This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
regen/regcharclass.pl: avoid autoderef feature
[perl5.git] / regen / regcharclass.pl
CommitLineData
1f00b0d6 1#!perl
e64b1bd1 2package CharClass::Matcher;
12b72891 3use strict;
8770da0e 4use 5.008;
12b72891 5use warnings;
e64b1bd1 6use warnings FATAL => 'all';
12b72891 7use Data::Dumper;
e64b1bd1
YO
8$Data::Dumper::Useqq= 1;
9our $hex_fmt= "0x%02X";
12b72891 10
75929b4b
KW
11sub DEBUG () { 0 }
12$|=1 if DEBUG;
13
8770da0e 14require 'regen/regen_lib.pl';
a1b2a50f 15require 'regen/charset_translations.pl';
09be8123 16require "regen/regcharclass_multi_char_folds.pl";
8770da0e 17
ab84f958 18=head1 NAME
0ccab2bc 19
e64b1bd1 20CharClass::Matcher -- Generate C macros that match character classes efficiently
12b72891 21
e64b1bd1
YO
22=head1 SYNOPSIS
23
ab84f958 24 perl Porting/regcharclass.pl
e64b1bd1
YO
25
26=head1 DESCRIPTION
12b72891
RGS
27
28Dynamically generates macros for detecting special charclasses
e64b1bd1 29in latin-1, utf8, and codepoint forms. Macros can be set to return
cc08b31c 30the length (in bytes) of the matched codepoint, and/or the codepoint itself.
12b72891 31
cc08b31c 32To regenerate F<regcharclass.h>, run this script from perl-root. No arguments
12b72891
RGS
33are necessary.
34
cc08b31c
KW
35Using WHATEVER as an example the following macros can be produced, depending
36on the input parameters (how to get each is described by internal comments at
37the C<__DATA__> line):
12b72891
RGS
38
39=over 4
40
cc08b31c 41=item C<is_WHATEVER(s,is_utf8)>
12b72891 42
cc08b31c 43=item C<is_WHATEVER_safe(s,e,is_utf8)>
12b72891 44
cc08b31c
KW
45Do a lookup as appropriate based on the C<is_utf8> flag. When possible
46comparisons involving octect<128 are done before checking the C<is_utf8>
12b72891
RGS
47flag, hopefully saving time.
48
cc08b31c
KW
49The version without the C<_safe> suffix should be used only when the input is
50known to be well-formed.
12b72891 51
cc08b31c
KW
52=item C<is_WHATEVER_utf8(s)>
53
54=item C<is_WHATEVER_utf8_safe(s,e)>
12b72891
RGS
55
56Do a lookup assuming the string is encoded in (normalized) UTF8.
57
cc08b31c
KW
58The version without the C<_safe> suffix should be used only when the input is
59known to be well-formed.
60
61=item C<is_WHATEVER_latin1(s)>
12b72891 62
cc08b31c 63=item C<is_WHATEVER_latin1_safe(s,e)>
12b72891
RGS
64
65Do a lookup assuming the string is encoded in latin-1 (aka plan octets).
66
cc08b31c
KW
67The version without the C<_safe> suffix should be used only when it is known
68that C<s> contains at least one character.
69
70=item C<is_WHATEVER_cp(cp)>
12b72891 71
47e01c32 72Check to see if the string matches a given codepoint (hypothetically a
b6a6e956 73U32). The condition is constructed as to "break out" as early as
12b72891
RGS
74possible if the codepoint is out of range of the condition.
75
76IOW:
77
78 (cp==X || (cp>X && (cp==Y || (cp>Y && ...))))
79
80Thus if the character is X+1 only two comparisons will be done. Making
81matching lookups slower, but non-matching faster.
82
cc08b31c
KW
83=item C<what_len_WHATEVER_FOO(arg1, ..., len)>
84
85A variant form of each of the macro types described above can be generated, in
86which the code point is returned by the macro, and an extra parameter (in the
87final position) is added, which is a pointer for the macro to set the byte
88length of the returned code point.
89
90These forms all have a C<what_len> prefix instead of the C<is_>, for example
91C<what_len_WHATEVER_safe(s,e,is_utf8,len)> and
92C<what_len_WHATEVER_utf8(s,len)>.
93
94These forms should not be used I<except> on small sets of mostly widely
95separated code points; otherwise the code generated is inefficient. For these
96cases, it is best to use the C<is_> forms, and then find the code point with
97C<utf8_to_uvchr_buf>(). This program can fail with a "deep recursion"
98message on the worst of the inappropriate sets. Examine the generated macro
99to see if it is acceptable.
12b72891 100
cc08b31c
KW
101=item C<what_WHATEVER_FOO(arg1, ...)>
102
103A variant form of each of the C<is_> macro types described above can be generated, in
104which the code point and not the length is returned by the macro. These have
105the same caveat as L</what_len_WHATEVER_FOO(arg1, ..., len)>, plus they should
106not be used where the set contains a NULL, as 0 is returned for two different
107cases: a) the set doesn't include the input code point; b) the set does
108include it, and it is a NULL.
109
110=back
e64b1bd1 111
40f914fd
KW
112The above isn't quite complete, as for specialized purposes one can get a
113macro like C<is_WHATEVER_utf8_no_length_checks(s)>, which assumes that it is
114already known that there is enough space to hold the character starting at
115C<s>, but otherwise checks that it is well-formed. In other words, this is
116intermediary in checking between C<is_WHATEVER_utf8(s)> and
117C<is_WHATEVER_utf8_safe(s,e)>.
118
e64b1bd1
YO
119=head2 CODE FORMAT
120
121perltidy -st -bt=1 -bbt=0 -pt=0 -sbt=1 -ce -nwls== "%f"
122
123
124=head1 AUTHOR
125
cc08b31c 126Author: Yves Orton (demerphq) 2007. Maintained by Perl5 Porters.
e64b1bd1
YO
127
128=head1 BUGS
129
130No tests directly here (although the regex engine will fail tests
131if this code is broken). Insufficient documentation and no Getopts
132handler for using the module as a script.
133
134=head1 LICENSE
135
136You may distribute under the terms of either the GNU General Public
137License or the Artistic License, as specified in the README file.
138
12b72891
RGS
139=cut
140
e64b1bd1
YO
141# Sub naming convention:
142# __func : private subroutine, can not be called as a method
143# _func : private method, not meant for external use
144# func : public method.
145
146# private subs
147#-------------------------------------------------------------------------------
148#
149# ($cp,$n,$l,$u)=__uni_latin($str);
150#
47e01c32 151# Return a list of arrays, each of which when interpreted correctly
e64b1bd1
YO
152# represent the string in some given encoding with specific conditions.
153#
154# $cp - list of codepoints that make up the string.
295bcca9
KW
155# $n - list of octets that make up the string if all codepoints are invariant
156# regardless of if the string is in UTF-8 or not.
e64b1bd1 157# $l - list of octets that make up the string in latin1 encoding if all
295bcca9
KW
158# codepoints < 256, and at least one codepoint is UTF-8 variant.
159# $u - list of octets that make up the string in utf8 if any codepoint is
160# UTF-8 variant
e64b1bd1
YO
161#
162# High CP | Defined
163#-----------+----------
295bcca9 164# 0 - 127 : $n (127/128 are the values for ASCII platforms)
e64b1bd1
YO
165# 128 - 255 : $l, $u
166# 256 - ... : $u
167#
168
169sub __uni_latin1 {
a1b2a50f 170 my $charset= shift;
e64b1bd1
YO
171 my $str= shift;
172 my $max= 0;
173 my @cp;
900c17f9 174 my @cp_high;
295bcca9 175 my $only_has_invariants = 1;
c30a0cf2 176 my $a2n = get_a2n($charset);
e64b1bd1
YO
177 for my $ch ( split //, $str ) {
178 my $cp= ord $ch;
e64b1bd1 179 $max= $cp if $max < $cp;
a1b2a50f
KW
180 if ($cp > 255) {
181 push @cp, $cp;
182 push @cp_high, $cp;
183 }
184 else {
c30a0cf2 185 push @cp, $a2n->[$cp];
295bcca9 186 }
e64b1bd1
YO
187 }
188 my ( $n, $l, $u );
a1b2a50f 189 $only_has_invariants = ($charset =~ /ascii/i) ? $max < 128 : $max < 160;
295bcca9 190 if ($only_has_invariants) {
e64b1bd1
YO
191 $n= [@cp];
192 } else {
193 $l= [@cp] if $max && $max < 256;
194
a1b2a50f
KW
195 my @u;
196 for my $ch ( split //, $str ) {
197 push @u, map { ord } split //, cp_2_utfbytes(ord $ch, $charset);
198 }
199 $u = \@u;
12b72891 200 }
900c17f9 201 return ( \@cp, \@cp_high, $n, $l, $u );
12b72891
RGS
202}
203
12b72891 204#
e64b1bd1
YO
205# $clean= __clean($expr);
206#
207# Cleanup a ternary expression, removing unnecessary parens and apply some
208# simplifications using regexes.
