowFuzzy => Boolean -- True if name can be accessed "fuzzily" |
102 | ## |
103 | ## No args are required. |
104 | ## |
105 | sub Table::New |
106 | { |
107 | my $class = shift; |
108 | my %Args = @_; |
109 | |
110 | my $Table = bless [], $class; |
111 | |
112 | my $AllowFuzzy = delete $Args{AllowFuzzy}; |
113 | |
114 | for my $Type ('Is', 'In') |
115 | { |
116 | if (my $Name = delete $Args{$Type}) { |
117 | New_Prop($Type => $Name, $Table, AllowFuzzy => $AllowFuzzy); |
118 | } |
119 | } |
120 | |
121 | ## shouldn't have any left over |
122 | if (%Args) { |
123 | confess "$0: bad args to Table->New" |
124 | } |
125 | |
126 | return $Table; |
d73e5302 |
127 | } |
128 | |
cf25bb62 |
129 | ## |
130 | ## Returns true if the Table has no code points |
131 | ## |
132 | sub Table::IsEmpty |
133 | { |
134 | my $Table = shift; #self |
135 | return not @$Table; |
d73e5302 |
136 | } |
137 | |
cf25bb62 |
138 | ## |
139 | ## Returns true if the Table has code points |
140 | ## |
141 | sub Table::NotEmpty |
142 | { |
143 | my $Table = shift; #self |
144 | return @$Table; |
d73e5302 |
145 | } |
146 | |
cf25bb62 |
147 | ## |
148 | ## Returns the maximum code point currently in the table. |
149 | ## |
150 | sub Table::Max |
151 | { |
152 | my $Table = shift; #self |
153 | confess "oops" if $Table->IsEmpty; ## must have code points to have a max |
154 | return $Table->[-1]->[RANGE_END]; |
155 | } |
d73e5302 |
156 | |
cf25bb62 |
157 | ## |
158 | ## Replaces the codepoints in the Table with those in the Table given |
159 | ## as an arg. (NOTE: this is not a "deep copy"). |
160 | ## |
161 | sub Table::Replace($$) |
162 | { |
163 | my $Table = shift; #self |
164 | my $New = shift; |
d73e5302 |
165 | |
cf25bb62 |
166 | @$Table = @$New; |
167 | } |
71d929cb |
168 | |
cf25bb62 |
169 | ## |
170 | ## Given a new code point, make the last range of the Table extend to |
171 | ## include the new (and all intervening) code points. |
172 | ## |
173 | sub Table::Extend |
174 | { |
175 | my $Table = shift; #self |
176 | my $codepoint = shift; |
d73e5302 |
177 | |
cf25bb62 |
178 | my $PrevMax = $Table->Max; |
e904f995 |
179 | |
cf25bb62 |
180 | confess "oops ($codepoint <= $PrevMax)" if $codepoint <= $PrevMax; |
e904f995 |
181 | |
cf25bb62 |
182 | $Table->[-1]->[RANGE_END] = $codepoint; |
183 | } |
c3a8a2b8 |
184 | |
cf25bb62 |
185 | ## |
186 | ## Given a code point range start and end (and optional name), blindly |
187 | ## append them to the list of ranges for the Table. |
188 | ## |
189 | ## NOTE: Code points must be added in strictly ascending numeric order. |
190 | ## |
191 | sub Table::RawAppendRange |
192 | { |
193 | my $Table = shift; #self |
194 | my $start = shift; |
195 | my $end = shift; |
196 | my $name = shift; |
197 | $name = "" if not defined $name; ## warning: $name can be "0" |
198 | |
199 | push @$Table, [ $start, # RANGE_START |
200 | $end, # RANGE_END |
201 | $name ]; # RANGE_NAME |
202 | } |
c3a8a2b8 |
203 | |
cf25bb62 |
204 | ## |
205 | ## Given a code point (and optional name), add it to the Table. |
206 | ## |
207 | ## NOTE: Code points must be added in strictly ascending numeric order. |
208 | ## |
209 | sub Table::Append |
210 | { |
211 | my $Table = shift; #self |
212 | my $codepoint = shift; |
213 | my $name = shift; |
214 | $name = "" if not defined $name; ## warning: $name can be "0" |
215 | |
216 | ## |
217 | ## If we've already got a range working, and this code point is the next |
218 | ## one in line, and if the name is the same, just extend the current range. |
219 | ## |
220 | if ($Table->NotEmpty |
221 | and |
222 | $Table->Max == $codepoint - 1 |
223 | and |
224 | $Table->[-1]->[RANGE_NAME] eq $name) |
225 | { |
226 | $Table->Extend($codepoint); |
c3a8a2b8 |
227 | } |
cf25bb62 |
228 | else |
229 | { |
230 | $Table->RawAppendRange($codepoint, $codepoint, $name); |
d73e5302 |
231 | } |
cf25bb62 |
232 | } |
d73e5302 |
233 | |
cf25bb62 |
234 | ## |
235 | ## Given a code point range starting value and ending value (and name), |
236 | ## Add the range to teh Table. |
237 | ## |
238 | ## NOTE: Code points must be added in strictly ascending numeric order. |
239 | ## |
240 | sub Table::AppendRange |
241 | { |
242 | my $Table = shift; #self |
243 | my $start = shift; |
244 | my $end = shift; |
245 | my $name = shift; |
246 | $name = "" if not defined $name; ## warning: $name can be "0" |
247 | |
248 | $Table->Append($start, $name); |
249 | $Table->Extend($end) if $end > $start; |
250 | } |
d73e5302 |
251 | |
cf25bb62 |
252 | ## |
253 | ## Return a new Table that represents all code points not in the Table. |
