xref: /openbsd/gnu/usr.bin/perl/t/re/pat.t (revision 3bef86f7) 
1#!./perl
2#
3# This is a home for regular expression tests that don't fit into
4# the format supported by re/regexp.t.  If you want to add a test
5# that does fit that format, add it to re/re_tests, not here.
6
7use strict;
8use warnings;
9no warnings 'experimental::vlb';
10use 5.010;
11
12sub run_tests;
13
14$| = 1;
15
16
17BEGIN {
18    chdir 't' if -d 't';
19    require './test.pl';
20    set_up_inc('../lib', '.', '../ext/re');
21    require Config; import Config;
22    require './charset_tools.pl';
23    require './loc_tools.pl';
24}
25
26skip_all_without_unicode_tables();
27
28my $has_locales = locales_enabled('LC_CTYPE');
29
30plan tests => 1046;  # Update this when adding/deleting tests.
31
32run_tests() unless caller;
33
34#
35# Tests start here.
36#
37sub run_tests {
38    my $sharp_s = uni_to_native("\xdf");
39
40    {
41        my $x = "abc\ndef\n";
42	(my $x_pretty = $x) =~ s/\n/\\n/g;
43
44        ok $x =~ /^abc/,  qq ["$x_pretty" =~ /^abc/];
45        ok $x !~ /^def/,  qq ["$x_pretty" !~ /^def/];
46
47        # used to be a test for $*
48        ok $x =~ /^def/m, qq ["$x_pretty" =~ /^def/m];
49
50        ok(!($x =~ /^xxx/), qq ["$x_pretty" =~ /^xxx/]);
51        ok(!($x !~ /^abc/), qq ["$x_pretty" !~ /^abc/]);
52
53         ok $x =~ /def/, qq ["$x_pretty" =~ /def/];
54        ok(!($x !~ /def/), qq ["$x_pretty" !~ /def/]);
55
56         ok $x !~ /.def/, qq ["$x_pretty" !~ /.def/];
57        ok(!($x =~ /.def/), qq ["$x_pretty" =~ /.def/]);
58
59         ok $x =~ /\ndef/, qq ["$x_pretty" =~ /\\ndef/];
60        ok(!($x !~ /\ndef/), qq ["$x_pretty" !~ /\\ndef/]);
61    }
62
63    {
64        $_ = '123';
65        ok /^([0-9][0-9]*)/, qq [\$_ = '$_'; /^([0-9][0-9]*)/];
66    }
67
68    {
69        $_ = 'aaabbbccc';
70         ok /(a*b*)(c*)/ && $1 eq 'aaabbb' && $2 eq 'ccc',
71                                             qq [\$_ = '$_'; /(a*b*)(c*)/];
72         ok /(a+b+c+)/ && $1 eq 'aaabbbccc', qq [\$_ = '$_'; /(a+b+c+)/];
73        unlike($_, qr/a+b?c+/, qq [\$_ = '$_'; /a+b?c+/]);
74
75        $_ = 'aaabccc';
76         ok /a+b?c+/, qq [\$_ = '$_'; /a+b?c+/];
77         ok /a*b?c*/, qq [\$_ = '$_'; /a*b?c*/];
78
79        $_ = 'aaaccc';
80         ok /a*b?c*/, qq [\$_ = '$_'; /a*b?c*/];
81        unlike($_, qr/a*b+c*/, qq [\$_ = '$_'; /a*b+c*/]);
82
83        $_ = 'abcdef';
84         ok /bcd|xyz/, qq [\$_ = '$_'; /bcd|xyz/];
85         ok /xyz|bcd/, qq [\$_ = '$_'; /xyz|bcd/];
86         ok m|bc/*d|,  qq [\$_ = '$_'; m|bc/*d|];
87         ok /^$_$/,    qq [\$_ = '$_'; /^\$_\$/];
88    }
89
90    {
91        # used to be a test for $*
92        ok "ab\ncd\n" =~ /^cd/m, q ["ab\ncd\n" =~ /^cd/m];
93    }
94
95    {
96        our %XXX = map {($_ => $_)} 123, 234, 345;
97
98        our @XXX = ('ok 1','not ok 1', 'ok 2','not ok 2','not ok 3');
99        while ($_ = shift(@XXX)) {
100            my $e = index ($_, 'not') >= 0 ? '' : 1;
101            my $r = m?(.*)?;
102            is($r, $e, "?(.*)?");
103            /not/ && reset;
104            if (/not ok 2/) {
105                if ($^O eq 'VMS') {
106                    $_ = shift(@XXX);
107                }
108                else {
109                    reset 'X';
110                }
111            }
112        }
113
114        SKIP: {
115            if ($^O eq 'VMS') {
116                skip "Reset 'X'", 1;
117            }
118            ok !keys %XXX, "%XXX is empty";
119        }
120
121    }
122
123    {
124        my $message = "Test empty pattern";
125        my $xyz = 'xyz';
126        my $cde = 'cde';
127
128        $cde =~ /[^ab]*/;
129        $xyz =~ //;
130        is($&, $xyz, $message);
131
132        my $foo = '[^ab]*';
133        $cde =~ /$foo/;
134        $xyz =~ //;
135        is($&, $xyz, $message);
136
137        $cde =~ /$foo/;
138        my $null;
139        no warnings 'uninitialized';
140        $xyz =~ /$null/;
141        is($&, $xyz, $message);
142
143        $null = "";
144        $xyz =~ /$null/;
145        is($&, $xyz, $message);
146
147        # each entry: regexp, match string, $&, //o match success
148        my @tests =
149          (
150           [ "", "xy", "x", 1 ],
151           [ "y", "yz", "y", !1 ],
152          );
153        for my $test (@tests) {
154            my ($re, $str, $matched, $omatch) = @$test;
155            $xyz =~ /x/o;
156            ok($str =~ /$re/, "$str matches /$re/");
157            is($&, $matched, "on $matched");
158            $xyz =~ /x/o;
159            is($str =~ /$re/o, $omatch, "$str matches /$re/o (or not)");
160        }
161    }
162
163    {
164        my $message = q !Check $`, $&, $'!;
165        $_ = 'abcdefghi';
166        /def/;        # optimized up to cmd
167        is("$`:$&:$'", 'abc:def:ghi', $message);
168
169        no warnings 'void';
170        /cde/ + 0;    # optimized only to spat
171        is("$`:$&:$'", 'ab:cde:fghi', $message);
172
173        /[d][e][f]/;    # not optimized
174        is("$`:$&:$'", 'abc:def:ghi', $message);
175    }
176
177    {
178        $_ = 'now is the {time for all} good men to come to.';
179        / \{([^}]*)}/;
180        is($1, 'time for all', "Match braces");
181    }
182
183    {
184        my $message = "{N,M} quantifier";
185        $_ = 'xxx {3,4}  yyy   zzz';
186        ok(/( {3,4})/, $message);
187        is($1, '   ', $message);
188        unlike($_, qr/( {4,})/, $message);
189        ok(/( {2,3}.)/, $message);
190        is($1, '  y', $message);
191        ok(/(y{2,3}.)/, $message);
192        is($1, 'yyy ', $message);
193        unlike($_, qr/x {3,4}/, $message);
194        unlike($_, qr/^xxx {3,4}/, $message);
195    }
196
197    {
198        my $message = "Test /g";
199        local $" = ":";
200        $_ = "now is the time for all good men to come to.";
201        my @words = /(\w+)/g;
202        my $exp   = "now:is:the:time:for:all:good:men:to:come:to";
203
204        is("@words", $exp, $message);
205
206        @words = ();
207        while (/\w+/g) {
208            push (@words, $&);
209        }
210        is("@words", $exp, $message);
211
212        @words = ();
213        pos = 0;
214        while (/to/g) {
215            push(@words, $&);
216        }
217        is("@words", "to:to", $message);
218
219        pos $_ = 0;
220        @words = /to/g;
221        is("@words", "to:to", $message);
222    }
223
224    {
225        $_ = "abcdefghi";
226
227        my $pat1 = 'def';
228        my $pat2 = '^def';
229        my $pat3 = '.def.';
230        my $pat4 = 'abc';
231        my $pat5 = '^abc';
232        my $pat6 = 'abc$';
233        my $pat7 = 'ghi';
234        my $pat8 = '\w*ghi';
235        my $pat9 = 'ghi$';
236
237        my $t1 = my $t2 = my $t3 = my $t4 = my $t5 =