209#
210
211sub __clean {
212 my ( $expr )= @_;
8fdb8a9d 213
9a3182e9
YO
214 #return $expr;
215
e64b1bd1
YO
216 our $parens;
217 $parens= qr/ (?> \( (?> (?: (?> [^()]+ ) | (??{ $parens }) )* ) \) ) /x;
218
8fdb8a9d 219 ## remove redundant parens
e64b1bd1 220 1 while $expr =~ s/ \( \s* ( $parens ) \s* \) /$1/gx;
8fdb8a9d
YO
221
222
223 # repeatedly simplify conditions like
224 # ( (cond1) ? ( (cond2) ? X : Y ) : Y )
225 # into
226 # ( ( (cond1) && (cond2) ) ? X : Y )
6c4f0678
YO
227 # Also similarly handles expressions like:
228 # : (cond1) ? ( (cond2) ? X : Y ) : Y )
229 # Note the inclusion of the close paren in ([:()]) and the open paren in ([()]) is
230 # purely to ensure we have a balanced set of parens in the expression which makes
231 # it easier to understand the pattern in an editor that understands paren's, we do
232 # not expect either of these cases to actually fire. - Yves
8fdb8a9d 233 1 while $expr =~ s/
6c4f0678 234 ([:()]) \s*
8fdb8a9d
YO
235 ($parens) \s*
236 \? \s*
237 \( \s* ($parens) \s*
6c4f0678
YO
238 \? \s* ($parens|[^()?:\s]+?) \s*
239 : \s* ($parens|[^()?:\s]+?) \s*
8fdb8a9d 240 \) \s*
6c4f0678
YO
241 : \s* \5 \s*
242 ([()])
243 /$1 ( $2 && $3 ) ? $4 : $5 $6/gx;
39a0f513
YO
244 #$expr=~s/\(\(U8\*\)s\)\[(\d+)\]/S$1/g if length $expr > 8000;
245 #$expr=~s/\s+//g if length $expr > 8000;
246
247 die "Expression too long" if length $expr > 8000;
8fdb8a9d 248
e64b1bd1 249 return $expr;
12b72891
RGS
250}
251
e64b1bd1
YO
252#
253# $text= __macro(@args);
254# Join args together by newlines, and then neatly add backslashes to the end
255# of every line as expected by the C pre-processor for #define's.
256#
257
258sub __macro {
259 my $str= join "\n", @_;
260 $str =~ s/\s*$//;
261 my @lines= map { s/\s+$//; s/\t/ /g; $_ } split /\n/, $str;
262 my $last= pop @lines;
263 $str= join "\n", ( map { sprintf "%-76s\\", $_ } @lines ), $last;
264 1 while $str =~ s/^(\t*) {8}/$1\t/gm;
265 return $str . "\n";
12b72891
RGS
266}
267
e64b1bd1
YO
268#
269# my $op=__incrdepth($op);
270#
271# take an 'op' hashref and add one to it and all its childrens depths.
272#
273
274sub __incrdepth {
275 my $op= shift;
276 return unless ref $op;
277 $op->{depth} += 1;
278 __incrdepth( $op->{yes} );
279 __incrdepth( $op->{no} );
280 return $op;
281}
282
283# join two branches of an opcode together with a condition, incrementing
284# the depth on the yes branch when we do so.
285# returns the new root opcode of the tree.
286sub __cond_join {
287 my ( $cond, $yes, $no )= @_;
5ab0c3af 288 if (ref $yes) {
3ff97bcf
KW
289 return {
290 test => $cond,
291 yes => __incrdepth( $yes ),
292 no => $no,
293 depth => 0,
294 };
5ab0c3af
KW
295 }
296 else {
297 return {
298 test => $cond,
299 yes => $yes,
300 no => __incrdepth($no),
301 depth => 0,
302 };
303 }
e64b1bd1
YO
304}
305
306# Methods
307
308# constructor
309#
310# my $obj=CLASS->new(op=>'SOMENAME',title=>'blah',txt=>[..]);
311#
312# Create a new CharClass::Matcher object by parsing the text in
313# the txt array. Currently applies the following rules:
314#
315# Element starts with C<0x>, line is evaled the result treated as
316# a number which is passed to chr().
317#
318# Element starts with C<">, line is evaled and the result treated
319# as a string.
320#
321# Each string is then stored in the 'strs' subhash as a hash record
322# made up of the results of __uni_latin1, using the keynames
b1af8fef 323# 'low','latin1','utf8', as well as the synthesized 'LATIN1', 'high', and
b6a6e956 324# 'UTF8' which hold a merge of 'low' and their lowercase equivalents.
e64b1bd1
YO
325#
326# Size data is tracked per type in the 'size' subhash.
327#
328# Return an object
329#
12b72891
RGS
330sub new {
331 my $class= shift;
e64b1bd1
YO
332 my %opt= @_;
333 for ( qw(op txt) ) {
334 die "in " . __PACKAGE__ . " constructor '$_;' is a mandatory field"
335 if !exists $opt{$_};
336 }
337
338 my $self= bless {
339 op => $opt{op},
340 title => $opt{title} || '',
341 }, $class;
342 foreach my $txt ( @{ $opt{txt} } ) {
343 my $str= $txt;
344 if ( $str =~ /^[""]/ ) {
345 $str= eval $str;
05b688d9
KW
346 } elsif ($str =~ / - /x ) { # A range: Replace this element on the
347 # list with its expansion
348 my ($lower, $upper) = $str =~ / 0x (.+?) \s* - \s* 0x (.+) /x;
349 die "Format must be like '0xDEAD - 0xBEAF'; instead was '$str'" if ! defined $lower || ! defined $upper;
350 foreach my $cp (hex $lower .. hex $upper) {
351 push @{$opt{txt}}, sprintf "0x%X", $cp;
352 }
353 next;
295bcca9
KW
354 } elsif ($str =~ s/ ^ N (?= 0x ) //x ) {
355 # Otherwise undocumented, a leading N means is already in the
356 # native character set; don't convert.
e64b1bd1 357 $str= chr eval $str;
295bcca9
KW
358 } elsif ( $str =~ /^0x/ ) {
359 $str= eval $str;
295bcca9 360 $str = chr $str;
05b688d9
KW
361 } elsif ( $str =~ / \s* \\p \{ ( .*? ) \} /x) {
362 my $property = $1;
363 use Unicode::UCD qw(prop_invlist);
364
365 my @invlist = prop_invlist($property, '_perl_core_internal_ok');
366 if (! @invlist) {
367
368 # An empty return could mean an unknown property, or merely
369 # that it is empty. Call in scalar context to differentiate
370 my $count = prop_invlist($property, '_perl_core_internal_ok');
371 die "$property not found" unless defined $count;
372 }
373
374 # Replace this element on the list with the property's expansion
375 for (my $i = 0; $i < @invlist; $i += 2) {
376 foreach my $cp ($invlist[$i] .. $invlist[$i+1] - 1) {
295bcca9
KW
377
378 # prop_invlist() returns native values; add leading 'N'
379 # to indicate that.
380 push @{$opt{txt}}, sprintf "N0x%X", $cp;
05b688d9
KW
381 }
382 }
383 next;
60910c93
KW
384 } elsif ($str =~ / ^ do \s+ ( .* ) /x) {
385 die "do '$1' failed: $!$@" if ! do $1 or $@;
386 next;
387 } elsif ($str =~ / ^ & \s* ( .* ) /x) { # user-furnished sub() call
388 my @results = eval "$1";
389 die "eval '$1' failed: $@" if $@;
390 push @{$opt{txt}}, @results;
391 next;
12b72891 392 } else {
5e6c6c1e 393 die "Unparsable line: $txt\n";
12b72891 394 }
a1b2a50f 395 my ( $cp, $cp_high, $low, $latin1, $utf8 )= __uni_latin1( $opt{charset}, $str );
e64b1bd1
YO
396 my $UTF8= $low || $utf8;
397 my $LATIN1= $low || $latin1;
b1af8fef 398 my $high = (scalar grep { $_ < 256 } @$cp) ? 0 : $utf8;
dda856b2
YO
399 #die Dumper($txt,$cp,$low,$latin1,$utf8)
400 # if $txt=~/NEL/ or $utf8 and @$utf8>3;
e64b1bd1 401
900c17f9
KW
402 @{ $self->{strs}{$str} }{qw( str txt low utf8 latin1 high cp cp_high UTF8 LATIN1 )}=
403 ( $str, $txt, $low, $utf8, $latin1, $high, $cp, $cp_high, $UTF8, $LATIN1 );
e64b1bd1 404 my $rec= $self->{strs}{$str};
900c17f9 405 foreach my $key ( qw(low utf8 latin1 high cp cp_high UTF8 LATIN1) ) {
e64b1bd1
YO
406 $self->{size}{$key}{ 0 + @{ $self->{strs}{$str}{$key} } }++
407 if $self->{strs}{$str}{$key};
12b72891 408 }
e64b1bd1
YO
409 $self->{has_multi} ||= @$cp > 1;
410 $self->{has_ascii} ||= $latin1 && @$latin1;
411 $self->{has_low} ||= $low && @$low;
412 $self->{has_high} ||= !$low && !$latin1;
12b72891 413 }
e64b1bd1
YO
414 $self->{val_fmt}= $hex_fmt;
415 $self->{count}= 0 + keys %{ $self->{strs} };
12b72891
RGS
416 return $self;
417}
418
e64b1bd1 419# my $trie = make_trie($type,$maxlen);
12b72891 420#
47e01c32 421# using the data stored in the object build a trie of a specific type,
e64b1bd1
YO
422# and with specific maximum depth. The trie is made up the elements of
423# the given types array for each string in the object (assuming it is
424# not too long.)
425#
47e01c32 426# returns the trie, or undef if there was no relevant data in the object.
e64b1bd1
YO
427#
428
429sub make_trie {
430 my ( $self, $type, $maxlen )= @_;
431
432 my $strs= $self->{strs};
433 my %trie;
434 foreach my $rec ( values %$strs ) {
435 die "panic: unknown type '$type'"
436 if !exists $rec->{$type};
437 my $dat= $rec->{$type};
438 next unless $dat;
439 next if $maxlen && @$dat > $maxlen;
440 my $node= \%trie;
441 foreach my $elem ( @$dat ) {
442 $node->{$elem} ||= {};
443 $node= $node->{$elem};
12b72891 444 }
e64b1bd1 445 $node->{''}= $rec->{str};
12b72891 446 }
e64b1bd1 447 return 0 + keys( %trie ) ? \%trie : undef;
12b72891
RGS
448}
449
2efb8143
KW
450sub pop_count ($) {
451 my $word = shift;
452
453 # This returns a list of the positions of the bits in the input word that
454 # are 1.
