2 package CharClass::Matcher;
6 use warnings FATAL => 'all';
7 use Text::Wrap qw(wrap);
9 $Data::Dumper::Useqq= 1;
10 our $hex_fmt= "0x%02X";
15 sub ASCII_PLATFORM { (ord('A') == 65) }
17 require 'regen/regen_lib.pl';
21 CharClass::Matcher -- Generate C macros that match character classes efficiently
25 perl Porting/regcharclass.pl
29 Dynamically generates macros for detecting special charclasses
30 in latin-1, utf8, and codepoint forms. Macros can be set to return
31 the length (in bytes) of the matched codepoint, and/or the codepoint itself.
33 To regenerate F<regcharclass.h>, run this script from perl-root. No arguments
36 Using WHATEVER as an example the following macros can be produced, depending
37 on the input parameters (how to get each is described by internal comments at
38 the C<__DATA__> line):
42 =item C<is_WHATEVER(s,is_utf8)>
44 =item C<is_WHATEVER_safe(s,e,is_utf8)>
46 Do a lookup as appropriate based on the C<is_utf8> flag. When possible
47 comparisons involving octect<128 are done before checking the C<is_utf8>
48 flag, hopefully saving time.
50 The version without the C<_safe> suffix should be used only when the input is
51 known to be well-formed.
53 =item C<is_WHATEVER_utf8(s)>
55 =item C<is_WHATEVER_utf8_safe(s,e)>
57 Do a lookup assuming the string is encoded in (normalized) UTF8.
59 The version without the C<_safe> suffix should be used only when the input is
60 known to be well-formed.
62 =item C<is_WHATEVER_latin1(s)>
64 =item C<is_WHATEVER_latin1_safe(s,e)>
66 Do a lookup assuming the string is encoded in latin-1 (aka plan octets).
68 The version without the C<_safe> suffix should be used only when it is known
69 that C<s> contains at least one character.
71 =item C<is_WHATEVER_cp(cp)>
73 Check to see if the string matches a given codepoint (hypothetically a
74 U32). The condition is constructed as to "break out" as early as
75 possible if the codepoint is out of range of the condition.
79 (cp==X || (cp>X && (cp==Y || (cp>Y && ...))))
81 Thus if the character is X+1 only two comparisons will be done. Making
82 matching lookups slower, but non-matching faster.
84 =item C<what_len_WHATEVER_FOO(arg1, ..., len)>
86 A variant form of each of the macro types described above can be generated, in
87 which the code point is returned by the macro, and an extra parameter (in the
88 final position) is added, which is a pointer for the macro to set the byte
89 length of the returned code point.
91 These forms all have a C<what_len> prefix instead of the C<is_>, for example
92 C<what_len_WHATEVER_safe(s,e,is_utf8,len)> and
93 C<what_len_WHATEVER_utf8(s,len)>.
95 These forms should not be used I<except> on small sets of mostly widely
96 separated code points; otherwise the code generated is inefficient. For these
97 cases, it is best to use the C<is_> forms, and then find the code point with
98 C<utf8_to_uvchr_buf>(). This program can fail with a "deep recursion"
99 message on the worst of the inappropriate sets. Examine the generated macro
100 to see if it is acceptable.
102 =item C<what_WHATEVER_FOO(arg1, ...)>
104 A variant form of each of the C<is_> macro types described above can be generated, in
105 which the code point and not the length is returned by the macro. These have
106 the same caveat as L</what_len_WHATEVER_FOO(arg1, ..., len)>, plus they should
107 not be used where the set contains a NULL, as 0 is returned for two different
108 cases: a) the set doesn't include the input code point; b) the set does
109 include it, and it is a NULL.
115 perltidy -st -bt=1 -bbt=0 -pt=0 -sbt=1 -ce -nwls== "%f"
120 Author: Yves Orton (demerphq) 2007. Maintained by Perl5 Porters.
124 No tests directly here (although the regex engine will fail tests
125 if this code is broken). Insufficient documentation and no Getopts
126 handler for using the module as a script.
130 You may distribute under the terms of either the GNU General Public
131 License or the Artistic License, as specified in the README file.
135 # Sub naming convention:
136 # __func : private subroutine, can not be called as a method
137 # _func : private method, not meant for external use
138 # func : public method.
141 #-------------------------------------------------------------------------------
143 # ($cp,$n,$l,$u)=__uni_latin($str);
145 # Return a list of arrays, each of which when interpreted correctly
146 # represent the string in some given encoding with specific conditions.
148 # $cp - list of codepoints that make up the string.
149 # $n - list of octets that make up the string if all codepoints are invariant
150 # regardless of if the string is in UTF-8 or not.
151 # $l - list of octets that make up the string in latin1 encoding if all
152 # codepoints < 256, and at least one codepoint is UTF-8 variant.
153 # $u - list of octets that make up the string in utf8 if any codepoint is
157 #-----------+----------
158 # 0 - 127 : $n (127/128 are the values for ASCII platforms)
168 my $only_has_invariants = 1;
169 for my $ch ( split //, $str ) {
172 push @cp_high, $cp if $cp > 255;
173 $max= $cp if $max < $cp;
174 if (! ASCII_PLATFORM && $only_has_invariants) {
176 $only_has_invariants = 0;
180 utf8::upgrade($temp);
181 my @utf8 = unpack "U0C*", $temp;
182 $only_has_invariants = (@utf8 == 1 && $utf8[0] == $cp);
187 $only_has_invariants = $max < 128 if ASCII_PLATFORM;
188 if ($only_has_invariants) {
191 $l= [@cp] if $max && $max < 256;
195 $u= [ unpack "U0C*", $u ] if defined $u;
197 return ( \@cp, \@cp_high, $n, $l, $u );
201 # $clean= __clean($expr);
203 # Cleanup a ternary expression, removing unnecessary parens and apply some
204 # simplifications using regexes.