238        my $t6 = my $t7 = my $t8 = my $t9 = 0;
239
240        for my $iter (1 .. 5) {
241            $t1++ if /$pat1/o;
242            $t2++ if /$pat2/o;
243            $t3++ if /$pat3/o;
244            $t4++ if /$pat4/o;
245            $t5++ if /$pat5/o;
246            $t6++ if /$pat6/o;
247            $t7++ if /$pat7/o;
248            $t8++ if /$pat8/o;
249            $t9++ if /$pat9/o;
250        }
251        my $x = "$t1$t2$t3$t4$t5$t6$t7$t8$t9";
252        is($x, '505550555', "Test /o");
253    }
254
255    {
256        my $xyz = 'xyz';
257        ok "abc" =~ /^abc$|$xyz/, "| after \$";
258
259        # perl 4.009 says "unmatched ()"
260        my $message = '$ inside ()';
261
262        my $result;
263        eval '"abc" =~ /a(bc$)|$xyz/; $result = "$&:$1"';
264        is($@, "", $message);
265        is($result, "abc:bc", $message);
266    }
267
268    {
269        my $message = "Scalar /g";
270        $_ = "abcfooabcbar";
271
272        ok( /abc/g && $` eq "", $message);
273        ok( /abc/g && $` eq "abcfoo", $message);
274        ok(!/abc/g, $message);
275
276        $message = "Scalar /gi";
277        pos = 0;
278        ok( /ABC/gi && $` eq "", $message);
279        ok( /ABC/gi && $` eq "abcfoo", $message);
280        ok(!/ABC/gi, $message);
281
282        $message = "Scalar /g";
283        pos = 0;
284        ok( /abc/g && $' eq "fooabcbar", $message);
285        ok( /abc/g && $' eq "bar", $message);
286
287        $_ .= '';
288        my @x = /abc/g;
289        is(@x, 2, "/g reset after assignment");
290    }
291
292    {
293        my $message = '/g, \G and pos';
294        $_ = "abdc";
295        pos $_ = 2;
296        /\Gc/gc;
297        is(pos $_, 2, $message);
298        /\Gc/g;
299        is(pos $_, undef, $message);
300    }
301
302    {
303        my $message = '(?{ })';
304        our $out = 1;
305        'abc' =~ m'a(?{ $out = 2 })b';
306        is($out, 2, $message);
307
308        $out = 1;
309        'abc' =~ m'a(?{ $out = 3 })c';
310        is($out, 1, $message);
311    }
312
313    {
314        $_ = 'foobar1 bar2 foobar3 barfoobar5 foobar6';
315        my @out = /(?<!foo)bar./g;
316        is("@out", 'bar2 barf', "Negative lookbehind");
317    }
318
319    {
320        my $message = "REG_INFTY tests";
321        # Tests which depend on REG_INFTY
322
323	#  Defaults assumed if this fails
324	eval { require Config; };
325        $::reg_infty   = $Config::Config{reg_infty} // 65535;
326        $::reg_infty_m = $::reg_infty - 1;
327        $::reg_infty_p = $::reg_infty + 1;
328        $::reg_infty_m = $::reg_infty_m;   # Suppress warning.
329
330        # As well as failing if the pattern matches do unexpected things, the
331        # next three tests will fail if you should have picked up a lower-than-
332        # default value for $reg_infty from Config.pm, but have not.
333
334        is(eval q{('aaa' =~ /(a{1,$::reg_infty_m})/)[0]}, 'aaa', $message);
335        is($@, '', $message);
336        is(eval q{('a' x $::reg_infty_m) =~ /a{$::reg_infty_m}/}, 1, $message);
337        is($@, '', $message);
338        isnt(q{('a' x ($::reg_infty_m - 1)) !~ /a{$::reg_infty_m}/}, 1, $message);
339        is($@, '', $message);
340
341        eval "'aaa' =~ /a{1,$::reg_infty}/";
342        like($@, qr/^\QQuantifier in {,} bigger than/, $message);
343        eval "'aaa' =~ /a{1,$::reg_infty_p}/";
344        like($@, qr/^\QQuantifier in {,} bigger than/, $message);
345
346        # It should be 'a' x 2147483647, but that exhausts memory on
347        # reasonably sized modern machines
348        like('a' x $::reg_infty_p, qr/a{1,}/,
349             "{1,} matches more times than REG_INFTY");
350    }
351
352    {
353        # Poke a couple more parse failures
354        my $context = 'x' x 256;
355        eval qq("${context}y" =~ /(?<=$context)y/);
356        ok $@ =~ /^\QLookbehind longer than 255 not/, "Lookbehind limit";
357    }
358
359  SKIP:
360    {   # Long Monsters
361
362        my @trials = (125, 140, 250, 270, 300000, 30);
363
364        skip('limited memory', @trials * 4) if $ENV{'PERL_SKIP_BIG_MEM_TESTS'};
365
366        for my $l (@trials) { # Ordered to free memory
367            my $a = 'a' x $l;
368	    my $message = "Long monster, length = $l";
369	    like("ba$a=", qr/a$a=/, $message);
370            unlike("b$a=", qr/a$a=/, $message);
371            like("b$a=", qr/ba+=/, $message);
372
373	    like("ba$a=", qr/b(?:a|b)+=/, $message);
374        }
375    }
376
377  SKIP:
378    {   # 20000 nodes, each taking 3 words per string, and 1 per branch
379
380        my %ans = ( 'ax13876y25677lbc' => 1,
381                    'ax13876y25677mcb' => 0, # not b.
382                    'ax13876y35677nbc' => 0, # Num too big
383                    'ax13876y25677y21378obc' => 1,
384                    'ax13876y25677y21378zbc' => 0,    # Not followed by [k-o]
385                    'ax13876y25677y21378y21378kbc' => 1,
386                    'ax13876y25677y21378y21378kcb' => 0, # Not b.
387                    'ax13876y25677y21378y21378y21378kbc' => 0, # 5 runs
388                  );
389
390        skip('limited memory', 2 * scalar keys %ans) if $ENV{'PERL_SKIP_BIG_MEM_TESTS'};
391
392        my $long_constant_len = join '|', 12120 .. 32645;
393        my $long_var_len = join '|', 8120 .. 28645;
394
395        for (keys %ans) {
396	    my $message = "20000 nodes, const-len '$_'";
397            ok !($ans{$_} xor /a(?=([yx]($long_constant_len)){2,4}[k-o]).*b./o), $message;
398
399	    $message = "20000 nodes, var-len '$_'";
400            ok !($ans{$_} xor /a(?=([yx]($long_var_len)){2,4}[k-o]).*b./o,), $message;
401        }
402    }
403
404    {
405        my $message = "Complicated backtracking";
406        $_ = " a (bla()) and x(y b((l)u((e))) and b(l(e)e)e";
407        my $expect = "(bla()) ((l)u((e))) (l(e)e)";
408
409        our $c;
410        sub matchit {
411          m/
412             (
413               \(
414               (?{ $c = 1 })    # Initialize
415               (?:
416                 (?(?{ $c == 0 })   # PREVIOUS iteration was OK, stop the loop
417                   (?!
418                   )        # Fail: will unwind one iteration back
419                 )
420                 (?:
421                   [^()]+        # Match a big chunk
422                   (?=
423                     [()]
424                   )        # Do not try to match subchunks
425                 |
426                   \(
427                   (?{ ++$c })
428                 |
429                   \)
430                   (?{ --$c })
431                 )
432               )+        # This may not match with different subblocks
433             )
434             (?(?{ $c != 0 })
435               (?!