455
456 my @positions;
457 my $position = 0;
458 while ($word) {
459 push @positions, $position if $word & 1;
460 $position++;
461 $word >>= 1;
462 }
463 return @positions;
464}
465
e64b1bd1
YO
466# my $optree= _optree()
467#
468# recursively convert a trie to an optree where every node represents
469# an if else branch.
12b72891 470#
12b72891 471#
12b72891 472
e64b1bd1
YO
473sub _optree {
474 my ( $self, $trie, $test_type, $ret_type, $else, $depth )= @_;
475 return unless defined $trie;
476 if ( $self->{has_multi} and $ret_type =~ /cp|both/ ) {
477 die "Can't do 'cp' optree from multi-codepoint strings";
12b72891 478 }
e64b1bd1
YO
479 $ret_type ||= 'len';
480 $else= 0 unless defined $else;
481 $depth= 0 unless defined $depth;
482
b6a6e956 483 # if we have an empty string as a key it means we are in an
e405c23a
YO
484 # accepting state and unless we can match further on should
485 # return the value of the '' key.
895e25a5 486 if (exists $trie->{''} ) {
e405c23a
YO
487 # we can now update the "else" value, anything failing to match
488 # after this point should return the value from this.
e64b1bd1
YO
489 if ( $ret_type eq 'cp' ) {
490 $else= $self->{strs}{ $trie->{''} }{cp}[0];
491 $else= sprintf "$self->{val_fmt}", $else if $else > 9;
492 } elsif ( $ret_type eq 'len' ) {
493 $else= $depth;
494 } elsif ( $ret_type eq 'both') {
495 $else= $self->{strs}{ $trie->{''} }{cp}[0];
496 $else= sprintf "$self->{val_fmt}", $else if $else > 9;
497 $else= "len=$depth, $else";
12b72891 498 }
e64b1bd1 499 }
e405c23a
YO
500 # extract the meaningful keys from the trie, filter out '' as
501 # it means we are an accepting state (end of sequence).
502 my @conds= sort { $a <=> $b } grep { length $_ } keys %$trie;
503
b6a6e956 504 # if we haven't any keys there is no further we can match and we
e405c23a 505 # can return the "else" value.
e64b1bd1 506 return $else if !@conds;
e405c23a 507
c7c8bf55 508 my $test = $test_type =~ /^cp/ ? "cp" : "((U8*)s)[$depth]";
e405c23a 509
c7c8bf55 510 # First we loop over the possible keys/conditions and find out what they
3ff97bcf 511 # look like; we group conditions with the same optree together.
9a3182e9
YO
512 my %dmp_res;
513 my @res_order;
e405c23a
YO
514 local $Data::Dumper::Sortkeys=1;
515 foreach my $cond ( @conds ) {
516
517 # get the optree for this child/condition
518 my $res= $self->_optree( $trie->{$cond}, $test_type, $ret_type, $else, $depth + 1 );
519 # convert it to a string with Dumper
e64b1bd1 520 my $res_code= Dumper( $res );
e405c23a 521
9a3182e9
YO
522 push @{$dmp_res{$res_code}{vals}}, $cond;
523 if (!$dmp_res{$res_code}{optree}) {
524 $dmp_res{$res_code}{optree}= $res;
525 push @res_order, $res_code;
526 }
527 }
528
529 # now that we have deduped the optrees we construct a new optree containing the merged
530 # results.
531 my %root;
532 my $node= \%root;
533 foreach my $res_code_idx (0 .. $#res_order) {
534 my $res_code= $res_order[$res_code_idx];
535 $node->{vals}= $dmp_res{$res_code}{vals};
536 $node->{test}= $test;
537 $node->{yes}= $dmp_res{$res_code}{optree};
538 $node->{depth}= $depth;
539 if ($res_code_idx < $#res_order) {
540 $node= $node->{no}= {};
12b72891 541 } else {
9a3182e9 542 $node->{no}= $else;
12b72891
RGS
543 }
544 }
e405c23a
YO
545
546 # return the optree.
547 return \%root;
12b72891
RGS
548}
549
e64b1bd1
YO
550# my $optree= optree(%opts);
551#
552# Convert a trie to an optree, wrapper for _optree
553
554sub optree {
555 my $self= shift;
556 my %opt= @_;
557 my $trie= $self->make_trie( $opt{type}, $opt{max_depth} );
558 $opt{ret_type} ||= 'len';
900c17f9 559 my $test_type= $opt{type} =~ /^cp/ ? 'cp' : 'depth';
e64b1bd1 560 return $self->_optree( $trie, $test_type, $opt{ret_type}, $opt{else}, 0 );
12b72891
RGS
561}
562
e64b1bd1
YO
563# my $optree= generic_optree(%opts);
564#
565# build a "generic" optree out of the three 'low', 'latin1', 'utf8'
566# sets of strings, including a branch for handling the string type check.
567#
568
569sub generic_optree {
570 my $self= shift;
571 my %opt= @_;
572
573 $opt{ret_type} ||= 'len';
574 my $test_type= 'depth';
575 my $else= $opt{else} || 0;
576
577 my $latin1= $self->make_trie( 'latin1', $opt{max_depth} );
578 my $utf8= $self->make_trie( 'utf8', $opt{max_depth} );
579
580 $_= $self->_optree( $_, $test_type, $opt{ret_type}, $else, 0 )
581 for $latin1, $utf8;
582
583 if ( $utf8 ) {
584 $else= __cond_join( "( is_utf8 )", $utf8, $latin1 || $else );
585 } elsif ( $latin1 ) {
586 $else= __cond_join( "!( is_utf8 )", $latin1, $else );
587 }
87894a24 588 if ($opt{type} eq 'generic') {
61de6bbc
KW
589 my $low= $self->make_trie( 'low', $opt{max_depth} );
590 if ( $low ) {
591 $else= $self->_optree( $low, $test_type, $opt{ret_type}, $else, 0 );
592 }
87894a24 593 }
e64b1bd1
YO
594
595 return $else;
12b72891
RGS
596}
597
e64b1bd1 598# length_optree()
12b72891 599#
e64b1bd1 600# create a string length guarded optree.
12b72891 601#
e64b1bd1
YO
602
603sub length_optree {
604 my $self= shift;
605 my %opt= @_;
606 my $type= $opt{type};
607
608 die "Can't do a length_optree on type 'cp', makes no sense."
900c17f9 609 if $type =~ /^cp/;
e64b1bd1 610
5ab0c3af
KW
611 my $else= ( $opt{else} ||= 0 );
612
613 my $method = $type =~ /generic/ ? 'generic_optree' : 'optree';
614 if ($method eq 'optree' && scalar keys %{$self->{size}{$type}} == 1) {
615
616 # Here is non-generic output (meaning that we are only generating one
617 # type), and all things that match have the same number ('size') of
618 # bytes. The length guard is simply that we have that number of
619 # bytes.
620 my @size = keys %{$self->{size}{$type}};
621 my $cond= "((e) - (s)) >= $size[0]";
622 my $optree = $self->$method(%opt);
623 $else= __cond_join( $cond, $optree, $else );
624 }
625 elsif ($self->{has_multi}) {
626 my @size;
e64b1bd1 627
5ab0c3af
KW
628 # Here, there can be a match of a multiple character string. We use
629 # the traditional method which is to have a branch for each possible
630 # size (longest first) and test for the legal values for that size.
e64b1bd1
YO
631 my %sizes= (
632 %{ $self->{size}{low} || {} },
633 %{ $self->{size}{latin1} || {} },
634 %{ $self->{size}{utf8} || {} }
635 );
5ab0c3af
KW
636 if ($method eq 'generic_optree') {
637 @size= sort { $a <=> $b } keys %sizes;
638 } else {
639 @size= sort { $a <=> $b } keys %{ $self->{size}{$type} };
640 }
641 for my $size ( @size ) {
642 my $optree= $self->$method( %opt, type => $type, max_depth => $size );
643 my $cond= "((e)-(s) > " . ( $size - 1 ).")";
644 $else= __cond_join( $cond, $optree, $else );
645 }
12b72891 646 }
5ab0c3af
KW
647 else {
648 my $utf8;
649
650 # Here, has more than one possible size, and only matches a single
651 # character. For non-utf8, the needed length is 1; for utf8, it is
652 # found by array lookup 'UTF8SKIP'.
653
654 # If want just the code points above 255, set up to look for those;
655 # otherwise assume will be looking for all non-UTF-8-invariant code
656 # poiints.
657 my $trie_type = ($type eq 'high') ? 'high' : 'utf8';
658
659 # If we do want more than the 0-255 range, find those, and if they
660 # exist...
661 if ($opt{type} !~ /latin1/i && ($utf8 = $self->make_trie($trie_type, 0))) {
662
663 # ... get them into an optree, and set them up as the 'else' clause
664 $utf8 = $self->_optree( $utf8, 'depth', $opt{ret_type}, 0, 0 );
665
666 # We could make this
667 # UTF8_IS_START(*s) && ((e) - (s)) >= UTF8SKIP(s))";
668 # to avoid doing the UTF8SKIP and subsequent branches for invariants
669 # that don't match. But the current macros that get generated
670 # have only a few things that can match past this, so I (khw)
671 # don't think it is worth it. (Even better would be to use
672 # calculate_mask(keys %$utf8) instead of UTF8_IS_START, and use it
dd9bc2b0
KW
673 # if it saves a bunch. We assume that input text likely to be
674 # well-formed .
675 my $cond = "LIKELY(((e) - (s)) >= UTF8SKIP(s))";
5ab0c3af
KW
676 $else = __cond_join($cond, $utf8, $else);
677
678 # For 'generic', we also will want the latin1 UTF-8 variants for
679 # the case where the input isn't UTF-8.