213 $parens= qr/ (?> \( (?> (?: (?> [^()]+ ) | (??{ $parens }) )* ) \) ) /x;
215 ## remove redundant parens
216 1 while $expr =~ s/ \( \s* ( $parens ) \s* \) /$1/gx;
219 # repeatedly simplify conditions like
220 # ( (cond1) ? ( (cond2) ? X : Y ) : Y )
222 # ( ( (cond1) && (cond2) ) ? X : Y )
223 # Also similarly handles expressions like:
224 # : (cond1) ? ( (cond2) ? X : Y ) : Y )
225 # Note the inclusion of the close paren in ([:()]) and the open paren in ([()]) is
226 # purely to ensure we have a balanced set of parens in the expression which makes
227 # it easier to understand the pattern in an editor that understands paren's, we do
228 # not expect either of these cases to actually fire. - Yves
234 \? \s* ($parens|[^()?:\s]+?) \s*
235 : \s* ($parens|[^()?:\s]+?) \s*
239 /$1 ( $2 && $3 ) ? $4 : $5 $6/gx;
240 #$expr=~s/\(\(U8\*\)s\)\[(\d+)\]/S$1/g if length $expr > 8000;
241 #$expr=~s/\s+//g if length $expr > 8000;
243 die "Expression too long" if length $expr > 8000;
249 # $text= __macro(@args);
250 # Join args together by newlines, and then neatly add backslashes to the end
251 # of every line as expected by the C pre-processor for #define's.
255 my $str= join "\n", @_;
257 my @lines= map { s/\s+$//; s/\t/ /g; $_ } split /\n/, $str;
258 my $last= pop @lines;
259 $str= join "\n", ( map { sprintf "%-76s\\", $_ } @lines ), $last;
260 1 while $str =~ s/^(\t*) {8}/$1\t/gm;
265 # my $op=__incrdepth($op);
267 # take an 'op' hashref and add one to it and all its childrens depths.
272 return unless ref $op;
274 __incrdepth( $op->{yes} );
275 __incrdepth( $op->{no} );
279 # join two branches of an opcode together with a condition, incrementing
280 # the depth on the yes branch when we do so.
281 # returns the new root opcode of the tree.
283 my ( $cond, $yes, $no )= @_;
286 yes => __incrdepth( $yes ),
296 # my $obj=CLASS->new(op=>'SOMENAME',title=>'blah',txt=>[..]);
298 # Create a new CharClass::Matcher object by parsing the text in
299 # the txt array. Currently applies the following rules:
301 # Element starts with C<0x>, line is evaled the result treated as
302 # a number which is passed to chr().
304 # Element starts with C<">, line is evaled and the result treated
307 # Each string is then stored in the 'strs' subhash as a hash record
308 # made up of the results of __uni_latin1, using the keynames
309 # 'low','latin1','utf8', as well as the synthesized 'LATIN1', 'high', and
310 # 'UTF8' which hold a merge of 'low' and their lowercase equivalents.
312 # Size data is tracked per type in the 'size' subhash.
320 die "in " . __PACKAGE__ . " constructor '$_;' is a mandatory field"
326 title => $opt{title} || '',
328 foreach my $txt ( @{ $opt{txt} } ) {
330 if ( $str =~ /^[""]/ ) {
332 } elsif ($str =~ / - /x ) { # A range: Replace this element on the
333 # list with its expansion
334 my ($lower, $upper) = $str =~ / 0x (.+?) \s* - \s* 0x (.+) /x;
335 die "Format must be like '0xDEAD - 0xBEAF'; instead was '$str'" if ! defined $lower || ! defined $upper;
336 foreach my $cp (hex $lower .. hex $upper) {
337 push @{$opt{txt}}, sprintf "0x%X", $cp;
340 } elsif ($str =~ s/ ^ N (?= 0x ) //x ) {
341 # Otherwise undocumented, a leading N means is already in the
342 # native character set; don't convert.
344 } elsif ( $str =~ /^0x/ ) {
347 # Convert from Unicode/ASCII to native, if necessary
348 $str = utf8::unicode_to_native($str) if ! ASCII_PLATFORM
351 } elsif ( $str =~ / \s* \\p \{ ( .*? ) \} /x) {
353 use Unicode::UCD qw(prop_invlist);
355 my @invlist = prop_invlist($property, '_perl_core_internal_ok');
358 # An empty return could mean an unknown property, or merely
359 # that it is empty. Call in scalar context to differentiate
360 my $count = prop_invlist($property, '_perl_core_internal_ok');
361 die "$property not found" unless defined $count;
364 # Replace this element on the list with the property's expansion
365 for (my $i = 0; $i < @invlist; $i += 2) {
366 foreach my $cp ($invlist[$i] .. $invlist[$i+1] - 1) {
368 # prop_invlist() returns native values; add leading 'N'
370 push @{$opt{txt}}, sprintf "N0x%X", $cp;
374 } elsif ($str =~ / ^ do \s+ ( .* ) /x) {
375 die "do '$1' failed: $!$@" if ! do $1 or $@;
377 } elsif ($str =~ / ^ & \s* ( .* ) /x) { # user-furnished sub() call
378 my @results = eval "$1";
379 die "eval '$1' failed: $@" if $@;
380 push @{$opt{txt}}, @results;
383 die "Unparsable line: $txt\n";
385 my ( $cp, $cp_high, $low, $latin1, $utf8 )= __uni_latin1( $str );
386 my $UTF8= $low || $utf8;
387 my $LATIN1= $low || $latin1;
388 my $high = (scalar grep { $_ < 256 } @$cp) ? 0 : $utf8;
389 #die Dumper($txt,$cp,$low,$latin1,$utf8)
390 # if $txt=~/NEL/ or $utf8 and @$utf8>3;
392 @{ $self->{strs}{$str} }{qw( str txt low utf8 latin1 high cp cp_high UTF8 LATIN1 )}=
393 ( $str, $txt, $low, $utf8, $latin1, $high, $cp, $cp_high, $UTF8, $LATIN1 );
394 my $rec= $self->{strs}{$str};
395 foreach my $key ( qw(low utf8 latin1 high cp cp_high UTF8 LATIN1) ) {
396 $self->{size}{$key}{ 0 + @{ $self->{strs}{$str}{$key} } }++
397 if $self->{strs}{$str}{$key};
399 $self->{has_multi} ||= @$cp > 1;
400 $self->{has_ascii} ||= $latin1 && @$latin1;
401 $self->{has_low} ||= $low && @$low;
402 $self->{has_high} ||= !$low && !$latin1;
404 $self->{val_fmt}= $hex_fmt;
405 $self->{count}= 0 + keys %{ $self->{strs} };
409 # my $trie = make_trie($type,$maxlen);
411 # using the data stored in the object build a trie of a specific type,
412 # and with specific maximum depth. The trie is made up the elements of
413 # the given types array for each string in the object (assuming it is
416 # returns the trie, or undef if there was no relevant data in the object.