436               )        # Fail
437             )            # Otherwise the chunk 1 may succeed with $c>0
438           /xg;
439        }
440
441        my @ans = ();
442        my $res;
443        push @ans, $res while $res = matchit;
444        is("@ans", "1 1 1", $message);
445
446        @ans = matchit;
447        is("@ans", $expect, $message);
448
449        $message = "Recursion with (??{ })";
450        our $matched;
451        $matched = qr/\((?:(?>[^()]+)|(??{$matched}))*\)/;
452
453        @ans = my @ans1 = ();
454        push (@ans, $res), push (@ans1, $&) while $res = m/$matched/g;
455
456        is("@ans", "1 1 1", $message);
457        is("@ans1", $expect, $message);
458
459        @ans = m/$matched/g;
460        is("@ans", $expect, $message);
461
462    }
463
464    {
465        ok "abc" =~ /^(??{"a"})b/, '"abc" =~ /^(??{"a"})b/';
466    }
467
468    {
469        my @ans = ('a/b' =~ m%(.*/)?(.*)%);    # Stack may be bad
470        is("@ans", 'a/ b', "Stack may be bad");
471    }
472
473    {
474        my $message = "Eval-group not allowed at runtime";
475        my $code = '{$blah = 45}';
476        our $blah = 12;
477        eval { /(?$code)/ };
478        ok($@ && $@ =~ /not allowed at runtime/ && $blah == 12, $message);
479
480	$blah = 12;
481	my $res = eval { "xx" =~ /(?$code)/o };
482	{
483	    no warnings 'uninitialized';
484	    chomp $@; my $message = "$message '$@', '$res', '$blah'";
485	    ok($@ && $@ =~ /not allowed at runtime/ && $blah == 12, $message);
486	}
487
488        $code = '=xx';
489	$blah = 12;
490	$res = eval { "xx" =~ /(?$code)/o };
491	{
492	    no warnings 'uninitialized';
493	    my $message = "$message '$@', '$res', '$blah'";
494	    ok(!$@ && $res, $message);
495	}
496
497        $code = '{$blah = 45}';
498        $blah = 12;
499        eval "/(?$code)/";
500        is($blah, 45, $message);
501
502        $blah = 12;
503        /(?{$blah = 45})/;
504        is($blah, 45, $message);
505    }
506
507    {
508        my $message = "Pos checks";
509        my $x = 'banana';
510        $x =~ /.a/g;
511        is(pos $x, 2, $message);
512
513        $x =~ /.z/gc;
514        is(pos $x, 2, $message);
515
516        sub f {
517            my $p = $_[0];
518            return $p;
519        }
520
521        $x =~ /.a/g;
522        is(f (pos $x), 4, $message);
523    }
524
525    {
526        my $message = 'Checking $^R';
527        our $x = $^R = 67;
528        'foot' =~ /foo(?{$x = 12; 75})[t]/;
529        is($^R, 75, $message);
530
531        $x = $^R = 67;
532        'foot' =~ /foo(?{$x = 12; 75})[xy]/;
533        ok($^R eq '67' && $x eq '12', $message);
534
535        $x = $^R = 67;
536        'foot' =~ /foo(?{ $^R + 12 })((?{ $x = 12; $^R + 17 })[xy])?/;
537        ok($^R eq '79' && $x eq '12', $message);
538    }
539
540    {
541        is(qr/\b\v$/i,    '(?^i:\b\v$)', 'qr/\b\v$/i');
542        is(qr/\b\v$/s,    '(?^s:\b\v$)', 'qr/\b\v$/s');
543        is(qr/\b\v$/m,    '(?^m:\b\v$)', 'qr/\b\v$/m');
544        is(qr/\b\v$/x,    '(?^x:\b\v$)', 'qr/\b\v$/x');
545        is(qr/\b\v$/xism, '(?^msix:\b\v$)',  'qr/\b\v$/xism');
546        is(qr/\b\v$/,     '(?^:\b\v$)', 'qr/\b\v$/');
547    }
548
549    {   # Test that charset modifier work, and are interpolated
550        is(qr/\b\v$/, '(?^:\b\v$)', 'Verify no locale, no unicode_strings gives default modifier');
551        is(qr/(?l:\b\v$)/, '(?^:(?l:\b\v$))', 'Verify infix l modifier compiles');
552        is(qr/(?u:\b\v$)/, '(?^:(?u:\b\v$))', 'Verify infix u modifier compiles');
553        is(qr/(?l)\b\v$/, '(?^:(?l)\b\v$)', 'Verify (?l) compiles');
554        is(qr/(?u)\b\v$/, '(?^:(?u)\b\v$)', 'Verify (?u) compiles');
555
556        my $dual = qr/\b\v$/;
557        my $locale;
558
559      SKIP: {
560            skip 'Locales not available', 1 unless $has_locales;
561
562            use locale;
563            $locale = qr/\b\v$/;
564            is($locale,    '(?^l:\b\v$)', 'Verify has l modifier when compiled under use locale');
565            no locale;
566        }
567
568        use feature 'unicode_strings';
569        my $unicode = qr/\b\v$/;
570        is($unicode,    '(?^u:\b\v$)', 'Verify has u modifier when compiled under unicode_strings');
571        is(qr/abc$dual/,    '(?^u:abc(?^:\b\v$))', 'Verify retains d meaning when interpolated under locale');
572
573      SKIP: {
574            skip 'Locales not available', 1 unless $has_locales;
575
576            is(qr/abc$locale/,    '(?^u:abc(?^l:\b\v$))', 'Verify retains l when interpolated under unicode_strings');
577        }
578
579        no feature 'unicode_strings';
580      SKIP: {
581            skip 'Locales not available', 1 unless $has_locales;
582            is(qr/abc$locale/,    '(?^:abc(?^l:\b\v$))', 'Verify retains l when interpolated outside locale and unicode strings');
583        }
584
585        is(qr/def$unicode/,    '(?^:def(?^u:\b\v$))', 'Verify retains u when interpolated outside locale and unicode strings');
586
587      SKIP: {
588            skip 'Locales not available', 2 unless $has_locales;
589
590             use locale;
591            is(qr/abc$dual/,    '(?^l:abc(?^:\b\v$))', 'Verify retains d meaning when interpolated under locale');
592            is(qr/abc$unicode/,    '(?^l:abc(?^u:\b\v$))', 'Verify retains u when interpolated under locale');
593        }
594    }
595
596    {
597        my $message = "Look around";
598        $_ = 'xabcx';
599        foreach my $ans ('', 'c') {
600            ok(/(?<=(?=a)..)((?=c)|.)/g, $message);
601            is($1, $ans, $message);
602        }
603    }
604
605    {
606        my $message = "Empty clause";
607        $_ = 'a';
608        foreach my $ans ('', 'a', '') {
609            ok(/^|a|$/g, $message);
610            is($&, $ans, $message);
611        }
612    }
613
614    {
615        sub prefixify {
616        my $message = "Prefixify";
617            {
618                my ($v, $a, $b, $res) = @_;
619                ok($v =~ s/\Q$a\E/$b/, $message);
620                is($v, $res, $message);
621            }
622        }
623
624        prefixify ('/a/b/lib/arch', "/a/b/lib", 'X/lib', 'X/lib/arch');
625        prefixify ('/a/b/man/arch', "/a/b/man", 'X/man', 'X/man/arch');
626    }
627
628    {
629        $_ = 'var="foo"';
630        /(\")/;
631        ok $1 && /$1/, "Capture a quote";
632    }
633
634    {
635        no warnings 'closure';
636        my $message = '(?{ $var } refers to package vars';
637        package aa;
638        our $c = 2;
639        $::c = 3;
640        '' =~ /(?{ $c = 4 })/;
