xref: /onnv-gate/usr/src/cmd/perl/5.8.4/distrib/lib/unicore/mktables (revision 0:68f95e015346)
1#!/usr/bin/perl -w
2use strict;
3use Carp;
4
5die "$0: Please run me as ./mktables to avoid unnecessary differences\n"
6    unless $0 eq "./mktables";
7
8##
9## mktables -- create the runtime Perl Unicode files (lib/unicore/**/*.pl)
10## from the Unicode database files (lib/unicore/*.txt).
11##
12
13mkdir("lib", 0755);
14mkdir("To",  0755);
15
16##
17## Process any args.
18##
19my $Verbose        = 0;
20my $MakeTestScript = 0;
21
22while (@ARGV)
23{
24    my $arg = shift @ARGV;
25    if ($arg eq '-v') {
26        $Verbose = 1;
27    } elsif ($arg eq '-q') {
28        $Verbose = 0;
29    } elsif ($arg eq '-maketest') {
30        $MakeTestScript = 1;
31    } else {
32        die "usage: $0 [-v|-q] [-maketest]";
33    }
34}
35
36my $LastUnicodeCodepoint = 0x10FFFF; # As of Unicode 3.1.1.
37
38my $HEADER=<<"EOF";
39# !!!!!!!   DO NOT EDIT THIS FILE   !!!!!!!
40# This file is built by $0 from e.g. UnicodeData.txt.
41# Any changes made here will be lost!
42
43EOF
44
45
46##
47## Given a filename and a reference to an array of lines,
48## write the lines to the file only if the contents have not changed.
49##
50sub WriteIfChanged($\@)
51{
52    my $file  = shift;
53    my $lines = shift;
54
55    my $TextToWrite = join '', @$lines;
56    if (open IN, $file) {
57        local($/) = undef;
58        my $PreviousText = <IN>;
59        close IN;
60        if ($PreviousText eq $TextToWrite) {
61            print "$file unchanged.\n" if $Verbose;
62            return;
63        }
64    }
65    if (not open OUT, ">$file") {
66        die "$0: can't open $file for output: $!\n";
67    }
68    print "$file written.\n" if $Verbose;
69
70    print OUT $TextToWrite;
71    close OUT;
72}
73
74##
75## The main datastructure (a "Table") represents a set of code points that
76## are part of a particular quality (that are part of \pL, \p{InGreek},
77## etc.). They are kept as ranges of code points (starting and ending of
78## each range).
79##
80## For example, a range ASCII LETTERS would be represented as:
81##   [ [ 0x41 => 0x5A, 'UPPER' ],
82##     [ 0x61 => 0x7A, 'LOWER, ] ]
83##
84sub RANGE_START() { 0 } ## index into range element
85sub RANGE_END()   { 1 } ## index into range element
86sub RANGE_NAME()  { 2 } ## index into range element
87
88## Conceptually, these should really be folded into the 'Table' objects
89my %TableInfo;
90my %TableDesc;
91my %FuzzyNames;
92my %AliasInfo;
93my %CanonicalToOrig;
94
95##
96## Turn something like
97##    OLD-ITALIC
98## into
99##    OldItalic
100##
101sub CanonicalName($)
102{
103    my $orig = shift;
104    my $name = lc $orig;
105    $name =~ s/(?<![a-z])(\w)/\u$1/g;
106    $name =~ s/[-_\s]+//g;
107
108    $CanonicalToOrig{$name} = $orig if not $CanonicalToOrig{$name};
109    return $name;
110}
111
112##
113## Associates a property ("Greek", "Lu", "Assigned",...) with a Table.
114##
115## Called like:
116##       New_Prop(In => 'Greek', $Table, Desc => 'Greek Block', Fuzzy => 1);
117##
118## Normally, these parameters are set when the Table is created (when the
119## Table->New constructor is called), but there are times when it needs to
120## be done after-the-fact...)
121##
122sub New_Prop($$$@)
123{
124    my $Type = shift; ## "Is" or "In";
125    my $Name = shift;
126    my $Table = shift;
127
128    ## remaining args are optional key/val
129    my %Args = @_;
130
131    my $Fuzzy = delete $Args{Fuzzy};
132    my $Desc  = delete $Args{Desc}; # description
133
134    $Name = CanonicalName($Name) if $Fuzzy;
135
136    ## sanity check a few args
137    if (%Args or ($Type ne 'Is' and $Type ne 'In') or not ref $Table) {
138        confess "$0: bad args to New_Prop"
139    }
140
141    if (not $TableInfo{$Type}->{$Name})
142    {
143        $TableInfo{$Type}->{$Name} = $Table;
144        $TableDesc{$Type}->{$Name} = $Desc;
145        if ($Fuzzy) {
146            $FuzzyNames{$Type}->{$Name} = $Name;
147        }
148    }
149}
150
151
152##
153## Creates a new Table object.
154##
155## Args are key/value pairs:
156##    In => Name         -- Name of "In" property to be associated with
157##    Is => Name         -- Name of "Is" property to be associated with
158##    Fuzzy => Boolean   -- True if name can be accessed "fuzzily"
159##    Desc  => String    -- Description of the property
160##
161## No args are required.
162##
163sub Table::New
164{
165    my $class = shift;
166    my %Args = @_;
167
168    my $Table = bless [], $class;
169
170    my $Fuzzy = delete $Args{Fuzzy};
171    my $Desc  = delete $Args{Desc};
172
173    for my $Type ('Is', 'In')
174    {
175        if (my $Name = delete $Args{$Type}) {
176            New_Prop($Type => $Name, $Table, Desc => $Desc, Fuzzy => $Fuzzy);
177        }
178    }
179
180    ## shouldn't have any left over
181    if (%Args) {
182        confess "$0: bad args to Table->New"
183    }
184
185    return $Table;
186}
187
188##
189## Returns true if the Table has no code points
190##
191sub Table::IsEmpty
192{
193    my $Table = shift; #self
194    return not @$Table;
195}
196
197##
198## Returns true if the Table has code points
199##
200sub Table::NotEmpty
201{
202    my $Table = shift; #self
203    return @$Table;
204}
205
206##
207## Returns the maximum code point currently in the table.
208##
209sub Table::Max
210{
211    my $Table = shift; #self
212    confess "oops" if $Table->IsEmpty; ## must have code points to have a max
213    return $Table->[-1]->[RANGE_END];
214}
215
216##
217## Replaces the codepoints in the Table with those in the Table given
218## as an arg. (NOTE: this is not a "deep copy").
219##
220sub Table::Replace($$)
221{
222    my $Table = shift; #self
223    my $New   = shift;
224
225    @$Table = @$New;
226}
227
228##
229## Given a new code point, make the last range of the Table extend to
230## include the new (and all intervening) code points.
231##
232sub Table::Extend
233{
234    my $Table = shift; #self
235    my $codepoint = shift;
236
237    my $PrevMax = $Table->Max;
238
239    confess "oops ($codepoint <= $PrevMax)" if $codepoint <= $PrevMax;
240
241    $Table->[-1]->[RANGE_END] = $codepoint;
242}
243
244##
245## Given a code point range start and end (and optional name), blindly
246## append them to the list of ranges for the Table.
247##
248## NOTE: Code points must be added in strictly ascending numeric order.
249##
250sub Table::RawAppendRange
251{
252    my $Table = shift; #self
253    my $start = shift;
254    my $end   = shift;
255    my $name  = shift;
256    $name = "" if not defined $name; ## warning: $name can be "0"
257
258    push @$Table, [ $start,    # RANGE_START
259                    $end,      # RANGE_END
260                    $name   ]; # RANGE_NAME
261}
262
263##
264## Given a code point (and optional name), add it to the Table.
265##
266## NOTE: Code points must be added in strictly ascending numeric order.
267##
268sub Table::Append
269{
270    my $Table     = shift; #self
271    my $codepoint = shift;
272    my $name      = shift;
273    $name = "" if not defined $name; ## warning: $name can be "0"
274
275    ##
276    ## If we've already got a range working, and this code point is the next
277    ## one in line, and if the name is the same, just extend the current range.
