Commit | Line | Data |
00a8df5c |
1 | ## !!!!!!!!!!!!!! IF YOU MODIFY THIS FILE !!!!!!!!!!!!!!!!!!!!!!!!! |
2 | ## Any files created or read by this program should be listed in 'mktables.lst' |
3 | |
d73e5302 |
4 | #!/usr/bin/perl -w |
4392c1cd |
5 | require 5.008; # Needs pack "U". Probably safest to run on 5.8.x |
d73e5302 |
6 | use strict; |
cf25bb62 |
7 | use Carp; |
d07a55ed |
8 | use File::Spec; |
5beb625e |
9 | |
cf25bb62 |
10 | ## |
11 | ## mktables -- create the runtime Perl Unicode files (lib/unicore/**/*.pl) |
12 | ## from the Unicode database files (lib/unicore/*.txt). |
13 | ## |
d2d499f5 |
14 | |
cb3d361a |
15 | ## "Fuzzy" means this section in Unicode TR18: |
16 | ## |
17 | ## The recommended names for UCD properties and property values are in |
18 | ## PropertyAliases.txt [Prop] and PropertyValueAliases.txt |
19 | ## [PropValue]. There are both abbreviated names and longer, more |
20 | ## descriptive names. It is strongly recommended that both names be |
21 | ## recognized, and that loose matching of property names be used, |
22 | ## whereby the case distinctions, whitespace, hyphens, and underbar |
23 | ## are ignored. |
24 | |
8e4b7420 |
25 | ## Base names already used in lib/gc_sc (for avoiding 8.3 conflicts) |
26 | my %BaseNames; |
27 | |
cf25bb62 |
28 | ## |
29 | ## Process any args. |
30 | ## |
5beb625e |
31 | my $Verbose = 0; |
32 | my $MakeTestScript = 0; |
00da7942 |
33 | my $AlwaysWrite = 0; |
02b1aeec |
34 | my $UseDir = ""; |
35 | my $FileList = "$0.lst"; |
97050450 |
36 | my $MakeList = 0; |
cf25bb62 |
37 | |
38 | while (@ARGV) |
39 | { |
40 | my $arg = shift @ARGV; |
41 | if ($arg eq '-v') { |
42 | $Verbose = 1; |
43 | } elsif ($arg eq '-q') { |
44 | $Verbose = 0; |
00da7942 |
45 | } elsif ($arg eq '-w') { |
46 | $AlwaysWrite = 1; # update the files even if they havent changed |
02b1aeec |
47 | $FileList = ""; |
6ae7e459 |
48 | } elsif ($arg eq '-check') { |
49 | my $this = shift @ARGV; |
50 | my $ok = shift @ARGV; |
51 | if ($this ne $ok) { |
52 | print "Skipping as check params are not the same.\n"; |
53 | exit(0); |
54 | } |
5beb625e |
55 | } elsif ($arg eq '-maketest') { |
56 | $MakeTestScript = 1; |
97050450 |
57 | } elsif ($arg eq '-makelist') { |
58 | $MakeList = 1; |
02b1aeec |
59 | } elsif ($arg eq '-C' && defined ($UseDir = shift)) { |
60 | -d $UseDir or die "Unknown directory '$UseDir'"; |
61 | } elsif ($arg eq '-L' && defined ($FileList = shift)) { |
62 | -e $FileList or die "Filelist '$FileList' doesn't appear to exist!"; |
d73e5302 |
63 | } else { |
97050450 |
64 | die "usage: $0 [-v|-q|-w|-C dir|-L filelist] [-maketest] [-makelist]\n", |
65 | " -v : Verbose Mode\n", |
66 | " -q : Quiet Mode\n", |
67 | " -w : Write files regardless\n", |
68 | " -maketest : Make test script\n", |
69 | " -makelist : Rewrite the file list based on current setup\n", |
70 | " -L filelist : Use this file list, (defaults to $0)\n", |
6ae7e459 |
71 | " -C dir : Change to this directory before proceeding\n", |
72 | " -check A B : Executes only if A and B are the same\n"; |
d73e5302 |
73 | } |
74 | } |
75 | |
02b1aeec |
76 | if ($FileList) { |
77 | print "Reading file list '$FileList'\n" |
00a8df5c |
78 | if $Verbose; |
02b1aeec |
79 | open my $fh,"<",$FileList or die "Failed to read '$FileList':$!"; |
80 | my @input; |
00a8df5c |
81 | my @output; |
82 | for my $list ( \@input, \@output ) { |
83 | while (<$fh>) { |
84 | s/^ \s+ | \s+ $//xg; |
85 | next if /^ \s* (?: \# .* )? $/x; |
86 | last if /^ =+ $/x; |
87 | my ( $file ) = split /\t/, $_; |
88 | push @$list, $file; |
89 | } |
90 | my %dupe; |
91 | @$list = grep !$dupe{ $_ }++, @$list; |
92 | } |
93 | close $fh; |
02b1aeec |
94 | die "No input or output files in '$FileList'!" |
00a8df5c |
95 | if !@input or !@output; |
97050450 |
96 | if ( $MakeList ) { |
97 | foreach my $file (@output) { |
98 | unlink $file; |
99 | } |
100 | } |
00a8df5c |
101 | if ( $Verbose ) { |
102 | print "Expecting ".scalar( @input )." input files. ", |
103 | "Checking ".scalar( @output )." output files.\n"; |
104 | } |
02b1aeec |
105 | # we set maxtime to be the youngest input file, including $0 itself. |
106 | my $maxtime = -M $0; # do this before the chdir! |
107 | if ($UseDir) { |
108 | chdir $UseDir or die "Failed to chdir to '$UseDir':$!"; |
109 | } |
00a8df5c |
110 | foreach my $in (@input) { |
111 | my $time = -M $in; |
112 | die "Missing input file '$in'" unless defined $time; |
113 | $maxtime = $time if $maxtime < $time; |
114 | } |
115 | |
02b1aeec |
116 | # now we check to see if any output files are older than maxtime, if |
117 | # they are we need to continue on, otherwise we can presumably bail. |
00a8df5c |
118 | my $ok = 1; |
119 | foreach my $out (@output) { |
120 | if ( ! -e $out ) { |
121 | print "'$out' is missing.\n" |
122 | if $Verbose; |
123 | $ok = 0; |
124 | last; |
125 | } |
126 | if ( -M $out > $maxtime ) { |
127 | print "'$out' is too old.\n" |
128 | if $Verbose; |
129 | $ok = 0; |
130 | last; |
131 | } |
132 | } |
133 | if ($ok) { |
134 | print "Files seem to be ok, not bothering to rebuild.\n"; |
135 | exit(0); |
136 | } |
137 | print "Must rebuild tables.\n" |
138 | if $Verbose; |
02b1aeec |
139 | } else { |
140 | if ($Verbose) { |
141 | print "Not checking filelist.\n"; |
142 | } |
143 | if ($UseDir) { |
144 | chdir $UseDir or die "Failed to chdir to '$UseDir':$!"; |
145 | } |
00a8df5c |
146 | } |
147 | |
ca12659b |
148 | foreach my $lib ('To', 'lib', |
149 | map {File::Spec->catdir("lib",$_)} |
150 | qw(gc_sc dt bc hst ea jt lb nt ccc)) { |
151 | next if -d $lib; |
152 | mkdir $lib, 0755 or die "mkdir '$lib': $!"; |
153 | } |
154 | |
cf25bb62 |
155 | my $LastUnicodeCodepoint = 0x10FFFF; # As of Unicode 3.1.1. |
d73e5302 |
156 | |
cf25bb62 |
157 | my $HEADER=<<"EOF"; |
d73e5302 |
158 | # !!!!!!! DO NOT EDIT THIS FILE !!!!!!! |
551b6b6f |
159 | # This file is built by $0 from e.g. UnicodeData.txt. |
d73e5302 |
160 | # Any changes made here will be lost! |
cf25bb62 |
161 | |
162 | EOF |
163 | |
d07a55ed |
164 | sub force_unlink { |
165 | my $filename = shift; |
166 | return unless -e $filename; |
167 | return if CORE::unlink($filename); |
168 | # We might need write permission |
169 | chmod 0777, $filename; |
170 | CORE::unlink($filename) or die "Couldn't unlink $filename: $!\n"; |
171 | } |
5beb625e |
172 | |
173 | ## |
174 | ## Given a filename and a reference to an array of lines, |
175 | ## write the lines to the file only if the contents have not changed. |
d07a55ed |
176 | ## Filename can be given as an arrayref of directory names |
5beb625e |
177 | ## |
178 | sub WriteIfChanged($\@) |
179 | { |
180 | my $file = shift; |
181 | my $lines = shift; |
182 | |
d07a55ed |
183 | $file = File::Spec->catfile(@$file) if ref $file; |
184 | |
5beb625e |
185 | my $TextToWrite = join '', @$lines; |
186 | if (open IN, $file) { |
187 | local($/) = undef; |
188 | my $PreviousText = <IN>; |
189 | close IN; |
190 | if ($PreviousText eq $TextToWrite) { |
191 | print "$file unchanged.\n" if $Verbose; |
00da7942 |
192 | return unless $AlwaysWrite; |
5beb625e |
193 | } |
194 | } |
d07a55ed |
195 | force_unlink ($file); |
5beb625e |
196 | if (not open OUT, ">$file") { |
197 | die "$0: can't open $file for output: $!\n"; |
198 | } |
199 | print "$file written.\n" if $Verbose; |
200 | |
201 | print OUT $TextToWrite; |
202 | close OUT; |
203 | } |
204 | |
cf25bb62 |
205 | ## |
206 | ## The main datastructure (a "Table") represents a set of code points that |
207 | ## are part of a particular quality (that are part of \pL, \p{InGreek}, |
208 | ## etc.). They are kept as ranges of code points (starting and ending of |
209 | ## each range). |
210 | ## |
211 | ## For example, a range ASCII LETTERS would be represented as: |
212 | ## [ [ 0x41 => 0x5A, 'UPPER' ], |
213 | ## [ 0x61 => 0x7A, 'LOWER, ] ] |
214 | ## |
215 | sub RANGE_START() { 0 } ## index into range element |
216 | sub RANGE_END() { 1 } ## index into range element |
217 | sub RANGE_NAME() { 2 } ## index into range element |
218 | |
99598c8c |
219 | ## Conceptually, these should really be folded into the 'Table' objects |
cf25bb62 |
220 | my %TableInfo; |
99598c8c |
221 | my %TableDesc; |
cf25bb62 |
222 | my %FuzzyNames; |
223 | my %AliasInfo; |
5beb625e |
224 | my %CanonicalToOrig; |
cf25bb62 |
225 | |
226 | ## |
99598c8c |
227 | ## Turn something like |
228 | ## OLD-ITALIC |
5beb625e |
229 | ## into |
99598c8c |
230 | ## OldItalic |
231 | ## |
232 | sub CanonicalName($) |
233 | { |
5beb625e |
234 | my $orig = shift; |
235 | my $name = lc $orig; |
99598c8c |
236 | $name =~ s/(?<![a-z])(\w)/\u$1/g; |
5beb625e |
237 | $name =~ s/[-_\s]+//g; |
99598c8c |
238 | |
5beb625e |
239 | $CanonicalToOrig{$name} = $orig if not $CanonicalToOrig{$name}; |
99598c8c |
240 | return $name; |
241 | } |
242 | |
12ac2576 |
243 | |
244 | ## |
245 | ## Store the alias definitions for later use. |
246 | ## |
247 | my %PropertyAlias; |
248 | my %PropValueAlias; |
249 | |
250 | my %PA_reverse; |
251 | my %PVA_reverse; |
252 | |
253 | sub Build_Aliases() |
254 | { |
255 | ## |
256 | ## Most of the work with aliases doesn't occur here, |
7ebf06b3 |
257 | ## but rather in utf8_heavy.pl, which uses PVA.pl, |
258 | |
259 | # Placate the warnings about used only once. (They are used again, but |
260 | # via a typeglob lookup) |
261 | %utf8::PropertyAlias = (); |
262 | %utf8::PA_reverse = (); |
263 | %utf8::PropValueAlias = (); |
264 | %utf8::PVA_reverse = (); |
265 | %utf8::PVA_abbr_map = (); |
12ac2576 |
266 | |
267 | open PA, "< PropertyAliases.txt" |
268 | or confess "Can't open PropertyAliases.txt: $!"; |
269 | while (<PA>) { |
270 | s/#.*//; |
271 | s/\s+$//; |