680 my $latin1;
681 if ($method eq 'generic_optree') {
682 $latin1 = $self->make_trie( 'latin1', 1);
683 $latin1= $self->_optree( $latin1, 'depth', $opt{ret_type}, 0, 0 );
684 }
e64b1bd1 685
5ab0c3af
KW
686 # If we want the UTF-8 invariants, get those.
687 my $low;
688 if ($opt{type} !~ /non_low|high/
689 && ($low= $self->make_trie( 'low', 1)))
690 {
691 $low= $self->_optree( $low, 'depth', $opt{ret_type}, 0, 0 );
692
693 # Expand out the UTF-8 invariants as a string so that we
694 # can use them as the conditional
695 $low = $self->_cond_as_str( $low, 0, \%opt);
696
697 # If there are Latin1 variants, add a test for them.
698 if ($latin1) {
699 $else = __cond_join("(! is_utf8 )", $latin1, $else);
700 }
701 elsif ($method eq 'generic_optree') {
702
703 # Otherwise for 'generic' only we know that what
704 # follows must be valid for just UTF-8 strings,
705 $else->{test} = "( is_utf8 && $else->{test} )";
706 }
707
708 # If the invariants match, we are done; otherwise we have
709 # to go to the 'else' clause.
710 $else = __cond_join($low, 1, $else);
711 }
712 elsif ($latin1) { # Here, didn't want or didn't have invariants,
713 # but we do have latin variants
714 $else = __cond_join("(! is_utf8)", $latin1, $else);
715 }
716
717 # We need at least one byte available to start off the tests
dd9bc2b0 718 $else = __cond_join("LIKELY((e) > (s))", $else, 0);
5ab0c3af
KW
719 }
720 else { # Here, we don't want or there aren't any variants. A single
721 # byte available is enough.
722 my $cond= "((e) > (s))";
723 my $optree = $self->$method(%opt);
724 $else= __cond_join( $cond, $optree, $else );
725 }
e64b1bd1 726 }
5ab0c3af 727
e64b1bd1 728 return $else;
12b72891
RGS
729}
730
2efb8143 731sub calculate_mask(@) {
75929b4b
KW
732 # Look at the input list of byte values. This routine returns an array of
733 # mask/base pairs to generate that list.
734
2efb8143
KW
735 my @list = @_;
736 my $list_count = @list;
737
75929b4b
KW
738 # Consider a set of byte values, A, B, C .... If we want to determine if
739 # <c> is one of them, we can write c==A || c==B || c==C .... If the
740 # values are consecutive, we can shorten that to A<=c && c<=Z, which uses
741 # far fewer branches. If only some of them are consecutive we can still
742 # save some branches by creating range tests for just those that are
743 # consecutive. _cond_as_str() does this work for looking for ranges.
744 #
745 # Another approach is to look at the bit patterns for A, B, C .... and see
746 # if they have some commonalities. That's what this function does. For
747 # example, consider a set consisting of the bytes
748 # 0xF0, 0xF1, 0xF2, and 0xF3. We could write:
2efb8143
KW
749 # 0xF0 <= c && c <= 0xF4
750 # But the following mask/compare also works, and has just one test:
75929b4b
KW
751 # (c & 0xFC) == 0xF0
752 # The reason it works is that the set consists of exactly those bytes
2efb8143 753 # whose first 4 bits are 1, and the next two are 0. (The value of the
75929b4b 754 # other 2 bits is immaterial in determining if a byte is in the set or
2efb8143 755 # not.) The mask masks out those 2 irrelevant bits, and the comparison
75929b4b
KW
756 # makes sure that the result matches all bytes which match those 6
757 # material bits exactly. In other words, the set of bytes contains
2efb8143
KW
758 # exactly those whose bottom two bit positions are either 0 or 1. The
759 # same principle applies to bit positions that are not necessarily
760 # adjacent. And it can be applied to bytes that differ in 1 through all 8
761 # bit positions. In order to be a candidate for this optimization, the
75929b4b
KW
762 # number of bytes in the set must be a power of 2.
763 #
764 # Consider a different example, the set 0x53, 0x54, 0x73, and 0x74. That
765 # requires 4 tests using either ranges or individual values, and even
766 # though the number in the set is a power of 2, it doesn't qualify for the
767 # mask optimization described above because the number of bits that are
768 # different is too large for that. However, the set can be expressed as
769 # two branches with masks thusly:
770 # (c & 0xDF) == 0x53 || (c & 0xDF) == 0x54
771 # a branch savings of 50%. This is done by splitting the set into two
772 # subsets each of which has 2 elements, and within each set the values
773 # differ by 1 byte.
774 #
775 # This function attempts to find some way to save some branches using the
776 # mask technique. If not, it returns an empty list; if so, it
777 # returns a list consisting of
778 # [ [compare1, mask1], [compare2, mask2], ...
779 # [compare_n, undef], [compare_m, undef], ...
780 # ]
781 # The <mask> is undef in the above for those bytes that must be tested
782 # for individually.
783 #
784 # This function does not attempt to find the optimal set. To do so would
785 # probably require testing all possible combinations, and keeping track of
786 # the current best one.
787 #
788 # There are probably much better algorithms, but this is the one I (khw)
789 # came up with. We start with doing a bit-wise compare of every byte in
790 # the set with every other byte. The results are sorted into arrays of
791 # all those that differ by the same bit positions. These are stored in a
792 # hash with the each key being the bits they differ in. Here is the hash
793 # for the 0x53, 0x54, 0x73, 0x74 set:
794 # {
795 # 4 => {
796 # "0,1,2,5" => [
797 # 83,
798 # 116,
799 # 84,
800 # 115
801 # ]
802 # },
803 # 3 => {
804 # "0,1,2" => [
805 # 83,
806 # 84,
807 # 115,
808 # 116
809 # ]
810 # }
811 # 1 => {
812 # 5 => [
813 # 83,
814 # 115,
815 # 84,
816 # 116
817 # ]
818 # },
819 # }
820 #
821 # The set consisting of values which differ in the 4 bit positions 0, 1,
822 # 2, and 5 from some other value in the set consists of all 4 values.
823 # Likewise all 4 values differ from some other value in the 3 bit
824 # positions 0, 1, and 2; and all 4 values differ from some other value in
825 # the single bit position 5. The keys at the uppermost level in the above
826 # hash, 1, 3, and 4, give the number of bit positions that each sub-key
827 # below it has. For example, the 4 key could have as its value an array
828 # consisting of "0,1,2,5", "0,1,2,6", and "3,4,6,7", if the inputs were
829 # such. The best optimization will group the most values into a single
830 # mask. The most values will be the ones that differ in the most
831 # positions, the ones with the largest value for the topmost key. These
832 # keys, are thus just for convenience of sorting by that number, and do
833 # not have any bearing on the core of the algorithm.
834 #
835 # We start with an element from largest number of differing bits. The
836 # largest in this case is 4 bits, and there is only one situation in this
837 # set which has 4 differing bits, "0,1,2,5". We look for any subset of
838 # this set which has 16 values that differ in these 4 bits. There aren't
839 # any, because there are only 4 values in the entire set. We then look at
840 # the next possible thing, which is 3 bits differing in positions "0,1,2".
841 # We look for a subset that has 8 values that differ in these 3 bits.
842 # Again there are none. So we go to look for the next possible thing,
843 # which is a subset of 2**1 values that differ only in bit position 5. 83
844 # and 115 do, so we calculate a mask and base for those and remove them
845 # from every set. Since there is only the one set remaining, we remove
846 # them from just this one. We then look to see if there is another set of
847 # 2 values that differ in bit position 5. 84 and 116 do, so we calculate
848 # a mask and base for those and remove them from every set (again only
849 # this set remains in this example). The set is now empty, and there are
850 # no more sets to look at, so we are done.
851
852 if ($list_count == 256) { # All 256 is trivially masked
2efb8143
KW
853 return (0, 0);
854 }
855
75929b4b
KW
856 my %hash;
857
858 # Generate bits-differing lists for each element compared against each
859 # other element
860 for my $i (0 .. $list_count - 2) {
861 for my $j ($i + 1 .. $list_count - 1) {
862 my @bits_that_differ = pop_count($list[$i] ^ $list[$j]);
863 my $differ_count = @bits_that_differ;
864 my $key = join ",", @bits_that_differ;
865 push @{$hash{$differ_count}{$key}}, $list[$i] unless grep { $_ == $list[$i] } @{$hash{$differ_count}{$key}};
866 push @{$hash{$differ_count}{$key}}, $list[$j];
867 }
868 }
2efb8143 869
75929b4b 870 print STDERR __LINE__, ": calculate_mask() called: List of values grouped by differing bits: ", Dumper \%hash if DEBUG;
2efb8143 871
75929b4b
KW
872 my @final_results;
873 foreach my $count (reverse sort { $a <=> $b } keys %hash) {
874 my $need = 2 ** $count; # Need 8 values for 3 differing bits, etc
de6cb0ab 875 foreach my $bits (sort keys $hash{$count}->%*) {
2efb8143 876
75929b4b 877 print STDERR __LINE__, ": For $count bit(s) difference ($bits), need $need; have ", scalar @{$hash{$count}{$bits}}, "\n" if DEBUG;
2efb8143 878
75929b4b
KW
879 # Look only as long as there are at least as many elements in the
880 # subset as are needed
881 while ((my $cur_count = @{$hash{$count}{$bits}}) >= $need) {
2efb8143 882
75929b4b 883 print STDERR __LINE__, ": Looking at bit positions ($bits): ", Dumper $hash{$count}{$bits} if DEBUG;
2efb8143 884
75929b4b
KW
885 # Start with the first element in it
886 my $try_base = $hash{$count}{$bits}[0];
887 my @subset = $try_base;
888
889 # If it succeeds, we return a mask and a base to compare
890 # against the masked value. That base will be the AND of
891 # every element in the subset. Initialize to the one element
892 # we have so far.
893 my $compare = $try_base;
894
895 # We are trying to find a subset of this that has <need>
896 # elements that differ in the bit positions given by the
897 # string $bits, which is comma separated.