254 | ## |
255 | sub Table::Invert |
256 | { |
257 | my $Table = shift; #self |
258 | |
259 | my $New = Table->New(); |
260 | my $max = -1; |
261 | for my $range (@$Table) |
262 | { |
263 | my $start = $range->[RANGE_START]; |
264 | my $end = $range->[RANGE_END]; |
265 | if ($start-1 >= $max+1) { |
266 | $New->AppendRange($max+1, $start-1, ""); |
267 | } |
268 | $max = $end; |
d73e5302 |
269 | } |
cf25bb62 |
270 | if ($max+1 < $LastUnicodeCodepoint) { |
271 | $New->AppendRange($max+1, $LastUnicodeCodepoint); |
d73e5302 |
272 | } |
cf25bb62 |
273 | return $New; |
274 | } |
d73e5302 |
275 | |
cf25bb62 |
276 | ## |
277 | ## Merges any number of other tables with $self, returning the new table. |
278 | ## (existing tables are not modified) |
279 | ## |
280 | ## Can be called as either a constructor or a method. |
281 | ## |
282 | sub Table::Merge |
283 | { |
284 | shift(@_) if not ref $_[0]; ## if called as a constructor, lose the class |
285 | my @Tables = @_; |
286 | |
287 | ## Accumulate all records from all tables |
288 | my @Records; |
289 | for my $Table (@Tables) { |
290 | push @Records, @$Table; |
d73e5302 |
291 | } |
292 | |
cf25bb62 |
293 | ## sort by range start, with longer ranges coming first. |
294 | my ($first, @Rest) = sort { |
295 | ($a->[RANGE_START] <=> $b->[RANGE_START]) |
296 | or |
297 | ($b->[RANGE_END] <=> $b->[RANGE_END]) |
298 | } @Records; |
299 | |
300 | my $New = Table->New(); |
301 | |
302 | ## Ensuring the first range is there makes the subsequent loop easier |
303 | $New->AppendRange($first->[RANGE_START], |
304 | $first->[RANGE_END]); |
305 | |
306 | ## Fold in records so long as they add new information. |
307 | for my $set (@Rest) |
308 | { |
309 | my $start = $set->[RANGE_START]; |
310 | my $end = $set->[RANGE_END]; |
311 | if ($start > $New->Max) { |
312 | $New->AppendRange($start, $end); |
313 | } elsif ($end > $New->Max) { |
314 | $New->Extend($end); |
315 | } |
d73e5302 |
316 | } |
d73e5302 |
317 | |
cf25bb62 |
318 | return $New; |
d73e5302 |
319 | } |
320 | |
cf25bb62 |
321 | ## |
322 | ## Given a filename, write a representation of the Table to a file. |
323 | ## |
324 | sub Table::Write |
325 | { |
326 | my $Table = shift; #self |
327 | my $filename = shift; |
d73e5302 |
328 | |
cf25bb62 |
329 | print "$filename\n" if $Verbose; |
d73e5302 |
330 | |
cf25bb62 |
331 | if (not open(OUT, ">$filename")) { |
332 | die "$0: can't write $filename: $!\n"; |
333 | } |
d73e5302 |
334 | |
cf25bb62 |
335 | print OUT $HEADER; |
336 | print OUT "return <<'END';\n"; |
d73e5302 |
337 | |
cf25bb62 |
338 | for my $set (@$Table) |
339 | { |
340 | my $start = $set->[RANGE_START]; |
341 | my $end = $set->[RANGE_END]; |
342 | my $name = $set->[RANGE_NAME]; |
d73e5302 |
343 | |
cf25bb62 |
344 | if ($start == $end) { |
345 | printf OUT "%04X\t\t%s\n", $start, $name; |
346 | } else { |
347 | printf OUT "%04X\t%04X\t%s\n", $start, $end, $name; |
348 | } |
349 | } |
d73e5302 |
350 | |
cf25bb62 |
351 | print OUT "END\n"; |
352 | close OUT; |
353 | } |
354 | |
355 | ########################################################################### |
356 | ########################################################################### |
357 | ########################################################################### |
358 | |
359 | |
360 | ## |
361 | ## Called like: |
362 | ## New_Alias(Is => 'All', SameAs => 'Any', AllowFuzzy => 1); |
363 | ## |
364 | ## The args must be in that order, although the AllowFuzzy pair may be omitted. |
365 | ## |
366 | ## This creates 'IsAll' as an alias for 'IsAny' |
367 | ## |
368 | sub New_Alias($$$@) |
369 | { |
370 | my $Type = shift; ## "Is" or "In" |
371 | my $Alias = shift; |
372 | my $SameAs = shift; |
373 | my $Name = shift; |
374 | |
375 | ## remaining args are optional key/val |
376 | my %Args = @_; |
377 | |
378 | my $AllowFuzzy = delete $Args{AllowFuzzy}; |
379 | |
380 | ## sanity check a few args |
381 | if (%Args or ($Type ne 'Is' and $Type ne 'In') or $SameAs ne 'SameAs') { |
382 | confess "$0: bad args to New_Alias" |
d73e5302 |
383 | } |
384 | |
cf25bb62 |
385 | if (not $TableInfo{$Type}->{$Name}) { |
386 | confess "$0: don't have orignial $Type => $Name to make alias" |
387 | } |
388 | if ($TableInfo{$Alias}) { |
389 | confess "$0: already have original $Type => $Alias; can't make alias"; |
d73e5302 |
390 | } |
cf25bb62 |
391 | $AliasInfo{$Type}->{$Name} = $Alias; |
392 | if ($AllowFuzzy) { |
393 | $FuzzyNames{$Type}->{$Alias} = $Name; |
394 | } |
395 | |
d73e5302 |
396 | } |
397 | |
cf25bb62 |