278    ##
279    if ($Table->NotEmpty
280        and
281        $Table->Max == $codepoint - 1
282        and
283        $Table->[-1]->[RANGE_NAME] eq $name)
284    {
285        $Table->Extend($codepoint);
286    }
287    else
288    {
289        $Table->RawAppendRange($codepoint, $codepoint, $name);
290    }
291}
292
293##
294## Given a code point range starting value and ending value (and name),
295## Add the range to teh Table.
296##
297## NOTE: Code points must be added in strictly ascending numeric order.
298##
299sub Table::AppendRange
300{
301    my $Table = shift; #self
302    my $start = shift;
303    my $end   = shift;
304    my $name  = shift;
305    $name = "" if not defined $name; ## warning: $name can be "0"
306
307    $Table->Append($start, $name);
308    $Table->Extend($end) if $end > $start;
309}
310
311##
312## Return a new Table that represents all code points not in the Table.
313##
314sub Table::Invert
315{
316    my $Table = shift; #self
317
318    my $New = Table->New();
319    my $max = -1;
320    for my $range (@$Table)
321    {
322        my $start = $range->[RANGE_START];
323        my $end   = $range->[RANGE_END];
324        if ($start-1 >= $max+1) {
325            $New->AppendRange($max+1, $start-1, "");
326        }
327        $max = $end;
328    }
329    if ($max+1 < $LastUnicodeCodepoint) {
330        $New->AppendRange($max+1, $LastUnicodeCodepoint);
331    }
332    return $New;
333}
334
335##
336## Merges any number of other tables with $self, returning the new table.
337## (existing tables are not modified)
338##
339##
340## Args may be Tables, or individual code points (as integers).
341##
342## Can be called as either a constructor or a method.
343##
344sub Table::Merge
345{
346    shift(@_) if not ref $_[0]; ## if called as a constructor, lose the class
347    my @Tables = @_;
348
349    ## Accumulate all records from all tables
350    my @Records;
351    for my $Arg (@Tables)
352    {
353        if (ref $Arg) {
354            ## arg is a table -- get its ranges
355            push @Records, @$Arg;
356        } else {
357            ## arg is a codepoint, make a range
358            push @Records, [ $Arg, $Arg ]
359        }
360    }
361
362    ## sort by range start, with longer ranges coming first.
363    my ($first, @Rest) = sort {
364        ($a->[RANGE_START] <=> $b->[RANGE_START])
365          or
366        ($b->[RANGE_END]   <=> $b->[RANGE_END])
367    } @Records;
368
369    my $New = Table->New();
370
371    ## Ensuring the first range is there makes the subsequent loop easier
372    $New->AppendRange($first->[RANGE_START],
373                      $first->[RANGE_END]);
374
375    ## Fold in records so long as they add new information.
376    for my $set (@Rest)
377    {
378        my $start = $set->[RANGE_START];
379        my $end   = $set->[RANGE_END];
380        if ($start > $New->Max) {
381            $New->AppendRange($start, $end);
382        } elsif ($end > $New->Max) {
383            $New->Extend($end);
384        }
385    }
386
387    return $New;
388}
389
390##
391## Given a filename, write a representation of the Table to a file.
392## May have an optional comment as a 2nd arg.
393##
394sub Table::Write
395{
396    my $Table    = shift; #self
397    my $filename = shift;
398    my $comment  = shift;
399
400    my @OUT = $HEADER;
401    if (defined $comment) {
402        $comment =~ s/\s+\Z//;
403        $comment =~ s/^/# /gm;
404        push @OUT, "#\n$comment\n#\n";
405    }
406    push @OUT, "return <<'END';\n";
407
408    for my $set (@$Table)
409    {
410        my $start = $set->[RANGE_START];
411        my $end   = $set->[RANGE_END];
412        my $name  = $set->[RANGE_NAME];
413
414        if ($start == $end) {
415            push @OUT, sprintf "%04X\t\t%s\n", $start, $name;
416        } else {
417            push @OUT, sprintf "%04X\t%04X\t%s\n", $start, $end, $name;
418        }
419    }
420
421    push @OUT, "END\n";
422
423    WriteIfChanged($filename, @OUT);
424}
425
426## This used only for making the test script.
427## helper function
428sub IsUsable($)
429{
430    my $code = shift;
431    return 0 if $code <= 0x0000;                       ## don't use null
432    return 0 if $code >= $LastUnicodeCodepoint;        ## keep in range
433    return 0 if ($code >= 0xD800 and $code <= 0xDFFF); ## no surrogates
434    return 0 if ($code >= 0xFDD0 and $code <= 0xFDEF); ## utf8.c says no good
435    return 0 if (($code & 0xFFFF) == 0xFFFE);          ## utf8.c says no good
436    return 0 if (($code & 0xFFFF) == 0xFFFF);          ## utf8.c says no good
437    return 1;
438}
439
440## Return a code point that's part of the table.
441## Returns nothing if the table is empty (or covers only surrogates).
442## This used only for making the test script.
443sub Table::ValidCode
444{
445    my $Table = shift; #self
446    for my $set (@$Table) {
447        return $set->[RANGE_END] if IsUsable($set->[RANGE_END]);
448    }
449    return ();
450}
451
452## Return a code point that's not part of the table
453## Returns nothing if the table covers all code points.
454## This used only for making the test script.
455sub Table::InvalidCode
456{
457    my $Table = shift; #self
458
459    return 0x1234 if $Table->IsEmpty();
460
461    for my $set (@$Table)
462    {
463        if (IsUsable($set->[RANGE_END] + 1))
464        {
465            return $set->[RANGE_END] + 1;
466        }
467
468        if (IsUsable($set->[RANGE_START] - 1))
469        {
470            return $set->[RANGE_START] - 1;
471        }
472    }
473    return ();
474}
475
476###########################################################################
477###########################################################################
478###########################################################################
479
480
481##
482## Called like:
483##     New_Alias(Is => 'All', SameAs => 'Any', Fuzzy => 1);
484##
485## The args must be in that order, although the Fuzzy pair may be omitted.
486##
487## This creates 'IsAll' as an alias for 'IsAny'
488##
489sub New_Alias($$$@)
490{
491    my $Type   = shift; ## "Is" or "In"
492    my $Alias  = shift;
493    my $SameAs = shift; # expecting "SameAs" -- just ignored
494    my $Name   = shift;
495
496    ## remaining args are optional key/val
497    my %Args = @_;
498
499    my $Fuzzy = delete $Args{Fuzzy};
500
501    ## sanity check a few args
502    if (%Args or ($Type ne 'Is' and $Type ne 'In') or $SameAs ne 'SameAs') {
503        confess "$0: bad args to New_Alias"
504    }
505
506    $Alias = CanonicalName($Alias) if $Fuzzy;
507
508    if (not $TableInfo{$Type}->{$Name})
509    {
510        my $CName = CanonicalName($Name);
511        if ($TableInfo{$Type}->{$CName}) {
512            confess "$0: Use canonical form '$CName' instead of '$Name' for alias.";
513        } else {
514            confess "$0: don't have orignial $Type => $Name to make alias";
515        }
516    }
517    if ($TableInfo{$Alias}) {
518        confess "$0: already have original $Type => $Alias; can't make alias";
519    }
520    $AliasInfo{$Type}->{$Name} = $Alias;
521    if ($Fuzzy) {
522        $FuzzyNames{$Type}->{$Alias} = $Name;
523    }
524
525}
526
527
528## All assigned code points
529my $Assigned = Table->New(Is    => 'Assigned',
530                          Desc  => "All assigned code points",
531                          Fuzzy => 0);
532
533my $Name     = Table->New(); ## all characters, individually by name
534my $General  = Table->New(); ## all characters, grouped by category
535my %General;
536my %Cat;
537
538##
539## Process UnicodeData.txt (Categories, etc.)