641        main::is($c, 4, $message);
642        main::is($::c, 3, $message);
643    }
644
645    {
646        is(eval 'q(a:[b]:) =~ /[x[:foo:]]/', undef);
647	like ($@, qr/POSIX class \[:[^:]+:\] unknown in regex/,
648	      'POSIX class [: :] must have valid name');
649
650        for my $d (qw [= .]) {
651            is(eval "/[[${d}foo${d}]]/", undef);
652	    like ($@, qr/\QPOSIX syntax [$d $d] is reserved for future extensions/,
653		  "POSIX syntax [[$d $d]] is an error");
654        }
655    }
656
657    {
658        # test if failure of patterns returns empty list
659        my $message = "Failed pattern returns empty list";
660        $_ = 'aaa';
661        @_ = /bbb/;
662        is("@_", "", $message);
663
664        @_ = /bbb/g;
665        is("@_", "", $message);
666
667        @_ = /(bbb)/;
668        is("@_", "", $message);
669
670        @_ = /(bbb)/g;
671        is("@_", "", $message);
672    }
673    {
674        my $message = 'ACCEPT and CLOSE - ';
675        $_ = "aced";
676        #12           3  4  5
677        /((a?(*ACCEPT)())())()/
678            or die "Failed to match";
679        is($1,"a",$message . "buffer 1 is defined with expected value");
680        is($2,"a",$message . "buffer 2 is defined with expected value");
681        ok(!defined($3),$message . "buffer 3 is not defined");
682        ok(!defined($4),$message . "buffer 4 is not defined");
683        ok(!defined($5),$message . "buffer 5 is not defined");
684        ok(!defined($6),$message . "buffer 6 is not defined");
685        $message= 'NO ACCEPT and CLOSE - ';
686        /((a?())())()/
687            or die "Failed to match";
688        is($1,"a",$message . "buffer 1 is defined with expected value");
689        is($2,"a",$message . "buffer 2 is defined with expected value");
690        is($3,"", $message . "buffer 3 is defined with expected value");
691        is($4,"", $message . "buffer 4 is defined with expected value");
692        is($5,"",$message . "buffer 5 is defined with expected value");
693        ok(!defined($6),$message . "buffer 6 is not defined");
694        #12           3  4  5
695        $message = 'ACCEPT and CLOSE - ';
696        /((a?(*ACCEPT)(c))(e))(d)/
697            or die "Failed to match";
698        is($1,"a",$message . "buffer 1 is defined with expected value");
699        is($2,"a",$message . "buffer 2 is defined with expected value");
700        ok(!defined($3),$message . "buffer 3 is not defined");
701        ok(!defined($4),$message . "buffer 4 is not defined");
702        ok(!defined($5),$message . "buffer 5 is not defined");
703        ok(!defined($6),$message . "buffer 6 is not defined");
704        $message= 'NO ACCEPT and CLOSE - ';
705        /((a?(c))(e))(d)/
706            or die "Failed to match";
707        is($1,"ace", $message . "buffer 1 is defined with expected value");
708        is($2,"ac", $message . "buffer 2 is defined with expected value");
709        is($3,"c", $message . "buffer 3 is defined with expected value");
710        is($4,"e", $message . "buffer 4 is defined with expected value");
711        is($5,"d", $message . "buffer 5 is defined with expected value");
712        ok(!defined($6),$message . "buffer 6 is not defined");
713    }
714    {
715        my $message = '@- and @+ and @{^CAPTURE} tests';
716
717        $_= "ace";
718        /c(?=.$)/;
719        is($#{^CAPTURE}, -1, $message);
720        is($#+, 0, $message);
721        is($#-, 0, $message);
722        is($+ [0], 2, $message);
723        is($- [0], 1, $message);
724        ok(!defined $+ [1] && !defined $- [1] &&
725           !defined $+ [2] && !defined $- [2], $message);
726
727        /a(c)(e)/;
728        is($#{^CAPTURE}, 1, $message); # one less than $#-
729        is($#+, 2, $message);
730        is($#-, 2, $message);
731        is($+ [0], 3, $message);
732        is($- [0], 0, $message);
733        is(${^CAPTURE}[0], "c", $message);
734        is($+ [1], 2, $message);
735        is($- [1], 1, $message);
736        is(${^CAPTURE}[1], "e", $message);
737        is($+ [2], 3, $message);
738        is($- [2], 2, $message);
739        ok(!defined $+ [3] && !defined $- [3] &&
740           !defined ${^CAPTURE}[2] && !defined ${^CAPTURE}[3] &&
741           !defined $+ [4] && !defined $- [4], $message);
742
743        # Exists has a special check for @-/@+ - bug 45147
744        ok(exists $-[0], $message);
745        ok(exists $+[0], $message);
746        ok(exists ${^CAPTURE}[0], $message);
747        ok(exists ${^CAPTURE}[1], $message);
748        ok(exists $-[2], $message);
749        ok(exists $+[2], $message);
750        ok(!exists ${^CAPTURE}[2], $message);
751        ok(!exists $-[3], $message);
752        ok(!exists $+[3], $message);
753        ok(exists ${^CAPTURE}[-1], $message);
754        ok(exists ${^CAPTURE}[-2], $message);
755        ok(exists $-[-1], $message);
756        ok(exists $+[-1], $message);
757        ok(exists $-[-3], $message);
758        ok(exists $+[-3], $message);
759        ok(!exists $-[-4], $message);
760        ok(!exists $+[-4], $message);
761        ok(!exists ${^CAPTURE}[-3], $message);
762
763
764        /.(c)(b)?(e)/;
765        is($#{^CAPTURE}, 2, $message); # one less than $#-
766        is($#+, 3, $message);
767        is($#-, 3, $message);
768        is(${^CAPTURE}[0], "c", $message);
769        is(${^CAPTURE}[2], "e", $message . "[$1 $3]");
770        is($+ [1], 2, $message);
771        is($- [1], 1, $message);
772        is($+ [3], 3, $message);
773        is($- [3], 2, $message);
774        ok(!defined $+ [2] && !defined $- [2] &&
775           !defined $+ [4] && !defined $- [4] &&
776           !defined ${^CAPTURE}[1], $message);
777
778        /.(c)/;
779        is($#{^CAPTURE}, 0, $message); # one less than $#-
780        is($#+, 1, $message);
781        is($#-, 1, $message);
782        is(${^CAPTURE}[0], "c", $message);
783        is($+ [0], 2, $message);
784        is($- [0], 0, $message);
785        is($+ [1], 2, $message);
786        is($- [1], 1, $message);
787        ok(!defined $+ [2] && !defined $- [2] &&
788           !defined $+ [3] && !defined $- [3] &&
789           !defined ${^CAPTURE}[1], $message);
790
791        /.(c)(ba*)?/;
792        is($#{^CAPTURE}, 0, $message); # one less than $#-
793        is($#+, 2, $message);
794        is($#-, 1, $message);
795
796        # Check that values don't stick
797        "     "=~/()()()(.)(..)/;
798        my($m,$p,$q) = (\$-[5], \$+[5], \${^CAPTURE}[4]);
799        () = "$$_" for $m, $p, $q; # FETCH (or eqv.)