420 my ( $self, $type, $maxlen )= @_;
422 my $strs= $self->{strs};
424 foreach my $rec ( values %$strs ) {
425 die "panic: unknown type '$type'"
426 if !exists $rec->{$type};
427 my $dat= $rec->{$type};
429 next if $maxlen && @$dat > $maxlen;
431 foreach my $elem ( @$dat ) {
432 $node->{$elem} ||= {};
433 $node= $node->{$elem};
435 $node->{''}= $rec->{str};
437 return 0 + keys( %trie ) ? \%trie : undef;
443 # This returns a list of the positions of the bits in the input word that
449 push @positions, $position if $word & 1;
456 # my $optree= _optree()
458 # recursively convert a trie to an optree where every node represents
464 my ( $self, $trie, $test_type, $ret_type, $else, $depth )= @_;
465 return unless defined $trie;
466 if ( $self->{has_multi} and $ret_type =~ /cp|both/ ) {
467 die "Can't do 'cp' optree from multi-codepoint strings";
470 $else= 0 unless defined $else;
471 $depth= 0 unless defined $depth;
473 # if we have an empty string as a key it means we are in an
474 # accepting state and unless we can match further on should
475 # return the value of the '' key.
476 if (exists $trie->{''} ) {
477 # we can now update the "else" value, anything failing to match
478 # after this point should return the value from this.
479 if ( $ret_type eq 'cp' ) {
480 $else= $self->{strs}{ $trie->{''} }{cp}[0];
481 $else= sprintf "$self->{val_fmt}", $else if $else > 9;
482 } elsif ( $ret_type eq 'len' ) {
484 } elsif ( $ret_type eq 'both') {
485 $else= $self->{strs}{ $trie->{''} }{cp}[0];
486 $else= sprintf "$self->{val_fmt}", $else if $else > 9;
487 $else= "len=$depth, $else";
490 # extract the meaningful keys from the trie, filter out '' as
491 # it means we are an accepting state (end of sequence).
492 my @conds= sort { $a <=> $b } grep { length $_ } keys %$trie;
494 # if we haven't any keys there is no further we can match and we
495 # can return the "else" value.
496 return $else if !@conds;
498 # Assuming Perl is being released from an ASCII platform, the below makes
499 # it work for non-UTF-8 out-of-the box when porting to non-ASCII, by
500 # adding a translation back to ASCII. This is the wrong thing to do for
501 # UTF-EBCDIC, as that is different from UTF-8. But the intent here is
502 # that this regen should be run on the target system, which will omit the
503 # translation, and generate the correct UTF-EBCDIC. On ASCII systems, the
504 # translation macros expand to just their argument, so there is no harm
505 # done nor performance penalty by including them.
507 if ($test_type =~ /^cp/) {
509 $test = "NATIVE_TO_UNI($test)" if ASCII_PLATFORM;
512 $test = "((U8*)s)[$depth]";
513 $test = "NATIVE_TO_LATIN1($test)" if ASCII_PLATFORM;
516 # first we loop over the possible keys/conditions and find out what they
517 # look like we group conditions with the same optree together.
520 local $Data::Dumper::Sortkeys=1;
521 foreach my $cond ( @conds ) {
523 # get the optree for this child/condition
524 my $res= $self->_optree( $trie->{$cond}, $test_type, $ret_type, $else, $depth + 1 );
525 # convert it to a string with Dumper
526 my $res_code= Dumper( $res );
528 push @{$dmp_res{$res_code}{vals}}, $cond;
529 if (!$dmp_res{$res_code}{optree}) {
530 $dmp_res{$res_code}{optree}= $res;
531 push @res_order, $res_code;
535 # now that we have deduped the optrees we construct a new optree containing the merged
539 foreach my $res_code_idx (0 .. $#res_order) {
540 my $res_code= $res_order[$res_code_idx];
541 $node->{vals}= $dmp_res{$res_code}{vals};
542 $node->{test}= $test;
543 $node->{yes}= $dmp_res{$res_code}{optree};
544 $node->{depth}= $depth;
545 if ($res_code_idx < $#res_order) {
546 $node= $node->{no}= {};
556 # my $optree= optree(%opts);
558 # Convert a trie to an optree, wrapper for _optree
563 my $trie= $self->make_trie( $opt{type}, $opt{max_depth} );
564 $opt{ret_type} ||= 'len';
565 my $test_type= $opt{type} =~ /^cp/ ? 'cp' : 'depth';
566 return $self->_optree( $trie, $test_type, $opt{ret_type}, $opt{else}, 0 );
569 # my $optree= generic_optree(%opts);
571 # build a "generic" optree out of the three 'low', 'latin1', 'utf8'
572 # sets of strings, including a branch for handling the string type check.
579 $opt{ret_type} ||= 'len';
580 my $test_type= 'depth';
581 my $else= $opt{else} || 0;
583 my $latin1= $self->make_trie( 'latin1', $opt{max_depth} );
584 my $utf8= $self->make_trie( 'utf8', $opt{max_depth} );
586 $_= $self->_optree( $_, $test_type, $opt{ret_type}, $else, 0 )
590 $else= __cond_join( "( is_utf8 )", $utf8, $latin1 || $else );
591 } elsif ( $latin1 ) {
592 $else= __cond_join( "!( is_utf8 )", $latin1, $else );
594 if ($opt{type} eq 'generic') {
595 my $low= $self->make_trie( 'low', $opt{max_depth} );
597 $else= $self->_optree( $low, $test_type, $opt{ret_type}, $else, 0 );
606 # create a string length guarded optree.
612 my $type= $opt{type};
614 die "Can't do a length_optree on type 'cp', makes no sense."
617 my ( @size, $method );
619 if ( $type =~ /generic/ ) {
620 $method= 'generic_optree';
622 %{ $self->{size}{low} || {} },
623 %{ $self->{size}{latin1} || {} },
624 %{ $self->{size}{utf8} || {} }
626 @size= sort { $a <=> $b } keys %sizes;
629 @size= sort { $a <=> $b } keys %{ $self->{size}{$type} };
632 my $else= ( $opt{else} ||= 0 );
633 for my $size ( @size ) {
634 my $optree= $self->$method( %opt, type => $type, max_depth => $size );
635 my $cond= "((e)-(s) > " . ( $size - 1 ).")";
636 $else= __cond_join( $cond, $optree, $else );
641 sub calculate_mask(@) {
642 # Look at the input list of byte values. This routine returns an array of
643 # mask/base pairs to generate that list.