540##
541sub UnicodeData_Txt()
542{
543    my $Bidi     = Table->New();
544    my $Deco     = Table->New();
545    my $Comb     = Table->New();
546    my $Number   = Table->New();
547    my $Mirrored = Table->New(Is    => 'Mirrored',
548                              Desc  => "Mirrored in bidirectional text",
549                              Fuzzy => 0);
550
551    my %DC;
552    my %Bidi;
553    my %Deco;
554    $Deco{Canon}   = Table->New(Is    => 'Canon',
555                                Desc  => 'Decomposes to multiple characters',
556                                Fuzzy => 0);
557    $Deco{Compat}  = Table->New(Is    => 'Compat',
558                                Desc  => 'Compatible with a more-basic character',
559                                Fuzzy => 0);
560
561    ## Initialize Perl-generated categories
562    ## (Categories from UnicodeData.txt are auto-initialized in gencat)
563    $Cat{Alnum}  =
564	Table->New(Is => 'Alnum',  Desc => "[[:Alnum:]]",  Fuzzy => 0);
565    $Cat{Alpha}  =
566	Table->New(Is => 'Alpha',  Desc => "[[:Alpha:]]",  Fuzzy => 0);
567    $Cat{ASCII}  =
568	Table->New(Is => 'ASCII',  Desc => "[[:ASCII:]]",  Fuzzy => 0);
569    $Cat{Blank}  =
570	Table->New(Is => 'Blank',  Desc => "[[:Blank:]]",  Fuzzy => 0);
571    $Cat{Cntrl}  =
572	Table->New(Is => 'Cntrl',  Desc => "[[:Cntrl:]]",  Fuzzy => 0);
573    $Cat{Digit}  =
574	Table->New(Is => 'Digit',  Desc => "[[:Digit:]]",  Fuzzy => 0);
575    $Cat{Graph}  =
576	Table->New(Is => 'Graph',  Desc => "[[:Graph:]]",  Fuzzy => 0);
577    $Cat{Lower}  =
578	Table->New(Is => 'Lower',  Desc => "[[:Lower:]]",  Fuzzy => 0);
579    $Cat{Print}  =
580	Table->New(Is => 'Print',  Desc => "[[:Print:]]",  Fuzzy => 0);
581    $Cat{Punct}  =
582	Table->New(Is => 'Punct',  Desc => "[[:Punct:]]",  Fuzzy => 0);
583    $Cat{Space}  =
584	Table->New(Is => 'Space',  Desc => "[[:Space:]]",  Fuzzy => 0);
585    $Cat{Title}  =
586	Table->New(Is => 'Title',  Desc => "[[:Title:]]",  Fuzzy => 0);
587    $Cat{Upper}  =
588	Table->New(Is => 'Upper',  Desc => "[[:Upper:]]",  Fuzzy => 0);
589    $Cat{XDigit} =
590	Table->New(Is => 'XDigit', Desc => "[[:XDigit:]]", Fuzzy => 0);
591    $Cat{Word}   =
592	Table->New(Is => 'Word',   Desc => "[[:Word:]]",   Fuzzy => 0);
593    $Cat{SpacePerl} =
594	Table->New(Is => 'SpacePerl', Desc => '\s', Fuzzy => 0);
595
596    my %To;
597    $To{Upper} = Table->New();
598    $To{Lower} = Table->New();
599    $To{Title} = Table->New();
600    $To{Digit} = Table->New();
601
602    sub gencat($$$$)
603    {
604        my ($name, ## Name ("LATIN CAPITAL LETTER A")
605            $cat,  ## Category ("Lu", "Zp", "Nd", etc.)
606            $code, ## Code point (as an integer)
607            $op) = @_;
608
609        my $MajorCat = substr($cat, 0, 1); ## L, M, Z, S, etc
610
611        $Assigned->$op($code);
612        $Name->$op($code, $name);
613        $General->$op($code, $cat);
614
615        ## add to the sub category (e.g. "Lu", "Nd", "Cf", ..)
616        $Cat{$cat}      ||= Table->New(Is   => $cat,
617                                       Desc => "General Category '$cat'",
618                                       Fuzzy => 0);
619        $Cat{$cat}->$op($code);
620
621        ## add to the major category (e.g. "L", "N", "C", ...)
622        $Cat{$MajorCat} ||= Table->New(Is => $MajorCat,
623                                       Desc => "Major Category '$MajorCat'",
624                                       Fuzzy => 0);
625        $Cat{$MajorCat}->$op($code);
626
627        ($General{$name} ||= Table->New)->$op($code, $name);
628
629        # 005F: SPACING UNDERSCORE
630        $Cat{Word}->$op($code)  if $cat =~ /^[LMN]|Pc/;
631        $Cat{Alnum}->$op($code) if $cat =~ /^[LM]|Nd/;
632        $Cat{Alpha}->$op($code) if $cat =~ /^[LM]/;
633
634	my $isspace =
635	    ($cat =~ /Zs|Zl|Zp/ &&
636	     $code != 0x200B) # 200B is ZWSP which is for line break control
637	     # and therefore it is not part of "space" even while it is "Zs".
638                                || $code == 0x0009  # 0009: HORIZONTAL TAB
639                                || $code == 0x000A  # 000A: LINE FEED
640                                || $code == 0x000B  # 000B: VERTICAL TAB
641                                || $code == 0x000C  # 000C: FORM FEED
642                                || $code == 0x000D  # 000D: CARRIAGE RETURN
643                                || $code == 0x0085  # 0085: NEL
644
645	    ;
646
647        $Cat{Space}->$op($code) if $isspace;
648
649        $Cat{SpacePerl}->$op($code) if $isspace
650	                               && $code != 0x000B; # Backward compat.
651
652        $Cat{Blank}->$op($code) if $isspace
653                                && !($code == 0x000A ||
654				     $code == 0x000B ||
655				     $code == 0x000C ||
656				     $code == 0x000D ||
657				     $code == 0x0085 ||
658				     $cat =~ /^Z[lp]/);
659
660        $Cat{Digit}->$op($code) if $cat eq "Nd";
661        $Cat{Upper}->$op($code) if $cat eq "Lu";
662        $Cat{Lower}->$op($code) if $cat eq "Ll";
663        $Cat{Title}->$op($code) if $cat eq "Lt";
664        $Cat{ASCII}->$op($code) if $code <= 0x007F;
665        $Cat{Cntrl}->$op($code) if $cat =~ /^C/;
666	my $isgraph = !$isspace && $cat !~ /Cc|Cs|Cn/;
667        $Cat{Graph}->$op($code) if $isgraph;
668        $Cat{Print}->$op($code) if $isgraph || $isspace;
669        $Cat{Punct}->$op($code) if $cat =~ /^P/;
670
671        $Cat{XDigit}->$op($code) if ($code >= 0x30 && $code <= 0x39)  ## 0..9
672                                 || ($code >= 0x41 && $code <= 0x46)  ## A..F
673                                 || ($code >= 0x61 && $code <= 0x66); ## a..f
674    }
675
676    ## open ane read file.....
677    if (not open IN, "UnicodeData.txt") {
678        die "$0: UnicodeData.txt: $!\n";
679    }
680
681    ##
682    ## For building \p{_CombAbove} and \p{_CanonDCIJ}
683    ##
684    my %_Above_HexCodes; ## Hexcodes for chars with $comb == 230 ("ABOVE")
685
686    my %CodeToDeco;      ## Maps code to decomp. list for chars with first
687                         ## decomp. char an "i" or "j" (for \p{_CanonDCIJ})
688
689    ## This is filled in as we go....
690    my $CombAbove = Table->New(Is   => '_CombAbove',
691                               Desc  => '(for internal casefolding use)',
692                               Fuzzy => 0);
693
694    while (<IN>)
695    {
696        next unless /^[0-9A-Fa-f]+;/;
697        s/\s+$//;
698
699        my ($hexcode,   ## code point in hex (e.g. "0041")
700            $name,      ## character name (e.g. "LATIN CAPITAL LETTER A")
701            $cat,       ## category (e.g. "Lu")
702            $comb,      ## Canonical combining class (e.t. "230")
703            $bidi,      ## directional category (e.g. "L")
704            $deco,      ## decomposition mapping
705            $decimal,   ## decimal digit value
706            $digit,     ## digit value
707            $number,    ## numeric value
708            $mirrored,  ## mirrored
709            $unicode10, ## name in Unicode 1.0
710            $comment,   ## comment field
711            $upper,     ## uppercase mapping
712            $lower,     ## lowercase mapping
713            $title,     ## titlecase mapping
714              ) = split(/\s*;\s*/);
715
716	# Note that in Unicode 3.2 there will be names like
717	# LINE FEED (LF), which probably means that \N{} needs
718	# to cope also with LINE FEED and LF.