398 | ## |
399 | ## Turn something like |
400 | ## OLD-ITALIC |
401 | ## to |
402 | ## Old_Italic |
403 | ## |
404 | sub CanonicalName($) |
405 | { |
406 | my $name = lc shift; |
407 | $name =~ s/\W+/_/; |
408 | $name =~ s/(?<![a-z])(\w)/\u$1/g; |
409 | return $name; |
410 | } |
d73e5302 |
411 | |
d73e5302 |
412 | |
cf25bb62 |
413 | ## All assigned code points |
414 | my $Assigned = Table->New(Is => 'Assigned', AllowFuzzy => 1); |
d2d499f5 |
415 | |
cf25bb62 |
416 | my $Name = Table->New(); ## all characters, individually by name |
417 | my $General = Table->New(); ## all characters, grouped by category |
418 | my %General; |
419 | my %Cat; |
d73e5302 |
420 | |
cf25bb62 |
421 | ## |
422 | ## Process Unicode.txt (Categories, etc.) |
423 | ## |
424 | sub Unicode_Txt() |
425 | { |
426 | my $Bidi = Table->New(); |
427 | my $Deco = Table->New(); |
428 | my $Comb = Table->New(); |
429 | my $Number = Table->New(); |
430 | my $Mirrored = Table->New(Is => 'Mirrored', AllowFuzzy => 0); |
d73e5302 |
431 | |
cf25bb62 |
432 | my %DC; |
433 | my %Bidi; |
434 | my %Deco; |
435 | $Deco{Canon} = Table->New(Is => 'Canon', AllowFuzzy => 0); |
436 | $Deco{Compat} = Table->New(Is => 'Compat', AllowFuzzy => 0); |
437 | |
438 | ## Initialize Perl-generated categories |
439 | $Cat{Alnum} = Table->New(Is => 'Alnum', AllowFuzzy => 0); |
440 | $Cat{Alpha} = Table->New(Is => 'Alpha', AllowFuzzy => 0); |
441 | $Cat{ASCII} = Table->New(Is => 'ASCII', AllowFuzzy => 0); |
442 | $Cat{Blank} = Table->New(Is => 'Blank', AllowFuzzy => 0); |
443 | $Cat{Cntrl} = Table->New(Is => 'Cntrl', AllowFuzzy => 0); |
444 | $Cat{Digit} = Table->New(Is => 'Digit', AllowFuzzy => 0); |
445 | $Cat{Graph} = Table->New(Is => 'Graph', AllowFuzzy => 0); |
446 | $Cat{Lower} = Table->New(Is => 'Lower', AllowFuzzy => 0); |
447 | $Cat{Print} = Table->New(Is => 'Print', AllowFuzzy => 0); |
448 | $Cat{Punct} = Table->New(Is => 'Punct', AllowFuzzy => 0); |
449 | $Cat{SpacePerl} = Table->New(Is => 'SpacePerl', AllowFuzzy => 0); |
450 | $Cat{Space} = Table->New(Is => 'Space', AllowFuzzy => 0); |
451 | $Cat{Title} = Table->New(Is => 'Title', AllowFuzzy => 0); |
452 | $Cat{Upper} = Table->New(Is => 'Upper', AllowFuzzy => 0); |
453 | $Cat{Word} = Table->New(Is => 'Word' , AllowFuzzy => 0); |
454 | $Cat{XDigit} = Table->New(Is => 'XDigit', AllowFuzzy => 0); |
455 | ## Categories from Unicode.txt are auto-initialized in gencat() |
d73e5302 |
456 | |
cf25bb62 |
457 | my %To; |
458 | $To{Upper} = Table->New(); |
459 | $To{Lower} = Table->New(); |
460 | $To{Title} = Table->New(); |
461 | $To{Digit} = Table->New(); |
462 | |
463 | sub gencat($$$$) |
464 | { |
465 | my ($name, ## Name ("LATIN CAPITAL LETTER A") |
466 | $cat, ## Category ("Lu", "Zp", "Nd", etc.) |
467 | $code, ## Code point (as an integer) |
468 | $op) = @_; |
469 | |
470 | my $MajorCat = substr($cat, 0, 1); ## L, M, Z, S, etc |
471 | |
472 | $Assigned->$op($code); |
473 | $Name->$op($code, $name); |
474 | $General->$op($code, $cat); |
475 | |
476 | ## add to the sub category (e.g. "Lu", "Nd", "Cf", ..) |
477 | $Cat{$cat} ||= Table->New(Is => $cat, AllowFuzzy => 0); |
478 | $Cat{$cat}->$op($code); |
479 | |
480 | ## add to the major category (e.g. "L", "N", "C", ...) |
481 | $Cat{$MajorCat} ||= Table->New(Is => $MajorCat, AllowFuzzy => 0); |
482 | $Cat{$MajorCat}->$op($code); |
483 | |
484 | ($General{$name} ||= Table->New)->$op($code, $name); |
485 | |
486 | # 005F: SPACING UNDERSCORE |
487 | $Cat{Word}->$op($code) if $cat =~ /^[LMN]/ || $code == 0x005F; |
488 | $Cat{Alnum}->$op($code) if $cat =~ /^[LMN]/; |
489 | $Cat{Alpha}->$op($code) if $cat =~ /^[LM]/; |
490 | |
491 | |
492 | |
493 | $Cat{Space}->$op($code) if $cat =~ /^Z/ |
494 | || $code == 0x0009 # 0009: HORIZONTAL TAB |
495 | || $code == 0x000A # 000A: LINE FEED |
496 | || $code == 0x000B # 000B: VERTICAL TAB |
497 | || $code == 0x000C # 000C: FORM FEED |
498 | || $code == 0x000D; # 000D: CARRIAGE RETURN |
499 | |
500 | |
501 | $Cat{SpacePerl}->$op($code) if $cat =~ /^Z/ |
502 | || $code == 0x0009 # 0009: HORIZONTAL TAB |
503 | || $code == 0x000A # 000A: LINE FEED |
504 | || $code == 0x000C # 000C: FORM FEED |
505 | || $code == 0x000D # 000D: CARRIAGE RETURN |
506 | || $code == 0x0085 # 0085: <NEXT LINE> |
507 | || $code == 0x2028 # 2028: LINE SEPARATOR |
508 | || $code == 0x2029;# 2029: PARAGRAPH SEP. |
509 | |
510 | $Cat{Blank}->$op($code) if $cat =~ /^Z[^lp]$/ |
511 | || $code == 0x0009 # 0009: HORIZONTAL TAB |
512 | || $code == 0x0020; # 0020: SPACE |
513 | |
514 | $Cat{Digit}->$op($code) if $cat eq "Nd"; |