898 my @bits = split ",", $bits;
899
900 TRY: # Look through the remainder of the list for other
901 # elements that differ only by these bit positions.
902
903 for (my $i = 1; $i < $cur_count; $i++) {
904 my $try_this = $hash{$count}{$bits}[$i];
905 my @positions = pop_count($try_base ^ $try_this);
906
907 print STDERR __LINE__, ": $try_base vs $try_this: is (", join(',', @positions), ") a subset of ($bits)?" if DEBUG;;
908
909 foreach my $pos (@positions) {
910 unless (grep { $pos == $_ } @bits) {
911 print STDERR " No\n" if DEBUG;
912 my $remaining = $cur_count - $i - 1;
913 if ($remaining && @subset + $remaining < $need) {
914 print STDERR __LINE__, ": Can stop trying $try_base, because even if all the remaining $remaining values work, they wouldn't add up to the needed $need when combined with the existing ", scalar @subset, " ones\n" if DEBUG;
915 last TRY;
916 }
917 next TRY;
918 }
919 }
920
921 print STDERR " Yes\n" if DEBUG;
922 push @subset, $try_this;
923
924 # Add this to the mask base, in case it ultimately
925 # succeeds,
926 $compare &= $try_this;
927 }
928
929 print STDERR __LINE__, ": subset (", join(", ", @subset), ") has ", scalar @subset, " elements; needs $need\n" if DEBUG;
930
931 if (@subset < $need) {
932 shift @{$hash{$count}{$bits}};
933 next; # Try with next value
934 }
2efb8143 935
75929b4b
KW
936 # Create the mask
937 my $mask = 0;
938 foreach my $position (@bits) {
939 $mask |= 1 << $position;
940 }
941 $mask = ~$mask & 0xFF;
942 push @final_results, [$compare, $mask];
943
944 printf STDERR "%d: Got it: compare=%d=0x%X; mask=%X\n", __LINE__, $compare, $compare, $mask if DEBUG;
945
946 # These values are now spoken for. Remove them from future
947 # consideration
122a2d8f
YO
948 foreach my $remove_count (sort keys %hash) {
949 foreach my $bits (sort keys %{$hash{$remove_count}}) {
75929b4b
KW
950 foreach my $to_remove (@subset) {
951 @{$hash{$remove_count}{$bits}} = grep { $_ != $to_remove } @{$hash{$remove_count}{$bits}};
952 }
953 }
954 }
955 }
956 }
2efb8143
KW
957 }
958
75929b4b
KW
959 # Any values that remain in the list are ones that have to be tested for
960 # individually.
961 my @individuals;
962 foreach my $count (reverse sort { $a <=> $b } keys %hash) {
de6cb0ab 963 foreach my $bits (sort keys $hash{$count}->%*) {
75929b4b
KW
964 foreach my $remaining (@{$hash{$count}{$bits}}) {
965
966 # If we already know about this value, just ignore it.
967 next if grep { $remaining == $_ } @individuals;
968
969 # Otherwise it needs to be returned as something to match
970 # individually
971 push @final_results, [$remaining, undef];
972 push @individuals, $remaining;
973 }
974 }
2efb8143 975 }
2efb8143 976
75929b4b
KW
977 # Sort by increasing numeric value
978 @final_results = sort { $a->[0] <=> $b->[0] } @final_results;
979
980 print STDERR __LINE__, ": Final return: ", Dumper \@final_results if DEBUG;
981
982 return @final_results;
2efb8143
KW
983}
984
e64b1bd1
YO
985# _cond_as_str
986# turn a list of conditions into a text expression
987# - merges ranges of conditions, and joins the result with ||
988sub _cond_as_str {
ba073cf2 989 my ( $self, $op, $combine, $opts_ref )= @_;
e64b1bd1
YO
990 my $cond= $op->{vals};
991 my $test= $op->{test};
2efb8143 992 my $is_cp_ret = $opts_ref->{ret_type} eq "cp";
e64b1bd1
YO
993 return "( $test )" if !defined $cond;
994
f5772832 995 # rangify the list.
e64b1bd1
YO
996 my @ranges;
997 my $Update= sub {
f5772832
KW
998 # We skip this if there are optimizations that
999 # we can apply (below) to the individual ranges
1000 if ( ($is_cp_ret || $combine) && @ranges && ref $ranges[-1]) {
e64b1bd1
YO
1001 if ( $ranges[-1][0] == $ranges[-1][1] ) {
1002 $ranges[-1]= $ranges[-1][0];
1003 } elsif ( $ranges[-1][0] + 1 == $ranges[-1][1] ) {
1004 $ranges[-1]= $ranges[-1][0];
1005 push @ranges, $ranges[-1] + 1;
1006 }
1007 }
1008 };
4a8ca70e
KW
1009 for my $condition ( @$cond ) {
1010 if ( !@ranges || $condition != $ranges[-1][1] + 1 ) {
e64b1bd1 1011 $Update->();
4a8ca70e 1012 push @ranges, [ $condition, $condition ];
e64b1bd1
YO
1013 } else {
1014 $ranges[-1][1]++;
1015 }
1016 }
1017 $Update->();
f5772832 1018
e64b1bd1
YO
1019 return $self->_combine( $test, @ranges )
1020 if $combine;
f5772832
KW
1021
1022 if ($is_cp_ret) {
1f063c57
KW
1023 @ranges= map {
1024 ref $_
1025 ? sprintf(
1026 "( $self->{val_fmt} <= $test && $test <= $self->{val_fmt} )",
1027 @$_ )
1028 : sprintf( "$self->{val_fmt} == $test", $_ );
1029 } @ranges;
6a52943c
KW
1030
1031 return "( " . join( " || ", @ranges ) . " )";
f5772832 1032 }
75929b4b 1033
2358c533
KW
1034 # If the input set has certain characteristics, we can optimize tests
1035 # for it. This doesn't apply if returning the code point, as we want
1036 # each element of the set individually. The code above is for this
1037 # simpler case.
1038
1039 return 1 if @$cond == 256; # If all bytes match, is trivially true
1040
75929b4b 1041 my @masks;
2358c533 1042 if (@ranges > 1) {
75929b4b 1043
b6a6e956 1044 # See if the entire set shares optimizable characteristics, and if so,
2358c533
KW
1045 # return the optimization. We delay checking for this on sets with
1046 # just a single range, as there may be better optimizations available
1047 # in that case.
75929b4b
KW
1048 @masks = calculate_mask(@$cond);
1049
1050 # Stringify the output of calculate_mask()
1051 if (@masks) {
1052 my @return;
1053 foreach my $mask_ref (@masks) {
1054 if (defined $mask_ref->[1]) {
1055 push @return, sprintf "( ( $test & $self->{val_fmt} ) == $self->{val_fmt} )", $mask_ref->[1], $mask_ref->[0];
1056 }
1057 else { # An undefined mask means to use the value as-is
1058 push @return, sprintf "$test == $self->{val_fmt}", $mask_ref->[0];
1059 }
1060 }
1061
1062 # The best possible case below for specifying this set of values via
1063 # ranges is 1 branch per range. If our mask method yielded better
1064 # results, there is no sense trying something that is bound to be
1065 # worse.
1066 if (@return < @ranges) {
1067 return "( " . join( " || ", @return ) . " )";
1068 }
1069
1070 @masks = @return;
6e130234 1071 }
2358c533 1072 }
f5772832 1073
75929b4b
KW
1074 # Here, there was no entire-class optimization that was clearly better
1075 # than doing things by ranges. Look at each range.
1076 my $range_count_extra = 0;
2358c533
KW
1077 for (my $i = 0; $i < @ranges; $i++) {
1078 if (! ref $ranges[$i]) { # Trivial case: no range
1079 $ranges[$i] = sprintf "$self->{val_fmt} == $test", $ranges[$i];
1080 }
1081 elsif ($ranges[$i]->[0] == $ranges[$i]->[1]) {
1082 $ranges[$i] = # Trivial case: single element range
1083 sprintf "$self->{val_fmt} == $test", $ranges[$i]->[0];
1084 }
726137b5
KW
1085 elsif ($ranges[$i]->[0] == 0) {
1086 # If the range matches all 256 possible bytes, it is trivially
1087 # true.
1088 return 1 if $ranges[0]->[1] == 0xFF; # @ranges must be 1 in
1089 # this case
1090 $ranges[$i] = sprintf "( $test <= $self->{val_fmt} )",
1091 $ranges[$i]->[1];
1092 }
1093 elsif ($ranges[$i]->[1] == 255) {
1094
1095 # Similarly the max possible is 255, so can omit an upper bound
1096 # test if the calculated max is the max possible one.
1097 $ranges[$i] = sprintf "( $test >= $self->{val_fmt} )",
1098 $ranges[0]->[0];
1099 }
2358c533
KW
1100 else {
1101 my $output = "";
1102
1103 # Well-formed UTF-8 continuation bytes on ascii platforms must be
1104 # in the range 0x80 .. 0xBF. If we know that the input is
1105 # well-formed (indicated by not trying to be 'safe'), we can omit
1106 # tests that verify that the input is within either of these
1107 # bounds. (No legal UTF-8 character can begin with anything in
1108 # this range, so we don't have to worry about this being a
1109 # continuation byte or not.)
a1b2a50f
KW
1110 if ($opts_ref->{charset} =~ /ascii/i
1111 && (! $opts_ref->{safe} && ! $opts_ref->{no_length_checks})
2358c533
KW
1112 && $opts_ref->{type} =~ / ^ (?: utf8 | high ) $ /xi)
1113 {
1114 my $lower_limit_is_80 = ($ranges[$i]->[0] == 0x80);
1115 my $upper_limit_is_BF = ($ranges[$i]->[1] == 0xBF);
1116
1117 # If the range is the entire legal range, it matches any legal
1118 # byte, so we can omit both tests. (This should happen only
1119 # if the number of ranges is 1.)