272 | next if /^$/; |
273 | |
274 | my ($abbrev, $name) = split /\s*;\s*/; |
275 | next if $abbrev eq "n/a"; |
276 | $PropertyAlias{$abbrev} = $name; |
277 | $PA_reverse{$name} = $abbrev; |
7ebf06b3 |
278 | |
279 | # The %utf8::... versions use japhy's code originally from utf8_pva.pl |
280 | # However, it's moved here so that we build the tables at runtime. |
281 | tr/ _-//d for $abbrev, $name; |
282 | $utf8::PropertyAlias{lc $abbrev} = $name; |
283 | $utf8::PA_reverse{lc $name} = $abbrev; |
12ac2576 |
284 | } |
285 | close PA; |
286 | |
287 | open PVA, "< PropValueAliases.txt" |
288 | or confess "Can't open PropValueAliases.txt: $!"; |
289 | while (<PVA>) { |
290 | s/#.*//; |
291 | s/\s+$//; |
292 | next if /^$/; |
293 | |
294 | my ($prop, @data) = split /\s*;\s*/; |
295 | |
296 | if ($prop eq 'ccc') { |
297 | $PropValueAlias{$prop}{$data[1]} = [ @data[0,2] ]; |
298 | $PVA_reverse{$prop}{$data[2]} = [ @data[0,1] ]; |
299 | } |
300 | else { |
301 | next if $data[0] eq "n/a"; |
302 | $PropValueAlias{$prop}{$data[0]} = $data[1]; |
303 | $PVA_reverse{$prop}{$data[1]} = $data[0]; |
304 | } |
7ebf06b3 |
305 | |
306 | shift @data if $prop eq 'ccc'; |
307 | next if $data[0] eq "n/a"; |
308 | |
309 | $data[1] =~ tr/ _-//d; |
310 | $utf8::PropValueAlias{$prop}{lc $data[0]} = $data[1]; |
311 | $utf8::PVA_reverse{$prop}{lc $data[1]} = $data[0]; |
312 | |
313 | my $abbr_class = ($prop eq 'gc' or $prop eq 'sc') ? 'gc_sc' : $prop; |
314 | $utf8::PVA_abbr_map{$abbr_class}{lc $data[0]} = $data[0]; |
12ac2576 |
315 | } |
316 | close PVA; |
7ebf06b3 |
317 | |
318 | # backwards compatibility for L& -> LC |
319 | $utf8::PropValueAlias{gc}{'l&'} = $utf8::PropValueAlias{gc}{lc}; |
320 | $utf8::PVA_abbr_map{gc_sc}{'l&'} = $utf8::PVA_abbr_map{gc_sc}{lc}; |
321 | |
12ac2576 |
322 | } |
323 | |
324 | |
99598c8c |
325 | ## |
cf25bb62 |
326 | ## Associates a property ("Greek", "Lu", "Assigned",...) with a Table. |
327 | ## |
328 | ## Called like: |
99598c8c |
329 | ## New_Prop(In => 'Greek', $Table, Desc => 'Greek Block', Fuzzy => 1); |
cf25bb62 |
330 | ## |
331 | ## Normally, these parameters are set when the Table is created (when the |
332 | ## Table->New constructor is called), but there are times when it needs to |
333 | ## be done after-the-fact...) |
334 | ## |
335 | sub New_Prop($$$@) |
336 | { |
337 | my $Type = shift; ## "Is" or "In"; |
338 | my $Name = shift; |
339 | my $Table = shift; |
340 | |
341 | ## remaining args are optional key/val |
342 | my %Args = @_; |
343 | |
99598c8c |
344 | my $Fuzzy = delete $Args{Fuzzy}; |
345 | my $Desc = delete $Args{Desc}; # description |
346 | |
5beb625e |
347 | $Name = CanonicalName($Name) if $Fuzzy; |
cf25bb62 |
348 | |
349 | ## sanity check a few args |
350 | if (%Args or ($Type ne 'Is' and $Type ne 'In') or not ref $Table) { |
351 | confess "$0: bad args to New_Prop" |
352 | } |
353 | |
354 | if (not $TableInfo{$Type}->{$Name}) |
355 | { |
356 | $TableInfo{$Type}->{$Name} = $Table; |
99598c8c |
357 | $TableDesc{$Type}->{$Name} = $Desc; |
358 | if ($Fuzzy) { |
cf25bb62 |
359 | $FuzzyNames{$Type}->{$Name} = $Name; |
360 | } |
361 | } |
d73e5302 |
362 | } |
363 | |
d73e5302 |
364 | |
cf25bb62 |
365 | ## |
366 | ## Creates a new Table object. |
367 | ## |
368 | ## Args are key/value pairs: |
99598c8c |
369 | ## In => Name -- Name of "In" property to be associated with |
370 | ## Is => Name -- Name of "Is" property to be associated with |
371 | ## Fuzzy => Boolean -- True if name can be accessed "fuzzily" |
372 | ## Desc => String -- Description of the property |
cf25bb62 |
373 | ## |
374 | ## No args are required. |
375 | ## |
376 | sub Table::New |
377 | { |
378 | my $class = shift; |
379 | my %Args = @_; |
380 | |
381 | my $Table = bless [], $class; |
382 | |
99598c8c |
383 | my $Fuzzy = delete $Args{Fuzzy}; |
384 | my $Desc = delete $Args{Desc}; |
cf25bb62 |
385 | |
386 | for my $Type ('Is', 'In') |
387 | { |
388 | if (my $Name = delete $Args{$Type}) { |
99598c8c |
389 | New_Prop($Type => $Name, $Table, Desc => $Desc, Fuzzy => $Fuzzy); |
cf25bb62 |
390 | } |
391 | } |
392 | |
393 | ## shouldn't have any left over |
394 | if (%Args) { |
395 | confess "$0: bad args to Table->New" |
396 | } |
397 | |
398 | return $Table; |
d73e5302 |
399 | } |
400 | |
d73e5302 |
401 | |
cf25bb62 |
402 | ## |
403 | ## Returns the maximum code point currently in the table. |
404 | ## |
405 | sub Table::Max |
406 | { |
683a4240 |
407 | my $last = $_[0]->[-1]; ## last code point |
408 | confess "oops" unless $last; ## must have code points to have a max |
409 | return $last->[RANGE_END]; |
cf25bb62 |
410 | } |
d73e5302 |
411 | |
cf25bb62 |
412 | ## |
413 | ## Replaces the codepoints in the Table with those in the Table given |
414 | ## as an arg. (NOTE: this is not a "deep copy"). |
415 | ## |
416 | sub Table::Replace($$) |
417 | { |
418 | my $Table = shift; #self |
419 | my $New = shift; |
d73e5302 |
420 | |
cf25bb62 |
421 | @$Table = @$New; |
422 | } |
71d929cb |
423 | |
cf25bb62 |
424 | ## |
425 | ## Given a new code point, make the last range of the Table extend to |
426 | ## include the new (and all intervening) code points. |
427 | ## |
683a4240 |
428 | ## Takes the time to make sure that the extension is valid. |
429 | ## |
cf25bb62 |
430 | sub Table::Extend |
431 | { |
432 | my $Table = shift; #self |
433 | my $codepoint = shift; |
d73e5302 |
434 | |
cf25bb62 |
435 | my $PrevMax = $Table->Max; |
e904f995 |
436 | |
cf25bb62 |
437 | confess "oops ($codepoint <= $PrevMax)" if $codepoint <= $PrevMax; |
e904f995 |
438 | |
683a4240 |
439 | $Table->ExtendNoCheck($codepoint); |
440 | } |
441 | |
442 | |
443 | ## |
444 | ## Given a new code point, make the last range of the Table extend to |
445 | ## include the new (and all intervening) code points. |
446 | ## |
447 | ## Does NOT check that the extension is valid. Assumes that the caller |
448 | ## has already made this check. |
449 | ## |
450 | sub Table::ExtendNoCheck |
451 | { |
452 | ## Optmized adding: Assumes $Table and $codepoint as parms |
453 | $_[0]->[-1]->[RANGE_END] = $_[1]; |
cf25bb62 |
454 | } |
c3a8a2b8 |
455 | |
cf25bb62 |
456 | ## |
457 | ## Given a code point range start and end (and optional name), blindly |
458 | ## append them to the list of ranges for the Table. |
459 | ## |
460 | ## NOTE: Code points must be added in strictly ascending numeric order. |
461 | ## |
462 | sub Table::RawAppendRange |
463 | { |
464 | my $Table = shift; #self |
465 | my $start = shift; |
466 | my $end = shift; |
467 | my $name = shift; |
468 | $name = "" if not defined $name; ## warning: $name can be "0" |
469 | |
470 | push @$Table, [ $start, # RANGE_START |
471 | $end, # RANGE_END |
472 | $name ]; # RANGE_NAME |
473 | } |
c3a8a2b8 |
474 | |
cf25bb62 |
475 | ## |
476 | ## Given a code point (and optional name), add it to the Table. |
477 | ## |
478 | ## NOTE: Code points must be added in strictly ascending numeric order. |
479 | ## |
480 | sub Table::Append |
481 | { |
482 | my $Table = shift; #self |
483 | my $codepoint = shift; |
484 | my $name = shift; |
485 | $name = "" if not defined $name; ## warning: $name can be "0" |
486 | |
487 | ## |
488 | ## If we've already got a range working, and this code point is the next |
489 | ## one in line, and if the name is the same, just extend the current range. |
490 | ## |
683a4240 |
491 | my $last = $Table->[-1]; |
492 | if ($last |
cf25bb62 |
493 | and |
683a4240 |
494 | $last->[RANGE_END] == $codepoint - 1 |
cf25bb62 |
495 | and |
683a4240 |
496 | $last->[RANGE_NAME] eq $name) |
cf25bb62 |
497 | { |
683a4240 |
498 | $Table->ExtendNoCheck($codepoint); |
c3a8a2b8 |
499 | } |
cf25bb62 |
500 | else |
501 | { |
502 | $Table->RawAppendRange($codepoint, $codepoint, $name); |
d73e5302 |
503 | } |
cf25bb62 |
504 | } |
d73e5302 |
505 | |
cf25bb62 |
506 | ## |
507 | ## Given a code point range starting value and ending value (and name), |
508 | ## Add the range to teh Table. |
509 | ## |
510 | ## NOTE: Code points must be added in strictly ascending numeric order. |
511 | ## |
512 | sub Table::AppendRange |
513 | { |
514 | my $Table = shift; #self |
515 | my $start = shift; |
516 | my $end = shift; |
517 | my $name = shift; |
518 | $name = "" if not defined $name; ## warning: $name can be "0" |
519 | |
520 | $Table->Append($start, $name); |
521 | $Table->Extend($end) if $end > $start; |
522 | } |
d73e5302 |
523 | |
cf25bb62 |
524 | ## |
525 | ## Return a new Table that represents all code points not in the Table. |
526 | ## |
527 | sub Table::Invert |
528 | { |
529 | my $Table = shift; #self |
530 | |
531 | my $New = Table->New(); |
532 | my $max = -1; |
533 | for my $range (@$Table) |
534 | { |
535 | my $start = $range->[RANGE_START]; |
536 | my $end = $range->[RANGE_END]; |
537 | if ($start-1 >= $max+1) { |
538 | $New->AppendRange($max+1, $start-1, ""); |
539 | } |
540 | $max = $end; |
d73e5302 |
541 | } |
cf25bb62 |
542 | if ($max+1 < $LastUnicodeCodepoint) { |
543 | $New->AppendRange($max+1, $LastUnicodeCodepoint); |
d73e5302 |
544 | } |
cf25bb62 |
545 | return $New; |
546 | } |
d73e5302 |
547 | |
cf25bb62 |
548 | ## |
549 | ## Merges any number of other tables with $self, returning the new table. |
550 | ## (existing tables are not modified) |
551 | ## |
a3a8c5f0 |
552 | ## |
553 | ## Args may be Tables, or individual code points (as integers). |
554 | ## |
cf25bb62 |
555 | ## Can be called as either a constructor or a method. |
556 | ## |
557 | sub Table::Merge |
558 | { |
559 | shift(@_) if not ref $_[0]; ## if called as a constructor, lose the class |
560 | my @Tables = @_; |
561 | |
562 | ## Accumulate all records from all tables |
563 | my @Records; |