515 | $Cat{Upper}->$op($code) if $cat eq "Lu"; |
516 | $Cat{Lower}->$op($code) if $cat eq "Ll"; |
517 | $Cat{Title}->$op($code) if $cat eq "Lt"; |
518 | $Cat{ASCII}->$op($code) if $code <= 0x007F; |
519 | $Cat{Cntrl}->$op($code) if $cat =~ /^C/; |
520 | $Cat{Graph}->$op($code) if $cat =~ /^([LMNPS]|Co)/; |
521 | $Cat{Print}->$op($code) if $cat =~ /^([LMNPS]|Co|Zs)/; |
522 | $Cat{Punct}->$op($code) if $cat =~ /^P/; |
523 | |
524 | $Cat{XDigit}->$op($code) if ($code >= 0x30 && $code <= 0x39) ## 0..9 |
525 | || ($code >= 0x41 && $code <= 0x46) ## A..F |
526 | || ($code >= 0x61 && $code <= 0x66); ## a..f |
527 | } |
d73e5302 |
528 | |
cf25bb62 |
529 | ## open ane read file..... |
530 | if (not open IN, "Unicode.txt") { |
531 | die "$0: Unicode.txt: $!\n"; |
532 | } |
d73e5302 |
533 | |
cf25bb62 |
534 | while (<IN>) |
535 | { |
536 | next unless /^[0-9A-Fa-f]+;/; |
537 | s/\s+$//; |
538 | |
539 | my ($hexcode, ## code point in hex (e.g. "0041") |
540 | $name, ## character name (e.g. "LATIN CAPITAL LETTER A") |
541 | $cat, ## category (e.g. "Lu") |
542 | $comb, ## Canonical combining class (e.t. "230") |
543 | $bidi, ## directional category (e.g. "L") |
544 | $deco, ## decomposition mapping |
545 | $decimal, ## decimal digit value |
546 | $digit, ## digit value |
547 | $number, ## numeric value |
548 | $mirrored, ## mirrored |
549 | $unicode10, ## name in Unicode 1.0 |
550 | $comment, ## comment field |
551 | $upper, ## uppercase mapping |
552 | $lower, ## lowercase mapping |
553 | $title, ## titlecase mapping |
554 | ) = split(/\s*;\s*/); |
555 | |
556 | my $code = hex($hexcode); |
557 | |
558 | ## |
559 | ## There are a few pairs of lines like: |
560 | ## AC00;<Hangul Syllable, First>;Lo;0;L;;;;;N;;;;; |
561 | ## D7A3;<Hangul Syllable, Last>;Lo;0;L;;;;;N;;;;; |
562 | ## that define ranges. |
563 | ## |
564 | if ($name =~ /^<(.+), (First|Last)>$/) |
565 | { |
566 | $name = $1; |
567 | gencat($name, $cat, $code, $2 eq 'First' ? 'Append' : 'Extend'); |
568 | #New_Prop(In => $name, $General{$name}, AllowFuzzy => 1); |
569 | } |
570 | else |
571 | { |
572 | ## normal (single-character) lines |
573 | gencat($name, $cat, $code, 'Append'); |
574 | |
575 | # No Append() here since since several codes may map into one. |
576 | $To{Upper}->RawAppendRange($code, $code, $upper) if $upper; |
577 | $To{Lower}->RawAppendRange($code, $code, $lower) if $lower; |
578 | $To{Title}->RawAppendRange($code, $code, $title) if $title; |
579 | $To{Digit}->Append($code, $decimal) if length $decimal; |
580 | |
581 | $Bidi->Append($code, $bidi); |
582 | $Comb->Append($code, $comb) if $comb; |
583 | $Number->Append($code, $number) if length $number; |
584 | |
585 | $Mirrored->Append($code) if $mirrored eq "Y"; |
586 | |
587 | $Bidi{$bidi} ||= Table->New(Is => "Bidi$bidi", AllowFuzzy => 0); |
588 | $Bidi{$bidi}->Append($code); |
589 | |
590 | if ($deco) |
591 | { |
592 | $Deco->Append($code, $deco); |
593 | if ($deco =~/^<(\w+)>/) |
594 | { |
595 | $Deco{Compat}->Append($code); |
596 | |
597 | $DC{$1} ||= Table->New(Is => "DC$1", AllowFuzzy => 0); |
598 | $DC{$1}->Append($code); |
599 | } |
600 | else |
601 | { |
602 | $Deco{Canon}->Append($code); |
603 | } |
604 | } |
605 | } |
606 | } |
607 | close IN; |
d2d499f5 |
608 | |
cf25bb62 |
609 | ## |
610 | ## Tidy up a few special cases.... |
611 | ## |
d73e5302 |
612 | |
cf25bb62 |
613 | $Cat{Cn} = $Assigned->Invert; ## Cn is everything that doesn't exist |
614 | New_Prop(Is => 'Cn', $Cat{Cn}, AllowFuzzy => 0); |
d73e5302 |
615 | |
cf25bb62 |
616 | ## Unassigned is the same as 'Cn' |
617 | New_Alias(Is => 'Unassigned', SameAs => 'Cn', AllowFuzzy => 1); |
d73e5302 |
618 | |
cf25bb62 |
619 | $Cat{C}->Replace($Cat{C}->Merge($Cat{Cn})); ## Now merge in Cn into C |
d73e5302 |
620 | |
d73e5302 |
621 | |
cf25bb62 |
622 | # L& is Ll, Lu, and Lt. |
623 | New_Prop(Is => 'L&', |
624 | Table->Merge(@Cat{qw[Ll Lu Lt]}), |
625 | AllowFuzzy => 0); |
d73e5302 |
626 | |
cf25bb62 |
627 | ## Any and All are all code points. |
628 | my $Any = Table->New(Is => 'Any', AllowFuzzy => 1); |
629 | $Any->RawAppendRange(0, $LastUnicodeCodepoint); |
d73e5302 |
630 | |
cf25bb62 |
631 | New_Alias(Is => 'All', SameAs => 'Any', AllowFuzzy => 1); |
d73e5302 |
632 | |
d73e5302 |
633 | |
cf25bb62 |
634 | ## |
635 | ## Now dump the files. |
636 | ## |
637 | $Name->Write("Name.pl"); |
638 | $Bidi->Write("Bidirectional.pl"); |
639 | $Comb->Write("CombiningClass.pl"); |
640 | $Deco->Write("Decomposition.pl"); |
641 | $Number->Write("Number.pl"); |
642 | $General->Write("Category.pl"); |
643 | |