719	$name = $unicode10 if $name eq '<control>' && $unicode10 ne '';
720
721        my $code = hex($hexcode);
722
723        if ($comb and $comb == 230) {
724            $CombAbove->Append($code);
725            $_Above_HexCodes{$hexcode} = 1;
726        }
727
728        ## Used in building \p{_CanonDCIJ}
729        if ($deco and $deco =~ m/^006[9A]\b/) {
730            $CodeToDeco{$code} = $deco;
731        }
732
733        ##
734        ## There are a few pairs of lines like:
735        ##   AC00;<Hangul Syllable, First>;Lo;0;L;;;;;N;;;;;
736        ##   D7A3;<Hangul Syllable, Last>;Lo;0;L;;;;;N;;;;;
737        ## that define ranges.
738        ##
739        if ($name =~ /^<(.+), (First|Last)>$/)
740        {
741            $name = $1;
742            gencat($name, $cat, $code, $2 eq 'First' ? 'Append' : 'Extend');
743            #New_Prop(In => $name, $General{$name}, Fuzzy => 1);
744        }
745        else
746        {
747            ## normal (single-character) lines
748            gencat($name, $cat, $code, 'Append');
749
750            # No Append() here since since several codes may map into one.
751            $To{Upper}->RawAppendRange($code, $code, $upper) if $upper;
752            $To{Lower}->RawAppendRange($code, $code, $lower) if $lower;
753            $To{Title}->RawAppendRange($code, $code, $title) if $title;
754            $To{Digit}->Append($code, $decimal) if length $decimal;
755
756            $Bidi->Append($code, $bidi);
757            $Comb->Append($code, $comb) if $comb;
758            $Number->Append($code, $number) if length $number;
759
760            $Mirrored->Append($code) if $mirrored eq "Y";
761
762            $Bidi{$bidi} ||= Table->New(Is    => "Bidi$bidi",
763                                        Desc  => "Bi-directional category '$bidi'",
764                                        Fuzzy => 0);
765            $Bidi{$bidi}->Append($code);
766
767            if ($deco)
768            {
769                $Deco->Append($code, $deco);
770                if ($deco =~/^<(\w+)>/)
771                {
772                    $Deco{Compat}->Append($code);
773
774                    $DC{$1} ||= Table->New(Is => "DC$1",
775                                           Desc  => "Compatible with '$1'",
776                                           Fuzzy => 0);
777                    $DC{$1}->Append($code);
778                }
779                else
780                {
781                    $Deco{Canon}->Append($code);
782                }
783            }
784        }
785    }
786    close IN;
787
788    ##
789    ## Tidy up a few special cases....
790    ##
791
792    $Cat{Cn} = $Assigned->Invert; ## Cn is everything that doesn't exist
793    New_Prop(Is => 'Cn',
794             $Cat{Cn},
795             Desc => "General Category 'Cn' [not functional in Perl]",
796             Fuzzy => 0);
797
798    ## Unassigned is the same as 'Cn'
799    New_Alias(Is => 'Unassigned', SameAs => 'Cn', Fuzzy => 0);
800
801    $Cat{C}->Replace($Cat{C}->Merge($Cat{Cn}));  ## Now merge in Cn into C
802
803
804    # L& is Ll, Lu, and Lt.
805    New_Prop(Is => 'L&',
806             Table->Merge(@Cat{qw[Ll Lu Lt]}),
807             Desc  => '[\p{Ll}\p{Lu}\p{Lt}]',
808             Fuzzy => 0);
809
810    ## Any and All are all code points.
811    my $Any = Table->New(Is    => 'Any',
812                         Desc  => sprintf("[\\x{0000}-\\x{%X}]",
813                                          $LastUnicodeCodepoint),
814                         Fuzzy => 0);
815    $Any->RawAppendRange(0, $LastUnicodeCodepoint);
816
817    New_Alias(Is => 'All', SameAs => 'Any', Fuzzy => 0);
818
819    ##
820    ## Build special properties for Perl's internal case-folding needs:
821    ##    \p{_CaseIgnorable}
822    ##    \p{_CanonDCIJ}
823    ##    \p{_CombAbove}
824    ## _CombAbove was built above. Others are built here....
825    ##
826
827    ## \p{_CaseIgnorable} is [\p{Mn}\0x00AD\x2010]
828    New_Prop(Is => '_CaseIgnorable',
829             Table->Merge($Cat{Mn},
830                          0x00AD,    #SOFT HYPHEN
831                          0x2010),   #HYPHEN
832             Desc  => '(for internal casefolding use)',
833             Fuzzy => 0);
834
835
836    ## \p{_CanonDCIJ} is fairly complex...
837    my $CanonCDIJ = Table->New(Is    => '_CanonDCIJ',
838                               Desc  => '(for internal casefolding use)',
839                               Fuzzy => 0);
840    ## It contains the ASCII 'i' and 'j'....
841    $CanonCDIJ->Append(0x0069); # ASCII ord("i")
842    $CanonCDIJ->Append(0x006A); # ASCII ord("j")
843    ## ...and any character with a decomposition that starts with either of
844    ## those code points, but only if the decomposition does not have any
845    ## combining character with the "ABOVE" canonical combining class.
846    for my $code (sort { $a <=> $b} keys %CodeToDeco)
847    {
848        ## Need to ensure that all decomposition characters do not have
849        ## a %HexCodeToComb in %AboveCombClasses.
850        my $want = 1;
851        for my $deco_hexcode (split / /, $CodeToDeco{$code})
852        {
853            if (exists $_Above_HexCodes{$deco_hexcode}) {
854                ## one of the decmposition chars has an ABOVE combination
855                ## class, so we're not interested in this one
856                $want = 0;
857                last;
858            }
859        }
860        if ($want) {
861            $CanonCDIJ->Append($code);
862        }
863    }
864
865
866
867    ##
868    ## Now dump the files.
869    ##
870    $Name->Write("Name.pl");
871    $Bidi->Write("Bidirectional.pl");
872    $Comb->Write("CombiningClass.pl");
873    $Deco->Write("Decomposition.pl");
874    $Number->Write("Number.pl");
875    $General->Write("Category.pl");
876
877    for my $to (sort keys %To) {
878        $To{$to}->Write("To/$to.pl");
879    }
880}
881
882##
883## Process LineBreak.txt
884##
885sub LineBreak_Txt()
886{
887    if (not open IN, "LineBreak.txt") {
888        die "$0: LineBreak.txt: $!\n";
889    }
890
891    my $Lbrk = Table->New();
892    my %Lbrk;
893
894    while (<IN>)
895    {
896        next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(\w+)/;
897
898	my ($first, $last, $lbrk) = (hex($1), hex($2||""), $3);
899
900	$Lbrk->Append($first, $lbrk);
901
902        $Lbrk{$lbrk} ||= Table->New(Is    => "Lbrk$lbrk",
903                                    Desc  => "Linebreak category '$lbrk'",
904                                    Fuzzy => 0);
905        $Lbrk{$lbrk}->Append($first);
906
907	if ($last) {
908	    $Lbrk->Extend($last);
909	    $Lbrk{$lbrk}->Extend($last);
910	}
911    }
912    close IN;
913
914    $Lbrk->Write("Lbrk.pl");
915}
916
917##
918## Process ArabicShaping.txt.
919##
920sub ArabicShaping_txt()
921{
922    if (not open IN, "ArabicShaping.txt") {
923        die "$0: ArabicShaping.txt: $!\n";
924    }
925
926    my $ArabLink      = Table->New();
927    my $ArabLinkGroup = Table->New();
928
929    while (<IN>)
930    {
931	next unless /^[0-9A-Fa-f]+;/;
932	s/\s+$//;
933
934	my ($hexcode, $name, $link, $linkgroup) = split(/\s*;\s*/);
935        my $code = hex($hexcode);
936	$ArabLink->Append($code, $link);
937	$ArabLinkGroup->Append($code, $linkgroup);
938    }
939    close IN;
940
941    $ArabLink->Write("ArabLink.pl");
942    $ArabLinkGroup->Write("ArabLnkGrp.pl");
943}
944
945##
946## Process Jamo.txt.