a3a8c5f0 |
564 | for my $Arg (@Tables) |
565 | { |
566 | if (ref $Arg) { |
567 | ## arg is a table -- get its ranges |
568 | push @Records, @$Arg; |
569 | } else { |
570 | ## arg is a codepoint, make a range |
571 | push @Records, [ $Arg, $Arg ] |
572 | } |
d73e5302 |
573 | } |
574 | |
cf25bb62 |
575 | ## sort by range start, with longer ranges coming first. |
576 | my ($first, @Rest) = sort { |
577 | ($a->[RANGE_START] <=> $b->[RANGE_START]) |
578 | or |
579 | ($b->[RANGE_END] <=> $b->[RANGE_END]) |
580 | } @Records; |
581 | |
582 | my $New = Table->New(); |
583 | |
584 | ## Ensuring the first range is there makes the subsequent loop easier |
585 | $New->AppendRange($first->[RANGE_START], |
586 | $first->[RANGE_END]); |
587 | |
588 | ## Fold in records so long as they add new information. |
589 | for my $set (@Rest) |
590 | { |
591 | my $start = $set->[RANGE_START]; |
592 | my $end = $set->[RANGE_END]; |
593 | if ($start > $New->Max) { |
594 | $New->AppendRange($start, $end); |
595 | } elsif ($end > $New->Max) { |
683a4240 |
596 | $New->ExtendNoCheck($end); |
cf25bb62 |
597 | } |
d73e5302 |
598 | } |
d73e5302 |
599 | |
cf25bb62 |
600 | return $New; |
d73e5302 |
601 | } |
602 | |
cf25bb62 |
603 | ## |
604 | ## Given a filename, write a representation of the Table to a file. |
99598c8c |
605 | ## May have an optional comment as a 2nd arg. |
d07a55ed |
606 | ## Filename may actually be an arrayref of directories |
cf25bb62 |
607 | ## |
608 | sub Table::Write |
609 | { |
99598c8c |
610 | my $Table = shift; #self |
cf25bb62 |
611 | my $filename = shift; |
99598c8c |
612 | my $comment = shift; |
d73e5302 |
613 | |
5beb625e |
614 | my @OUT = $HEADER; |
99598c8c |
615 | if (defined $comment) { |
616 | $comment =~ s/\s+\Z//; |
617 | $comment =~ s/^/# /gm; |
5beb625e |
618 | push @OUT, "#\n$comment\n#\n"; |
99598c8c |
619 | } |
5beb625e |
620 | push @OUT, "return <<'END';\n"; |
d73e5302 |
621 | |
cf25bb62 |
622 | for my $set (@$Table) |
623 | { |
624 | my $start = $set->[RANGE_START]; |
625 | my $end = $set->[RANGE_END]; |
626 | my $name = $set->[RANGE_NAME]; |
d73e5302 |
627 | |
cf25bb62 |
628 | if ($start == $end) { |
5beb625e |
629 | push @OUT, sprintf "%04X\t\t%s\n", $start, $name; |
cf25bb62 |
630 | } else { |
5beb625e |
631 | push @OUT, sprintf "%04X\t%04X\t%s\n", $start, $end, $name; |
cf25bb62 |
632 | } |
633 | } |
d73e5302 |
634 | |
5beb625e |
635 | push @OUT, "END\n"; |
636 | |
637 | WriteIfChanged($filename, @OUT); |
638 | } |
639 | |
640 | ## This used only for making the test script. |
641 | ## helper function |
642 | sub IsUsable($) |
643 | { |
644 | my $code = shift; |
645 | return 0 if $code <= 0x0000; ## don't use null |
646 | return 0 if $code >= $LastUnicodeCodepoint; ## keep in range |
647 | return 0 if ($code >= 0xD800 and $code <= 0xDFFF); ## no surrogates |
648 | return 0 if ($code >= 0xFDD0 and $code <= 0xFDEF); ## utf8.c says no good |
649 | return 0 if (($code & 0xFFFF) == 0xFFFE); ## utf8.c says no good |
650 | return 0 if (($code & 0xFFFF) == 0xFFFF); ## utf8.c says no good |
651 | return 1; |
652 | } |
653 | |
654 | ## Return a code point that's part of the table. |
655 | ## Returns nothing if the table is empty (or covers only surrogates). |
656 | ## This used only for making the test script. |
657 | sub Table::ValidCode |
658 | { |
659 | my $Table = shift; #self |
660 | for my $set (@$Table) { |
661 | return $set->[RANGE_END] if IsUsable($set->[RANGE_END]); |
662 | } |
663 | return (); |
664 | } |
665 | |
666 | ## Return a code point that's not part of the table |
667 | ## Returns nothing if the table covers all code points. |
668 | ## This used only for making the test script. |
669 | sub Table::InvalidCode |
670 | { |
671 | my $Table = shift; #self |
672 | |
683a4240 |
673 | return 0x1234 if not @$Table; |
5beb625e |
674 | |
675 | for my $set (@$Table) |
676 | { |
677 | if (IsUsable($set->[RANGE_END] + 1)) |
678 | { |
679 | return $set->[RANGE_END] + 1; |
680 | } |
681 | |
682 | if (IsUsable($set->[RANGE_START] - 1)) |
683 | { |
684 | return $set->[RANGE_START] - 1; |
685 | } |
686 | } |
687 | return (); |
cf25bb62 |
688 | } |
689 | |
690 | ########################################################################### |
691 | ########################################################################### |
692 | ########################################################################### |
693 | |
694 | |
695 | ## |
696 | ## Called like: |
99598c8c |
697 | ## New_Alias(Is => 'All', SameAs => 'Any', Fuzzy => 1); |
cf25bb62 |
698 | ## |
99598c8c |
699 | ## The args must be in that order, although the Fuzzy pair may be omitted. |
cf25bb62 |
700 | ## |
701 | ## This creates 'IsAll' as an alias for 'IsAny' |
702 | ## |
703 | sub New_Alias($$$@) |
704 | { |
705 | my $Type = shift; ## "Is" or "In" |
706 | my $Alias = shift; |
99598c8c |
707 | my $SameAs = shift; # expecting "SameAs" -- just ignored |
cf25bb62 |
708 | my $Name = shift; |
709 | |
710 | ## remaining args are optional key/val |
711 | my %Args = @_; |
712 | |
99598c8c |
713 | my $Fuzzy = delete $Args{Fuzzy}; |
cf25bb62 |
714 | |
715 | ## sanity check a few args |
716 | if (%Args or ($Type ne 'Is' and $Type ne 'In') or $SameAs ne 'SameAs') { |
717 | confess "$0: bad args to New_Alias" |
d73e5302 |
718 | } |
719 | |
5beb625e |
720 | $Alias = CanonicalName($Alias) if $Fuzzy; |
721 | |
722 | if (not $TableInfo{$Type}->{$Name}) |
723 | { |
724 | my $CName = CanonicalName($Name); |
725 | if ($TableInfo{$Type}->{$CName}) { |
726 | confess "$0: Use canonical form '$CName' instead of '$Name' for alias."; |
727 | } else { |
12ac2576 |
728 | confess "$0: don't have original $Type => $Name to make alias\n"; |
5beb625e |
729 | } |
cf25bb62 |
730 | } |
731 | if ($TableInfo{$Alias}) { |
732 | confess "$0: already have original $Type => $Alias; can't make alias"; |
d73e5302 |
733 | } |
cf25bb62 |
734 | $AliasInfo{$Type}->{$Name} = $Alias; |
99598c8c |
735 | if ($Fuzzy) { |
cf25bb62 |
736 | $FuzzyNames{$Type}->{$Alias} = $Name; |
737 | } |
738 | |
d73e5302 |
739 | } |
740 | |
d73e5302 |
741 | |
cf25bb62 |
742 | ## All assigned code points |
99598c8c |
743 | my $Assigned = Table->New(Is => 'Assigned', |
744 | Desc => "All assigned code points", |
5beb625e |
745 | Fuzzy => 0); |
d2d499f5 |
746 | |
cf25bb62 |
747 | my $Name = Table->New(); ## all characters, individually by name |
748 | my $General = Table->New(); ## all characters, grouped by category |
749 | my %General; |
750 | my %Cat; |
d73e5302 |
751 | |
7ebf06b3 |
752 | ## Simple Data::Dumper alike. Good enough for our needs. We can't use the real |
753 | ## thing as we have to run under miniperl |
754 | sub simple_dumper { |
755 | my @lines; |
756 | my $item; |
757 | foreach $item (@_) { |
758 | if (ref $item) { |
759 | if (ref $item eq 'ARRAY') { |
760 | push @lines, "[\n", simple_dumper (@$item), "],\n"; |
761 | } elsif (ref $item eq 'HASH') { |
762 | push @lines, "{\n", simple_dumper (%$item), "},\n"; |
763 | } else { |
764 | die "Can't cope with $item"; |
765 | } |
766 | } else { |
767 | if (defined $item) { |
768 | my $copy = $item; |
769 | $copy =~ s/([\'\\])/\\$1/gs; |
770 | push @lines, "'$copy',\n"; |
771 | } else { |
772 | push @lines, "undef,\n"; |
773 | } |
774 | } |
775 | } |
776 | @lines; |
777 | } |
778 | |
cf25bb62 |
779 | ## |
551b6b6f |
780 | ## Process UnicodeData.txt (Categories, etc.) |
cf25bb62 |
781 | ## |
44da8cae |
782 | sub UnicodeData_Txt() |
cf25bb62 |
783 | { |
784 | my $Bidi = Table->New(); |
785 | my $Deco = Table->New(); |
786 | my $Comb = Table->New(); |
787 | my $Number = Table->New(); |
12ac2576 |
788 | my $Mirrored = Table->New();#Is => 'Mirrored', |
789 | #Desc => "Mirrored in bidirectional text", |
790 | #Fuzzy => 0); |
d73e5302 |
791 | |
cf25bb62 |
792 | my %DC; |
793 | my %Bidi; |
12ac2576 |
794 | my %Number; |
795 | $DC{can} = Table->New(); |
796 | $DC{com} = Table->New(); |
cf25bb62 |
797 | |
798 | ## Initialize Perl-generated categories |
551b6b6f |
799 | ## (Categories from UnicodeData.txt are auto-initialized in gencat) |
44da8cae |
800 | $Cat{Alnum} = |
801 | Table->New(Is => 'Alnum', Desc => "[[:Alnum:]]", Fuzzy => 0); |
802 | $Cat{Alpha} = |
803 | Table->New(Is => 'Alpha', Desc => "[[:Alpha:]]", Fuzzy => 0); |
804 | $Cat{ASCII} = |
805 | Table->New(Is => 'ASCII', Desc => "[[:ASCII:]]", Fuzzy => 0); |
806 | $Cat{Blank} = |
807 | Table->New(Is => 'Blank', Desc => "[[:Blank:]]", Fuzzy => 0); |
808 | $Cat{Cntrl} = |
809 | Table->New(Is => 'Cntrl', Desc => "[[:Cntrl:]]", Fuzzy => 0); |
810 | $Cat{Digit} = |
811 | Table->New(Is => 'Digit', Desc => "[[:Digit:]]", Fuzzy => 0); |
812 | $Cat{Graph} = |
813 | Table->New(Is => 'Graph', Desc => "[[:Graph:]]", Fuzzy => 0); |
814 | $Cat{Lower} = |
815 | Table->New(Is => 'Lower', Desc => "[[:Lower:]]", Fuzzy => 0); |
816 | $Cat{Print} = |
817 | Table->New(Is => 'Print', Desc => "[[:Print:]]", Fuzzy => 0); |
818 | $Cat{Punct} = |
819 | Table->New(Is => 'Punct', Desc => "[[:Punct:]]", Fuzzy => 0); |
820 | $Cat{Space} = |
821 | Table->New(Is => 'Space', Desc => "[[:Space:]]", Fuzzy => 0); |
822 | $Cat{Title} = |
823 | Table->New(Is => 'Title', Desc => "[[:Title:]]", Fuzzy => 0); |
824 | $Cat{Upper} = |
825 | Table->New(Is => 'Upper', Desc => "[[:Upper:]]", Fuzzy => 0); |
826 | $Cat{XDigit} = |
827 | Table->New(Is => 'XDigit', Desc => "[[:XDigit:]]", Fuzzy => 0); |
828 | $Cat{Word} = |
829 | Table->New(Is => 'Word', Desc => "[[:Word:]]", Fuzzy => 0); |
830 | $Cat{SpacePerl} = |
831 | Table->New(Is => 'SpacePerl', Desc => '\s', Fuzzy => 0); |
e1d1eefb |
832 | $Cat{VertSpace} = |
833 | Table->New(Is => 'VertSpace', Desc => '\v', Fuzzy => 0); |
834 | $Cat{HorizSpace} = |