646 my $list_count = @list;
648 # Consider a set of byte values, A, B, C .... If we want to determine if
649 # <c> is one of them, we can write c==A || c==B || c==C .... If the
650 # values are consecutive, we can shorten that to A<=c && c<=Z, which uses
651 # far fewer branches. If only some of them are consecutive we can still
652 # save some branches by creating range tests for just those that are
653 # consecutive. _cond_as_str() does this work for looking for ranges.
655 # Another approach is to look at the bit patterns for A, B, C .... and see
656 # if they have some commonalities. That's what this function does. For
657 # example, consider a set consisting of the bytes
658 # 0xF0, 0xF1, 0xF2, and 0xF3. We could write:
659 # 0xF0 <= c && c <= 0xF4
660 # But the following mask/compare also works, and has just one test:
662 # The reason it works is that the set consists of exactly those bytes
663 # whose first 4 bits are 1, and the next two are 0. (The value of the
664 # other 2 bits is immaterial in determining if a byte is in the set or
665 # not.) The mask masks out those 2 irrelevant bits, and the comparison
666 # makes sure that the result matches all bytes which match those 6
667 # material bits exactly. In other words, the set of bytes contains
668 # exactly those whose bottom two bit positions are either 0 or 1. The
669 # same principle applies to bit positions that are not necessarily
670 # adjacent. And it can be applied to bytes that differ in 1 through all 8
671 # bit positions. In order to be a candidate for this optimization, the
672 # number of bytes in the set must be a power of 2.
674 # Consider a different example, the set 0x53, 0x54, 0x73, and 0x74. That
675 # requires 4 tests using either ranges or individual values, and even
676 # though the number in the set is a power of 2, it doesn't qualify for the
677 # mask optimization described above because the number of bits that are
678 # different is too large for that. However, the set can be expressed as
679 # two branches with masks thusly:
680 # (c & 0xDF) == 0x53 || (c & 0xDF) == 0x54
681 # a branch savings of 50%. This is done by splitting the set into two
682 # subsets each of which has 2 elements, and within each set the values
685 # This function attempts to find some way to save some branches using the
686 # mask technique. If not, it returns an empty list; if so, it
687 # returns a list consisting of
688 # [ [compare1, mask1], [compare2, mask2], ...
689 # [compare_n, undef], [compare_m, undef], ...
691 # The <mask> is undef in the above for those bytes that must be tested
694 # This function does not attempt to find the optimal set. To do so would
695 # probably require testing all possible combinations, and keeping track of
696 # the current best one.
698 # There are probably much better algorithms, but this is the one I (khw)
699 # came up with. We start with doing a bit-wise compare of every byte in
700 # the set with every other byte. The results are sorted into arrays of
701 # all those that differ by the same bit positions. These are stored in a
702 # hash with the each key being the bits they differ in. Here is the hash
703 # for the 0x53, 0x54, 0x73, 0x74 set:
731 # The set consisting of values which differ in the 4 bit positions 0, 1,
732 # 2, and 5 from some other value in the set consists of all 4 values.
733 # Likewise all 4 values differ from some other value in the 3 bit
734 # positions 0, 1, and 2; and all 4 values differ from some other value in
735 # the single bit position 5. The keys at the uppermost level in the above
736 # hash, 1, 3, and 4, give the number of bit positions that each sub-key
737 # below it has. For example, the 4 key could have as its value an array
738 # consisting of "0,1,2,5", "0,1,2,6", and "3,4,6,7", if the inputs were
739 # such. The best optimization will group the most values into a single
740 # mask. The most values will be the ones that differ in the most
741 # positions, the ones with the largest value for the topmost key. These
742 # keys, are thus just for convenience of sorting by that number, and do
743 # not have any bearing on the core of the algorithm.
745 # We start with an element from largest number of differing bits. The
746 # largest in this case is 4 bits, and there is only one situation in this
747 # set which has 4 differing bits, "0,1,2,5". We look for any subset of
748 # this set which has 16 values that differ in these 4 bits. There aren't
749 # any, because there are only 4 values in the entire set. We then look at
750 # the next possible thing, which is 3 bits differing in positions "0,1,2".
751 # We look for a subset that has 8 values that differ in these 3 bits.
752 # Again there are none. So we go to look for the next possible thing,
753 # which is a subset of 2**1 values that differ only in bit position 5. 83
754 # and 115 do, so we calculate a mask and base for those and remove them
755 # from every set. Since there is only the one set remaining, we remove
756 # them from just this one. We then look to see if there is another set of
757 # 2 values that differ in bit position 5. 84 and 116 do, so we calculate
758 # a mask and base for those and remove them from every set (again only
759 # this set remains in this example). The set is now empty, and there are
760 # no more sets to look at, so we are done.
762 if ($list_count == 256) { # All 256 is trivially masked
768 # Generate bits-differing lists for each element compared against each
770 for my $i (0 .. $list_count - 2) {
771 for my $j ($i + 1 .. $list_count - 1) {
772 my @bits_that_differ = pop_count($list[$i] ^ $list[$j]);
773 my $differ_count = @bits_that_differ;
774 my $key = join ",", @bits_that_differ;
775 push @{$hash{$differ_count}{$key}}, $list[$i] unless grep { $_ == $list[$i] } @{$hash{$differ_count}{$key}};
776 push @{$hash{$differ_count}{$key}}, $list[$j];
780 print STDERR __LINE__, ": calculate_mask() called: List of values grouped by differing bits: ", Dumper \%hash if DEBUG;
783 foreach my $count (reverse sort { $a <=> $b } keys %hash) {
784 my $need = 2 ** $count; # Need 8 values for 3 differing bits, etc
785 foreach my $bits (sort keys $hash{$count}) {
787 print STDERR __LINE__, ": For $count bit(s) difference ($bits), need $need; have ", scalar @{$hash{$count}{$bits}}, "\n" if DEBUG;
789 # Look only as long as there are at least as many elements in the
790 # subset as are needed
791 while ((my $cur_count = @{$hash{$count}{$bits}}) >= $need) {
793 print STDERR __LINE__, ": Looking at bit positions ($bits): ", Dumper $hash{$count}{$bits} if DEBUG;
795 # Start with the first element in it
796 my $try_base = $hash{$count}{$bits}[0];
797 my @subset = $try_base;
799 # If it succeeds, we return a mask and a base to compare
800 # against the masked value. That base will be the AND of
801 # every element in the subset. Initialize to the one element
803 my $compare = $try_base;
805 # We are trying to find a subset of this that has <need>
806 # elements that differ in the bit positions given by the
807 # string $bits, which is comma separated.