644 | for my $to (sort keys %To) { |
645 | $To{$to}->Write("To/$to.pl"); |
d73e5302 |
646 | } |
647 | } |
648 | |
cf25bb62 |
649 | ## |
650 | ## Process LineBrk.txt |
651 | ## |
652 | sub LineBrk_Txt() |
653 | { |
654 | if (not open IN, "LineBrk.txt") { |
655 | die "$0: LineBrk.txt: $!\n"; |
656 | } |
d73e5302 |
657 | |
cf25bb62 |
658 | my $Lbrk = Table->New(); |
659 | my %Lbrk; |
d73e5302 |
660 | |
cf25bb62 |
661 | while (<IN>) |
662 | { |
663 | next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(\w+)/; |
d73e5302 |
664 | |
cf25bb62 |
665 | my ($first, $last, $lbrk) = (hex($1), hex($2||""), $3); |
d73e5302 |
666 | |
cf25bb62 |
667 | $Lbrk->Append($first, $lbrk); |
d73e5302 |
668 | |
cf25bb62 |
669 | $Lbrk{$lbrk} ||= Table->New(Is => "Lbrk$lbrk", AllowFuzzy => 0); |
670 | $Lbrk{$lbrk}->Append($first); |
d73e5302 |
671 | |
cf25bb62 |
672 | if ($last) { |
673 | $Lbrk->Extend($last); |
674 | $Lbrk{$lbrk}->Extend($last); |
d73e5302 |
675 | } |
676 | } |
cf25bb62 |
677 | close IN; |
d73e5302 |
678 | |
cf25bb62 |
679 | $Lbrk->Write("Lbrk.pl"); |
680 | } |
d73e5302 |
681 | |
cf25bb62 |
682 | ## |
683 | ## Process ArabShap.txt. |
684 | ## |
685 | sub ArabShap_txt() |
686 | { |
687 | if (not open IN, "ArabShap.txt") { |
688 | die "$0: ArabShap.txt: $!\n"; |
689 | } |
d73e5302 |
690 | |
cf25bb62 |
691 | my $ArabLink = Table->New(); |
692 | my $ArabLinkGroup = Table->New(); |
d73e5302 |
693 | |
cf25bb62 |
694 | while (<IN>) |
695 | { |
696 | next unless /^[0-9A-Fa-f]+;/; |
697 | s/\s+$//; |
d73e5302 |
698 | |
cf25bb62 |
699 | my ($hexcode, $name, $link, $linkgroup) = split(/\s*;\s*/); |
700 | my $code = hex($hexcode); |
701 | $ArabLink->Append($code, $link); |
702 | $ArabLinkGroup->Append($code, $linkgroup); |
d73e5302 |
703 | } |
cf25bb62 |
704 | close IN; |
705 | |
706 | $ArabLink->Write("ArabLink.pl"); |
707 | $ArabLinkGroup->Write("ArabLnkGrp.pl"); |
d73e5302 |
708 | } |
709 | |
cf25bb62 |
710 | ## |
711 | ## Process Jamo.txt. |
712 | ## |
713 | sub Jamo_txt() |
714 | { |
715 | if (not open IN, "Jamo.txt") { |
716 | die "$0: Jamo.txt: $!\n"; |
717 | } |
718 | my $Short = Table->New(); |
d73e5302 |
719 | |
cf25bb62 |
720 | while (<IN>) |
721 | { |
722 | next unless /^([0-9A-Fa-f]+)\s*;\s*(\w*)/; |
723 | my ($code, $short) = (hex($1), $2); |
d73e5302 |
724 | |
cf25bb62 |
725 | $Short->Append($code, $short); |
d73e5302 |
726 | } |
cf25bb62 |
727 | close IN; |
728 | $Short->Write("JamoShort.pl"); |
d73e5302 |
729 | } |
730 | |
cf25bb62 |
731 | ## |
732 | ## Process Scripts.txt. |
733 | ## |
734 | sub Scripts_txt() |
735 | { |
736 | my @ScriptInfo; |
d73e5302 |
737 | |
cf25bb62 |
738 | if (not open(IN, "Scripts.txt")) { |
739 | die "$0: Scripts.txt: $!\n"; |
740 | } |
741 | while (<IN>) { |
742 | next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(.+?)\s*\#/; |
d73e5302 |
743 | |
cf25bb62 |
744 | # Wait until all the scripts have been read since |
745 | # they are not listed in numeric order. |
746 | push @ScriptInfo, [ hex($1), hex($2||""), $3 ]; |
747 | } |
748 | close IN; |
d73e5302 |
749 | |
cf25bb62 |
750 | # Now append the scripts properties in their code point order. |
d73e5302 |
751 | |
cf25bb62 |
752 | my %Script; |
753 | my $Scripts = Table->New(); |
d73e5302 |
754 | |
cf25bb62 |
755 | for my $script (sort { $a->[0] <=> $b->[0] } @ScriptInfo) |
756 | { |
757 | my ($first, $last, $name) = @$script; |
758 | $Scripts->Append($first, $name); |
d73e5302 |
759 | |
cf25bb62 |
760 | $Script{$name} ||= Table->New(Is => CanonicalName($name), |
761 | AllowFuzzy => 1); |
762 | $Script{$name}->Append($first, $name); |
d73e5302 |
763 | |
cf25bb62 |
764 | if ($last) { |
765 | $Scripts->Extend($last); |
766 | $Script{$name}->Extend($last); |
767 | } |
768 | } |
d73e5302 |
769 | |
cf25bb62 |
770 | $Scripts->Write("Scripts.pl"); |
d73e5302 |
771 | |
cf25bb62 |
772 | ## Common is everything not explicitly assigned to a Script |
773 | ## |
774 | ## ***shouldn't this be intersected with \p{Assigned}? ****** |
775 | ## |
776 | New_Prop(Is => 'Common', $Scripts->Invert, AllowFuzzy => 1); |
777 | } |
d73e5302 |
778 | |
cf25bb62 |
779 | ## |
780 | ## Given a name like "Close Punctuation", return a regex (that when applied |
781 | ## with /i) matches any valid form of that name (e.g. "ClosePunctuation", |
782 | ## "Close-Punctuation", etc.) |
783 | ## |
784 | ## Accept any space, dash, or underbar where in the official name there is |
785 | ## space or a dash (or underbar, but there never is). |
786 | ## |
787 | ## |
788 | sub NameToRegex($) |
789 | { |
790 | my $Name = shift; |
791 | $Name =~ s/[- _]/(?:[-_]|\\s+)?/g; |