835 | Table->New(Is => 'HorizSpace', Desc => '\h', Fuzzy => 0); |
cf25bb62 |
836 | my %To; |
837 | $To{Upper} = Table->New(); |
838 | $To{Lower} = Table->New(); |
839 | $To{Title} = Table->New(); |
840 | $To{Digit} = Table->New(); |
841 | |
842 | sub gencat($$$$) |
843 | { |
844 | my ($name, ## Name ("LATIN CAPITAL LETTER A") |
845 | $cat, ## Category ("Lu", "Zp", "Nd", etc.) |
846 | $code, ## Code point (as an integer) |
847 | $op) = @_; |
848 | |
849 | my $MajorCat = substr($cat, 0, 1); ## L, M, Z, S, etc |
850 | |
851 | $Assigned->$op($code); |
852 | $Name->$op($code, $name); |
853 | $General->$op($code, $cat); |
854 | |
855 | ## add to the sub category (e.g. "Lu", "Nd", "Cf", ..) |
99598c8c |
856 | $Cat{$cat} ||= Table->New(Is => $cat, |
857 | Desc => "General Category '$cat'", |
858 | Fuzzy => 0); |
cf25bb62 |
859 | $Cat{$cat}->$op($code); |
860 | |
861 | ## add to the major category (e.g. "L", "N", "C", ...) |
99598c8c |
862 | $Cat{$MajorCat} ||= Table->New(Is => $MajorCat, |
863 | Desc => "Major Category '$MajorCat'", |
864 | Fuzzy => 0); |
cf25bb62 |
865 | $Cat{$MajorCat}->$op($code); |
866 | |
867 | ($General{$name} ||= Table->New)->$op($code, $name); |
868 | |
869 | # 005F: SPACING UNDERSCORE |
d75d706f |
870 | $Cat{Word}->$op($code) if $cat =~ /^[LMN]|Pc/; |
c65e4d19 |
871 | $Cat{Alnum}->$op($code) if $cat =~ /^[LM]|Nd/; |
cf25bb62 |
872 | $Cat{Alpha}->$op($code) if $cat =~ /^[LM]/; |
873 | |
d75d706f |
874 | my $isspace = |
875 | ($cat =~ /Zs|Zl|Zp/ && |
876 | $code != 0x200B) # 200B is ZWSP which is for line break control |
877 | # and therefore it is not part of "space" even while it is "Zs". |
cf25bb62 |
878 | || $code == 0x0009 # 0009: HORIZONTAL TAB |
879 | || $code == 0x000A # 000A: LINE FEED |
880 | || $code == 0x000B # 000B: VERTICAL TAB |
881 | || $code == 0x000C # 000C: FORM FEED |
e6f58734 |
882 | || $code == 0x000D # 000D: CARRIAGE RETURN |
d75d706f |
883 | || $code == 0x0085 # 0085: NEL |
884 | |
885 | ; |
cf25bb62 |
886 | |
d75d706f |
887 | $Cat{Space}->$op($code) if $isspace; |
cf25bb62 |
888 | |
d75d706f |
889 | $Cat{SpacePerl}->$op($code) if $isspace |
890 | && $code != 0x000B; # Backward compat. |
cf25bb62 |
891 | |
e1d1eefb |
892 | $Cat{VertSpace}->$op($code) if grep {$code == $_} |
893 | ( 0x0A..0x0D,0x85,0x2028,0x2029 ); |
894 | |
895 | $Cat{HorizSpace}->$op($code) if grep {$code == $_} ( |
896 | 0x09, 0x20, 0xa0, 0x1680, 0x180e, 0x2000, 0x2001, 0x2002, |
897 | 0x2003, 0x2004, 0x2005, 0x2006, 0x2007, 0x2008, 0x2009, 0x200a, |
898 | 0x202f, 0x205f, 0x3000 |
899 | ); |
900 | |
d75d706f |
901 | $Cat{Blank}->$op($code) if $isspace |
902 | && !($code == 0x000A || |
903 | $code == 0x000B || |
904 | $code == 0x000C || |
905 | $code == 0x000D || |
906 | $code == 0x0085 || |
907 | $cat =~ /^Z[lp]/); |
cf25bb62 |
908 | |
909 | $Cat{Digit}->$op($code) if $cat eq "Nd"; |
910 | $Cat{Upper}->$op($code) if $cat eq "Lu"; |
911 | $Cat{Lower}->$op($code) if $cat eq "Ll"; |
912 | $Cat{Title}->$op($code) if $cat eq "Lt"; |
913 | $Cat{ASCII}->$op($code) if $code <= 0x007F; |
914 | $Cat{Cntrl}->$op($code) if $cat =~ /^C/; |
d75d706f |
915 | my $isgraph = !$isspace && $cat !~ /Cc|Cs|Cn/; |
916 | $Cat{Graph}->$op($code) if $isgraph; |
917 | $Cat{Print}->$op($code) if $isgraph || $isspace; |
cf25bb62 |
918 | $Cat{Punct}->$op($code) if $cat =~ /^P/; |
919 | |
920 | $Cat{XDigit}->$op($code) if ($code >= 0x30 && $code <= 0x39) ## 0..9 |
921 | || ($code >= 0x41 && $code <= 0x46) ## A..F |
922 | || ($code >= 0x61 && $code <= 0x66); ## a..f |
923 | } |
d73e5302 |
924 | |
cf25bb62 |
925 | ## open ane read file..... |
551b6b6f |
926 | if (not open IN, "UnicodeData.txt") { |
927 | die "$0: UnicodeData.txt: $!\n"; |
cf25bb62 |
928 | } |
d73e5302 |
929 | |
a3a8c5f0 |
930 | ## |
931 | ## For building \p{_CombAbove} and \p{_CanonDCIJ} |
932 | ## |
933 | my %_Above_HexCodes; ## Hexcodes for chars with $comb == 230 ("ABOVE") |
934 | |
935 | my %CodeToDeco; ## Maps code to decomp. list for chars with first |
936 | ## decomp. char an "i" or "j" (for \p{_CanonDCIJ}) |
937 | |
938 | ## This is filled in as we go.... |
99598c8c |
939 | my $CombAbove = Table->New(Is => '_CombAbove', |
940 | Desc => '(for internal casefolding use)', |
941 | Fuzzy => 0); |
a3a8c5f0 |
942 | |
cf25bb62 |
943 | while (<IN>) |
944 | { |
945 | next unless /^[0-9A-Fa-f]+;/; |
946 | s/\s+$//; |
947 | |
948 | my ($hexcode, ## code point in hex (e.g. "0041") |
949 | $name, ## character name (e.g. "LATIN CAPITAL LETTER A") |
950 | $cat, ## category (e.g. "Lu") |
951 | $comb, ## Canonical combining class (e.t. "230") |
952 | $bidi, ## directional category (e.g. "L") |
953 | $deco, ## decomposition mapping |
954 | $decimal, ## decimal digit value |
955 | $digit, ## digit value |
956 | $number, ## numeric value |
957 | $mirrored, ## mirrored |
958 | $unicode10, ## name in Unicode 1.0 |
959 | $comment, ## comment field |
960 | $upper, ## uppercase mapping |
961 | $lower, ## lowercase mapping |
962 | $title, ## titlecase mapping |
963 | ) = split(/\s*;\s*/); |
964 | |
2eb5892f |
965 | # Note that in Unicode 3.2 there will be names like |
966 | # LINE FEED (LF), which probably means that \N{} needs |
967 | # to cope also with LINE FEED and LF. |
968 | $name = $unicode10 if $name eq '<control>' && $unicode10 ne ''; |
969 | |
cf25bb62 |
970 | my $code = hex($hexcode); |
971 | |
a3a8c5f0 |
972 | if ($comb and $comb == 230) { |
973 | $CombAbove->Append($code); |
974 | $_Above_HexCodes{$hexcode} = 1; |
975 | } |
976 | |
977 | ## Used in building \p{_CanonDCIJ} |
978 | if ($deco and $deco =~ m/^006[9A]\b/) { |
979 | $CodeToDeco{$code} = $deco; |
980 | } |
981 | |
cf25bb62 |
982 | ## |
983 | ## There are a few pairs of lines like: |
984 | ## AC00;<Hangul Syllable, First>;Lo;0;L;;;;;N;;;;; |
985 | ## D7A3;<Hangul Syllable, Last>;Lo;0;L;;;;;N;;;;; |
986 | ## that define ranges. |
987 | ## |
988 | if ($name =~ /^<(.+), (First|Last)>$/) |
989 | { |
990 | $name = $1; |
991 | gencat($name, $cat, $code, $2 eq 'First' ? 'Append' : 'Extend'); |
99598c8c |
992 | #New_Prop(In => $name, $General{$name}, Fuzzy => 1); |
cf25bb62 |
993 | } |
994 | else |
995 | { |
996 | ## normal (single-character) lines |
997 | gencat($name, $cat, $code, 'Append'); |
998 | |
999 | # No Append() here since since several codes may map into one. |
1000 | $To{Upper}->RawAppendRange($code, $code, $upper) if $upper; |
1001 | $To{Lower}->RawAppendRange($code, $code, $lower) if $lower; |
1002 | $To{Title}->RawAppendRange($code, $code, $title) if $title; |
1003 | $To{Digit}->Append($code, $decimal) if length $decimal; |
1004 | |
1005 | $Bidi->Append($code, $bidi); |
1006 | $Comb->Append($code, $comb) if $comb; |
1007 | $Number->Append($code, $number) if length $number; |
1008 | |
12ac2576 |
1009 | length($decimal) and ($Number{De} ||= Table->New())->Append($code) |
1010 | or |
1011 | length($digit) and ($Number{Di} ||= Table->New())->Append($code) |
1012 | or |
1013 | length($number) and ($Number{Nu} ||= Table->New())->Append($code); |
1014 | |
cf25bb62 |
1015 | $Mirrored->Append($code) if $mirrored eq "Y"; |
1016 | |
12ac2576 |
1017 | $Bidi{$bidi} ||= Table->New();#Is => "bt/$bidi", |
1018 | #Desc => "Bi-directional category '$bidi'", |
1019 | #Fuzzy => 0); |
cf25bb62 |
1020 | $Bidi{$bidi}->Append($code); |
1021 | |
1022 | if ($deco) |
1023 | { |
1024 | $Deco->Append($code, $deco); |
1025 | if ($deco =~/^<(\w+)>/) |
1026 | { |
12ac2576 |
1027 | my $dshort = $PVA_reverse{dt}{ucfirst lc $1}; |
1028 | $DC{com}->Append($code); |
cf25bb62 |
1029 | |
12ac2576 |
1030 | $DC{$dshort} ||= Table->New(); |
1031 | $DC{$dshort}->Append($code); |
cf25bb62 |
1032 | } |
1033 | else |
1034 | { |
12ac2576 |
1035 | $DC{can}->Append($code); |
cf25bb62 |
1036 | } |
1037 | } |
1038 | } |
1039 | } |
1040 | close IN; |
d2d499f5 |
1041 | |
cf25bb62 |
1042 | ## |
1043 | ## Tidy up a few special cases.... |
1044 | ## |
d73e5302 |
1045 | |
cf25bb62 |
1046 | $Cat{Cn} = $Assigned->Invert; ## Cn is everything that doesn't exist |
99598c8c |
1047 | New_Prop(Is => 'Cn', |
1048 | $Cat{Cn}, |
1049 | Desc => "General Category 'Cn' [not functional in Perl]", |
1050 | Fuzzy => 0); |
d73e5302 |
1051 | |
cf25bb62 |
1052 | ## Unassigned is the same as 'Cn' |
5beb625e |
1053 | New_Alias(Is => 'Unassigned', SameAs => 'Cn', Fuzzy => 0); |
d73e5302 |
1054 | |
cf25bb62 |
1055 | $Cat{C}->Replace($Cat{C}->Merge($Cat{Cn})); ## Now merge in Cn into C |
d73e5302 |
1056 | |
d73e5302 |
1057 | |
12ac2576 |
1058 | # LC is Ll, Lu, and Lt. |
1059 | # (used to be L& or L_, but PropValueAliases.txt defines it as LC) |
1060 | New_Prop(Is => 'LC', |
cf25bb62 |
1061 | Table->Merge(@Cat{qw[Ll Lu Lt]}), |
99598c8c |
1062 | Desc => '[\p{Ll}\p{Lu}\p{Lt}]', |
1063 | Fuzzy => 0); |
d73e5302 |
1064 | |
cf25bb62 |
1065 | ## Any and All are all code points. |
99598c8c |
1066 | my $Any = Table->New(Is => 'Any', |
1067 | Desc => sprintf("[\\x{0000}-\\x{%X}]", |
1068 | $LastUnicodeCodepoint), |
5beb625e |
1069 | Fuzzy => 0); |
cf25bb62 |
1070 | $Any->RawAppendRange(0, $LastUnicodeCodepoint); |
d73e5302 |
1071 | |
5beb625e |
1072 | New_Alias(Is => 'All', SameAs => 'Any', Fuzzy => 0); |
d73e5302 |
1073 | |
a3a8c5f0 |
1074 | ## |
1075 | ## Build special properties for Perl's internal case-folding needs: |
1076 | ## \p{_CaseIgnorable} |
1077 | ## \p{_CanonDCIJ} |
1078 | ## \p{_CombAbove} |
1079 | ## _CombAbove was built above. Others are built here.... |
1080 | ## |
1081 | |
1082 | ## \p{_CaseIgnorable} is [\p{Mn}\0x00AD\x2010] |
1083 | New_Prop(Is => '_CaseIgnorable', |
1084 | Table->Merge($Cat{Mn}, |
1085 | 0x00AD, #SOFT HYPHEN |
1086 | 0x2010), #HYPHEN |
99598c8c |
1087 | Desc => '(for internal casefolding use)', |