808 my @bits = split ",", $bits;
810 TRY: # Look through the remainder of the list for other
811 # elements that differ only by these bit positions.
813 for (my $i = 1; $i < $cur_count; $i++) {
814 my $try_this = $hash{$count}{$bits}[$i];
815 my @positions = pop_count($try_base ^ $try_this);
817 print STDERR __LINE__, ": $try_base vs $try_this: is (", join(',', @positions), ") a subset of ($bits)?" if DEBUG;;
819 foreach my $pos (@positions) {
820 unless (grep { $pos == $_ } @bits) {
821 print STDERR " No\n" if DEBUG;
822 my $remaining = $cur_count - $i - 1;
823 if ($remaining && @subset + $remaining < $need) {
824 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;
831 print STDERR " Yes\n" if DEBUG;
832 push @subset, $try_this;
834 # Add this to the mask base, in case it ultimately
836 $compare &= $try_this;
839 print STDERR __LINE__, ": subset (", join(", ", @subset), ") has ", scalar @subset, " elements; needs $need\n" if DEBUG;
841 if (@subset < $need) {
842 shift @{$hash{$count}{$bits}};
843 next; # Try with next value
848 foreach my $position (@bits) {
849 $mask |= 1 << $position;
851 $mask = ~$mask & 0xFF;
852 push @final_results, [$compare, $mask];
854 printf STDERR "%d: Got it: compare=%d=0x%X; mask=%X\n", __LINE__, $compare, $compare, $mask if DEBUG;
856 # These values are now spoken for. Remove them from future
858 foreach my $remove_count (sort keys %hash) {
859 foreach my $bits (sort keys %{$hash{$remove_count}}) {
860 foreach my $to_remove (@subset) {
861 @{$hash{$remove_count}{$bits}} = grep { $_ != $to_remove } @{$hash{$remove_count}{$bits}};
869 # Any values that remain in the list are ones that have to be tested for
872 foreach my $count (reverse sort { $a <=> $b } keys %hash) {
873 foreach my $bits (sort keys $hash{$count}) {
874 foreach my $remaining (@{$hash{$count}{$bits}}) {
876 # If we already know about this value, just ignore it.
877 next if grep { $remaining == $_ } @individuals;
879 # Otherwise it needs to be returned as something to match
881 push @final_results, [$remaining, undef];
882 push @individuals, $remaining;
887 # Sort by increasing numeric value
888 @final_results = sort { $a->[0] <=> $b->[0] } @final_results;
890 print STDERR __LINE__, ": Final return: ", Dumper \@final_results if DEBUG;
892 return @final_results;
896 # turn a list of conditions into a text expression
897 # - merges ranges of conditions, and joins the result with ||
899 my ( $self, $op, $combine, $opts_ref )= @_;
900 my $cond= $op->{vals};
901 my $test= $op->{test};
902 my $is_cp_ret = $opts_ref->{ret_type} eq "cp";
903 return "( $test )" if !defined $cond;
908 # We skip this if there are optimizations that
909 # we can apply (below) to the individual ranges
910 if ( ($is_cp_ret || $combine) && @ranges && ref $ranges[-1]) {
911 if ( $ranges[-1][0] == $ranges[-1][1] ) {
912 $ranges[-1]= $ranges[-1][0];
913 } elsif ( $ranges[-1][0] + 1 == $ranges[-1][1] ) {
914 $ranges[-1]= $ranges[-1][0];
915 push @ranges, $ranges[-1] + 1;
919 for my $condition ( @$cond ) {
920 if ( !@ranges || $condition != $ranges[-1][1] + 1 ) {
922 push @ranges, [ $condition, $condition ];
929 return $self->_combine( $test, @ranges )
936 "( $self->{val_fmt} <= $test && $test <= $self->{val_fmt} )",
938 : sprintf( "$self->{val_fmt} == $test", $_ );
941 return "( " . join( " || ", @ranges ) . " )";
944 # If the input set has certain characteristics, we can optimize tests
945 # for it. This doesn't apply if returning the code point, as we want
946 # each element of the set individually. The code above is for this
949 return 1 if @$cond == 256; # If all bytes match, is trivially true
954 # See if the entire set shares optimizable characteristics, and if so,
955 # return the optimization. We delay checking for this on sets with
956 # just a single range, as there may be better optimizations available
958 @masks = calculate_mask(@$cond);
960 # Stringify the output of calculate_mask()
963 foreach my $mask_ref (@masks) {
964 if (defined $mask_ref->[1]) {
965 push @return, sprintf "( ( $test & $self->{val_fmt} ) == $self->{val_fmt} )", $mask_ref->[1], $mask_ref->[0];
967 else { # An undefined mask means to use the value as-is
968 push @return, sprintf "$test == $self->{val_fmt}", $mask_ref->[0];
972 # The best possible case below for specifying this set of values via
973 # ranges is 1 branch per range. If our mask method yielded better
974 # results, there is no sense trying something that is bound to be
976 if (@return < @ranges) {
977 return "( " . join( " || ", @return ) . " )";
984 # Here, there was no entire-class optimization that was clearly better
985 # than doing things by ranges. Look at each range.
986 my $range_count_extra = 0;
987 for (my $i = 0; $i < @ranges; $i++) {
988 if (! ref $ranges[$i]) { # Trivial case: no range
989 $ranges[$i] = sprintf "$self->{val_fmt} == $test", $ranges[$i];
991 elsif ($ranges[$i]->[0] == $ranges[$i]->[1]) {
992 $ranges[$i] = # Trivial case: single element range
993 sprintf "$self->{val_fmt} == $test", $ranges[$i]->[0];
998 # Well-formed UTF-8 continuation bytes on ascii platforms must be
999 # in the range 0x80 .. 0xBF. If we know that the input is
1000 # well-formed (indicated by not trying to be 'safe'), we can omit
1001 # tests that verify that the input is within either of these
1002 # bounds. (No legal UTF-8 character can begin with anything in
1003 # this range, so we don't have to worry about this being a
1004 # continuation byte or not.)