792 | return $Name; |
793 | } |
d73e5302 |
794 | |
cf25bb62 |
795 | ## |
796 | ## Process Blocks.txt. |
797 | ## |
798 | sub Blocks_txt() |
799 | { |
800 | my $Blocks = Table->New(); |
801 | my %Blocks; |
d73e5302 |
802 | |
cf25bb62 |
803 | if (not open IN, "Blocks.txt") { |
804 | die "$0: Blocks.txt: $!\n"; |
805 | } |
d73e5302 |
806 | |
cf25bb62 |
807 | while (<IN>) |
808 | { |
809 | #next if not /Private Use$/; |
810 | next if not /^([0-9A-Fa-f]+)\.\.([0-9A-Fa-f]+)\s*;\s*(.+?)\s*$/; |
d73e5302 |
811 | |
cf25bb62 |
812 | my ($first, $last, $name) = (hex($1), hex($2), $3); |
d73e5302 |
813 | |
cf25bb62 |
814 | $Blocks->Append($first, $name); |
76ccdbe2 |
815 | |
cf25bb62 |
816 | $Blocks{$name} ||= Table->New(In=>CanonicalName($name), AllowFuzzy=>1); |
817 | $Blocks{$name}->Append($first, $name); |
76ccdbe2 |
818 | |
cf25bb62 |
819 | if ($last and $last != $first) { |
820 | $Blocks->Extend($last); |
821 | $Blocks{$name}->Extend($last); |
d73e5302 |
822 | } |
d73e5302 |
823 | } |
cf25bb62 |
824 | close IN; |
825 | |
826 | $Blocks->Write("Blocks.pl"); |
d73e5302 |
827 | } |
828 | |
cf25bb62 |
829 | ## |
830 | ## Read in the PropList.txt. It contains extended properties not |
831 | ## listed in the Unicode.txt, such as 'Other_Alphabetic': |
832 | ## alphabetic but not of the general category L; many modifiers |
833 | ## belong to this extended property category: while they are not |
834 | ## alphabets, they are alphabetic in nature. |
835 | ## |
836 | sub PropList_txt() |
837 | { |
838 | my @PropInfo; |
839 | |
840 | if (not open IN, "PropList.txt") { |
841 | die "$0: PropList.txt: $!\n"; |
842 | } |
d73e5302 |
843 | |
cf25bb62 |
844 | while (<IN>) |
845 | { |
846 | next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(.+?)\s*\#/; |
d73e5302 |
847 | |
cf25bb62 |
848 | # Wait until all the extended properties have been read since |
849 | # they are not listed in numeric order. |
850 | push @PropInfo, [ hex($1), hex($2||""), $3 ]; |
851 | } |
852 | close IN; |
71d929cb |
853 | |
cf25bb62 |
854 | # Now append the extended properties in their code point order. |
855 | my $Props = Table->New(); |
856 | my %Prop; |
71d929cb |
857 | |
cf25bb62 |
858 | for my $prop (sort { $a->[0] <=> $b->[0] } @PropInfo) |
859 | { |
860 | my ($first, $last, $name) = @$prop; |
861 | $Props->Append($first, $name); |
71d929cb |
862 | |
cf25bb62 |
863 | $Prop{$name} ||= Table->New(Is => $name, AllowFuzzy => 1); |
864 | $Prop{$name}->Append($first, $name); |
71d929cb |
865 | |
cf25bb62 |
866 | if ($last) { |
867 | $Props->Extend($last); |
868 | $Prop{$name}->Extend($last); |
869 | } |
71d929cb |
870 | } |
d73e5302 |
871 | |
cf25bb62 |
872 | # Alphabetic is L and Other_Alphabetic. |
873 | New_Prop(Is => 'Alphabetic', |
874 | Table->Merge($Cat{L}, $Prop{Other_Alphabetic}), |
875 | AllowFuzzy => 1); |
876 | |
877 | # Lowercase is Ll and Other_Lowercase. |
878 | New_Prop(Is => 'Lowercase', |
879 | Table->Merge($Cat{Ll}, $Prop{Other_Lowercase}), |
880 | AllowFuzzy => 1); |
881 | |
882 | # Uppercase is Lu and Other_Uppercase. |
883 | New_Prop(Is => 'Uppercase', |
884 | Table->Merge($Cat{Lu}, $Prop{Other_Uppercase}), |
885 | AllowFuzzy => 1); |
886 | |
887 | # Math is Sm and Other_Math. |
888 | New_Prop(Is => 'Math', |
889 | Table->Merge($Cat{Sm}, $Prop{Other_Math}), |
890 | AllowFuzzy => 1); |
891 | |
892 | # ID_Start is Ll, Lu, Lt, Lm, Lo, and Nl. |
893 | New_Prop(Is => 'ID_Start', |
894 | Table->Merge(@Cat{qw[Ll Lu Lt Lm Lo Nl]}), |
895 | AllowFuzzy => 1); |
896 | |
897 | # ID_Continue is ID_Start, Mn, Mc, Nd, and Pc. |
898 | New_Prop(Is => 'ID_Continue', |
899 | Table->Merge(@Cat{qw[Ll Lu Lt Lm Lo Nl Mn Mc Nd Pc ]}), |
900 | AllowFuzzy => 1); |
d73e5302 |
901 | } |
902 | |
cf25bb62 |
903 | sub Make_GC_Aliases() |
904 | { |
905 | ## |
906 | ## The mapping from General Category long forms to short forms is |
907 | ## currently hardwired here since no simple data file in the UCD |
908 | ## seems to do that. Unicode 3.2 will assumedly correct this. |
909 | ## |
910 | my %Is = ( |
d73e5302 |
911 | 'Letter' => 'L', |
e150c829 |
912 | 'Uppercase_Letter' => 'Lu', |
913 | 'Lowercase_Letter' => 'Ll', |
914 | 'Titlecase_Letter' => 'Lt', |
915 | 'Modifier_Letter' => 'Lm', |
916 | 'Other_Letter' => 'Lo', |
d73e5302 |
917 | |
918 | 'Mark' => 'M', |
e150c829 |
919 | 'Non_Spacing_Mark' => 'Mn', |
920 | 'Spacing_Mark' => 'Mc', |
921 | 'Enclosing_Mark' => 'Me', |
d73e5302 |
922 | |
923 | 'Separator' => 'Z', |
e150c829 |
924 | 'Space_Separator' => 'Zs', |
925 | 'Line_Separator' => 'Zl', |