800        " " =~ /()/;
801        is $$m, undef, 'values do not stick to @- elements';
802        is $$p, undef, 'values do not stick to @+ elements';
803        is $$q, undef, 'values do not stick to @{^CAPTURE} elements';
804    }
805
806    foreach ('$+[0] = 13', '$-[0] = 13', '@+ = (7, 6, 5)',
807             '${^CAPTURE}[0] = 13',
808	     '@- = qw (foo bar)', '$^N = 42') {
809	is(eval $_, undef);
810        like($@, qr/^Modification of a read-only value attempted/,
811	     '$^N, @- and @+ are read-only');
812    }
813
814    {
815        my $message = '\G testing';
816        $_ = 'aaa';
817        pos = 1;
818        my @a = /\Ga/g;
819        is("@a", "a a", $message);
820
821        my $str = 'abcde';
822        pos $str = 2;
823        unlike($str, qr/^\G/, $message);
824        unlike($str, qr/^.\G/, $message);
825        like($str, qr/^..\G/, $message);
826        unlike($str, qr/^...\G/, $message);
827        ok($str =~ /\G../ && $& eq 'cd', $message);
828        ok($str =~ /.\G./ && $& eq 'bc', $message);
829
830    }
831
832    {
833        my $message = '\G and intuit and anchoring';
834	$_ = "abcdef";
835	pos = 0;
836	ok($_ =~ /\Gabc/, $message);
837	ok($_ =~ /^\Gabc/, $message);
838
839	pos = 3;
840	ok($_ =~ /\Gdef/, $message);
841	pos = 3;
842	ok($_ =~ /\Gdef$/, $message);
843	pos = 3;
844	ok($_ =~ /abc\Gdef$/, $message);
845	pos = 3;
846	ok($_ =~ /^abc\Gdef$/, $message);
847	pos = 3;
848	ok($_ =~ /c\Gd/, $message);
849	pos = 3;
850	ok($_ =~ /..\GX?def/, $message);
851    }
852
853    {
854        my $s = '123';
855        pos($s) = 1;
856        my @a = $s =~ /(\d)\G/g; # this infinitely looped up till 5.19.1
857        is("@a", "1", '\G looping');
858    }
859
860
861    {
862        my $message = 'pos inside (?{ })';
863        my $str = 'abcde';
864        our ($foo, $bar);
865        like($str, qr/b(?{$foo = $_; $bar = pos})c/, $message);
866        is($foo, $str, $message);
867        is($bar, 2, $message);
868        is(pos $str, undef, $message);
869
870        undef $foo;
871        undef $bar;
872        pos $str = undef;
873        ok($str =~ /b(?{$foo = $_; $bar = pos})c/g, $message);
874        is($foo, $str, $message);
875        is($bar, 2, $message);
876        is(pos $str, 3, $message);
877
878        $_ = $str;
879        undef $foo;
880        undef $bar;
881        like($_, qr/b(?{$foo = $_; $bar = pos})c/, $message);
882        is($foo, $str, $message);
883        is($bar, 2, $message);
884
885        undef $foo;
886        undef $bar;
887        ok(/b(?{$foo = $_; $bar = pos})c/g, $message);
888        is($foo, $str, $message);
889        is($bar, 2, $message);
890        is(pos, 3, $message);
891
892        undef $foo;
893        undef $bar;
894        pos = undef;
895        1 while /b(?{$foo = $_; $bar = pos})c/g;
896        is($foo, $str, $message);
897        is($bar, 2, $message);
898        is(pos, undef, $message);
899
900        undef $foo;
901        undef $bar;
902        $_ = 'abcde|abcde';
903        ok(s/b(?{$foo = $_; $bar = pos})c/x/g, $message);
904        is($foo, 'abcde|abcde', $message);
905        is($bar, 8, $message);
906        is($_, 'axde|axde', $message);
907
908        # List context:
909        $_ = 'abcde|abcde';
910        our @res;
911        () = /([ace]).(?{push @res, $1,$2})([ce])(?{push @res, $1,$2})/g;
912        @res = map {defined $_ ? "'$_'" : 'undef'} @res;
913        is("@res", "'a' undef 'a' 'c' 'e' undef 'a' undef 'a' 'c'", $message);
914
915        @res = ();
916        () = /([ace]).(?{push @res, $`,$&,$'})([ce])(?{push @res, $`,$&,$'})/g;
917        @res = map {defined $_ ? "'$_'" : 'undef'} @res;
918        is("@res", "'' 'ab' 'cde|abcde' " .
919                     "'' 'abc' 'de|abcde' " .
920                     "'abcd' 'e|' 'abcde' " .
921                     "'abcde|' 'ab' 'cde' " .
922                     "'abcde|' 'abc' 'de'", $message);
923    }
924
925    {
926        my $message = '\G anchor checks';
927        my $foo = 'aabbccddeeffgg';
928        pos ($foo) = 1;
929
930	ok($foo =~ /.\G(..)/g, $message);
931	is($1, 'ab', $message);
932
933	pos ($foo) += 1;
934	ok($foo =~ /.\G(..)/g, $message);
935	is($1, 'cc', $message);
936
937	pos ($foo) += 1;
938	ok($foo =~ /.\G(..)/g, $message);
939	is($1, 'de', $message);
940
941	ok($foo =~ /\Gef/g, $message);
942
943        undef pos $foo;
944        ok($foo =~ /\G(..)/g, $message);
945        is($1, 'aa', $message);
946
947        ok($foo =~ /\G(..)/g, $message);
948        is($1, 'bb', $message);
949
950        pos ($foo) = 5;
951        ok($foo =~ /\G(..)/g, $message);
952        is($1, 'cd', $message);
953    }
954
955    {
956        my $message = 'basic \G floating checks';
957        my $foo = 'aabbccddeeffgg';
958        pos ($foo) = 1;
959
960	ok($foo =~ /a+\G(..)/g, "$message: a+\\G");
961	is($1, 'ab', "$message: ab");
962
963	pos ($foo) += 1;
964	ok($foo =~ /b+\G(..)/g, "$message: b+\\G");
965	is($1, 'cc', "$message: cc");
966
967	pos ($foo) += 1;
968	ok($foo =~ /d+\G(..)/g, "$message: d+\\G");
969	is($1, 'de', "$message: de");
970
971	ok($foo =~ /\Gef/g, "$message: \\Gef");
972
973        pos ($foo) = 1;
974
975	ok($foo =~ /(?=a+\G)(..)/g, "$message: (?a+\\G)");
976	is($1, 'aa', "$message: aa");
977
978        pos ($foo) = 2;
979
980	ok($foo =~ /a(?=a+\G)(..)/g, "$message: a(?=a+\\G)");
981	is($1, 'ab', "$message: ab");
982
983    }
984
985    {
986        $_ = '123x123';
987        my @res = /(\d*|x)/g;
988        local $" = '|';
989        is("@res", "123||x|123|", "0 match in alternation");
990    }
991
992    {
993        my $message = "Match against temporaries (created via pp_helem())" .
994                         " is safe";
995        ok({foo => "bar\n" . $^X} -> {foo} =~ /^(.*)\n/g, $message);
996        is($1, "bar", $message);
997    }
998
999    {
1000        my $message = 'package $i inside (?{ }), ' .
1001                         'saved substrings and changing $_';
1002        our @a = qw [foo bar];
1003        our @b = ();
1004        s/(\w)(?{push @b, $1})/,$1,/g for @a;
1005        is("@b", "f o o b a r", $message);
1006        is("@a", ",f,,o,,o, ,b,,a,,r,", $message);
1007
1008        $message = 'lexical $i inside (?{ }), ' .
1009                         'saved substrings and changing $_';
1010        no warnings 'closure';
1011        my @c = qw [foo bar];
1012        my @d = ();
1013        s/(\w)(?{push @d, $1})/,$1,/g for @c;
1014        is("@d", "f o o b a r", $message);
1015        is("@c", ",f,,o,,o, ,b,,a,,r,", $message);
1016    }
1017
1018    {
1019        my $message = 'Brackets';
1020        our $brackets;
1021        $brackets = qr {
1022            {  (?> [^{}]+ | (??{ $brackets }) )* }
1023        }x;
1024
1025        ok("{{}" =~ $brackets, $message);
1026        is($&, "{}", $message);
1027        ok("something { long { and } hairy" =~ $brackets, $message);
1028        is($&, "{ and }", $message);
1029        ok("something { long { and } hairy" =~ m/((??{ $brackets }))/, $message);
1030        is($&, "{ and }", $message);
1031    }
1032
1033    {
1034        $_ = "a-a\nxbb";
1035        pos = 1;
1036        ok(!m/^-.*bb/mg, '$_ = "a-a\nxbb"; m/^-.*bb/mg');
1037    }
1038
1039    {
1040        my $message = '\G anchor checks';
1041        my $text = "aaXbXcc";
1042        pos ($text) = 0;
1043        ok($text !~ /\GXb*X/g, $message);
1044    }
1045
1046    {
1047        $_ = "xA\n" x 500;
1048        unlike($_, qr/^\s*A/m, '$_ = "xA\n" x 500; /^\s*A/m"');
1049
1050        my $text = "abc dbf";
1051        my @res = ($text =~ /.*?(b).*?\b/g);
1052        is("@res", "b b", '\b is not special');
1053    }
1054
1055    {
1056        my $message = '\S, [\S], \s, [\s]';
1057        my @a = map chr, 0 .. 255;
1058        my @b = grep m/\S/, @a;
1059        my @c = grep m/[^\s]/, @a;
1060        is("@b", "@c", $message);
1061
1062        @b = grep /\S/, @a;
1063        @c = grep /[\S]/, @a;
1064        is("@b", "@c", $message);
1065
1066        @b = grep /\s/, @a;
1067        @c = grep /[^\S]/, @a;
1068        is("@b", "@c", $message);
1069
1070        @b = grep /\s/, @a;
1071        @c = grep /[\s]/, @a;
1072        is("@b", "@c", $message);
1073
1074        # Test an inverted posix class with a char also in the class.