1006 && ! $opts_ref->{safe}
1007 && $opts_ref->{type} =~ / ^ (?: utf8 | high ) $ /xi)
1009 my $lower_limit_is_80 = ($ranges[$i]->[0] == 0x80);
1010 my $upper_limit_is_BF = ($ranges[$i]->[1] == 0xBF);
1012 # If the range is the entire legal range, it matches any legal
1013 # byte, so we can omit both tests. (This should happen only
1014 # if the number of ranges is 1.)
1015 if ($lower_limit_is_80 && $upper_limit_is_BF) {
1018 elsif ($lower_limit_is_80) { # Just use the upper limit test
1019 $output = sprintf("( $test <= $self->{val_fmt} )",
1022 elsif ($upper_limit_is_BF) { # Just use the lower limit test
1023 $output = sprintf("( $test >= $self->{val_fmt} )",
1028 # If we didn't change to omit a test above, see if the number of
1029 # elements is a power of 2 (only a single bit in the
1030 # representation of its count will be set) and if so, it may be
1031 # that a mask/compare optimization is possible.
1033 && pop_count($ranges[$i]->[1] - $ranges[$i]->[0] + 1) == 1)
1036 push @list, $_ for ($ranges[$i]->[0] .. $ranges[$i]->[1]);
1037 my @this_masks = calculate_mask(@list);
1039 # Use the mask if there is just one for the whole range.
1040 # Otherwise there is no savings over the two branches that can
1042 if (@this_masks == 1 && defined $this_masks[0][1]) {
1043 $output = sprintf "( $test & $self->{val_fmt} ) == $self->{val_fmt}", $this_masks[0][1], $this_masks[0][0];
1047 if ($output ne "") { # Prefer any optimization
1048 $ranges[$i] = $output;
1051 # No optimization happened. We need a test that the code
1052 # point is within both bounds. But, if the bounds are
1053 # adjacent code points, it is cleaner to say
1054 # 'first == test || second == test'
1056 # 'first <= test && test <= second'
1058 $range_count_extra++; # This range requires 2 branches to
1060 if ($ranges[$i]->[0] + 1 == $ranges[$i]->[1]) {
1062 . join( " || ", ( map
1063 { sprintf "$self->{val_fmt} == $test", $_ }
1067 else { # Full bounds checking
1068 $ranges[$i] = sprintf("( $self->{val_fmt} <= $test && $test <= $self->{val_fmt} )", $ranges[$i]->[0], $ranges[$i]->[1]);
1074 # We have generated the list of bytes in two ways; one trying to use masks
1075 # to cut the number of branches down, and the other to look at individual
1076 # ranges (some of which could be cut down by using a mask for just it).
1077 # We return whichever method uses the fewest branches.
1079 . join( " || ", (@masks && @masks < @ranges + $range_count_extra)
1086 # recursively turn a list of conditions into a fast break-out condition
1087 # used by _cond_as_str() for 'cp' type macros.
1089 my ( $self, $test, @cond )= @_;
1091 my $item= shift @cond;
1095 sprintf( "( $self->{val_fmt} <= $test && $test <= $self->{val_fmt} )",
1097 $gtv= sprintf "$self->{val_fmt}", $item->[1];
1099 $cstr= sprintf( "$self->{val_fmt} == $test", $item );
1100 $gtv= sprintf "$self->{val_fmt}", $item;
1103 my $combine= $self->_combine( $test, @cond );
1105 return "( $cstr || ( $gtv < $test &&\n"
1106 . $combine . " ) )";
1108 return "( $cstr || $combine )";
1116 # recursively convert an optree to text with reasonably neat formatting
1118 my ( $self, $op, $combine, $brace, $opts_ref, $def, $submacros )= @_;
1119 return 0 if ! defined $op; # The set is empty
1123 my $cond= $self->_cond_as_str( $op, $combine, $opts_ref );
1124 #no warnings 'recursion'; # This would allow really really inefficient
1125 # code to be generated. See pod
1126 my $yes= $self->_render( $op->{yes}, $combine, 1, $opts_ref, $def, $submacros );
1127 return $yes if $cond eq '1';
1129 my $no= $self->_render( $op->{no}, $combine, 0, $opts_ref, $def, $submacros );
1130 return "( $cond )" if $yes eq '1' and $no eq '0';
1131 my ( $lb, $rb )= $brace ? ( "( ", " )" ) : ( "", "" );
1132 return "$lb$cond ? $yes : $no$rb"
1133 if !ref( $op->{yes} ) && !ref( $op->{no} );
1135 my $ind= "\n" . ( $ind1 x $op->{depth} );
1137 if ( ref $op->{yes} ) {
1138 $yes= $ind . $ind1 . $yes;
1143 my $str= "$lb$cond ?$yes$ind: $no$rb";
1144 if (length $str > 6000) {
1145 push @$submacros, sprintf "#define $def\n( %s )", "_part" . (my $yes_idx= 0+@$submacros), $yes;
1146 push @$submacros, sprintf "#define $def\n( %s )", "_part" . (my $no_idx= 0+@$submacros), $no;
1147 return sprintf "%s%s ? $def : $def%s", $lb, $cond, "_part$yes_idx", "_part$no_idx", $rb;
1152 # $expr=render($op,$combine)
1154 # convert an optree to text with reasonably neat formatting. If $combine
1155 # is true then the condition is created using "fast breakouts" which
1156 # produce uglier expressions that are more efficient for common case,
1157 # longer lists such as that resulting from type 'cp' output.
1158 # Currently only used for type 'cp' macros.
1160 my ( $self, $op, $combine, $opts_ref, $def_fmt )= @_;
1163 my $macro= sprintf "#define $def_fmt\n( %s )", "", $self->_render( $op, $combine, 0, $opts_ref, $def_fmt, \@submacros );
1165 return join "\n\n", map { "/*** GENERATED CODE ***/\n" . __macro( __clean( $_ ) ) } @submacros, $macro;
1169 # make a macro of a given type.
1170 # calls into make_trie and (generic_|length_)optree as needed
1172 # type : 'cp','cp_high', 'generic','high','low','latin1','utf8','LATIN1','UTF8'
1173 # ret_type : 'cp' or 'len'
1174 # safe : add length guards to macro
1176 # type defaults to 'generic', and ret_type to 'len' unless type is 'cp'
1177 # in which case it defaults to 'cp' as well.
1179 # it is illegal to do a type 'cp' macro on a pattern with multi-codepoint
1180 # sequences in it, as the generated macro will accept only a single codepoint
1183 # returns the macro.