1120 if ($lower_limit_is_80 && $upper_limit_is_BF) {
1121 return 1;
6e130234 1122 }
2358c533
KW
1123 elsif ($lower_limit_is_80) { # Just use the upper limit test
1124 $output = sprintf("( $test <= $self->{val_fmt} )",
1125 $ranges[$i]->[1]);
f5772832 1126 }
2358c533
KW
1127 elsif ($upper_limit_is_BF) { # Just use the lower limit test
1128 $output = sprintf("( $test >= $self->{val_fmt} )",
1129 $ranges[$i]->[0]);
f5772832 1130 }
2358c533
KW
1131 }
1132
1133 # If we didn't change to omit a test above, see if the number of
1134 # elements is a power of 2 (only a single bit in the
1135 # representation of its count will be set) and if so, it may be
1136 # that a mask/compare optimization is possible.
1137 if ($output eq ""
1138 && pop_count($ranges[$i]->[1] - $ranges[$i]->[0] + 1) == 1)
1139 {
1140 my @list;
1141 push @list, $_ for ($ranges[$i]->[0] .. $ranges[$i]->[1]);
75929b4b
KW
1142 my @this_masks = calculate_mask(@list);
1143
1144 # Use the mask if there is just one for the whole range.
1145 # Otherwise there is no savings over the two branches that can
1146 # define the range.
1147 if (@this_masks == 1 && defined $this_masks[0][1]) {
1148 $output = sprintf "( $test & $self->{val_fmt} ) == $self->{val_fmt}", $this_masks[0][1], $this_masks[0][0];
f5772832
KW
1149 }
1150 }
2358c533
KW
1151
1152 if ($output ne "") { # Prefer any optimization
1153 $ranges[$i] = $output;
1154 }
75929b4b 1155 else {
2358c533
KW
1156 # No optimization happened. We need a test that the code
1157 # point is within both bounds. But, if the bounds are
1158 # adjacent code points, it is cleaner to say
1159 # 'first == test || second == test'
1160 # than it is to say
1161 # 'first <= test && test <= second'
75929b4b
KW
1162
1163 $range_count_extra++; # This range requires 2 branches to
1164 # represent
e2a80cb5
KW
1165 if ($ranges[$i]->[0] + 1 == $ranges[$i]->[1]) {
1166 $ranges[$i] = "( "
1167 . join( " || ", ( map
1168 { sprintf "$self->{val_fmt} == $test", $_ }
1169 @{$ranges[$i]} ) )
1170 . " )";
1171 }
1172 else { # Full bounds checking
1173 $ranges[$i] = sprintf("( $self->{val_fmt} <= $test && $test <= $self->{val_fmt} )", $ranges[$i]->[0], $ranges[$i]->[1]);
1174 }
75929b4b 1175 }
f5772832 1176 }
2358c533 1177 }
f5772832 1178
75929b4b
KW
1179 # We have generated the list of bytes in two ways; one trying to use masks
1180 # to cut the number of branches down, and the other to look at individual
1181 # ranges (some of which could be cut down by using a mask for just it).
1182 # We return whichever method uses the fewest branches.
1183 return "( "
1184 . join( " || ", (@masks && @masks < @ranges + $range_count_extra)
1185 ? @masks
1186 : @ranges)
1187 . " )";
12b72891
RGS
1188}
1189
e64b1bd1
YO
1190# _combine
1191# recursively turn a list of conditions into a fast break-out condition
1192# used by _cond_as_str() for 'cp' type macros.
1193sub _combine {
1194 my ( $self, $test, @cond )= @_;
1195 return if !@cond;
1196 my $item= shift @cond;
1197 my ( $cstr, $gtv );
6c62bf0f
KW
1198 if ( ref $item ) { # @item should be a 2-element array giving range start
1199 # and end
1200 if ($item->[0] == 0) { # UV's are never negative, so skip "0 <= "
1201 # test which could generate a compiler warning
1202 # that test is always true
1203 $cstr= sprintf( "$test <= $self->{val_fmt}", $item->[1] );
1204 }
1205 else {
1206 $cstr=
e64b1bd1 1207 sprintf( "( $self->{val_fmt} <= $test && $test <= $self->{val_fmt} )",
6c62bf0f
KW
1208 @$item );
1209 }
e64b1bd1 1210 $gtv= sprintf "$self->{val_fmt}", $item->[1];
12b72891 1211 } else {
e64b1bd1
YO
1212 $cstr= sprintf( "$self->{val_fmt} == $test", $item );
1213 $gtv= sprintf "$self->{val_fmt}", $item;
12b72891 1214 }
e64b1bd1 1215 if ( @cond ) {
ee98d22d
YO
1216 my $combine= $self->_combine( $test, @cond );
1217 if (@cond >1) {
1218 return "( $cstr || ( $gtv < $test &&\n"
1219 . $combine . " ) )";
1220 } else {
1221 return "( $cstr || $combine )";
1222 }
12b72891 1223 } else {
e64b1bd1 1224 return $cstr;
12b72891 1225 }
e64b1bd1 1226}
12b72891 1227
e64b1bd1
YO
1228# _render()
1229# recursively convert an optree to text with reasonably neat formatting
1230sub _render {
39a0f513 1231 my ( $self, $op, $combine, $brace, $opts_ref, $def, $submacros )= @_;
2e39f0c2 1232 return 0 if ! defined $op; # The set is empty
e64b1bd1
YO
1233 if ( !ref $op ) {
1234 return $op;
12b72891 1235 }
ba073cf2 1236 my $cond= $self->_cond_as_str( $op, $combine, $opts_ref );
cc08b31c
KW
1237 #no warnings 'recursion'; # This would allow really really inefficient
1238 # code to be generated. See pod
39a0f513 1239 my $yes= $self->_render( $op->{yes}, $combine, 1, $opts_ref, $def, $submacros );
30188af7
KW
1240 return $yes if $cond eq '1';
1241
39a0f513 1242 my $no= $self->_render( $op->{no}, $combine, 0, $opts_ref, $def, $submacros );
e64b1bd1
YO
1243 return "( $cond )" if $yes eq '1' and $no eq '0';
1244 my ( $lb, $rb )= $brace ? ( "( ", " )" ) : ( "", "" );
1245 return "$lb$cond ? $yes : $no$rb"
1246 if !ref( $op->{yes} ) && !ref( $op->{no} );
1247 my $ind1= " " x 4;
1248 my $ind= "\n" . ( $ind1 x $op->{depth} );
1249
1250 if ( ref $op->{yes} ) {
1251 $yes= $ind . $ind1 . $yes;
1252 } else {
1253 $yes= " " . $yes;
1254 }
1255
39a0f513
YO
1256 my $str= "$lb$cond ?$yes$ind: $no$rb";
1257 if (length $str > 6000) {
1258 push @$submacros, sprintf "#define $def\n( %s )", "_part" . (my $yes_idx= 0+@$submacros), $yes;
1259 push @$submacros, sprintf "#define $def\n( %s )", "_part" . (my $no_idx= 0+@$submacros), $no;
1260 return sprintf "%s%s ? $def : $def%s", $lb, $cond, "_part$yes_idx", "_part$no_idx", $rb;
1261 }
1262 return $str;
12b72891 1263}
32e6a07c 1264
e64b1bd1
YO
1265# $expr=render($op,$combine)
1266#
1267# convert an optree to text with reasonably neat formatting. If $combine
1268# is true then the condition is created using "fast breakouts" which
1269# produce uglier expressions that are more efficient for common case,
1270# longer lists such as that resulting from type 'cp' output.
1271# Currently only used for type 'cp' macros.
1272sub render {
39a0f513
YO
1273 my ( $self, $op, $combine, $opts_ref, $def_fmt )= @_;
1274
1275 my @submacros;
1276 my $macro= sprintf "#define $def_fmt\n( %s )", "", $self->_render( $op, $combine, 0, $opts_ref, $def_fmt, \@submacros );
1277
1278 return join "\n\n", map { "/*** GENERATED CODE ***/\n" . __macro( __clean( $_ ) ) } @submacros, $macro;
12b72891 1279}
e64b1bd1
YO
1280
1281# make_macro
1282# make a macro of a given type.
1283# calls into make_trie and (generic_|length_)optree as needed
1284# Opts are:
40f914fd
KW
1285# type : 'cp','cp_high', 'generic','high','low','latin1','utf8','LATIN1','UTF8'
1286# ret_type : 'cp' or 'len'
1287# safe : don't assume is well-formed UTF-8, so don't skip any range
1288# checks, and add length guards to macro
1289# no_length_checks : like safe, but don't add length guards.
e64b1bd1
YO
1290#
1291# type defaults to 'generic', and ret_type to 'len' unless type is 'cp'
1292# in which case it defaults to 'cp' as well.
1293#
3ff97bcf 1294# It is illegal to do a type 'cp' macro on a pattern with multi-codepoint
e64b1bd1
YO
1295# sequences in it, as the generated macro will accept only a single codepoint
1296# as an argument.
1297#
6b94381d
KW
1298# It is also illegal to do a non-safe macro on a pattern with multi-codepoint
1299# sequences in it, as even if it is known to be well-formed, we need to not
91e83b73 1300# run off the end of the buffer when, say, the buffer ends with the first two
6b94381d
KW
1301# characters, but three are looked at by the macro.
1302#
e64b1bd1
YO
1303# returns the macro.