1075        my $nbsp = chr utf8::unicode_to_native(0xA0);
1076        my $non_s = chr utf8::unicode_to_native(0xA1);
1077        my $pat_string = "[^\\S ]";
1078        unlike(" ", qr/$pat_string/, "Verify ' ' !~ /$pat_string/");
1079        like("\t", qr/$pat_string/, "Verify '\\t =~ /$pat_string/");
1080        unlike($nbsp, qr/$pat_string/, "Verify non-utf8-NBSP !~ /$pat_string/");
1081        utf8::upgrade($nbsp);
1082        like($nbsp, qr/$pat_string/, "Verify utf8-NBSP =~ /$pat_string/");
1083        unlike($non_s, qr/$pat_string/, "Verify non-utf8-inverted-bang !~ /$pat_string/");
1084        utf8::upgrade($non_s);
1085        unlike($non_s, qr/$pat_string/, "Verify utf8-inverted-bang !~ /$pat_string/");
1086    }
1087    {
1088        my $message = '\D, [\D], \d, [\d]';
1089        my @a = map chr, 0 .. 255;
1090        my @b = grep /\D/, @a;
1091        my @c = grep /[^\d]/, @a;
1092        is("@b", "@c", $message);
1093
1094        @b = grep /\D/, @a;
1095        @c = grep /[\D]/, @a;
1096        is("@b", "@c", $message);
1097
1098        @b = grep /\d/, @a;
1099        @c = grep /[^\D]/, @a;
1100        is("@b", "@c", $message);
1101
1102        @b = grep /\d/, @a;
1103        @c = grep /[\d]/, @a;
1104        is("@b", "@c", $message);
1105    }
1106    {
1107        my $message = '\W, [\W], \w, [\w]';
1108        my @a = map chr, 0 .. 255;
1109        my @b = grep /\W/, @a;
1110        my @c = grep /[^\w]/, @a;
1111        is("@b", "@c", $message);
1112
1113        @b = grep /\W/, @a;
1114        @c = grep /[\W]/, @a;
1115        is("@b", "@c", $message);
1116
1117        @b = grep /\w/, @a;
1118        @c = grep /[^\W]/, @a;
1119        is("@b", "@c", $message);
1120
1121        @b = grep /\w/, @a;
1122        @c = grep /[\w]/, @a;
1123        is("@b", "@c", $message);
1124    }
1125
1126    {
1127        # see if backtracking optimization works correctly
1128        my $message = 'Backtrack optimization';
1129        like("\n\n", qr/\n   $ \n/x, $message);
1130        like("\n\n", qr/\n*  $ \n/x, $message);
1131        like("\n\n", qr/\n+  $ \n/x, $message);
1132        like("\n\n", qr/\n?  $ \n/x, $message);
1133        like("\n\n", qr/\n*? $ \n/x, $message);
1134        like("\n\n", qr/\n+? $ \n/x, $message);
1135        like("\n\n", qr/\n?? $ \n/x, $message);
1136        unlike("\n\n", qr/\n*+ $ \n/x, $message);
1137        unlike("\n\n", qr/\n++ $ \n/x, $message);
1138        like("\n\n", qr/\n?+ $ \n/x, $message);
1139    }
1140
1141    {
1142        package S;
1143        use overload '""' => sub {'Object S'};
1144        sub new {bless []}
1145
1146        my $message  = "Ref stringification";
1147      ::ok(do { \my $v} =~ /^SCALAR/,   "Scalar ref stringification") or diag($message);
1148      ::ok(do {\\my $v} =~ /^REF/,      "Ref ref stringification") or diag($message);
1149      ::ok([]           =~ /^ARRAY/,    "Array ref stringification") or diag($message);
1150      ::ok({}           =~ /^HASH/,     "Hash ref stringification") or diag($message);
1151      ::ok('S' -> new   =~ /^Object S/, "Object stringification") or diag($message);
1152    }
1153
1154    {
1155        my $message = "Test result of match used as match";
1156        ok('a1b' =~ ('xyz' =~ /y/), $message);
1157        is($`, 'a', $message);
1158        ok('a1b' =~ ('xyz' =~ /t/), $message);
1159        is($`, 'a', $message);
1160    }
1161
1162    {
1163        my $message = '"1" is not \s';
1164        warning_is(sub {unlike("1\n" x 102, qr/^\s*\n/m, $message)},
1165		   undef, "$message (did not warn)");
1166    }
1167
1168    {
1169        my $message = '\s, [[:space:]] and [[:blank:]]';
1170        my %space = (spc   => " ",
1171                     tab   => "\t",
1172                     cr    => "\r",
1173                     lf    => "\n",
1174                     ff    => "\f",
1175        # There's no \v but the vertical tabulator seems miraculously
1176        # be 11 both in ASCII and EBCDIC.
1177                     vt    => chr(11),
1178                     false => "space");
1179
1180        my @space0 = sort grep {$space {$_} =~ /\s/         } keys %space;
1181        my @space1 = sort grep {$space {$_} =~ /[[:space:]]/} keys %space;
1182        my @space2 = sort grep {$space {$_} =~ /[[:blank:]]/} keys %space;
1183
1184        is("@space0", "cr ff lf spc tab vt", $message);
1185        is("@space1", "cr ff lf spc tab vt", $message);
1186        is("@space2", "spc tab", $message);
1187    }
1188
1189    {
1190        my $n= 50;
1191        # this must be a high number and go from 0 to N, as the bug we are looking for doesn't
1192        # seem to be predictable. Slight changes to the test make it fail earlier or later.
1193        foreach my $i (0 .. $n)
1194        {
1195            my $str= "\n" x $i;
1196            ok $str=~/.*\z/, "implicit MBOL check string disable does not break things length=$i";
1197        }
1198    }
1199    {
1200        # we are actually testing that we dont die when executing these patterns
1201        use utf8;
1202        my $e = "Böck";
1203        ok(utf8::is_utf8($e),"got a unicode string - rt75680");
1204
1205        ok($e !~ m/.*?[x]$/, "unicode string against /.*?[x]\$/ - rt75680");
1206        ok($e !~ m/.*?\p{Space}$/i, "unicode string against /.*?\\p{space}\$/i - rt75680");
1207        ok($e !~ m/.*?[xyz]$/, "unicode string against /.*?[xyz]\$/ - rt75680");
1208        ok($e !~ m/(.*?)[,\p{isSpace}]+((?:\p{isAlpha}[\p{isSpace}\.]{1,2})+)\p{isSpace}*$/, "unicode string against big pattern - rt75680");
1209    }
1210    {
1211        # we are actually testing that we dont die when executing these patterns
1212        my $e = "B" . uni_to_native("\x{f6}") . "ck";
1213        ok(!utf8::is_utf8($e), "got a latin string - rt75680");
1214
1215        ok($e !~ m/.*?[x]$/, "latin string against /.*?[x]\$/ - rt75680");
1216        ok($e !~ m/.*?\p{Space}$/i, "latin string against /.*?\\p{space}\$/i - rt75680");
1217        ok($e !~ m/.*?[xyz]$/,"latin string against /.*?[xyz]\$/ - rt75680");
1218        ok($e !~ m/(.*?)[,\p{isSpace}]+((?:\p{isAlpha}[\p{isSpace}\.]{1,2})+)\p{isSpace}*$/,"latin string against big pattern - rt75680");
1219    }
1220
1221    {
1222        #
1223        # Tests for bug 77414.