1189 my $type= $opts{type} || 'generic';
1190 die "Can't do a 'cp' on multi-codepoint character class '$self->{op}'"
1192 and $self->{has_multi};
1193 my $ret_type= $opts{ret_type} || ( $opts{type} =~ /^cp/ ? 'cp' : 'len' );
1195 if ( $opts{safe} ) {
1196 $method= 'length_optree';
1197 } elsif ( $type =~ /generic/ ) {
1198 $method= 'generic_optree';
1202 my @args= $type =~ /^cp/ ? 'cp' : 's';
1203 push @args, "e" if $opts{safe};
1204 push @args, "is_utf8" if $type =~ /generic/;
1205 push @args, "len" if $ret_type eq 'both';
1206 my $pfx= $ret_type eq 'both' ? 'what_len_' :
1207 $ret_type eq 'cp' ? 'what_' : 'is_';
1208 my $ext= $type =~ /generic/ ? '' : '_' . lc( $type );
1209 $ext .= '_non_low' if $type eq 'generic_non_low';
1210 $ext .= "_safe" if $opts{safe};
1211 my $argstr= join ",", @args;
1212 my $def_fmt="$pfx$self->{op}$ext%s($argstr)";
1213 my $optree= $self->$method( %opts, type => $type, ret_type => $ret_type );
1214 return $self->render( $optree, ($type =~ /^cp/) ? 1 : 0, \%opts, $def_fmt );
1217 # if we aren't being used as a module (highly likely) then process
1218 # the __DATA__ below and produce macros in regcharclass.h
1219 # if an argument is provided to the script then it is assumed to
1220 # be the path of the file to output to, if the arg is '-' outputs
1224 my $path= shift @ARGV || "regcharclass.h";
1226 if ( $path eq '-' ) {
1229 $out_fh = open_new( $path );
1231 print $out_fh read_only_top( lang => 'C', by => $0,
1232 file => 'regcharclass.h', style => '*',
1233 copyright => [2007, 2011] );
1234 print $out_fh "\n#ifndef H_REGCHARCLASS /* Guard against nested #includes */\n#define H_REGCHARCLASS 1\n\n";
1236 my ( $op, $title, @txt, @types, %mods );
1240 # Skip if to compile on a different platform.
1241 return if delete $mods{only_ascii_platform} && ! ASCII_PLATFORM;
1242 return if delete $mods{only_ebcdic_platform} && ord 'A' != 193;
1244 print $out_fh "/*\n\t$op: $title\n\n";
1245 print $out_fh join "\n", ( map { "\t$_" } @txt ), "*/", "";
1246 my $obj= __PACKAGE__->new( op => $op, title => $title, txt => \@txt );
1248 #die Dumper(\@types,\%mods);
1251 push @mods, 'safe' if delete $mods{safe};
1252 unshift @mods, 'fast' if delete $mods{fast} || ! @mods; # Default to 'fast'
1257 die "Unknown modifiers: ", join ", ", map { "'$_'" } sort keys %mods;
1260 foreach my $type_spec ( @types ) {
1261 my ( $type, $ret )= split /-/, $type_spec;
1263 foreach my $mod ( @mods ) {
1264 next if $mod eq 'safe' and $type =~ /^cp/;
1266 my $macro= $obj->make_macro(
1269 safe => $mod eq 'safe'
1271 print $out_fh $macro, "\n";
1277 s/^ \s* (?: \# .* ) ? $ //x; # squeeze out comment and blanks
1281 $doit->(); # This starts a new definition; do the previous one
1282 ( $op, $title )= split /\s*:\s*/, $_, 2;
1284 } elsif ( s/^=>// ) {
1285 my ( $type, $modifier )= split /:/, $_;
1286 @types= split ' ', $type;
1288 map { $mods{$_} = 1 } split ' ', $modifier;
1295 print $out_fh "\n#endif /* H_REGCHARCLASS */\n";
1298 print $out_fh "/* ex: set ro: */\n";
1300 read_only_bottom_close_and_rename($out_fh)
1304 # The form of the input is a series of definitions to make macros for.
1305 # The first line gives the base name of the macro, followed by a colon, and
1306 # then text to be used in comments associated with the macro that are its
1307 # title or description. In all cases the first (perhaps only) parameter to
1308 # the macro is a pointer to the first byte of the code point it is to test to
1309 # see if it is in the class determined by the macro. In the case of non-UTF8,
1310 # the code point consists only of a single byte.
1312 # The second line must begin with a '=>' and be followed by the types of
1313 # macro(s) to be generated; these are specified below. A colon follows the
1314 # types, followed by the modifiers, also specified below. At least one
1315 # modifier is required.
1317 # The subsequent lines give what code points go into the class defined by the
1318 # macro. Multiple characters may be specified via a string like "\x0D\x0A",
1319 # enclosed in quotes. Otherwise the lines consist of one of:
1320 # 1) a single Unicode code point, prefaced by 0x
1321 # 2) a single range of Unicode code points separated by a minus (and
1323 # 3) a single Unicode property specified in the standard Perl form
1325 # 4) a line like 'do path'. This will do a 'do' on the file given by
1326 # 'path'. It is assumed that this does nothing but load subroutines
1327 # (See item 5 below). The reason 'require path' is not used instead is
1328 # because 'do' doesn't assume that path is in @INC.
1329 # 5) a subroutine call
1330 # &pkg::foo(arg1, ...)
1331 # where pkg::foo was loaded by a 'do' line (item 4). The subroutine
1332 # returns an array of entries of forms like items 1-3 above. This
1333 # allows more complex inputs than achievable from the other input types.
1335 # A blank line or one whose first non-blank character is '#' is a comment.
1336 # The definition of the macro is terminated by a line unlike those described.
1339 # low generate a macro whose name is 'is_BASE_low' and defines a
1340 # class that includes only ASCII-range chars. (BASE is the
1341 # input macro base name.)
1342 # latin1 generate a macro whose name is 'is_BASE_latin1' and defines a
1343 # class that includes only upper-Latin1-range chars. It is not
1344 # designed to take a UTF-8 input parameter.
1345 # high generate a macro whose name is 'is_BASE_high' and defines a
1346 # class that includes all relevant code points that are above
1347 # the Latin1 range. This is for very specialized uses only.
1348 # It is designed to take only an input UTF-8 parameter.