947##
948sub Jamo_txt()
949{
950    if (not open IN, "Jamo.txt") {
951        die "$0: Jamo.txt: $!\n";
952    }
953    my $Short = Table->New();
954
955    while (<IN>)
956    {
957	next unless /^([0-9A-Fa-f]+)\s*;\s*(\w*)/;
958	my ($code, $short) = (hex($1), $2);
959
960	$Short->Append($code, $short);
961    }
962    close IN;
963    $Short->Write("JamoShort.pl");
964}
965
966##
967## Process Scripts.txt.
968##
969sub Scripts_txt()
970{
971    my @ScriptInfo;
972
973    if (not open(IN, "Scripts.txt")) {
974        die "$0: Scripts.txt: $!\n";
975    }
976    while (<IN>) {
977        next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(.+?)\s*\#/;
978
979        # Wait until all the scripts have been read since
980        # they are not listed in numeric order.
981        push @ScriptInfo, [ hex($1), hex($2||""), $3 ];
982    }
983    close IN;
984
985    # Now append the scripts properties in their code point order.
986
987    my %Script;
988    my $Scripts = Table->New();
989
990    for my $script (sort { $a->[0] <=> $b->[0] } @ScriptInfo)
991    {
992        my ($first, $last, $name) = @$script;
993        $Scripts->Append($first, $name);
994
995        $Script{$name} ||= Table->New(Is    => $name,
996                                      Desc  => "Script '$name'",
997                                      Fuzzy => 1);
998        $Script{$name}->Append($first, $name);
999
1000        if ($last) {
1001            $Scripts->Extend($last);
1002            $Script{$name}->Extend($last);
1003        }
1004    }
1005
1006    $Scripts->Write("Scripts.pl");
1007
1008    ## Common is everything not explicitly assigned to a Script
1009    ##
1010    ##    ***shouldn't this be intersected with \p{Assigned}? ******
1011    ##
1012    New_Prop(Is => 'Common',
1013             $Scripts->Invert,
1014             Desc  => 'Pseudo-Script of codepoints not in other Unicode scripts',
1015             Fuzzy => 1);
1016}
1017
1018##
1019## Given a name like "Close Punctuation", return a regex (that when applied
1020## with /i) matches any valid form of that name (e.g. "ClosePunctuation",
1021## "Close-Punctuation", etc.)
1022##
1023## Accept any space, dash, or underbar where in the official name there is
1024## space or a dash (or underbar, but there never is).
1025##
1026##
1027sub NameToRegex($)
1028{
1029    my $Name = shift;
1030    $Name =~ s/[- _]/(?:[-_]|\\s+)?/g;
1031    return $Name;
1032}
1033
1034##
1035## Process Blocks.txt.
1036##
1037sub Blocks_txt()
1038{
1039    my $Blocks = Table->New();
1040    my %Blocks;
1041
1042    if (not open IN, "Blocks.txt") {
1043        die "$0: Blocks.txt: $!\n";
1044    }
1045
1046    while (<IN>)
1047    {
1048        #next if not /Private Use$/;
1049	next if not /^([0-9A-Fa-f]+)\.\.([0-9A-Fa-f]+)\s*;\s*(.+?)\s*$/;
1050
1051	my ($first, $last, $name) = (hex($1), hex($2), $3);
1052
1053	$Blocks->Append($first, $name);
1054
1055        $Blocks{$name} ||= Table->New(In    => $name,
1056                                      Desc  => "Block '$name'",
1057                                      Fuzzy => 1);
1058        $Blocks{$name}->Append($first, $name);
1059
1060	if ($last and $last != $first) {
1061	    $Blocks->Extend($last);
1062	    $Blocks{$name}->Extend($last);
1063	}
1064    }
1065    close IN;
1066
1067    $Blocks->Write("Blocks.pl");
1068}
1069
1070##
1071## Read in the PropList.txt.  It contains extended properties not
1072## listed in the UnicodeData.txt, such as 'Other_Alphabetic':
1073## alphabetic but not of the general category L; many modifiers
1074## belong to this extended property category: while they are not
1075## alphabets, they are alphabetic in nature.
1076##
1077sub PropList_txt()
1078{
1079    my @PropInfo;
1080
1081    if (not open IN, "PropList.txt") {
1082        die "$0: PropList.txt: $!\n";
1083    }
1084
1085    while (<IN>)
1086    {
1087	next unless /^([0-9A-Fa-f]+)(?:\.\.([0-9A-Fa-f]+))?\s*;\s*(.+?)\s*\#/;
1088
1089	# Wait until all the extended properties have been read since
1090	# they are not listed in numeric order.
1091	push @PropInfo, [ hex($1), hex($2||""), $3 ];
1092    }
1093    close IN;
1094
1095    # Now append the extended properties in their code point order.
1096    my $Props = Table->New();
1097    my %Prop;
1098
1099    for my $prop (sort { $a->[0] <=> $b->[0] } @PropInfo)
1100    {
1101        my ($first, $last, $name) = @$prop;
1102        $Props->Append($first, $name);
1103
1104        $Prop{$name} ||= Table->New(Is    => $name,
1105                                    Desc  => "Extended property '$name'",
1106                                    Fuzzy => 1);
1107        $Prop{$name}->Append($first, $name);
1108
1109        if ($last) {
1110            $Props->Extend($last);
1111            $Prop{$name}->Extend($last);
1112        }
1113    }
1114
1115    # Alphabetic is L and Other_Alphabetic.
1116    New_Prop(Is    => 'Alphabetic',
1117             Table->Merge($Cat{L}, $Prop{Other_Alphabetic}),
1118             Desc  => '[\p{L}\p{OtherAlphabetic}]', # use canonical names here
1119             Fuzzy => 1);
1120
1121    # Lowercase is Ll and Other_Lowercase.
1122    New_Prop(Is    => 'Lowercase',
1123             Table->Merge($Cat{Ll}, $Prop{Other_Lowercase}),
1124             Desc  => '[\p{Ll}\p{OtherLowercase}]', # use canonical names here
1125             Fuzzy => 1);
1126
1127    # Uppercase is Lu and Other_Uppercase.
1128    New_Prop(Is => 'Uppercase',
1129             Table->Merge($Cat{Lu}, $Prop{Other_Uppercase}),
1130             Desc  => '[\p{Lu}\p{Other_Uppercase}]', # use canonical names here
1131             Fuzzy => 1);
1132
1133    # Math is Sm and Other_Math.
1134    New_Prop(Is => 'Math',
1135             Table->Merge($Cat{Sm}, $Prop{Other_Math}),
1136             Desc  => '[\p{Sm}\p{OtherMath}]', # use canonical names here
1137             Fuzzy => 1);
1138
1139    # ID_Start is Ll, Lu, Lt, Lm, Lo, and Nl.
1140    New_Prop(Is => 'ID_Start',
1141             Table->Merge(@Cat{qw[Ll Lu Lt Lm Lo Nl]}),
1142             Desc  => '[\p{Ll}\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{Nl}]',
1143             Fuzzy => 1);
1144
1145    # ID_Continue is ID_Start, Mn, Mc, Nd, and Pc.
1146    New_Prop(Is => 'ID_Continue',
1147             Table->Merge(@Cat{qw[Ll Lu Lt Lm Lo Nl Mn Mc Nd Pc ]}),
1148             Desc  => '[\p{ID_Start}\p{Mn}\p{Mc}\p{Nd}\p{Pc}]',
1149             Fuzzy => 1);
1150}
1151
1152sub Make_GC_Aliases()
1153{
1154    ##
1155    ## The mapping from General Category long forms to short forms is
1156    ## currently hardwired here since no simple data file in the UCD
1157    ## seems to do that.  Unicode 3.2 will assumedly correct this.