1088 | Fuzzy => 0); |
a3a8c5f0 |
1089 | |
1090 | |
1091 | ## \p{_CanonDCIJ} is fairly complex... |
99598c8c |
1092 | my $CanonCDIJ = Table->New(Is => '_CanonDCIJ', |
1093 | Desc => '(for internal casefolding use)', |
1094 | Fuzzy => 0); |
a3a8c5f0 |
1095 | ## It contains the ASCII 'i' and 'j'.... |
1096 | $CanonCDIJ->Append(0x0069); # ASCII ord("i") |
1097 | $CanonCDIJ->Append(0x006A); # ASCII ord("j") |
1098 | ## ...and any character with a decomposition that starts with either of |
1099 | ## those code points, but only if the decomposition does not have any |
1100 | ## combining character with the "ABOVE" canonical combining class. |
1101 | for my $code (sort { $a <=> $b} keys %CodeToDeco) |
1102 | { |
1103 | ## Need to ensure that all decomposition characters do not have |
1104 | ## a %HexCodeToComb in %AboveCombClasses. |
1105 | my $want = 1; |
1106 | for my $deco_hexcode (split / /, $CodeToDeco{$code}) |
1107 | { |
1108 | if (exists $_Above_HexCodes{$deco_hexcode}) { |
1109 | ## one of the decmposition chars has an ABOVE combination |
1110 | ## class, so we're not interested in this one |
1111 | $want = 0; |
1112 | last; |
1113 | } |
1114 | } |
1115 | if ($want) { |
1116 | $CanonCDIJ->Append($code); |
1117 | } |
1118 | } |
1119 | |
1120 | |
d73e5302 |
1121 | |
cf25bb62 |
1122 | ## |
1123 | ## Now dump the files. |
1124 | ## |
1125 | $Name->Write("Name.pl"); |
12ac2576 |
1126 | |
7ebf06b3 |
1127 | { |
1128 | my @PVA = $HEADER; |
1129 | foreach my $name (qw (PropertyAlias PA_reverse PropValueAlias |
1130 | PVA_reverse PVA_abbr_map)) { |
1131 | # Should I really jump through typeglob hoops just to avoid a |
1132 | # symbolic reference? (%{"utf8::$name}) |
1133 | push @PVA, "\n", "\%utf8::$name = (\n", |
1134 | simple_dumper (%{$utf8::{$name}}), ");\n"; |
1135 | } |
d403599d |
1136 | push @PVA, "1;\n"; |
7ebf06b3 |
1137 | WriteIfChanged("PVA.pl", @PVA); |
1138 | } |
1139 | |
12ac2576 |
1140 | # $Bidi->Write("Bidirectional.pl"); |
12ac2576 |
1141 | for (keys %Bidi) { |
1142 | $Bidi{$_}->Write( |
d07a55ed |
1143 | ["lib","bc","$_.pl"], |
12ac2576 |
1144 | "BidiClass category '$PropValueAlias{bc}{$_}'" |
1145 | ); |
1146 | } |
1147 | |
cf25bb62 |
1148 | $Comb->Write("CombiningClass.pl"); |
12ac2576 |
1149 | for (keys %{ $PropValueAlias{ccc} }) { |
1150 | my ($code, $name) = @{ $PropValueAlias{ccc}{$_} }; |
1151 | (my $c = Table->New())->Append($code); |
1152 | $c->Write( |
d07a55ed |
1153 | ["lib","ccc","$_.pl"], |
12ac2576 |
1154 | "CombiningClass category '$name'" |
1155 | ); |
1156 | } |
1157 | |
cf25bb62 |
1158 | $Deco->Write("Decomposition.pl"); |
12ac2576 |
1159 | for (keys %DC) { |
1160 | $DC{$_}->Write( |
d07a55ed |
1161 | ["lib","dt","$_.pl"], |
12ac2576 |
1162 | "DecompositionType category '$PropValueAlias{dt}{$_}'" |
1163 | ); |
1164 | } |
1165 | |
1166 | # $Number->Write("Number.pl"); |
12ac2576 |
1167 | for (keys %Number) { |
1168 | $Number{$_}->Write( |
d07a55ed |
1169 | ["lib","nt","$_.pl"], |
12ac2576 |
1170 | "NumericType category '$PropValueAlias{nt}{$_}'" |
1171 | ); |
1172 | } |
1173 | |
1174 | # $General->Write("Category.pl"); |
cf25bb62 |
1175 | |
1176 | for my $to (sort keys %To) { |
d07a55ed |
1177 | $To{$to}->Write(["To","$to.pl"]); |
d73e5302 |
1178 | } |
12ac2576 |
1179 | |
1180 | for (keys %{ $PropValueAlias{gc} }) { |
1181 | New_Alias(Is => $PropValueAlias{gc}{$_}, SameAs => $_, Fuzzy => 1); |
1182 | } |
d73e5302 |
1183 | } |
1184 | |
cf25bb62 |
1185 | ## |
551b6b6f |
1186 | ## Process LineBreak.txt |
cf25bb62 |
1187 | ## |
551b6b6f |
1188 | sub LineBreak_Txt() |
cf25bb62 |
1189 | { |
551b6b6f |
1190 | if (not open IN, "LineBreak.txt") { |
1191 | die "$0: LineBreak.txt: $!\n"; |
cf25bb62 |
1192 | } |
d73e5302 |
1193 | |
cf25bb62 |
1194 | my $Lbrk = Table->New(); |
1195 | my %Lbrk; |
d73e5302 |
1196 | |
cf25bb62 |
1197 | while (<IN>) |
1198 | { |
1199 | next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(\w+)/; |
d73e5302 |
1200 | |
cf25bb62 |
1201 | my ($first, $last, $lbrk) = (hex($1), hex($2||""), $3); |
d73e5302 |
1202 | |
cf25bb62 |
1203 | $Lbrk->Append($first, $lbrk); |
d73e5302 |
1204 | |
12ac2576 |
1205 | $Lbrk{$lbrk} ||= Table->New(); |
cf25bb62 |
1206 | $Lbrk{$lbrk}->Append($first); |
d73e5302 |
1207 | |
cf25bb62 |
1208 | if ($last) { |
1209 | $Lbrk->Extend($last); |
1210 | $Lbrk{$lbrk}->Extend($last); |
d73e5302 |
1211 | } |
1212 | } |
cf25bb62 |
1213 | close IN; |
d73e5302 |
1214 | |
12ac2576 |
1215 | # $Lbrk->Write("Lbrk.pl"); |
1216 | |
12ac2576 |
1217 | |
1218 | for (keys %Lbrk) { |
1219 | $Lbrk{$_}->Write( |
d07a55ed |
1220 | ["lib","lb","$_.pl"], |
12ac2576 |
1221 | "Linebreak category '$PropValueAlias{lb}{$_}'" |
1222 | ); |
1223 | } |
cf25bb62 |
1224 | } |
d73e5302 |
1225 | |
cf25bb62 |
1226 | ## |
551b6b6f |
1227 | ## Process ArabicShaping.txt. |
cf25bb62 |
1228 | ## |
551b6b6f |
1229 | sub ArabicShaping_txt() |
cf25bb62 |
1230 | { |
551b6b6f |
1231 | if (not open IN, "ArabicShaping.txt") { |
1232 | die "$0: ArabicShaping.txt: $!\n"; |
cf25bb62 |
1233 | } |
d73e5302 |
1234 | |
cf25bb62 |
1235 | my $ArabLink = Table->New(); |
1236 | my $ArabLinkGroup = Table->New(); |
d73e5302 |
1237 | |
12ac2576 |
1238 | my %JoinType; |
1239 | |
cf25bb62 |
1240 | while (<IN>) |
1241 | { |
1242 | next unless /^[0-9A-Fa-f]+;/; |
1243 | s/\s+$//; |
d73e5302 |
1244 | |
cf25bb62 |
1245 | my ($hexcode, $name, $link, $linkgroup) = split(/\s*;\s*/); |
1246 | my $code = hex($hexcode); |
1247 | $ArabLink->Append($code, $link); |
1248 | $ArabLinkGroup->Append($code, $linkgroup); |
12ac2576 |
1249 | |
1250 | $JoinType{$link} ||= Table->New(Is => "JoinType$link"); |
1251 | $JoinType{$link}->Append($code); |
d73e5302 |
1252 | } |
cf25bb62 |
1253 | close IN; |
1254 | |
12ac2576 |
1255 | # $ArabLink->Write("ArabLink.pl"); |
1256 | # $ArabLinkGroup->Write("ArabLnkGrp.pl"); |
1257 | |
12ac2576 |
1258 | |
1259 | for (keys %JoinType) { |
1260 | $JoinType{$_}->Write( |
d07a55ed |
1261 | ["lib","jt","$_.pl"], |
12ac2576 |
1262 | "JoiningType category '$PropValueAlias{jt}{$_}'" |
1263 | ); |
1264 | } |
1265 | } |
1266 | |
1267 | ## |
1268 | ## Process EastAsianWidth.txt. |
1269 | ## |
1270 | sub EastAsianWidth_txt() |
1271 | { |
1272 | if (not open IN, "EastAsianWidth.txt") { |
1273 | die "$0: EastAsianWidth.txt: $!\n"; |
1274 | } |
1275 | |
1276 | my %EAW; |
1277 | |
1278 | while (<IN>) |
1279 | { |
adc6a8bd |
1280 | next unless /^[0-9A-Fa-f]+(\.\.[0-9A-Fa-f]+)?;/; |
12ac2576 |
1281 | s/#.*//; |
1282 | s/\s+$//; |
1283 | |
adc6a8bd |
1284 | my ($hexcodes, $pv) = split(/\s*;\s*/); |
12ac2576 |
1285 | $EAW{$pv} ||= Table->New(Is => "EastAsianWidth$pv"); |
adc6a8bd |
1286 | my ($start, $end) = split(/\.\./, $hexcodes); |
1287 | if (defined $end) { |
1288 | $EAW{$pv}->AppendRange(hex($start), hex($end)); |
1289 | } else { |
1290 | $EAW{$pv}->Append(hex($start)); |
1291 | } |
12ac2576 |
1292 | } |
1293 | close IN; |
1294 | |
12ac2576 |
1295 | |
1296 | for (keys %EAW) { |
1297 | $EAW{$_}->Write( |
d07a55ed |
1298 | ["lib","ea","$_.pl"], |
12ac2576 |
1299 | "EastAsianWidth category '$PropValueAlias{ea}{$_}'" |
1300 | ); |
1301 | } |
1302 | } |
1303 | |
1304 | ## |
1305 | ## Process HangulSyllableType.txt. |
1306 | ## |
1307 | sub HangulSyllableType_txt() |
1308 | { |
1309 | if (not open IN, "HangulSyllableType.txt") { |
1310 | die "$0: HangulSyllableType.txt: $!\n"; |
1311 | } |
1312 | |
1313 | my %HST; |
1314 | |
1315 | while (<IN>) |
1316 | { |
1317 | next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(\w+)/; |
1318 | my ($first, $last, $pv) = (hex($1), hex($2||""), $3); |
1319 | |
1320 | $HST{$pv} ||= Table->New(Is => "HangulSyllableType$pv"); |
1321 | $HST{$pv}->Append($first); |
1322 | |
1323 | if ($last) { $HST{$pv}->Extend($last) } |
1324 | } |
1325 | close IN; |
1326 | |
12ac2576 |
1327 | for (keys %HST) { |
1328 | $HST{$_}->Write( |
d07a55ed |
1329 | ["lib","hst","$_.pl"], |
12ac2576 |
1330 | "HangulSyllableType category '$PropValueAlias{hst}{$_}'" |
1331 | ); |
1332 | } |
d73e5302 |
1333 | } |
1334 | |
cf25bb62 |
1335 | ## |
1336 | ## Process Jamo.txt. |
1337 | ## |
1338 | sub Jamo_txt() |
1339 | { |
1340 | if (not open IN, "Jamo.txt") { |
1341 | die "$0: Jamo.txt: $!\n"; |
1342 | } |
1343 | my $Short = Table->New(); |
d73e5302 |
1344 | |
cf25bb62 |
1345 | while (<IN>) |
1346 | { |
1347 | next unless /^([0-9A-Fa-f]+)\s*;\s*(\w*)/; |
1348 | my ($code, $short) = (hex($1), $2); |
d73e5302 |
1349 | |
cf25bb62 |
1350 | $Short->Append($code, $short); |
d73e5302 |
1351 | } |
cf25bb62 |
1352 | close IN; |
12ac2576 |
1353 | # $Short->Write("JamoShort.pl"); |
d73e5302 |
1354 | } |
1355 | |
cf25bb62 |
1356 | ## |
1357 | ## Process Scripts.txt. |
1358 | ## |
1359 | sub Scripts_txt() |
1360 | { |
1361 | my @ScriptInfo; |
d73e5302 |
1362 | |
cf25bb62 |
1363 | if (not open(IN, "Scripts.txt")) { |
1364 | die "$0: Scripts.txt: $!\n"; |
1365 | } |
1366 | while (<IN>) { |
1367 | next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(.+?)\s*\#/; |
d73e5302 |
1368 | |
cf25bb62 |
1369 | # Wait until all the scripts have been read since |
1370 | # they are not listed in numeric order. |
1371 | push @ScriptInfo, [ hex($1), hex($2||""), $3 ]; |
1372 | } |
1373 | close IN; |
d73e5302 |
1374 | |
cf25bb62 |
1375 | # Now append the scripts properties in their code point order. |
d73e5302 |
1376 | |
cf25bb62 |
1377 | my %Script; |
1378 | my $Scripts = Table->New(); |
d73e5302 |
1379 | |
cf25bb62 |
1380 | for my $script (sort { $a->[0] <=> $b->[0] } @ScriptInfo) |
1381 | { |
1382 | my ($first, $last, $name) = @$script; |
1383 | $Scripts->Append($first, $name); |
d73e5302 |
1384 | |
99598c8c |
1385 | $Script{$name} ||= Table->New(Is => $name, |