1349 # utf8 generate a macro whose name is 'is_BASE_utf8' and defines a
1350 # class that includes all relevant characters that aren't ASCII.
1351 # It is designed to take only an input UTF-8 parameter.
1352 # LATIN1 generate a macro whose name is 'is_BASE_latin1' and defines a
1353 # class that includes both ASCII and upper-Latin1-range chars.
1354 # It is not designed to take a UTF-8 input parameter.
1355 # UTF8 generate a macro whose name is 'is_BASE_utf8' and defines a
1356 # class that can include any code point, adding the 'low' ones
1357 # to what 'utf8' works on. It is designed to take only an input
1359 # generic generate a macro whose name is 'is_BASE". It has a 2nd,
1360 # boolean, parameter which indicates if the first one points to
1361 # a UTF-8 string or not. Thus it works in all circumstances.
1362 # generic_non_low generate a macro whose name is 'is_BASE_non_low". It has
1363 # a 2nd, boolean, parameter which indicates if the first one
1364 # points to a UTF-8 string or not. It excludes any ASCII-range
1365 # matches, but otherwise it works in all circumstances.
1366 # cp generate a macro whose name is 'is_BASE_cp' and defines a
1367 # class that returns true if the UV parameter is a member of the
1368 # class; false if not.
1369 # cp_high like cp, but it is assumed that it is known that the UV
1370 # parameter is above Latin1. The name of the generated macro is
1371 # 'is_BASE_cp_high'. This is different from high-cp, derived
1373 # A macro of the given type is generated for each type listed in the input.
1374 # The default return value is the number of octets read to generate the match.
1375 # Append "-cp" to the type to have it instead return the matched codepoint.
1376 # The macro name is changed to 'what_BASE...'. See pod for
1378 # Appending '-both" instead adds an extra parameter to the end of the argument
1379 # list, which is a pointer as to where to store the number of
1380 # bytes matched, while also returning the code point. The macro
1381 # name is changed to 'what_len_BASE...'. See pod for caveats
1384 # safe The input string is not necessarily valid UTF-8. In
1385 # particular an extra parameter (always the 2nd) to the macro is
1386 # required, which points to one beyond the end of the string.
1387 # The macro will make sure not to read off the end of the
1388 # string. In the case of non-UTF8, it makes sure that the
1389 # string has at least one byte in it. The macro name has
1390 # '_safe' appended to it.
1391 # fast The input string is valid UTF-8. No bounds checking is done,
1392 # and the macro can make assumptions that lead to faster
1394 # only_ascii_platform Skip this definition if this program is being run on
1395 # a non-ASCII platform.
1396 # only_ebcdic_platform Skip this definition if this program is being run on
1397 # a non-EBCDIC platform.
1398 # No modifier need be specified; fast is assumed for this case. If both
1399 # 'fast', and 'safe' are specified, two macros will be created for each
1402 # If run on a non-ASCII platform will automatically convert the Unicode input
1403 # to native. The documentation above is slightly wrong in this case. 'low'
1404 # actually refers to code points whose UTF-8 representation is the same as the
1405 # non-UTF-8 version (invariants); and 'latin1' refers to all the rest of the
1406 # code points less than 256.
1408 1; # in the unlikely case we are being used as a module
1411 # This is no longer used, but retained in case it is needed some day.
1412 # TRICKYFOLD: Problematic fold case letters. When adding to this list, also should add them to regcomp.c and fold_grind.t
1413 # => generic cp generic-cp generic-both :fast safe
1414 # 0x00DF # LATIN SMALL LETTER SHARP S
1415 # 0x0390 # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
1416 # 0x03B0 # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
1417 # 0x1E9E # LATIN CAPITAL LETTER SHARP S, because maps to same as 00DF
1418 # 0x1FD3 # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND OXIA; maps same as 0390
1419 # 0x1FE3 # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND OXIA; maps same as 03B0
1421 LNBREAK: Line Break: \R
1422 => generic UTF8 LATIN1 :fast safe
1423 "\x0D\x0A" # CRLF - Network (Windows) line ending
1426 HORIZWS: Horizontal Whitespace: \h \H
1427 => generic UTF8 LATIN1 high cp cp_high :fast safe
1430 VERTWS: Vertical Whitespace: \v \V
1431 => generic UTF8 high LATIN1 cp cp_high :fast safe
1434 XDIGIT: Hexadecimal digits
1435 => UTF8 high cp_high :fast
1438 XPERLSPACE: \p{XPerlSpace}
1439 => generic UTF8 high cp_high :fast
1442 REPLACEMENT: Unicode REPLACEMENT CHARACTER
1446 NONCHAR: Non character code points
1450 SURROGATE: Surrogate characters
1454 GCB_L: Grapheme_Cluster_Break=L
1458 GCB_LV_LVT_V: Grapheme_Cluster_Break=(LV or LVT or V)
1462 GCB_Prepend: Grapheme_Cluster_Break=Prepend
1466 GCB_RI: Grapheme_Cluster_Break=RI
1470 GCB_SPECIAL_BEGIN_START: Grapheme_Cluster_Break=special_begin_starts
1472 \p{_X_Special_Begin_Start}
1474 GCB_T: Grapheme_Cluster_Break=T
1478 GCB_V: Grapheme_Cluster_Break=V
1482 # This program was run with this enabled, and the results copied to utf8.h;
1483 # then this was commented out because it takes so long to figure out these 2
1484 # million code points. The results would not change unless utf8.h decides it
1485 # wants a maximum other than 4 bytes, or this program creates better
1487 #UTF8_CHAR: Matches utf8 from 1 to 4 bytes
1488 #=> UTF8 :safe only_ascii_platform
1491 # This hasn't been commented out, because we haven't an EBCDIC platform to run
1492 # it on, and the 3 types of EBCDIC allegedly supported by Perl would have
1494 UTF8_CHAR: Matches utf8 from 1 to 5 bytes
1495 => UTF8 :safe only_ebcdic_platform
1498 QUOTEMETA: Meta-characters that \Q should quote
1502 MULTI_CHAR_FOLD: multi-char strings that are folded to by a single character
1504 do regen/regcharclass_multi_char_folds.pl
1507 ®charclass_multi_char_folds::multi_char_folds(1)
1509 MULTI_CHAR_FOLD: multi-char strings that are folded to by a single character
1512 ®charclass_multi_char_folds::multi_char_folds(0)
1515 PATWS: pattern white space
1516 => generic generic_non_low cp : fast safe