1304
1305
1306sub make_macro {
1307 my $self= shift;
1308 my %opts= @_;
1309 my $type= $opts{type} || 'generic';
6b94381d
KW
1310 if ($self->{has_multi}) {
1311 if ($type =~ /^cp/) {
1312 die "Can't do a 'cp' on multi-codepoint character class '$self->{op}'"
1313 }
1314 elsif (! $opts{safe}) {
1315 die "'safe' is required on multi-codepoint character class '$self->{op}'"
1316 }
1317 }
900c17f9 1318 my $ret_type= $opts{ret_type} || ( $opts{type} =~ /^cp/ ? 'cp' : 'len' );
e64b1bd1
YO
1319 my $method;
1320 if ( $opts{safe} ) {
1321 $method= 'length_optree';
87894a24 1322 } elsif ( $type =~ /generic/ ) {
e64b1bd1
YO
1323 $method= 'generic_optree';
1324 } else {
1325 $method= 'optree';
1326 }
900c17f9 1327 my @args= $type =~ /^cp/ ? 'cp' : 's';
e64b1bd1 1328 push @args, "e" if $opts{safe};
87894a24 1329 push @args, "is_utf8" if $type =~ /generic/;
e64b1bd1
YO
1330 push @args, "len" if $ret_type eq 'both';
1331 my $pfx= $ret_type eq 'both' ? 'what_len_' :
1332 $ret_type eq 'cp' ? 'what_' : 'is_';
87894a24
KW
1333 my $ext= $type =~ /generic/ ? '' : '_' . lc( $type );
1334 $ext .= '_non_low' if $type eq 'generic_non_low';
e64b1bd1 1335 $ext .= "_safe" if $opts{safe};
40f914fd 1336 $ext .= "_no_length_checks" if $opts{no_length_checks};
e64b1bd1 1337 my $argstr= join ",", @args;
39a0f513
YO
1338 my $def_fmt="$pfx$self->{op}$ext%s($argstr)";
1339 my $optree= $self->$method( %opts, type => $type, ret_type => $ret_type );
1340 return $self->render( $optree, ($type =~ /^cp/) ? 1 : 0, \%opts, $def_fmt );
32e6a07c 1341}
e64b1bd1 1342
b6a6e956 1343# if we aren't being used as a module (highly likely) then process
e64b1bd1
YO
1344# the __DATA__ below and produce macros in regcharclass.h
1345# if an argument is provided to the script then it is assumed to
1346# be the path of the file to output to, if the arg is '-' outputs
1347# to STDOUT.
1348if ( !caller ) {
e64b1bd1 1349 $|++;
8770da0e 1350 my $path= shift @ARGV || "regcharclass.h";
e64b1bd1
YO
1351 my $out_fh;
1352 if ( $path eq '-' ) {
1353 $out_fh= \*STDOUT;
1354 } else {
29c22b52 1355 $out_fh = open_new( $path );
e64b1bd1 1356 }
8770da0e
NC
1357 print $out_fh read_only_top( lang => 'C', by => $0,
1358 file => 'regcharclass.h', style => '*',
212b6c86
KW
1359 copyright => [2007, 2011],
1360 final => <<EOF,
1361WARNING: These macros are for internal Perl core use only, and may be
1362changed or removed without notice.
1363EOF
1364 );
a1b2a50f 1365 print $out_fh "\n#ifndef H_REGCHARCLASS /* Guard against nested #includes */\n#define H_REGCHARCLASS 1\n";
12b72891 1366
bb949220 1367 my ( $op, $title, @txt, @types, %mods );
a1b2a50f 1368 my $doit= sub ($) {
e64b1bd1 1369 return unless $op;
ae1d4929 1370
a1b2a50f
KW
1371 my $charset = shift;
1372
ae1d4929 1373 # Skip if to compile on a different platform.
a1b2a50f
KW
1374 return if delete $mods{only_ascii_platform} && $charset !~ /ascii/i;
1375 return if delete $mods{only_ebcdic_platform} && $charset !~ /ebcdic/i;
ae1d4929 1376
e64b1bd1
YO
1377 print $out_fh "/*\n\t$op: $title\n\n";
1378 print $out_fh join "\n", ( map { "\t$_" } @txt ), "*/", "";
a1b2a50f 1379 my $obj= __PACKAGE__->new( op => $op, title => $title, txt => \@txt, charset => $charset);
e64b1bd1 1380
bb949220
KW
1381 #die Dumper(\@types,\%mods);
1382
1383 my @mods;
1384 push @mods, 'safe' if delete $mods{safe};
40f914fd 1385 push @mods, 'no_length_checks' if delete $mods{no_length_checks};
bb949220
KW
1386 unshift @mods, 'fast' if delete $mods{fast} || ! @mods; # Default to 'fast'
1387 # do this one
1388 # first, as
1389 # traditional
1390 if (%mods) {
122a2d8f 1391 die "Unknown modifiers: ", join ", ", map { "'$_'" } sort keys %mods;
bb949220 1392 }
e64b1bd1
YO
1393
1394 foreach my $type_spec ( @types ) {
1395 my ( $type, $ret )= split /-/, $type_spec;
1396 $ret ||= 'len';
1397 foreach my $mod ( @mods ) {
f71bd789
KW
1398
1399 # 'safe' is irrelevant with code point macros, so skip if
1400 # there is also a 'fast', but don't skip if this is the only
1401 # way a cp macro will get generated. Below we convert 'safe'
1402 # to 'fast' in this instance
1403 next if $type =~ /^cp/
40f914fd
KW
1404 && ($mod eq 'safe' || $mod eq 'no_length_checks')
1405 && grep { 'fast' =~ $_ } @mods;
bb949220 1406 delete $mods{$mod};
e64b1bd1
YO
1407 my $macro= $obj->make_macro(
1408 type => $type,
1409 ret_type => $ret,
81200454 1410 safe => $mod eq 'safe' && $type !~ /^cp/,
a1b2a50f 1411 charset => $charset,
40f914fd 1412 no_length_checks => $mod eq 'no_length_checks' && $type !~ /^cp/,
e64b1bd1
YO
1413 );
1414 print $out_fh $macro, "\n";
1415 }
32e6a07c 1416 }
e64b1bd1
YO
1417 };
1418
a1b2a50f
KW
1419 my @data = <DATA>;
1420 foreach my $charset (get_supported_code_pages()) {
1421 my $first_time = 1;
1422 undef $op;
1423 undef $title;
1424 undef @txt;
1425 undef @types;
1426 undef %mods;
1427 print $out_fh "\n", get_conditional_compile_line_start($charset);
1428 my @data_copy = @data;
1429 for (@data_copy) {
91e83b73
KW
1430 s/^ \s* (?: \# .* ) ? $ //x; # squeeze out comment and blanks
1431 next unless /\S/;
1432 chomp;
1433 if ( /^[A-Z]/ ) {
1434 $doit->($charset) unless $first_time; # This starts a new
1435 # definition; do the
1436 # previous one
1437 $first_time = 0;
1438 ( $op, $title )= split /\s*:\s*/, $_, 2;
1439 @txt= ();
1440 } elsif ( s/^=>// ) {
1441 my ( $type, $modifier )= split /:/, $_;
1442 @types= split ' ', $type;
1443 undef %mods;
1444 map { $mods{$_} = 1 } split ' ', $modifier;
1445 } else {
1446 push @txt, "$_";
1447 }
12b72891 1448 }
91e83b73 1449 $doit->($charset);
a1b2a50f
KW
1450 print $out_fh get_conditional_compile_line_end();
1451 }
d10c72f2
KW
1452
1453 print $out_fh "\n#endif /* H_REGCHARCLASS */\n";
1454
8770da0e
NC
1455 if($path eq '-') {
1456 print $out_fh "/* ex: set ro: */\n";
1457 } else {
0c36c41b
KW
1458 # Some of the sources for these macros come from Unicode tables
1459 my $sources_list = "lib/unicore/mktables.lst";
b60dc4b9
KW
1460 my @sources = ($0, qw(lib/unicore/mktables
1461 lib/Unicode/UCD.pm
1462 regen/regcharclass_multi_char_folds.pl
1463 regen/charset_translations.pl
1464 ));
0c36c41b
KW
1465 {
1466 # Depend on mktables’ own sources. It’s a shorter list of files than
1467 # those that Unicode::UCD uses.
1468 if (! open my $mktables_list, $sources_list) {
1469
1470 # This should force a rebuild once $sources_list exists
1471 push @sources, $sources_list;
1472 }
1473 else {
1474 while(<$mktables_list>) {
1475 last if /===/;
1476 chomp;
1477 push @sources, "lib/unicore/$_" if /^[^#]/;
1478 }
1479 }
1480 }
1481 read_only_bottom_close_and_rename($out_fh, \@sources)
8770da0e 1482 }
12b72891 1483}
e64b1bd1 1484
cc08b31c
KW
1485# The form of the input is a series of definitions to make macros for.
1486# The first line gives the base name of the macro, followed by a colon, and
1487# then text to be used in comments associated with the macro that are its
1488# title or description. In all cases the first (perhaps only) parameter to
1489# the macro is a pointer to the first byte of the code point it is to test to
1490# see if it is in the class determined by the macro. In the case of non-UTF8,
1491# the code point consists only of a single byte.
1492#
1493# The second line must begin with a '=>' and be followed by the types of
1494# macro(s) to be generated; these are specified below. A colon follows the
1495# types, followed by the modifiers, also specified below. At least one
1496# modifier is required.
1497#
1498# The subsequent lines give what code points go into the class defined by the
1499# macro. Multiple characters may be specified via a string like "\x0D\x0A",
60910c93
KW
1500# enclosed in quotes. Otherwise the lines consist of one of:
1501# 1) a single Unicode code point, prefaced by 0x
1502# 2) a single range of Unicode code points separated by a minus (and
1503# optional space)
1504# 3) a single Unicode property specified in the standard Perl form
1505# "\p{...}"
1506# 4) a line like 'do path'. This will do a 'do' on the file given by
1507# 'path'. It is assumed that this does nothing but load subroutines
1508# (See item 5 below). The reason 'require path' is not used instead is
1509# because 'do' doesn't assume that path is in @INC.
1510# 5) a subroutine call
1511# &pkg::foo(arg1, ...)