1386 | Desc => "Script '$name'", |
1387 | Fuzzy => 1); |
cf25bb62 |
1388 | $Script{$name}->Append($first, $name); |
d73e5302 |
1389 | |
cf25bb62 |
1390 | if ($last) { |
1391 | $Scripts->Extend($last); |
1392 | $Script{$name}->Extend($last); |
1393 | } |
1394 | } |
d73e5302 |
1395 | |
12ac2576 |
1396 | # $Scripts->Write("Scripts.pl"); |
d73e5302 |
1397 | |
cf25bb62 |
1398 | ## Common is everything not explicitly assigned to a Script |
1399 | ## |
1400 | ## ***shouldn't this be intersected with \p{Assigned}? ****** |
1401 | ## |
99598c8c |
1402 | New_Prop(Is => 'Common', |
1403 | $Scripts->Invert, |
1404 | Desc => 'Pseudo-Script of codepoints not in other Unicode scripts', |
1405 | Fuzzy => 1); |
cf25bb62 |
1406 | } |
d73e5302 |
1407 | |
cf25bb62 |
1408 | ## |
1409 | ## Given a name like "Close Punctuation", return a regex (that when applied |
1410 | ## with /i) matches any valid form of that name (e.g. "ClosePunctuation", |
1411 | ## "Close-Punctuation", etc.) |
1412 | ## |
1413 | ## Accept any space, dash, or underbar where in the official name there is |
1414 | ## space or a dash (or underbar, but there never is). |
1415 | ## |
1416 | ## |
1417 | sub NameToRegex($) |
1418 | { |
1419 | my $Name = shift; |
1420 | $Name =~ s/[- _]/(?:[-_]|\\s+)?/g; |
1421 | return $Name; |
1422 | } |
d73e5302 |
1423 | |
cf25bb62 |
1424 | ## |
1425 | ## Process Blocks.txt. |
1426 | ## |
1427 | sub Blocks_txt() |
1428 | { |
1429 | my $Blocks = Table->New(); |
1430 | my %Blocks; |
d73e5302 |
1431 | |
cf25bb62 |
1432 | if (not open IN, "Blocks.txt") { |
1433 | die "$0: Blocks.txt: $!\n"; |
1434 | } |
d73e5302 |
1435 | |
cf25bb62 |
1436 | while (<IN>) |
1437 | { |
1438 | #next if not /Private Use$/; |
1439 | next if not /^([0-9A-Fa-f]+)\.\.([0-9A-Fa-f]+)\s*;\s*(.+?)\s*$/; |
d73e5302 |
1440 | |
cf25bb62 |
1441 | my ($first, $last, $name) = (hex($1), hex($2), $3); |
d73e5302 |
1442 | |
cf25bb62 |
1443 | $Blocks->Append($first, $name); |
76ccdbe2 |
1444 | |
99598c8c |
1445 | $Blocks{$name} ||= Table->New(In => $name, |
1446 | Desc => "Block '$name'", |
1447 | Fuzzy => 1); |
cf25bb62 |
1448 | $Blocks{$name}->Append($first, $name); |
76ccdbe2 |
1449 | |
cf25bb62 |
1450 | if ($last and $last != $first) { |
1451 | $Blocks->Extend($last); |
1452 | $Blocks{$name}->Extend($last); |
d73e5302 |
1453 | } |
d73e5302 |
1454 | } |
cf25bb62 |
1455 | close IN; |
1456 | |
12ac2576 |
1457 | # $Blocks->Write("Blocks.pl"); |
d73e5302 |
1458 | } |
1459 | |
cf25bb62 |
1460 | ## |
1461 | ## Read in the PropList.txt. It contains extended properties not |
551b6b6f |
1462 | ## listed in the UnicodeData.txt, such as 'Other_Alphabetic': |
cf25bb62 |
1463 | ## alphabetic but not of the general category L; many modifiers |
1464 | ## belong to this extended property category: while they are not |
1465 | ## alphabets, they are alphabetic in nature. |
1466 | ## |
1467 | sub PropList_txt() |
1468 | { |
1469 | my @PropInfo; |
1470 | |
1471 | if (not open IN, "PropList.txt") { |
1472 | die "$0: PropList.txt: $!\n"; |
1473 | } |
d73e5302 |
1474 | |
cf25bb62 |
1475 | while (<IN>) |
1476 | { |
1477 | next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(.+?)\s*\#/; |
d73e5302 |
1478 | |
cf25bb62 |
1479 | # Wait until all the extended properties have been read since |
1480 | # they are not listed in numeric order. |
1481 | push @PropInfo, [ hex($1), hex($2||""), $3 ]; |
1482 | } |
1483 | close IN; |
71d929cb |
1484 | |
cf25bb62 |
1485 | # Now append the extended properties in their code point order. |
1486 | my $Props = Table->New(); |
1487 | my %Prop; |
71d929cb |
1488 | |
cf25bb62 |
1489 | for my $prop (sort { $a->[0] <=> $b->[0] } @PropInfo) |
1490 | { |
1491 | my ($first, $last, $name) = @$prop; |
1492 | $Props->Append($first, $name); |
71d929cb |
1493 | |
99598c8c |
1494 | $Prop{$name} ||= Table->New(Is => $name, |
1495 | Desc => "Extended property '$name'", |
1496 | Fuzzy => 1); |
cf25bb62 |
1497 | $Prop{$name}->Append($first, $name); |
71d929cb |
1498 | |
cf25bb62 |
1499 | if ($last) { |
1500 | $Props->Extend($last); |
1501 | $Prop{$name}->Extend($last); |
1502 | } |
71d929cb |
1503 | } |
d73e5302 |
1504 | |
12ac2576 |
1505 | for (keys %Prop) { |
1506 | (my $file = $PA_reverse{$_}) =~ tr/_//d; |
8e4b7420 |
1507 | # XXX I'm assuming that the names from %Prop don't suffer 8.3 clashes. |
1508 | $BaseNames{lc $file}++; |
12ac2576 |
1509 | $Prop{$_}->Write( |
d07a55ed |
1510 | ["lib","gc_sc","$file.pl"], |
12ac2576 |
1511 | "Binary property '$_'" |
1512 | ); |
1513 | } |
1514 | |
8158862b |
1515 | # Alphabetic is L, Nl, and Other_Alphabetic. |
99598c8c |
1516 | New_Prop(Is => 'Alphabetic', |
8158862b |
1517 | Table->Merge($Cat{L}, $Cat{Nl}, $Prop{Other_Alphabetic}), |
1518 | Desc => '[\p{L}\p{Nl}\p{OtherAlphabetic}]', # canonical names |
99598c8c |
1519 | Fuzzy => 1); |
cf25bb62 |
1520 | |
1521 | # Lowercase is Ll and Other_Lowercase. |
99598c8c |
1522 | New_Prop(Is => 'Lowercase', |
cf25bb62 |
1523 | Table->Merge($Cat{Ll}, $Prop{Other_Lowercase}), |
8158862b |
1524 | Desc => '[\p{Ll}\p{OtherLowercase}]', # canonical names |
99598c8c |
1525 | Fuzzy => 1); |
cf25bb62 |
1526 | |
1527 | # Uppercase is Lu and Other_Uppercase. |
1528 | New_Prop(Is => 'Uppercase', |
1529 | Table->Merge($Cat{Lu}, $Prop{Other_Uppercase}), |
8158862b |
1530 | Desc => '[\p{Lu}\p{OtherUppercase}]', # canonical names |
99598c8c |
1531 | Fuzzy => 1); |
cf25bb62 |
1532 | |
1533 | # Math is Sm and Other_Math. |
1534 | New_Prop(Is => 'Math', |
1535 | Table->Merge($Cat{Sm}, $Prop{Other_Math}), |
8158862b |
1536 | Desc => '[\p{Sm}\p{OtherMath}]', # canonical names |
99598c8c |
1537 | Fuzzy => 1); |
cf25bb62 |
1538 | |
8158862b |
1539 | # ID_Start is Ll, Lu, Lt, Lm, Lo, Nl, and Other_ID_Start. |
cf25bb62 |
1540 | New_Prop(Is => 'ID_Start', |
8158862b |
1541 | Table->Merge(@Cat{qw[Ll Lu Lt Lm Lo Nl]}, $Prop{Other_ID_Start}), |
1542 | Desc => '[\p{Ll}\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{Nl}\p{OtherIDStart}]', |
99598c8c |
1543 | Fuzzy => 1); |
cf25bb62 |
1544 | |
8158862b |
1545 | # ID_Continue is ID_Start, Mn, Mc, Nd, Pc, and Other_ID_Continue. |
cf25bb62 |
1546 | New_Prop(Is => 'ID_Continue', |
8158862b |
1547 | Table->Merge(@Cat{qw[Ll Lu Lt Lm Lo Nl Mn Mc Nd Pc ]}, |
1548 | @Prop{qw[Other_ID_Start Other_ID_Continue]}), |
1549 | Desc => '[\p{ID_Start}\p{Mn}\p{Mc}\p{Nd}\p{Pc}\p{OtherIDContinue}]', |
99598c8c |
1550 | Fuzzy => 1); |
8158862b |
1551 | |
1552 | # Default_Ignorable_Code_Point = Other_Default_Ignorable_Code_Point |
1553 | # + Cf + Cc + Cs + Noncharacter + Variation_Selector |
1554 | # - WhiteSpace - FFF9..FFFB (Annotation Characters) |
1555 | |
1556 | my $Annotation = Table->New(); |
1557 | $Annotation->RawAppendRange(0xFFF9, 0xFFFB); |
1558 | |
1559 | New_Prop(Is => 'Default_Ignorable_Code_Point', |
1560 | Table->Merge(@Cat{qw[Cf Cc Cs]}, |
1561 | $Prop{Noncharacter_Code_Point}, |
1562 | $Prop{Variation_Selector}, |
1563 | $Prop{Other_Default_Ignorable_Code_Point}) |
1564 | ->Invert |
1565 | ->Merge($Prop{White_Space}, $Annotation) |
1566 | ->Invert, |
1567 | Desc => '(?![\p{WhiteSpace}\x{FFF9}-\x{FFFB}])[\p{Cf}\p{Cc}'. |
1568 | '\p{Cs}\p{NoncharacterCodePoint}\p{VariationSelector}'. |
1569 | '\p{OtherDefaultIgnorableCodePoint}]', |
1570 | Fuzzy => 1); |
1571 | |
d73e5302 |
1572 | } |
1573 | |
5beb625e |
1574 | |
1575 | ## |
1576 | ## These are used in: |
1577 | ## MakePropTestScript() |
1578 | ## WriteAllMappings() |
1579 | ## for making the test script. |
1580 | ## |
1581 | my %FuzzyNameToTest; |
1582 | my %ExactNameToTest; |
1583 | |
1584 | |
1585 | ## This used only for making the test script |
1586 | sub GenTests($$$$) |
1587 | { |
1588 | my $FH = shift; |
1589 | my $Prop = shift; |
1590 | my $MatchCode = shift; |
1591 | my $FailCode = shift; |
1592 | |
1593 | if (defined $MatchCode) { |
1594 | printf $FH qq/Expect(1, "\\x{%04X}", '\\p{$Prop}' );\n/, $MatchCode; |
1595 | printf $FH qq/Expect(0, "\\x{%04X}", '\\p{^$Prop}');\n/, $MatchCode; |
1596 | printf $FH qq/Expect(0, "\\x{%04X}", '\\P{$Prop}' );\n/, $MatchCode; |
1597 | printf $FH qq/Expect(1, "\\x{%04X}", '\\P{^$Prop}');\n/, $MatchCode; |
1598 | } |
1599 | if (defined $FailCode) { |
1600 | printf $FH qq/Expect(0, "\\x{%04X}", '\\p{$Prop}' );\n/, $FailCode; |
1601 | printf $FH qq/Expect(1, "\\x{%04X}", '\\p{^$Prop}');\n/, $FailCode; |
1602 | printf $FH qq/Expect(1, "\\x{%04X}", '\\P{$Prop}' );\n/, $FailCode; |
1603 | printf $FH qq/Expect(0, "\\x{%04X}", '\\P{^$Prop}');\n/, $FailCode; |
1604 | } |
1605 | } |
1606 | |
1607 | ## This used only for making the test script |
1608 | sub ExpectError($$) |
1609 | { |
1610 | my $FH = shift; |
1611 | my $prop = shift; |
1612 | |
1613 | print $FH qq/Error('\\p{$prop}');\n/; |
1614 | print $FH qq/Error('\\P{$prop}');\n/; |
1615 | } |
1616 | |
1617 | ## This used only for making the test script |
1618 | my @GoodSeps = ( |
1619 | " ", |
1620 | "-", |
1621 | " \t ", |
1622 | "", |
1623 | "", |
1624 | "_", |
1625 | ); |
1626 | my @BadSeps = ( |
1627 | "--", |
1628 | "__", |
1629 | " _", |
1630 | "/" |
1631 | ); |
1632 | |
1633 | ## This used only for making the test script |
1634 | sub RandomlyFuzzifyName($;$) |
1635 | { |
1636 | my $Name = shift; |
1637 | my $WantError = shift; ## if true, make an error |
1638 | |
1639 | my @parts; |
1640 | for my $part (split /[-\s_]+/, $Name) |
1641 | { |
1642 | if (@parts) { |
1643 | if ($WantError and rand() < 0.3) { |
1644 | push @parts, $BadSeps[rand(@BadSeps)]; |
1645 | $WantError = 0; |
1646 | } else { |
1647 | push @parts, $GoodSeps[rand(@GoodSeps)]; |
1648 | } |
1649 | } |
1650 | my $switch = int rand(4); |
1651 | if ($switch == 0) { |
1652 | push @parts, uc $part; |