1224        #
1225
1226        my $message = '\p property after empty * match';
1227        {
1228            like("1", qr/\s*\pN/, $message);
1229            like("-", qr/\s*\p{Dash}/, $message);
1230            like(" ", qr/\w*\p{Blank}/, $message);
1231        }
1232
1233        like("1", qr/\s*\pN+/, $message);
1234        like("-", qr/\s*\p{Dash}{1}/, $message);
1235        like(" ", qr/\w*\p{Blank}{1,4}/, $message);
1236
1237    }
1238
1239    {   # Some constructs with Latin1 characters cause a utf8 string not
1240        # to match itself in non-utf8
1241        my $c = uni_to_native("\xc0");
1242        my $pattern = my $utf8_pattern = qr/(($c)+,?)/;
1243        utf8::upgrade($utf8_pattern);
1244        ok $c =~ $pattern, "\\xc0 =~ $pattern; Neither pattern nor target utf8";
1245        ok $c =~ /$pattern/i, "\\xc0 =~ /$pattern/i; Neither pattern nor target utf8";
1246        ok $c =~ $utf8_pattern, "\\xc0 =~ $pattern; pattern utf8, target not";
1247        ok $c =~ /$utf8_pattern/i, "\\xc0 =~ /$pattern/i; pattern utf8, target not";
1248        utf8::upgrade($c);
1249        ok $c =~ $pattern, "\\xc0 =~ $pattern; target utf8, pattern not";
1250        ok $c =~ /$pattern/i, "\\xc0 =~ /$pattern/i; target utf8, pattern not";
1251        ok $c =~ $utf8_pattern, "\\xc0 =~ $pattern; Both target and pattern utf8";
1252        ok $c =~ /$utf8_pattern/i, "\\xc0 =~ /$pattern/i; Both target and pattern utf8";
1253    }
1254
1255    {   # Make sure can override the formatting
1256        use feature 'unicode_strings';
1257        ok uni_to_native("\xc0") =~ /\w/, 'Under unicode_strings: "\xc0" =~ /\w/';
1258        ok uni_to_native("\xc0") !~ /(?d:\w)/, 'Under unicode_strings: "\xc0" !~ /(?d:\w)/';
1259    }
1260
1261    {
1262        my $str= "\x{100}";
1263        chop $str;
1264        my $qr= qr/$str/;
1265        is("$qr", "(?^:)", "Empty pattern qr// stringifies to (?^:) with unicode flag enabled - Bug #80212");
1266        $str= "";
1267        $qr= qr/$str/;
1268        is("$qr", "(?^:)", "Empty pattern qr// stringifies to (?^:) with unicode flag disabled - Bug #80212");
1269
1270    }
1271
1272    {
1273        local $::TODO = "[perl #38133]";
1274
1275        "A" =~ /(((?:A))?)+/;
1276        my $first = $2;
1277
1278        "A" =~ /(((A))?)+/;
1279        my $second = $2;
1280
1281        is($first, $second);
1282    }
1283
1284    {
1285	# RT #3516: \G in a m//g expression causes problems
1286	my $count = 0;
1287	while ("abc" =~ m/(\G[ac])?/g) {
1288	    last if $count++ > 10;
1289	}
1290	ok($count < 10, 'RT #3516 A');
1291
1292	$count = 0;
1293	while ("abc" =~ m/(\G|.)[ac]/g) {
1294	    last if $count++ > 10;
1295	}
1296	ok($count < 10, 'RT #3516 B');
1297
1298	$count = 0;
1299	while ("abc" =~ m/(\G?[ac])?/g) {
1300	    last if $count++ > 10;
1301	}
1302	ok($count < 10, 'RT #3516 C');
1303    }
1304    {
1305        # RT #84294: Is this a bug in the simple Perl regex?
1306        #          : Nested buffers and (?{...}) dont play nicely on partial matches
1307        our @got= ();
1308        ok("ab" =~ /((\w+)(?{ push @got, $2 })){2}/,"RT #84294: Pattern should match");
1309        my $want= "'ab', 'a', 'b'";
1310        my $got= join(", ", map { defined($_) ? "'$_'" : "undef" } @got);
1311        is($got,$want,'RT #84294: check that "ab" =~ /((\w+)(?{ push @got, $2 })){2}/ leaves @got in the correct state');
1312    }
1313
1314    {
1315        # Suppress warnings, as the non-unicode one comes out even if turn off
1316        # warnings here (because the execution is done in another scope).
1317        local $SIG{__WARN__} = sub {};
1318        my $str = "\x{110000}";
1319
1320        unlike($str, qr/\p{ASCII_Hex_Digit=True}/, "Non-Unicode doesn't match \\p{AHEX=True}");
1321        like($str, qr/\p{ASCII_Hex_Digit=False}/, "Non-Unicode matches \\p{AHEX=False}");
1322        like($str, qr/\P{ASCII_Hex_Digit=True}/, "Non-Unicode matches \\P{AHEX=True}");
1323        unlike($str, qr/\P{ASCII_Hex_Digit=False}/, "Non-Unicode matches \\P{AHEX=FALSE}");
1324    }
1325
1326    {
1327        # Test that IDstart works, but because the author (khw) knows
1328        # regexes much better than the rest of the core, it is being done here
1329        # in the context of a regex which relies on buffer names beginng with
1330        # IDStarts.
1331        use utf8;
1332        my $str = "abc";
1333        like($str, qr/(?<a>abc)/, "'a' is legal IDStart");
1334        like($str, qr/(?<_>abc)/, "'_' is legal IDStart");
1335        like($str, qr/(?<ß>abc)/, "U+00DF is legal IDStart");
1336        like($str, qr/(?<ℕ>abc)/, "U+2115' is legal IDStart");
1337
1338        # This test works on Unicode 6.0 in which U+2118 and U+212E are legal
1339        # IDStarts there, but are not Word characters, and therefore Perl
1340        # doesn't allow them to be IDStarts.  But there is no guarantee that
1341        # Unicode won't change things around in the future so that at some
1342        # future Unicode revision these tests would need to be revised.
1343        foreach my $char ("%", "×", chr(0x2118), chr(0x212E)) {
1344            my $prog = <<"EOP";
1345use utf8;;
1346"abc" =~ qr/(?<$char>abc)/;
1347EOP
1348            utf8::encode($prog);
1349            fresh_perl_like($prog, qr!Group name must start with a non-digit word character!, {},
1350                        sprintf("'U+%04X not legal IDFirst'", ord($char)));
1351        }
1352    }
1353
1354    { # [perl #101710]
1355        my $pat = "b";
1356        utf8::upgrade($pat);
1357        like("\xffb", qr/$pat/i, "/i: utf8 pattern, non-utf8 string, latin1-char preceding matching char in string");
1358    }
1359
1360    { # Crash with @a =~ // warning
1361	local $SIG{__WARN__} = sub {
1362             pass 'no crash for @a =~ // warning'
1363        };
1364	eval ' sub { my @a =~ // } ';
1365    }
1366
1367    { # Concat overloading and qr// thingies
1368	my @refs;
1369	my $qr = qr//;
1370        package Cat {
1371            require overload;
1372            overload->import(
1373		'""' => sub { ${$_[0]} },
1374		'.' => sub {
1375		    push @refs, ref $_[1] if ref $_[1];
1376		    bless $_[2] ? \"$_[1]${$_[0]}" : \"${$_[0]}$_[1]"
1377		}
1378            );
1379	}
1380	my $s = "foo";
1381	my $o = bless \$s, Cat::;
1382	/$o$qr/;
1383	is "@refs", "Regexp", '/$o$qr/ passes qr ref to cat overload meth';
1384    }
1385
1386    {
1387        my $count=0;
1388        my $str="\n";
1389        $count++ while $str=~/.*/g;
1390        is $count, 2, 'test that ANCH_MBOL works properly. We should get 2 from $count++ while "\n"=~/.*/g';
1391        my $class_count= 0;
1392        $class_count++ while $str=~/[^\n]*/g;
1393        is $class_count, $count, 'while "\n"=~/.*/g and while "\n"=~/[^\n]*/g should behave the same';
1394        my $anch_count= 0;
1395        $anch_count++ while $str=~/^.*/mg;
1396        is $anch_count, 1, 'while "\n"=~/^.*/mg should match only once';
1397    }
1398
1399    { # [perl #111174]
1400        use re '/u';
1401        my $A_grave = uni_to_native("\xc0");
1402        like uni_to_native("\xe0"), qr/(?i:$A_grave)/, "(?i: shouldn't lose the passed in /u";
1403        use re '/a';
1404        unlike "\x{100}", qr/(?i:\w)/, "(?i: shouldn't lose the passed in /a";
1405        use re '/aa';
1406        unlike 'k', qr/(?i:\N{KELVIN SIGN})/, "(?i: shouldn't lose the passed in /aa";
1407        unlike 'k', qr'(?i:\N{KELVIN SIGN})', "(?i: shouldn't lose the passed in /aa";
1408    }
1409
1410    {
1411	# the test for whether the pattern should be re-compiled should
1412	# consider the UTF8ness of the previous and current pattern
1413	# string, as well as the physical bytes of the pattern string
1414
1415	for my $s (byte_utf8a_to_utf8n("\xc4\x80"), "\x{100}") {
1416	    ok($s =~ /^$s$/, "re-compile check is UTF8-aware");
1417	}
1418    }
1419
1420    #  #113682 more overloading and qr//
1421    # when doing /foo$overloaded/, if $overloaded returns
1422    # a qr/(?{})/ via qr or "" overloading, then 'use re 'eval'
1423    # shouldn't be required. Via '.', it still is.