1512# where pkg::foo was loaded by a 'do' line (item 4). The subroutine
1513# returns an array of entries of forms like items 1-3 above. This
1514# allows more complex inputs than achievable from the other input types.
cc08b31c
KW
1515#
1516# A blank line or one whose first non-blank character is '#' is a comment.
1517# The definition of the macro is terminated by a line unlike those described.
1518#
1519# Valid types:
1520# low generate a macro whose name is 'is_BASE_low' and defines a
1521# class that includes only ASCII-range chars. (BASE is the
1522# input macro base name.)
1523# latin1 generate a macro whose name is 'is_BASE_latin1' and defines a
1524# class that includes only upper-Latin1-range chars. It is not
1525# designed to take a UTF-8 input parameter.
b1af8fef
KW
1526# high generate a macro whose name is 'is_BASE_high' and defines a
1527# class that includes all relevant code points that are above
1528# the Latin1 range. This is for very specialized uses only.
1529# It is designed to take only an input UTF-8 parameter.
cc08b31c
KW
1530# utf8 generate a macro whose name is 'is_BASE_utf8' and defines a
1531# class that includes all relevant characters that aren't ASCII.
1532# It is designed to take only an input UTF-8 parameter.
1533# LATIN1 generate a macro whose name is 'is_BASE_latin1' and defines a
1534# class that includes both ASCII and upper-Latin1-range chars.
1535# It is not designed to take a UTF-8 input parameter.
1536# UTF8 generate a macro whose name is 'is_BASE_utf8' and defines a
1537# class that can include any code point, adding the 'low' ones
1538# to what 'utf8' works on. It is designed to take only an input
1539# UTF-8 parameter.
1540# generic generate a macro whose name is 'is_BASE". It has a 2nd,
1541# boolean, parameter which indicates if the first one points to
1542# a UTF-8 string or not. Thus it works in all circumstances.
87894a24
KW
1543# generic_non_low generate a macro whose name is 'is_BASE_non_low". It has
1544# a 2nd, boolean, parameter which indicates if the first one
1545# points to a UTF-8 string or not. It excludes any ASCII-range
1546# matches, but otherwise it works in all circumstances.
cc08b31c
KW
1547# cp generate a macro whose name is 'is_BASE_cp' and defines a
1548# class that returns true if the UV parameter is a member of the
1549# class; false if not.
900c17f9
KW
1550# cp_high like cp, but it is assumed that it is known that the UV
1551# parameter is above Latin1. The name of the generated macro is
1552# 'is_BASE_cp_high'. This is different from high-cp, derived
1553# below.
cc08b31c
KW
1554# A macro of the given type is generated for each type listed in the input.
1555# The default return value is the number of octets read to generate the match.
1556# Append "-cp" to the type to have it instead return the matched codepoint.
1557# The macro name is changed to 'what_BASE...'. See pod for
1558# caveats
1559# Appending '-both" instead adds an extra parameter to the end of the argument
1560# list, which is a pointer as to where to store the number of
1561# bytes matched, while also returning the code point. The macro
1562# name is changed to 'what_len_BASE...'. See pod for caveats
1563#
1564# Valid modifiers:
1565# safe The input string is not necessarily valid UTF-8. In
1566# particular an extra parameter (always the 2nd) to the macro is
1567# required, which points to one beyond the end of the string.
1568# The macro will make sure not to read off the end of the
1569# string. In the case of non-UTF8, it makes sure that the
1570# string has at least one byte in it. The macro name has
1571# '_safe' appended to it.
40f914fd
KW
1572# no_length_checks The input string is not necessarily valid UTF-8, but it
1573# is to be assumed that the length has already been checked and
1574# found to be valid
cc08b31c
KW
1575# fast The input string is valid UTF-8. No bounds checking is done,
1576# and the macro can make assumptions that lead to faster
1577# execution.
a1b2a50f 1578# only_ascii_platform Skip this definition if the character set is for
ae1d4929 1579# a non-ASCII platform.
a1b2a50f 1580# only_ebcdic_platform Skip this definition if the character set is for
ae1d4929 1581# a non-EBCDIC platform.
cc08b31c
KW
1582# No modifier need be specified; fast is assumed for this case. If both
1583# 'fast', and 'safe' are specified, two macros will be created for each
1584# 'type'.
e90ac8de 1585#
295bcca9 1586# If run on a non-ASCII platform will automatically convert the Unicode input
cc08b31c
KW
1587# to native. The documentation above is slightly wrong in this case. 'low'
1588# actually refers to code points whose UTF-8 representation is the same as the
1589# non-UTF-8 version (invariants); and 'latin1' refers to all the rest of the
1590# code points less than 256.
5e6c6c1e
KW
1591
15921; # in the unlikely case we are being used as a module
1593
1594__DATA__
1595# This is no longer used, but retained in case it is needed some day.
e90ac8de
KW
1596# TRICKYFOLD: Problematic fold case letters. When adding to this list, also should add them to regcomp.c and fold_grind.t
1597# => generic cp generic-cp generic-both :fast safe
1598# 0x00DF # LATIN SMALL LETTER SHARP S
1599# 0x0390 # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
1600# 0x03B0 # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
1601# 0x1E9E # LATIN CAPITAL LETTER SHARP S, because maps to same as 00DF
1602# 0x1FD3 # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND OXIA; maps same as 0390
1603# 0x1FE3 # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND OXIA; maps same as 03B0
1604
12b72891 1605LNBREAK: Line Break: \R
5c025f03 1606=> generic UTF8 LATIN1 : safe
12b72891 1607"\x0D\x0A" # CRLF - Network (Windows) line ending
05b688d9 1608\p{VertSpace}
12b72891
RGS
1609
1610HORIZWS: Horizontal Whitespace: \h \H
507ce328 1611=> high cp_high : fast
05b688d9 1612\p{HorizSpace}
12b72891
RGS
1613
1614VERTWS: Vertical Whitespace: \v \V
507ce328 1615=> high cp_high : fast
05b688d9 1616\p{VertSpace}
612ead59 1617
4ac6419d 1618XDIGIT: Hexadecimal digits
507ce328 1619=> high cp_high : fast
4ac6419d
KW
1620\p{XDigit}
1621
bedac28b 1622XPERLSPACE: \p{XPerlSpace}
507ce328 1623=> high cp_high : fast
bedac28b
KW
1624\p{XPerlSpace}
1625
b96a92fb
KW
1626REPLACEMENT: Unicode REPLACEMENT CHARACTER
1627=> UTF8 :safe
16280xFFFD
1629
1630NONCHAR: Non character code points
1631=> UTF8 :fast
1632\p{Nchar}
1633
1634SURROGATE: Surrogate characters
1635=> UTF8 :fast
1636\p{Gc=Cs}
1637
4d646140
KW
1638# This program was run with this enabled, and the results copied to utf8.h;
1639# then this was commented out because it takes so long to figure out these 2
1640# million code points. The results would not change unless utf8.h decides it
1641# wants a maximum other than 4 bytes, or this program creates better
40f914fd
KW
1642# optimizations. Trying with 5 bytes used too much memory to calculate.
1643#
5dca9278
KW
1644# We don't generate code for invariants here because the EBCDIC form is too
1645# complicated and would slow things down; instead the user should test for
1646# invariants first.
1647#
40f914fd
KW
1648# NOTE: The number of bytes generated here must match the value in
1649# IS_UTF8_CHAR_FAST in utf8.h
1650#
5dca9278 1651#UTF8_CHAR: Matches legal UTF-8 encoded characters from 2 through 4 bytes
40f914fd 1652#=> UTF8 :no_length_checks only_ascii_platform
5dca9278 1653#0x80 - 0x1FFFFF
4d646140 1654
40f914fd
KW
1655# This hasn't been commented out, but the number of bytes it works on has been
1656# cut down to 3, so it doesn't cover the full legal Unicode range. Making it
1657# 5 bytes would cover beyond the full range, but takes quite a bit of time and
1658# memory to calculate. The generated table varies depending on the EBCDIC
1659# code page.
1660
1661# NOTE: The number of bytes generated here must match the value in
1662# IS_UTF8_CHAR_FAST in utf8.h
1663#
5dca9278 1664UTF8_CHAR: Matches legal UTF-EBCDIC encoded characters from 2 through 3 bytes
40f914fd 1665=> UTF8 :no_length_checks only_ebcdic_platform
5dca9278 16660xA0 - 0x3FFF
4d646140 1667
685289b5
KW
1668QUOTEMETA: Meta-characters that \Q should quote
1669=> high :fast
1670\p{_Perl_Quotemeta}
8769f413
KW
1671
1672MULTI_CHAR_FOLD: multi-char strings that are folded to by a single character
251b239f 1673=> UTF8 :safe
8769f413
KW
1674
1675# 1 => All folds
1676&regcharclass_multi_char_folds::multi_char_folds(1)
1677
40b1ba4f 1678MULTI_CHAR_FOLD: multi-char strings that are folded to by a single character
251b239f 1679=> LATIN1 : safe
8769f413 1680
8769f413 1681&regcharclass_multi_char_folds::multi_char_folds(0)
40b1ba4f 1682# 0 => Latin1-only
0b50d62a 1683
1a27eb96
KW
1684FOLDS_TO_MULTI: characters that fold to multi-char strings
1685=> UTF8 :fast
1686\p{_Perl_Folds_To_Multi_Char}
1687
31f05a37
KW
1688PROBLEMATIC_LOCALE_FOLD : characters whose fold is problematic under locale
1689=> UTF8 cp :fast
1690\p{_Perl_Problematic_Locale_Folds}
1691
1692PROBLEMATIC_LOCALE_FOLDEDS_START : The first folded character of folds which are problematic under locale
1693=> UTF8 cp :fast
1694\p{_Perl_Problematic_Locale_Foldeds_Start}
1695
0b50d62a 1696PATWS: pattern white space
8373491a 1697=> generic cp : safe
0b50d62a 1698\p{PatWS}