1653 | } elsif ($switch == 1) { |
1654 | push @parts, lc $part; |
1655 | } elsif ($switch == 2) { |
1656 | push @parts, ucfirst $part; |
1657 | } else { |
1658 | push @parts, $part; |
1659 | } |
1660 | } |
1661 | my $new = join('', @parts); |
1662 | |
1663 | if ($WantError) { |
1664 | if (rand() >= 0.5) { |
1665 | $new .= $BadSeps[rand(@BadSeps)]; |
1666 | } else { |
1667 | $new = $BadSeps[rand(@BadSeps)] . $new; |
1668 | } |
1669 | } |
1670 | return $new; |
1671 | } |
1672 | |
1673 | ## This used only for making the test script |
1674 | sub MakePropTestScript() |
1675 | { |
1676 | ## this written directly -- it's huge. |
d07a55ed |
1677 | force_unlink ("TestProp.pl"); |
5beb625e |
1678 | if (not open OUT, ">TestProp.pl") { |
1679 | die "$0: TestProp.pl: $!\n"; |
1680 | } |
1681 | print OUT <DATA>; |
1682 | |
1683 | while (my ($Name, $Table) = each %ExactNameToTest) |
1684 | { |
1685 | GenTests(*OUT, $Name, $Table->ValidCode, $Table->InvalidCode); |
1686 | ExpectError(*OUT, uc $Name) if uc $Name ne $Name; |
1687 | ExpectError(*OUT, lc $Name) if lc $Name ne $Name; |
1688 | } |
1689 | |
1690 | |
1691 | while (my ($Name, $Table) = each %FuzzyNameToTest) |
1692 | { |
1693 | my $Orig = $CanonicalToOrig{$Name}; |
1694 | my %Names = ( |
1695 | $Name => 1, |
1696 | $Orig => 1, |
1697 | RandomlyFuzzifyName($Orig) => 1 |
1698 | ); |
1699 | |
1700 | for my $N (keys %Names) { |
1701 | GenTests(*OUT, $N, $Table->ValidCode, $Table->InvalidCode); |
1702 | } |
1703 | |
1704 | ExpectError(*OUT, RandomlyFuzzifyName($Orig, 'ERROR')); |
1705 | } |
1706 | |
1707 | print OUT "Finished();\n"; |
1708 | close OUT; |
1709 | } |
1710 | |
1711 | |
1712 | ## |
1713 | ## These are used only in: |
1714 | ## RegisterFileForName() |
1715 | ## WriteAllMappings() |
1716 | ## |
1717 | my %Exact; ## will become %utf8::Exact; |
1718 | my %Canonical; ## will become %utf8::Canonical; |
1719 | my %CaComment; ## Comment for %Canonical entry of same key |
1720 | |
1721 | ## |
1722 | ## Given info about a name and a datafile that it should be associated with, |
1723 | ## register that assocation in %Exact and %Canonical. |
1724 | sub RegisterFileForName($$$$) |
1725 | { |
1726 | my $Type = shift; |
1727 | my $Name = shift; |
1728 | my $IsFuzzy = shift; |
1729 | my $filename = shift; |
1730 | |
1731 | ## |
1732 | ## Now in details for the mapping. $Type eq 'Is' has the |
1733 | ## Is removed, as it will be removed in utf8_heavy when this |
1734 | ## data is being checked. In keeps its "In", but a second |
1735 | ## sans-In record is written if it doesn't conflict with |
1736 | ## anything already there. |
1737 | ## |
1738 | if (not $IsFuzzy) |
1739 | { |
1740 | if ($Type eq 'Is') { |
1741 | die "oops[$Name]" if $Exact{$Name}; |
1742 | $Exact{$Name} = $filename; |
1743 | } else { |
1744 | die "oops[$Type$Name]" if $Exact{"$Type$Name"}; |
1745 | $Exact{"$Type$Name"} = $filename; |
1746 | $Exact{$Name} = $filename if not $Exact{$Name}; |
1747 | } |
1748 | } |
1749 | else |
1750 | { |
1751 | my $CName = lc $Name; |
1752 | if ($Type eq 'Is') { |
1753 | die "oops[$CName]" if $Canonical{$CName}; |
1754 | $Canonical{$CName} = $filename; |
1755 | $CaComment{$CName} = $Name if $Name =~ tr/A-Z// >= 2; |
1756 | } else { |
1757 | die "oops[$Type$CName]" if $Canonical{lc "$Type$CName"}; |
1758 | $Canonical{lc "$Type$CName"} = $filename; |
1759 | $CaComment{lc "$Type$CName"} = "$Type$Name"; |
1760 | if (not $Canonical{$CName}) { |
1761 | $Canonical{$CName} = $filename; |
1762 | $CaComment{$CName} = "$Type$Name"; |
1763 | } |
1764 | } |
1765 | } |
1766 | } |
1767 | |
cf25bb62 |
1768 | ## |
1769 | ## Writes the info accumulated in |
1770 | ## |
1771 | ## %TableInfo; |
1772 | ## %FuzzyNames; |
1773 | ## %AliasInfo; |
1774 | ## |
1775 | ## |
1776 | sub WriteAllMappings() |
1777 | { |
99598c8c |
1778 | my @MAP; |
1779 | |
5beb625e |
1780 | ## 'Is' *MUST* come first, so its names have precidence over 'In's |
1781 | for my $Type ('Is', 'In') |
1782 | { |
1783 | my %RawNameToFile; ## a per-$Type cache |
cf25bb62 |
1784 | |
5beb625e |
1785 | for my $Name (sort {length $a <=> length $b} keys %{$TableInfo{$Type}}) |
cf25bb62 |
1786 | { |
5beb625e |
1787 | ## Note: $Name is already canonical |
99598c8c |
1788 | my $Table = $TableInfo{$Type}->{$Name}; |
5beb625e |
1789 | my $IsFuzzy = $FuzzyNames{$Type}->{$Name}; |
99598c8c |
1790 | |
1791 | ## Need an 8.3 safe filename (which means "an 8 safe" $filename) |
5beb625e |
1792 | my $filename; |
cf25bb62 |
1793 | { |
5beb625e |
1794 | ## 'Is' items lose 'Is' from the basename. |
12ac2576 |
1795 | $filename = $Type eq 'Is' ? |
1796 | ($PVA_reverse{sc}{$Name} || $Name) : |
1797 | "$Type$Name"; |
5beb625e |
1798 | |
1799 | $filename =~ s/[^\w_]+/_/g; # "L&" -> "L_" |
1800 | substr($filename, 8) = '' if length($filename) > 8; |
1801 | |
1802 | ## |
1803 | ## Make sure the basename doesn't conflict with something we |
1804 | ## might have already written. If we have, say, |
1805 | ## InGreekExtended1 |
1806 | ## InGreekExtended2 |
1807 | ## they become |
1808 | ## InGreekE |
1809 | ## InGreek2 |
1810 | ## |
1811 | while (my $num = $BaseNames{lc $filename}++) |
1812 | { |
1813 | $num++; ## so basenames with numbers start with '2', which |
1814 | ## just looks more natural. |
1815 | ## Want to append $num, but if it'll make the basename longer |
1816 | ## than 8 characters, pre-truncate $filename so that the result |
1817 | ## is acceptable. |
1818 | my $delta = length($filename) + length($num) - 8; |
1819 | if ($delta > 0) { |
1820 | substr($filename, -$delta) = $num; |
1821 | } else { |
1822 | $filename .= $num; |
1823 | } |
99598c8c |
1824 | } |
5beb625e |
1825 | }; |
99598c8c |
1826 | |
1827 | ## |
1828 | ## Construct a nice comment to add to the file, and build data |
1829 | ## for the "./Properties" file along the way. |
1830 | ## |
1831 | my $Comment; |
1832 | { |
1833 | my $Desc = $TableDesc{$Type}->{$Name} || ""; |
1834 | ## get list of names this table is reference by |
1835 | my @Supported = $Name; |
1836 | while (my ($Orig, $Alias) = each %{ $AliasInfo{$Type} }) |
1837 | { |
1838 | if ($Orig eq $Name) { |
1839 | push @Supported, $Alias; |
1840 | } |
1841 | } |
1842 | |
1843 | my $TypeToShow = $Type eq 'Is' ? "" : $Type; |
1844 | my $OrigProp; |
1845 | |
1846 | $Comment = "This file supports:\n"; |
1847 | for my $N (@Supported) |
1848 | { |
1849 | my $IsFuzzy = $FuzzyNames{$Type}->{$N}; |
5beb625e |
1850 | my $Prop = "\\p{$TypeToShow$Name}"; |
99598c8c |
1851 | $OrigProp = $Prop if not $OrigProp; #cache for aliases |
1852 | if ($IsFuzzy) { |
1853 | $Comment .= "\t$Prop (and fuzzy permutations)\n"; |
1854 | } else { |
1855 | $Comment .= "\t$Prop\n"; |
1856 | } |
1857 | my $MyDesc = ($N eq $Name) ? $Desc : "Alias for $OrigProp ($Desc)"; |
1858 | |
1859 | push @MAP, sprintf("%s %-42s %s\n", |
1860 | $IsFuzzy ? '*' : ' ', $Prop, $MyDesc); |
1861 | } |
1862 | if ($Desc) { |
1863 | $Comment .= "\nMeaning: $Desc\n"; |
1864 | } |
1865 | |
1866 | } |
cf25bb62 |
1867 | ## |
1868 | ## Okay, write the file... |
1869 | ## |
d07a55ed |
1870 | $Table->Write(["lib","gc_sc","$filename.pl"], $Comment); |
99598c8c |
1871 | |
5beb625e |
1872 | ## and register it |
1873 | $RawNameToFile{$Name} = $filename; |
1874 | RegisterFileForName($Type => $Name, $IsFuzzy, $filename); |
cf25bb62 |
1875 | |
5beb625e |
1876 | if ($IsFuzzy) |
1877 | { |
1878 | my $CName = CanonicalName($Type . '_'. $Name); |
1879 | $FuzzyNameToTest{$Name} = $Table if !$FuzzyNameToTest{$Name}; |
1880 | $FuzzyNameToTest{$CName} = $Table if !$FuzzyNameToTest{$CName}; |
1881 | } else { |
1882 | $ExactNameToTest{$Name} = $Table; |
cf25bb62 |
1883 | } |
1884 | |
cf25bb62 |
1885 | } |
1886 | |
5beb625e |
1887 | ## Register aliase info |
1888 | for my $Name (sort {length $a <=> length $b} keys %{$AliasInfo{$Type}}) |
cf25bb62 |
1889 | { |
5beb625e |
1890 | my $Alias = $AliasInfo{$Type}->{$Name}; |
1891 | my $IsFuzzy = $FuzzyNames{$Type}->{$Alias}; |
1892 | my $filename = $RawNameToFile{$Name}; |
1893 | die "oops [$Alias]->[$Name]" if not $filename; |
1894 | RegisterFileForName($Type => $Alias, $IsFuzzy, $filename); |
1895 | |
1896 | my $Table = $TableInfo{$Type}->{$Name}; |
1897 | die "oops" if not $Table; |
1898 | if ($IsFuzzy) |
1899 | { |
1900 | my $CName = CanonicalName($Type .'_'. $Alias); |
1901 | $FuzzyNameToTest{$Alias} = $Table if !$FuzzyNameToTest{$Alias}; |
1902 | $FuzzyNameToTest{$CName} = $Table if !$FuzzyNameToTest{$CName}; |
1903 | } else { |
1904 | $ExactNameToTest{$Alias} = $Table; |
1905 | } |
cf25bb62 |
1906 | } |
5beb625e |
1907 | } |
cf25bb62 |
1908 | |
5beb625e |
1909 | ## |
1910 | ## Write out the property list |
1911 | ## |
1912 | { |
1913 | my @OUT = ( |
1914 | "##\n", |
1915 | "## This file created by $0\n", |
1916 | "## List of built-in \\p{...}/\\P{...} properties.\n", |
1917 | "##\n", |
1918 | "## '*' means name may be 'fuzzy'\n", |
1919 | "##\n\n", |
1920 | sort { substr($a,2) cmp substr($b, 2) } @MAP, |
1921 | ); |
1922 | WriteIfChanged('Properties', @OUT); |
1923 | } |
cf25bb62 |
1924 | |
5beb625e |
1925 | use Text::Tabs (); ## using this makes the files about half the size |
1926 | |
1927 | ## Write Exact.pl |
1928 | { |
1929 | my @OUT = ( |
1930 | $HEADER, |
1931 | "##\n", |
1932 | "## Data in this file used by ../utf8_heavy.pl\n", |
1933 | "##\n\n", |
12ac2576 |
1934 | "## Mapping from name to filename in ./lib/gc_sc\n", |
5beb625e |
1935 | "%utf8::Exact = (\n", |
1936 | ); |
cf25bb62 |
1937 | |
12ac2576 |
1938 | $Exact{InGreek} = 'InGreekA'; # this is evil kludge |
cf25bb62 |
1939 | for my $Name (sort keys %Exact) |
1940 | { |
1941 | my $File = $Exact{$Name}; |
5beb625e |