1158    ##
1159    my %Is = (
1160	'Letter'			=>	'L',
1161	'Uppercase_Letter'		=>	'Lu',
1162	'Lowercase_Letter'		=>	'Ll',
1163	'Titlecase_Letter'		=>	'Lt',
1164	'Modifier_Letter'		=>	'Lm',
1165	'Other_Letter'			=>	'Lo',
1166
1167	'Mark'				=>	'M',
1168	'Non_Spacing_Mark'		=>	'Mn',
1169	'Spacing_Mark'			=>	'Mc',
1170	'Enclosing_Mark'		=>	'Me',
1171
1172	'Separator'			=>	'Z',
1173	'Space_Separator'		=>	'Zs',
1174	'Line_Separator'		=>	'Zl',
1175	'Paragraph_Separator'		=>	'Zp',
1176
1177	'Number'			=>	'N',
1178	'Decimal_Number'		=>	'Nd',
1179	'Letter_Number'			=>	'Nl',
1180	'Other_Number'			=>	'No',
1181
1182	'Punctuation'			=>	'P',
1183	'Connector_Punctuation'		=>	'Pc',
1184	'Dash_Punctuation'		=>	'Pd',
1185	'Open_Punctuation'		=>	'Ps',
1186	'Close_Punctuation'		=>	'Pe',
1187	'Initial_Punctuation'		=>	'Pi',
1188	'Final_Punctuation'		=>	'Pf',
1189	'Other_Punctuation'		=>	'Po',
1190
1191	'Symbol'			=>	'S',
1192	'Math_Symbol'			=>	'Sm',
1193	'Currency_Symbol'		=>	'Sc',
1194	'Modifier_Symbol'		=>	'Sk',
1195	'Other_Symbol'			=>	'So',
1196
1197	'Other'				=>	'C',
1198	'Control'			=>	'Cc',
1199	'Format'			=>	'Cf',
1200	'Surrogate'			=>	'Cs',
1201	'Private Use'			=>	'Co',
1202	'Unassigned'			=>	'Cn',
1203    );
1204
1205    ## make the aliases....
1206    while (my ($Alias, $Name) = each %Is) {
1207        New_Alias(Is => $Alias, SameAs => $Name, Fuzzy => 1);
1208    }
1209}
1210
1211
1212##
1213## These are used in:
1214##   MakePropTestScript()
1215##   WriteAllMappings()
1216## for making the test script.
1217##
1218my %FuzzyNameToTest;
1219my %ExactNameToTest;
1220
1221
1222## This used only for making the test script
1223sub GenTests($$$$)
1224{
1225    my $FH = shift;
1226    my $Prop = shift;
1227    my $MatchCode = shift;
1228    my $FailCode = shift;
1229
1230    if (defined $MatchCode) {
1231        printf $FH qq/Expect(1, "\\x{%04X}", '\\p{$Prop}' );\n/, $MatchCode;
1232        printf $FH qq/Expect(0, "\\x{%04X}", '\\p{^$Prop}');\n/, $MatchCode;
1233        printf $FH qq/Expect(0, "\\x{%04X}", '\\P{$Prop}' );\n/, $MatchCode;
1234        printf $FH qq/Expect(1, "\\x{%04X}", '\\P{^$Prop}');\n/, $MatchCode;
1235    }
1236    if (defined $FailCode) {
1237        printf $FH qq/Expect(0, "\\x{%04X}", '\\p{$Prop}' );\n/, $FailCode;
1238        printf $FH qq/Expect(1, "\\x{%04X}", '\\p{^$Prop}');\n/, $FailCode;
1239        printf $FH qq/Expect(1, "\\x{%04X}", '\\P{$Prop}' );\n/, $FailCode;
1240        printf $FH qq/Expect(0, "\\x{%04X}", '\\P{^$Prop}');\n/, $FailCode;
1241    }
1242}
1243
1244## This used only for making the test script
1245sub ExpectError($$)
1246{
1247    my $FH = shift;
1248    my $prop = shift;
1249
1250    print $FH qq/Error('\\p{$prop}');\n/;
1251    print $FH qq/Error('\\P{$prop}');\n/;
1252}
1253
1254## This used only for making the test script
1255my @GoodSeps = (
1256                " ",
1257                "-",
1258                " \t ",
1259                "",
1260                "",
1261                "_",
1262               );
1263my @BadSeps = (
1264               "--",
1265               "__",
1266               " _",
1267               "/"
1268              );
1269
1270## This used only for making the test script
1271sub RandomlyFuzzifyName($;$)
1272{
1273    my $Name = shift;
1274    my $WantError = shift;  ## if true, make an error
1275
1276    my @parts;
1277    for my $part (split /[-\s_]+/, $Name)
1278    {
1279        if (@parts) {
1280            if ($WantError and rand() < 0.3) {
1281                push @parts, $BadSeps[rand(@BadSeps)];
1282                $WantError = 0;
1283            } else {
1284                push @parts, $GoodSeps[rand(@GoodSeps)];
1285            }
1286        }
1287        my $switch = int rand(4);
1288        if ($switch == 0) {
1289            push @parts, uc $part;
1290        } elsif ($switch == 1) {
1291            push @parts, lc $part;
1292        } elsif ($switch == 2) {
1293            push @parts, ucfirst $part;
1294        } else {
1295            push @parts, $part;
1296        }
1297    }
1298    my $new = join('', @parts);
1299
1300    if ($WantError) {
1301        if (rand() >= 0.5) {
1302            $new .= $BadSeps[rand(@BadSeps)];
1303        } else {
1304            $new = $BadSeps[rand(@BadSeps)] . $new;
1305        }
1306    }
1307    return $new;
1308}
1309
1310## This used only for making the test script
1311sub MakePropTestScript()
1312{
1313    ## this written directly -- it's huge.
1314    if (not open OUT, ">TestProp.pl") {
1315        die "$0: TestProp.pl: $!\n";
1316    }
1317    print OUT <DATA>;
1318
1319    while (my ($Name, $Table) = each %ExactNameToTest)
1320    {
1321        GenTests(*OUT, $Name, $Table->ValidCode, $Table->InvalidCode);
1322        ExpectError(*OUT, uc $Name) if uc $Name ne $Name;
1323        ExpectError(*OUT, lc $Name) if lc $Name ne $Name;
1324    }
1325
1326
1327    while (my ($Name, $Table) = each %FuzzyNameToTest)
1328    {
1329        my $Orig  = $CanonicalToOrig{$Name};
1330        my %Names = (
1331                     $Name => 1,
1332                     $Orig => 1,
1333                     RandomlyFuzzifyName($Orig) => 1
1334                    );
1335
1336        for my $N (keys %Names) {
1337            GenTests(*OUT, $N, $Table->ValidCode, $Table->InvalidCode);
1338        }
1339
1340        ExpectError(*OUT, RandomlyFuzzifyName($Orig, 'ERROR'));
1341    }
1342
1343    print OUT "Finished();\n";
1344    close OUT;
1345}
1346
1347
1348##
1349## These are used only in:
1350##   RegisterFileForName()
1351##   WriteAllMappings()
1352##
1353my %Exact;      ## will become %utf8::Exact;
1354my %Canonical;  ## will become %utf8::Canonical;
1355my %CaComment;  ## Comment for %Canonical entry of same key
1356
1357##
1358## Given info about a name and a datafile that it should be associated with,
1359## register that assocation in %Exact and %Canonical.
1360sub RegisterFileForName($$$$)
1361{
1362    my $Type     = shift;
1363    my $Name     = shift;
1364    my $IsFuzzy  = shift;
1365    my $filename = shift;
1366
1367    ##
1368    ## Now in details for the mapping. $Type eq 'Is' has the
1369    ## Is removed, as it will be removed in utf8_heavy when this
1370    ## data is being checked. In keeps its "In", but a second
1371    ## sans-In record is written if it doesn't conflict with
1372    ## anything already there.