926 | 'Paragraph_Separator' => 'Zp', |
d73e5302 |
927 | |
928 | 'Number' => 'N', |
e150c829 |
929 | 'Decimal_Number' => 'Nd', |
930 | 'Letter_Number' => 'Nl', |
931 | 'Other_Number' => 'No', |
d73e5302 |
932 | |
933 | 'Punctuation' => 'P', |
e150c829 |
934 | 'Connector_Punctuation' => 'Pc', |
935 | 'Dash_Punctuation' => 'Pd', |
936 | 'Open_Punctuation' => 'Ps', |
937 | 'Close_Punctuation' => 'Pe', |
938 | 'Initial_Punctuation' => 'Pi', |
939 | 'Final_Punctuation' => 'Pf', |
940 | 'Other_Punctuation' => 'Po', |
d73e5302 |
941 | |
942 | 'Symbol' => 'S', |
e150c829 |
943 | 'Math_Symbol' => 'Sm', |
944 | 'Currency_Symbol' => 'Sc', |
945 | 'Modifier_Symbol' => 'Sk', |
946 | 'Other_Symbol' => 'So', |
d73e5302 |
947 | |
948 | 'Other' => 'C', |
949 | 'Control' => 'Cc', |
950 | 'Format' => 'Cf', |
951 | 'Surrogate' => 'Cs', |
952 | 'Private Use' => 'Co', |
e150c829 |
953 | 'Unassigned' => 'Cn', |
cf25bb62 |
954 | ); |
d2d499f5 |
955 | |
cf25bb62 |
956 | ## make the aliases.... |
957 | while (my ($Alias, $Name) = each %Is) { |
958 | New_Alias(Is => $Alias, SameAs => $Name, AllowFuzzy => 1); |
959 | } |
960 | } |
d2d499f5 |
961 | |
cf25bb62 |
962 | ## |
963 | ## Writes the info accumulated in |
964 | ## |
965 | ## %TableInfo; |
966 | ## %FuzzyNames; |
967 | ## %AliasInfo; |
968 | ## |
969 | ## |
970 | sub WriteAllMappings() |
971 | { |
972 | for my $Type ('In', 'Is') |
973 | { |
974 | my %Filenames; |
975 | my %NameToFile; |
976 | |
977 | my %Exact; ## will become %utf8::Is or %utf8::In |
978 | my %Pat; ## will become %utf8::IsPat or %utf8::InPat |
979 | |
980 | ## |
981 | ## First write all the files to the $Type/ directory |
982 | ## |
983 | while (my ($Name, $Table) = each %{$TableInfo{$Type}}) |
984 | { |
985 | ## Need an 8.3 safe filename. |
986 | my $filename = $Name; |
987 | $filename =~ s/[_\W]+(\w*)/\u$1/g; |
988 | substr($filename, 8) = '' if length($filename) > 8; |
989 | |
990 | ## |
991 | ## Make sure the filename doesn't conflict with something we |
992 | ## might have already written. If we have, say, |
993 | ## Greek_Extended1 |
994 | ## Greek_Extended2 |
995 | ## they become |
996 | ## Greek_Ex |
997 | ## Greek_E2 |
998 | ## |
999 | while (my $num = $Filenames{lc $filename}++) |
1000 | { |
1001 | $num++; ## so filenames with numbers start with '2', which |
1002 | ## just looks more natural. |
1003 | substr($filename, -length($num)) = $num; |
1004 | } |
1005 | |
1006 | ## |
1007 | ## Okay, write the file... |
1008 | ## |
1009 | $Exact{$Name} = $filename; |
1010 | $Table->Write("$Type/$filename.pl"); |
1011 | } |
1012 | |
1013 | ## |
1014 | ## Build %Pat |
1015 | ## |
1016 | while (my ($Fuzzy, $Real) = each %{$FuzzyNames{$Type}}) |
1017 | { |
1018 | my $File = $Exact{$Real}; |
1019 | |
1020 | if (not $File) { |
1021 | die "$0: oops [$Real]"; |
1022 | } |
1023 | |
1024 | ## The prefix length of 2 is enough spread, |
1025 | ## and besides, we have 'Yi' as an In category. |
1026 | my $Prefix = lc(substr($Fuzzy, 0, 2)); |
1027 | my $Regex = NameToRegex($Fuzzy); |
1028 | |
1029 | if ($Pat{$Prefix}->{$Regex}) { |
1030 | warn "WHOA, conflict with /$Regex/: $Pat{$Prefix}->{$Regex} vs $File\n"; |
1031 | } |
1032 | |
1033 | $Pat{$Prefix}->{$Regex} = $File; |
1034 | } |
1035 | |
1036 | ## |
1037 | ## Since the fuzzy method will provide for a way to match $Fuzzy, |
1038 | ## there's no need for $Fuzzy to be in %Exact as well. |
1039 | ## This can't be done in the loop above because there could be |
1040 | ## multiple $Fuzzys pointing at the same $Real, and we don't want |
1041 | ## the first to delete the exact mapping out from under the second. |
1042 | ## |
1043 | for my $Fuzzy (keys %{$FuzzyNames{$Type}}) |
1044 | { |
1045 | delete $Exact{$Fuzzy}; |
1046 | } |
1047 | |
1048 | |
1049 | |
1050 | ## |
1051 | ## Now write In.pl / Is.pl |
1052 | ## |
1053 | if (not open OUT, ">$Type.pl") { |
1054 | die "$0: $Type.pl: $!\n"; |
1055 | } |
1056 | print OUT $HEADER; |
1057 | print OUT "##\n"; |
1058 | print OUT "## Data in this file used by ../utf8_heavy.pl\n"; |
1059 | print OUT "##\n"; |
1060 | print OUT "\n"; |
1061 | print OUT "## Mapping from name to filename in ./$Type\n"; |
1062 | print OUT "%utf8::$Type = (\n"; |
1063 | for my $Name (sort keys %Exact) |
1064 | { |
1065 | my $File = $Exact{$Name}; |
1066 | printf OUT " %-41s => %s,\n", "'$Name'", "'$File'"; |
1067 | } |
1068 | print OUT ");\n\n"; |
1069 | |
1070 | print OUT "## Mappings from regex to filename in ./$Type/\n"; |
1071 | print OUT "%utf8::${Type}Pat = (\n"; |
1072 | for my $Prefix (sort keys %Pat) |
1073 | { |