1942 | $Name = $Name =~ m/\W/ ? qq/'$Name'/ : " $Name "; |
1943 | my $Text = sprintf("%-15s => %s,\n", $Name, qq/'$File'/); |
1944 | push @OUT, Text::Tabs::unexpand($Text); |
cf25bb62 |
1945 | } |
5beb625e |
1946 | push @OUT, ");\n1;\n"; |
1947 | |
1948 | WriteIfChanged('Exact.pl', @OUT); |
1949 | } |
cf25bb62 |
1950 | |
5beb625e |
1951 | ## Write Canonical.pl |
1952 | { |
1953 | my @OUT = ( |
1954 | $HEADER, |
1955 | "##\n", |
1956 | "## Data in this file used by ../utf8_heavy.pl\n", |
1957 | "##\n\n", |
1958 | "## Mapping from lc(canonical name) to filename in ./lib\n", |
1959 | "%utf8::Canonical = (\n", |
1960 | ); |
1961 | my $Trail = ""; ## used just to keep the spacing pretty |
1962 | for my $Name (sort keys %Canonical) |
cf25bb62 |
1963 | { |
5beb625e |
1964 | my $File = $Canonical{$Name}; |
1965 | if ($CaComment{$Name}) { |
1966 | push @OUT, "\n" if not $Trail; |
1967 | push @OUT, " # $CaComment{$Name}\n"; |
1968 | $Trail = "\n"; |
1969 | } else { |
1970 | $Trail = ""; |
cf25bb62 |
1971 | } |
5beb625e |
1972 | $Name = $Name =~ m/\W/ ? qq/'$Name'/ : " $Name "; |
1973 | my $Text = sprintf(" %-41s => %s,\n$Trail", $Name, qq/'$File'/); |
1974 | push @OUT, Text::Tabs::unexpand($Text); |
cf25bb62 |
1975 | } |
5beb625e |
1976 | push @OUT, ");\n1\n"; |
1977 | WriteIfChanged('Canonical.pl', @OUT); |
d2d499f5 |
1978 | } |
5beb625e |
1979 | |
1980 | MakePropTestScript() if $MakeTestScript; |
d2d499f5 |
1981 | } |
1982 | |
5beb625e |
1983 | |
551b6b6f |
1984 | sub SpecialCasing_txt() |
cf25bb62 |
1985 | { |
1986 | # |
1987 | # Read in the special cases. |
1988 | # |
983ffd37 |
1989 | |
cf25bb62 |
1990 | my %CaseInfo; |
1991 | |
551b6b6f |
1992 | if (not open IN, "SpecialCasing.txt") { |
1993 | die "$0: SpecialCasing.txt: $!\n"; |
cf25bb62 |
1994 | } |
1995 | while (<IN>) { |
1996 | next unless /^[0-9A-Fa-f]+;/; |
1997 | s/\#.*//; |
1998 | s/\s+$//; |
1999 | |
2000 | my ($code, $lower, $title, $upper, $condition) = split(/\s*;\s*/); |
2001 | |
2002 | if ($condition) { # not implemented yet |
2003 | print "# SKIPPING $_\n" if $Verbose; |
2004 | next; |
2005 | } |
2006 | |
2007 | # Wait until all the special cases have been read since |
2008 | # they are not listed in numeric order. |
2009 | my $ix = hex($code); |
a6da3000 |
2010 | push @{$CaseInfo{Lower}}, [ $ix, $code, $lower ] |
2011 | unless $code eq $lower; |
2012 | push @{$CaseInfo{Title}}, [ $ix, $code, $title ] |
2013 | unless $code eq $title; |
2014 | push @{$CaseInfo{Upper}}, [ $ix, $code, $upper ] |
2015 | unless $code eq $upper; |
cf25bb62 |
2016 | } |
2017 | close IN; |
2018 | |
2019 | # Now write out the special cases properties in their code point order. |
2020 | # Prepend them to the To/{Upper,Lower,Title}.pl. |
2021 | |
2022 | for my $case (qw(Lower Title Upper)) |
2023 | { |
2024 | my $NormalCase = do "To/$case.pl" || die "$0: $@\n"; |
cf25bb62 |
2025 | |
b08cf34e |
2026 | my @OUT = |
2027 | ( |
2028 | $HEADER, "\n", |
2029 | "# The key UTF-8 _bytes_, the value UTF-8 (speed hack)\n", |
2030 | "%utf8::ToSpec$case =\n(\n", |
2031 | ); |
cf25bb62 |
2032 | |
2033 | for my $prop (sort { $a->[0] <=> $b->[0] } @{$CaseInfo{$case}}) { |
2034 | my ($ix, $code, $to) = @$prop; |
2035 | my $tostr = |
2036 | join "", map { sprintf "\\x{%s}", $_ } split ' ', $to; |
b08cf34e |
2037 | push @OUT, sprintf qq["%s" => "$tostr",\n], join("", map { sprintf "\\x%02X", $_ } unpack("U0C*", pack("U", $ix))); |
5cb851a6 |
2038 | # Remove any single-character mappings for |
2039 | # the same character since we are going for |
2040 | # the special casing rules. |
2041 | $NormalCase =~ s/^$code\t\t\w+\n//m; |
cf25bb62 |
2042 | } |
5beb625e |
2043 | push @OUT, ( |
2044 | ");\n\n", |
2045 | "return <<'END';\n", |
2046 | $NormalCase, |
2047 | "END\n" |
2048 | ); |
d07a55ed |
2049 | WriteIfChanged(["To","$case.pl"], @OUT); |
d2d499f5 |
2050 | } |
d2d499f5 |
2051 | } |
2052 | |
c4051cc5 |
2053 | # |
2054 | # Read in the case foldings. |
2055 | # |
551b6b6f |
2056 | # We will do full case folding, C + F + I (see CaseFolding.txt). |
c4051cc5 |
2057 | # |
551b6b6f |
2058 | sub CaseFolding_txt() |
cf25bb62 |
2059 | { |
551b6b6f |
2060 | if (not open IN, "CaseFolding.txt") { |
2061 | die "$0: CaseFolding.txt: $!\n"; |
cf25bb62 |
2062 | } |
c4051cc5 |
2063 | |
cf25bb62 |
2064 | my $Fold = Table->New(); |
c4051cc5 |
2065 | my %Fold; |
2066 | |
cf25bb62 |
2067 | while (<IN>) { |
254ba52a |
2068 | # Skip status 'S', simple case folding |
c4051cc5 |
2069 | next unless /^([0-9A-Fa-f]+)\s*;\s*([CFI])\s*;\s*([0-9A-Fa-f]+(?: [0-9A-Fa-f]+)*)\s*;/; |
2070 | |
cf25bb62 |
2071 | my ($code, $status, $fold) = (hex($1), $2, $3); |
c4051cc5 |
2072 | |
2073 | if ($status eq 'C') { # Common: one-to-one folding |
254ba52a |
2074 | # No append() since several codes may fold into one. |
cf25bb62 |
2075 | $Fold->RawAppendRange($code, $code, $fold); |
c4051cc5 |
2076 | } else { # F: full, or I: dotted uppercase I -> dotless lowercase I |
cf25bb62 |
2077 | $Fold{$code} = $fold; |
c4051cc5 |
2078 | } |
2079 | } |
cf25bb62 |
2080 | close IN; |
c4051cc5 |
2081 | |
cf25bb62 |
2082 | $Fold->Write("To/Fold.pl"); |
c4051cc5 |
2083 | |
2084 | # |
2085 | # Prepend the special foldings to the common foldings. |
2086 | # |
c4051cc5 |
2087 | my $CommonFold = do "To/Fold.pl" || die "$0: To/Fold.pl: $!\n"; |
5beb625e |
2088 | |
b08cf34e |
2089 | my @OUT = |
2090 | ( |
2091 | $HEADER, "\n", |
2092 | "# The ke UTF-8 _bytes_, the value UTF-8 (speed hack)\n", |
2093 | "%utf8::ToSpecFold =\n(\n", |
2094 | ); |
cf25bb62 |
2095 | for my $code (sort { $a <=> $b } keys %Fold) { |
2096 | my $foldstr = |
2097 | join "", map { sprintf "\\x{%s}", $_ } split ' ', $Fold{$code}; |
b08cf34e |
2098 | push @OUT, sprintf qq["%s" => "$foldstr",\n], join("", map { sprintf "\\x%02X", $_ } unpack("U0C*", pack("U", $code))); |
c4051cc5 |
2099 | } |
5beb625e |
2100 | push @OUT, ( |
2101 | ");\n\n", |
2102 | "return <<'END';\n", |
2103 | $CommonFold, |
2104 | "END\n", |
2105 | ); |
2106 | |
d07a55ed |
2107 | WriteIfChanged(["To","Fold.pl"], @OUT); |
c4051cc5 |
2108 | } |
2109 | |
cf25bb62 |
2110 | ## Do it.... |
2111 | |
12ac2576 |
2112 | Build_Aliases(); |
44da8cae |
2113 | UnicodeData_Txt(); |
cf25bb62 |
2114 | PropList_txt(); |
2115 | |
2116 | Scripts_txt(); |
2117 | Blocks_txt(); |
2118 | |
5beb625e |
2119 | WriteAllMappings(); |
2120 | |
551b6b6f |
2121 | LineBreak_Txt(); |
2122 | ArabicShaping_txt(); |
12ac2576 |
2123 | EastAsianWidth_txt(); |
2124 | HangulSyllableType_txt(); |
cf25bb62 |
2125 | Jamo_txt(); |
551b6b6f |
2126 | SpecialCasing_txt(); |
2127 | CaseFolding_txt(); |
cf25bb62 |
2128 | |
97050450 |
2129 | if ( $FileList and $MakeList ) { |
2130 | |
2131 | print "Updating '$FileList'\n" |
2132 | if ($Verbose); |
2133 | |
2134 | open my $ofh,">",$FileList |
2135 | or die "Can't write to '$FileList':$!"; |
2136 | print $ofh <<"EOFHEADER"; |
2137 | # |
2138 | # mktables.lst -- File list for mktables. |
2139 | # |
2140 | # Autogenerated on @{[scalar localtime]} |
2141 | # |
2142 | # - First section is input files |
2143 | # (mktables itself is automatically included) |
2144 | # - Section seperator is /^=+\$/ |
2145 | # - Second section is a list of output files. |
2146 | # - Lines matching /^\\s*#/ are treated as comments |
2147 | # which along with blank lines are ignored. |
2148 | # |
2149 | |
2150 | # Input files: |
2151 | |
2152 | EOFHEADER |
2153 | my @input=("version",glob('*.txt')); |
2154 | print $ofh "$_\n" for |
2155 | @input, |
2156 | "\n=================================\n", |
2157 | "# Output files:\n", |
2158 | # special files |
2159 | "Properties"; |
2160 | |
2161 | |
2162 | require File::Find; |
2163 | my $count=0; |
2164 | File::Find::find({ |
2165 | no_chdir=>1, |
2166 | wanted=>sub { |
2167 | if (/\.pl$/) { |
2168 | s!^\./!!; |
2169 | print $ofh "$_\n"; |
2170 | $count++; |
2171 | } |
2172 | }, |
2173 | },"."); |
2174 | |
2175 | print $ofh "\n# ",scalar(@input)," input files\n", |
2176 | "# ",scalar($count+1)," output files\n\n", |
2177 | "# End list\n"; |
2178 | close $ofh |
2179 | or warn "Failed to close $ofh: $!"; |
2180 | |
2181 | print "Filelist has ",scalar(@input)," input files and ", |
2182 | scalar($count+1)," output files\n" |
2183 | if $Verbose; |
2184 | } |
2185 | print "All done\n" if $Verbose; |
5beb625e |
2186 | exit(0); |
cf25bb62 |
2187 | |
5beb625e |
2188 | ## TRAILING CODE IS USED BY MakePropTestScript() |
2189 | __DATA__ |
2190 | use strict; |
2191 | use warnings; |
2192 | |
2193 | my $Tests = 0; |
2194 | my $Fails = 0; |
cf25bb62 |
2195 | |
5beb625e |
2196 | sub Expect($$$) |
2197 | { |
2198 | my $Expect = shift; |
2199 | my $String = shift; |
2200 | my $Regex = shift; |
2201 | my $Line = (caller)[2]; |
2202 | |
2203 | $Tests++; |
2204 | my $RegObj; |
2205 | my $result = eval { |
2206 | $RegObj = qr/$Regex/; |
2207 | $String =~ $RegObj ? 1 : 0 |
2208 | }; |
2209 | |
2210 | if (not defined $result) { |
2211 | print "couldn't compile /$Regex/ on $0 line $Line: $@\n"; |
2212 | $Fails++; |
2213 | } elsif ($result ^ $Expect) { |
2214 | print "bad result (expected $Expect) on $0 line $Line: $@\n"; |
2215 | $Fails++; |
2216 | } |
2217 | } |
d73e5302 |
2218 | |
5beb625e |
2219 | sub Error($) |
2220 | { |
2221 | my $Regex = shift; |
2222 | $Tests++; |
2223 | if (eval { 'x' =~ qr/$Regex/; 1 }) { |
2224 | $Fails++; |
2225 | my $Line = (caller)[2]; |
2226 | print "expected error for /$Regex/ on $0 line $Line: $@\n"; |
2227 | } |
2228 | } |
2229 | |
2230 | sub Finished() |
2231 | { |
2232 | if ($Fails == 0) { |
2233 | print "All $Tests tests passed.\n"; |
2234 | exit(0); |
2235 | } else { |
2236 | print "$Tests tests, $Fails failed!\n"; |
2237 | exit(-1); |
2238 | } |
2239 | } |