1373    ##
1374    if (not $IsFuzzy)
1375    {
1376        if ($Type eq 'Is') {
1377            die "oops[$Name]" if $Exact{$Name};
1378            $Exact{$Name} = $filename;
1379        } else {
1380            die "oops[$Type$Name]" if $Exact{"$Type$Name"};
1381            $Exact{"$Type$Name"} = $filename;
1382            $Exact{$Name} = $filename if not $Exact{$Name};
1383        }
1384    }
1385    else
1386    {
1387        my $CName = lc $Name;
1388        if ($Type eq 'Is') {
1389            die "oops[$CName]" if $Canonical{$CName};
1390            $Canonical{$CName} = $filename;
1391            $CaComment{$CName} = $Name if $Name =~ tr/A-Z// >= 2;
1392        } else {
1393            die "oops[$Type$CName]" if $Canonical{lc "$Type$CName"};
1394            $Canonical{lc "$Type$CName"} = $filename;
1395            $CaComment{lc "$Type$CName"} = "$Type$Name";
1396            if (not $Canonical{$CName}) {
1397                $Canonical{$CName} = $filename;
1398                $CaComment{$CName} = "$Type$Name";
1399            }
1400        }
1401    }
1402}
1403
1404##
1405## Writes the info accumulated in
1406##
1407##       %TableInfo;
1408##       %FuzzyNames;
1409##       %AliasInfo;
1410##
1411##
1412sub WriteAllMappings()
1413{
1414    my @MAP;
1415
1416    my %BaseNames;  ## Base names already used (for avoiding 8.3 conflicts)
1417
1418    ## 'Is' *MUST* come first, so its names have precidence over 'In's
1419    for my $Type ('Is', 'In')
1420    {
1421        my %RawNameToFile; ## a per-$Type cache
1422
1423        for my $Name (sort {length $a <=> length $b} keys %{$TableInfo{$Type}})
1424        {
1425            ## Note: $Name is already canonical
1426            my $Table   = $TableInfo{$Type}->{$Name};
1427            my $IsFuzzy = $FuzzyNames{$Type}->{$Name};
1428
1429            ## Need an 8.3 safe filename (which means "an 8 safe" $filename)
1430            my $filename;
1431            {
1432                ## 'Is' items lose 'Is' from the basename.
1433                $filename = $Type eq 'Is' ? $Name : "$Type$Name";
1434
1435                $filename =~ s/[^\w_]+/_/g; # "L&" -> "L_"
1436                substr($filename, 8) = '' if length($filename) > 8;
1437
1438                ##
1439                ## Make sure the basename doesn't conflict with something we
1440                ## might have already written. If we have, say,
1441                ##     InGreekExtended1
1442                ##     InGreekExtended2
1443                ## they become
1444                ##     InGreekE
1445                ##     InGreek2
1446                ##
1447                while (my $num = $BaseNames{lc $filename}++)
1448                {
1449                    $num++; ## so basenames with numbers start with '2', which
1450                            ## just looks more natural.
1451                    ## Want to append $num, but if it'll make the basename longer
1452                    ## than 8 characters, pre-truncate $filename so that the result
1453                    ## is acceptable.
1454                    my $delta = length($filename) + length($num) - 8;
1455                    if ($delta > 0) {
1456                        substr($filename, -$delta) = $num;
1457                    } else {
1458                        $filename .= $num;
1459                    }
1460                }
1461            };
1462
1463            ##
1464            ## Construct a nice comment to add to the file, and build data
1465            ## for the "./Properties" file along the way.
1466            ##
1467            my $Comment;
1468            {
1469                my $Desc = $TableDesc{$Type}->{$Name} || "";
1470                ## get list of names this table is reference by
1471                my @Supported = $Name;
1472                while (my ($Orig, $Alias) = each %{ $AliasInfo{$Type} })
1473                {
1474                    if ($Orig eq $Name) {
1475                        push @Supported, $Alias;
1476                    }
1477                }
1478
1479                my $TypeToShow = $Type eq 'Is' ? "" : $Type;
1480                my $OrigProp;
1481
1482                $Comment = "This file supports:\n";
1483                for my $N (@Supported)
1484                {
1485                    my $IsFuzzy = $FuzzyNames{$Type}->{$N};
1486                    my $Prop    = "\\p{$TypeToShow$Name}";
1487                    $OrigProp = $Prop if not $OrigProp; #cache for aliases
1488                    if ($IsFuzzy) {
1489                        $Comment .= "\t$Prop (and fuzzy permutations)\n";
1490                    } else {
1491                        $Comment .= "\t$Prop\n";
1492                    }
1493                    my $MyDesc = ($N eq $Name) ? $Desc : "Alias for $OrigProp ($Desc)";
1494
1495                    push @MAP, sprintf("%s %-42s %s\n",
1496                                       $IsFuzzy ? '*' : ' ', $Prop, $MyDesc);
1497                }
1498                if ($Desc) {
1499                    $Comment .= "\nMeaning: $Desc\n";
1500                }
1501
1502            }
1503            ##
1504            ## Okay, write the file...
1505            ##
1506            $Table->Write("lib/$filename.pl", $Comment);
1507
1508            ## and register it
1509            $RawNameToFile{$Name} = $filename;
1510            RegisterFileForName($Type => $Name, $IsFuzzy, $filename);
1511
1512            if ($IsFuzzy)
1513            {
1514                my $CName = CanonicalName($Type . '_'. $Name);
1515                $FuzzyNameToTest{$Name}  = $Table if !$FuzzyNameToTest{$Name};
1516                $FuzzyNameToTest{$CName} = $Table if !$FuzzyNameToTest{$CName};
1517            } else {
1518                $ExactNameToTest{$Name} = $Table;
1519            }
1520
1521        }
1522
1523        ## Register aliase info
1524        for my $Name (sort {length $a <=> length $b} keys %{$AliasInfo{$Type}})
1525        {
1526            my $Alias    = $AliasInfo{$Type}->{$Name};
1527            my $IsFuzzy  = $FuzzyNames{$Type}->{$Alias};
1528            my $filename = $RawNameToFile{$Name};
1529            die "oops [$Alias]->[$Name]" if not $filename;
1530            RegisterFileForName($Type => $Alias, $IsFuzzy, $filename);
1531
1532            my $Table = $TableInfo{$Type}->{$Name};
1533            die "oops" if not $Table;
1534            if ($IsFuzzy)
1535            {
1536                my $CName = CanonicalName($Type .'_'. $Alias);
1537                $FuzzyNameToTest{$Alias} = $Table if !$FuzzyNameToTest{$Alias};
1538                $FuzzyNameToTest{$CName} = $Table if !$FuzzyNameToTest{$CName};
1539            } else {
1540                $ExactNameToTest{$Alias} = $Table;
1541            }
1542        }
1543    }
1544
1545    ##
1546    ## Write out the property list
1547    ##
1548    {
1549        my @OUT = (
1550                   "##\n",
1551                   "## This file created by $0\n",
1552                   "## List of built-in \\p{...}/\\P{...} properties.\n",
1553                   "##\n",
1554                   "## '*' means name may be 'fuzzy'\n",
1555                   "##\n\n",
1556                   sort { substr($a,2) cmp substr($b, 2) } @MAP,
1557                  );
1558        WriteIfChanged('Properties', @OUT);
1559    }
1560
1561    use Text::Tabs ();  ## using this makes the files about half the size
1562
1563    ## Write Exact.pl
1564    {
1565        my @OUT = (
1566                   $HEADER,
1567                   "##\n",
1568                   "## Data in this file used by ../utf8_heavy.pl\n",
1569                   "##\n\n",
1570                   "## Mapping from name to filename in ./lib\n",
1571                   "%utf8::Exact = (\n",
1572                  );
1573
1574        for my $Name (sort keys %Exact)
1575        {
1576            my $File = $Exact{$Name};
1577            $Name = $Name =~ m/\W/ ? qq/'$Name'/ : " $Name ";
1578            my $Text = sprintf("%-15s => %s,\n", $Name, qq/'$File'/);
1579            push @OUT, Text::Tabs::unexpand($Text);
1580        }
1581        push @OUT, ");\n1;\n";
1582
1583        WriteIfChanged('Exact.pl', @OUT);
1584    }
1585
1586    ## Write Canonical.pl
1587    {
1588        my @OUT = (
1589                   $HEADER,
1590                   "##\n",
1591                   "## Data in this file used by ../utf8_heavy.pl\n",
1592                   "##\n\n",
1593                   "## Mapping from lc(canonical name) to filename in ./lib\n",
1594                   "%utf8::Canonical = (\n",
1595                  );
1596        my $Trail = ""; ## used just to keep the spacing pretty
1597        for my $Name (sort keys %Canonical)
1598        {
1599            my $File = $Canonical{$Name};
1600            if ($CaComment{$Name}) {
1601                push @OUT, "\n" if not $Trail;
1602                push @OUT, " # $CaComment{$Name}\n";
1603                $Trail = "\n";
1604            } else {
1605                $Trail = "";
1606            }
1607            $Name = $Name =~ m/\W/ ? qq/'$Name'/ : " $Name ";
1608            my $Text = sprintf("  %-41s => %s,\n$Trail", $Name, qq/'$File'/);
1609            push @OUT, Text::Tabs::unexpand($Text);
1610        }
1611        push @OUT, ");\n1\n";
1612        WriteIfChanged('Canonical.pl', @OUT);
1613    }
1614
1615    MakePropTestScript() if $MakeTestScript;
1616}
1617
1618
1619sub SpecialCasing_txt()
1620{
1621    #
1622    # Read in the special cases.