1074 | print OUT " '$Prefix' => {\n"; |
1075 | while (my ($Regex, $File) = each %{ $Pat{$Prefix} }) { |
1076 | print OUT "\t'$Regex' => '$File',\n"; |
1077 | } |
1078 | print OUT " },\n"; |
1079 | } |
1080 | print OUT ");\n"; |
1081 | |
1082 | close(OUT); |
d2d499f5 |
1083 | } |
d2d499f5 |
1084 | } |
1085 | |
cf25bb62 |
1086 | sub SpecCase_txt() |
1087 | { |
1088 | # |
1089 | # Read in the special cases. |
1090 | # |
983ffd37 |
1091 | |
cf25bb62 |
1092 | my %CaseInfo; |
1093 | |
1094 | if (not open IN, "SpecCase.txt") { |
1095 | die "$0: SpecCase.txt: $!\n"; |
1096 | } |
1097 | while (<IN>) { |
1098 | next unless /^[0-9A-Fa-f]+;/; |
1099 | s/\#.*//; |
1100 | s/\s+$//; |
1101 | |
1102 | my ($code, $lower, $title, $upper, $condition) = split(/\s*;\s*/); |
1103 | |
1104 | if ($condition) { # not implemented yet |
1105 | print "# SKIPPING $_\n" if $Verbose; |
1106 | next; |
1107 | } |
1108 | |
1109 | # Wait until all the special cases have been read since |
1110 | # they are not listed in numeric order. |
1111 | my $ix = hex($code); |
1112 | push @{$CaseInfo{Lower}}, [ $ix, $code, $lower ]; |
1113 | push @{$CaseInfo{Title}}, [ $ix, $code, $title ]; |
1114 | push @{$CaseInfo{Upper}}, [ $ix, $code, $upper ]; |
1115 | } |
1116 | close IN; |
1117 | |
1118 | # Now write out the special cases properties in their code point order. |
1119 | # Prepend them to the To/{Upper,Lower,Title}.pl. |
1120 | |
1121 | for my $case (qw(Lower Title Upper)) |
1122 | { |
1123 | my $NormalCase = do "To/$case.pl" || die "$0: $@\n"; |
1124 | if (not open OUT, ">To/$case.pl") { |
1125 | die "$0: To/$case.txt: $!"; |
1126 | } |
1127 | |
1128 | print OUT $HEADER, "\n"; |
1129 | print OUT "%utf8::ToSpec$case =\n(\n"; |
1130 | |
1131 | for my $prop (sort { $a->[0] <=> $b->[0] } @{$CaseInfo{$case}}) { |
1132 | my ($ix, $code, $to) = @$prop; |
1133 | my $tostr = |
1134 | join "", map { sprintf "\\x{%s}", $_ } split ' ', $to; |
1135 | printf OUT qq['%04X' => "$tostr",\n], $ix; |
1136 | } |
1137 | print OUT ");\n\n"; |
1138 | print OUT "return <<'END';\n"; |
1139 | print OUT $NormalCase; |
1140 | print OUT "END\n"; |
1141 | close OUT; |
d2d499f5 |
1142 | } |
d2d499f5 |
1143 | } |
1144 | |
c4051cc5 |
1145 | # |
1146 | # Read in the case foldings. |
1147 | # |
1148 | # We will do full case folding, C + F + I (see CaseFold.txt). |
1149 | # |
cf25bb62 |
1150 | sub CaseFold_txt() |
1151 | { |
1152 | if (not open IN, "CaseFold.txt") { |
1153 | die "$0: To/Fold.pl: $!\n"; |
1154 | } |
c4051cc5 |
1155 | |
cf25bb62 |
1156 | my $Fold = Table->New(); |
c4051cc5 |
1157 | my %Fold; |
1158 | |
cf25bb62 |
1159 | while (<IN>) { |
254ba52a |
1160 | # Skip status 'S', simple case folding |
c4051cc5 |
1161 | next unless /^([0-9A-Fa-f]+)\s*;\s*([CFI])\s*;\s*([0-9A-Fa-f]+(?: [0-9A-Fa-f]+)*)\s*;/; |
1162 | |
cf25bb62 |
1163 | my ($code, $status, $fold) = (hex($1), $2, $3); |
c4051cc5 |
1164 | |
1165 | if ($status eq 'C') { # Common: one-to-one folding |
254ba52a |
1166 | # No append() since several codes may fold into one. |
cf25bb62 |
1167 | $Fold->RawAppendRange($code, $code, $fold); |
c4051cc5 |
1168 | } else { # F: full, or I: dotted uppercase I -> dotless lowercase I |
cf25bb62 |
1169 | $Fold{$code} = $fold; |
c4051cc5 |
1170 | } |
1171 | } |
cf25bb62 |
1172 | close IN; |
c4051cc5 |
1173 | |
cf25bb62 |
1174 | $Fold->Write("To/Fold.pl"); |
c4051cc5 |
1175 | |
1176 | # |
1177 | # Prepend the special foldings to the common foldings. |
1178 | # |
1179 | |
1180 | my $CommonFold = do "To/Fold.pl" || die "$0: To/Fold.pl: $!\n"; |
cf25bb62 |
1181 | if (not open OUT, ">To/Fold.pl") { |
1182 | die "$0: To/Fold.pl: $!\n"; |
1183 | } |
1184 | print OUT $HEADER, "\n"; |
1185 | print OUT "%utf8::ToSpecFold =\n(\n"; |
1186 | for my $code (sort { $a <=> $b } keys %Fold) { |
1187 | my $foldstr = |
1188 | join "", map { sprintf "\\x{%s}", $_ } split ' ', $Fold{$code}; |
1189 | printf OUT qq['%04X' => "$foldstr",\n], $code; |
c4051cc5 |
1190 | } |
cf25bb62 |
1191 | print OUT ");\n\n"; |
1192 | print OUT "return <<'END';\n"; |
1193 | print OUT $CommonFold; |
1194 | print OUT "END\n"; |
1195 | close OUT; |
c4051cc5 |
1196 | } |
1197 | |
cf25bb62 |
1198 | ## Do it.... |
1199 | |
1200 | Unicode_Txt(); |
1201 | Make_GC_Aliases(); |
1202 | PropList_txt(); |
1203 | |
1204 | Scripts_txt(); |
1205 | Blocks_txt(); |
1206 | |
1207 | LineBrk_Txt(); |
1208 | ArabShap_txt(); |
1209 | Jamo_txt(); |
1210 | SpecCase_txt(); |
1211 | |
1212 | WriteAllMappings(); |
1213 | |
1214 | CaseFold_txt(); |
1215 | |
d73e5302 |
1216 | # That's all, folks! |
1217 | |
cf25bb62 |
1218 | __END__ |