1623    #
1624
1625    my %CaseInfo;
1626
1627    if (not open IN, "SpecialCasing.txt") {
1628        die "$0: SpecialCasing.txt: $!\n";
1629    }
1630    while (<IN>) {
1631        next unless /^[0-9A-Fa-f]+;/;
1632        s/\#.*//;
1633        s/\s+$//;
1634
1635        my ($code, $lower, $title, $upper, $condition) = split(/\s*;\s*/);
1636
1637        if ($condition) { # not implemented yet
1638            print "# SKIPPING $_\n" if $Verbose;
1639            next;
1640        }
1641
1642        # Wait until all the special cases have been read since
1643        # they are not listed in numeric order.
1644        my $ix = hex($code);
1645        push @{$CaseInfo{Lower}}, [ $ix, $code, $lower ]
1646	    unless $code eq $lower;
1647        push @{$CaseInfo{Title}}, [ $ix, $code, $title ]
1648	    unless $code eq $title;
1649        push @{$CaseInfo{Upper}}, [ $ix, $code, $upper ]
1650	    unless $code eq $upper;
1651    }
1652    close IN;
1653
1654    # Now write out the special cases properties in their code point order.
1655    # Prepend them to the To/{Upper,Lower,Title}.pl.
1656
1657    for my $case (qw(Lower Title Upper))
1658    {
1659        my $NormalCase = do "To/$case.pl" || die "$0: $@\n";
1660
1661        my @OUT =
1662	    (
1663	     $HEADER, "\n",
1664	     "# The key UTF-8 _bytes_, the value UTF-8 (speed hack)\n",
1665	     "%utf8::ToSpec$case =\n(\n",
1666	    );
1667
1668        for my $prop (sort { $a->[0] <=> $b->[0] } @{$CaseInfo{$case}}) {
1669            my ($ix, $code, $to) = @$prop;
1670            my $tostr =
1671              join "", map { sprintf "\\x{%s}", $_ } split ' ', $to;
1672            push @OUT, sprintf qq["%s" => "$tostr",\n], join("", map { sprintf "\\x%02X", $_ } unpack("U0C*", pack("U", $ix)));
1673	    # Remove any single-character mappings for
1674	    # the same character since we are going for
1675	    # the special casing rules.
1676	    $NormalCase =~ s/^$code\t\t\w+\n//m;
1677        }
1678        push @OUT, (
1679                    ");\n\n",
1680                    "return <<'END';\n",
1681                    $NormalCase,
1682                    "END\n"
1683                    );
1684        WriteIfChanged("To/$case.pl", @OUT);
1685    }
1686}
1687
1688#
1689# Read in the case foldings.
1690#
1691# We will do full case folding, C + F + I (see CaseFolding.txt).
1692#
1693sub CaseFolding_txt()
1694{
1695    if (not open IN, "CaseFolding.txt") {
1696	die "$0: CaseFolding.txt: $!\n";
1697    }
1698
1699    my $Fold = Table->New();
1700    my %Fold;
1701
1702    while (<IN>) {
1703	# Skip status 'S', simple case folding
1704	next unless /^([0-9A-Fa-f]+)\s*;\s*([CFI])\s*;\s*([0-9A-Fa-f]+(?: [0-9A-Fa-f]+)*)\s*;/;
1705
1706	my ($code, $status, $fold) = (hex($1), $2, $3);
1707
1708	if ($status eq 'C') { # Common: one-to-one folding
1709	    # No append() since several codes may fold into one.
1710	    $Fold->RawAppendRange($code, $code, $fold);
1711	} else { # F: full, or I: dotted uppercase I -> dotless lowercase I
1712	    $Fold{$code} = $fold;
1713	}
1714    }
1715    close IN;
1716
1717    $Fold->Write("To/Fold.pl");
1718
1719    #
1720    # Prepend the special foldings to the common foldings.
1721    #
1722    my $CommonFold = do "To/Fold.pl" || die "$0: To/Fold.pl: $!\n";
1723
1724    my @OUT =
1725	(
1726	 $HEADER, "\n",
1727	 "#  The ke UTF-8 _bytes_, the value UTF-8 (speed hack)\n",
1728	 "%utf8::ToSpecFold =\n(\n",
1729	);
1730    for my $code (sort { $a <=> $b } keys %Fold) {
1731        my $foldstr =
1732          join "", map { sprintf "\\x{%s}", $_ } split ' ', $Fold{$code};
1733        push @OUT, sprintf qq["%s" => "$foldstr",\n], join("", map { sprintf "\\x%02X", $_ } unpack("U0C*", pack("U", $code)));
1734    }
1735    push @OUT, (
1736                ");\n\n",
1737                "return <<'END';\n",
1738                $CommonFold,
1739                "END\n",
1740               );
1741
1742    WriteIfChanged("To/Fold.pl", @OUT);
1743}
1744
1745## Do it....
1746
1747UnicodeData_Txt();
1748Make_GC_Aliases();
1749PropList_txt();
1750
1751Scripts_txt();
1752Blocks_txt();
1753
1754WriteAllMappings();
1755
1756LineBreak_Txt();
1757ArabicShaping_txt();
1758Jamo_txt();
1759SpecialCasing_txt();
1760CaseFolding_txt();
1761
1762exit(0);
1763
1764## TRAILING CODE IS USED BY MakePropTestScript()
1765__DATA__
1766use strict;
1767use warnings;
1768
1769my $Tests = 0;
1770my $Fails = 0;
1771
1772sub Expect($$$)
1773{
1774    my $Expect = shift;
1775    my $String = shift;
1776    my $Regex  = shift;
1777    my $Line   = (caller)[2];
1778
1779    $Tests++;
1780    my $RegObj;
1781    my $result = eval {
1782        $RegObj = qr/$Regex/;
1783        $String =~ $RegObj ? 1 : 0
1784    };
1785
1786    if (not defined $result) {
1787        print "couldn't compile /$Regex/ on $0 line $Line: $@\n";
1788        $Fails++;
1789    } elsif ($result ^ $Expect) {
1790        print "bad result (expected $Expect) on $0 line $Line: $@\n";
1791        $Fails++;
1792    }
1793}
1794
1795sub Error($)
1796{
1797    my $Regex  = shift;
1798    $Tests++;
1799    if (eval { 'x' =~ qr/$Regex/; 1 }) {
1800        $Fails++;
1801        my $Line = (caller)[2];
1802        print "expected error for /$Regex/ on $0 line $Line: $@\n";
1803    }
1804}
1805
1806sub Finished()
1807{
1808   if ($Fails == 0) {
1809      print "All $Tests tests passed.\n";
1810      exit(0);
1811   } else {
1812      print "$Tests tests, $Fails failed!\n";
1813      exit(-1);
1814   }
1815}
1816