1424    {
1425        package Qr0;
1426	use overload 'qr' => sub { qr/(??{50})/ };
1427
1428        package Qr1;
1429	use overload '""' => sub { qr/(??{51})/ };
1430
1431        package Qr2;
1432	use overload '.'  => sub { $_[1] . qr/(??{52})/ };
1433
1434        package Qr3;
1435	use overload '""' => sub { qr/(??{7})/ },
1436		     '.'  => sub { $_[1] . qr/(??{53})/ };
1437
1438        package Qr_indirect;
1439	use overload '""'  => sub { $_[0][0] };
1440
1441	package main;
1442
1443	for my $i (0..3) {
1444	    my $o = bless [], "Qr$i";
1445	    if ((0,0,1,1)[$i]) {
1446		eval { "A5$i" =~ /^A$o$/ };
1447		like($@, qr/Eval-group not allowed/, "Qr$i");
1448		eval { "5$i" =~ /$o/ };
1449		like($@, ($i == 3 ? qr/^$/ : qr/no method found,/),
1450			"Qr$i bare");
1451		{
1452		    use re 'eval';
1453		    ok("A5$i" =~ /^A$o$/, "Qr$i - with use re eval");
1454		    eval { "5$i" =~ /$o/ };
1455		    like($@, ($i == 3 ? qr/^$/ : qr/no method found,/),
1456			    "Qr$i bare - with use re eval");
1457		}
1458	    }
1459	    else {
1460		ok("A5$i" =~ /^A$o$/, "Qr$i");
1461		ok("5$i" =~ /$o/, "Qr$i bare");
1462	    }
1463	}
1464
1465	my $o = bless [ bless [], "Qr1" ], 'Qr_indirect';
1466	ok("A51" =~ /^A$o/, "Qr_indirect");
1467	ok("51" =~ /$o/, "Qr_indirect bare");
1468    }
1469
1470    {   # Various flags weren't being set when a [] is optimized into an
1471        # EXACTish node
1472        ok("\x{017F}\x{017F}" =~ qr/^[$sharp_s]?$/i, "[] to EXACTish optimization");
1473    }
1474
1475    {   # Test that it avoids spllitting a multi-char fold across nodes.
1476        # These all fold to things that are like 'ss', which, if split across
1477        # nodes could fail to match a single character that folds to the
1478        # combination.  1F0 byte expands when folded;
1479        my $utf8_locale = find_utf8_ctype_locale();
1480        for my $char('F', $sharp_s, "\x{1F0}", "\x{FB00}") {
1481            my $length = 260;    # Long enough to overflow an EXACTFish regnode
1482            my $p = $char x $length;
1483            my $s = ($char eq $sharp_s) ? 'ss'
1484                                        : $char eq "\x{1F0}"
1485                                          ? "j\x{30c}"
1486                                          : 'ff';
1487            $s = $s x $length;
1488            for my $charset (qw(u d l aa)) {
1489                for my $utf8 (0..1) {
1490                    for my $locale ('C', $utf8_locale) {
1491                      SKIP:
1492                        {
1493                            skip "test skipped for non-C locales", 2
1494                                    if $charset ne 'l'
1495                                    && (! defined $locale || $locale ne 'C');
1496                            if ($charset eq 'l') {
1497                                skip 'Locales not available', 2
1498                                                            unless $has_locales;
1499                                if (! defined $locale) {
1500                                    skip "No UTF-8 locale", 2;
1501                                }
1502                                skip "Can't test in miniperl",2
1503                                  if is_miniperl();
1504
1505                                require POSIX;
1506                                POSIX::setlocale(&LC_CTYPE, $locale);
1507                            }
1508
1509                            my $pat = $p;
1510                            utf8::upgrade($pat) if $utf8;
1511                            my $should_pass =
1512                              (    $charset eq 'u'
1513                               || ($charset eq 'd' && $utf8)
1514                               || ($charset eq 'd' && (   $char =~ /[[:ascii:]]/
1515                                                       || ord $char > 255))
1516                               || ($charset eq 'aa' && $char =~ /[[:ascii:]]/)
1517                               || ($charset eq 'l' && $locale ne 'C')
1518                               || ($charset eq 'l' && $char =~ /[[:ascii:]]/)
1519                              );
1520                            my $name = "(?i$charset), utf8=$utf8, locale=$locale,"
1521                              . " char=" . sprintf "%x", ord $char;
1522                            no warnings 'locale';
1523                            is (eval " '$s' =~ qr/(?i$charset)$pat/;",
1524                                $should_pass, $name);
1525                            fail "$name: $@" if $@;
1526                            is (eval " 'a$s' =~ qr/(?i$charset)a$pat/;",
1527                                $should_pass, "extra a, $name");
1528                            fail "$name: $@" if $@;
1529                        }
1530                    }
1531                }
1532            }
1533        }
1534    }
1535
1536    SKIP:
1537    {
1538        skip "no re debug", 5 if is_miniperl;
1539        my $s = ("0123456789" x 26214) x 2; # Should fill 2 LEXACTS, plus
1540                                            # small change
1541        my $pattern_prefix = "use utf8; use re qw(Debug COMPILE)";
1542        my $pattern = "$pattern_prefix; qr/$s/;";
1543        my $result = fresh_perl($pattern);
1544        if ($? != 0) {  # Re-run so as to display STDERR.
1545            fail($pattern);
1546            fresh_perl($pattern, { stderr => 0, verbose => 1 });
1547        }
1548        like($result, qr/Final program[^X]*\bLEXACT\b[^X]*\bLEXACT\b[^X]*\bEXACT\b[^X]*\bEND\b/s,
1549             "Check that LEXACT nodes are generated");
1550        like($s, qr/$s/, "Check that LEXACT nodes match");
1551        like("a$s", qr/a$s/, "Previous test preceded by an 'a'");
1552        substr($s, 260000, 1) = "\x{100}";
1553        $pattern = "$pattern_prefix; qr/$s/;";
1554        $result = fresh_perl($pattern, { 'wide_chars' => 1 } );
1555        if ($? != 0) {  # Re-run so as to display STDERR.
1556            fail($pattern);
1557            fresh_perl($pattern, { stderr => 0, verbose => 1 });
1558        }
1559        like($result, qr/Final program[^X]*\bLEXACT_REQ8\b[^X]*\bLEXACT\b[^X]*\bEXACT\b[^X]*\bEND\b/s,
1560             "Check that an LEXACT_ONLY node is generated with a \\x{100}");
1561        like($s, qr/$s/, "Check that LEXACT_REQ8 nodes match");
1562    }
1563
1564    {
1565        for my $char (":", uni_to_native("\x{f7}"), "\x{2010}") {
1566            my $utf8_char = $char;
1567            utf8::upgrade($utf8_char);
1568            my $display = $char;
1569            $display = display($display);
1570            my $utf8_display = "utf8::upgrade(\"$display\")";
1571
1572            like($char, qr/^$char?$/, "\"$display\" =~ /^$display?\$/");
1573            like($char, qr/^$utf8_char?$/, "my \$p = \"$display\"; utf8::upgrade(\$p); \"$display\" =~ /^\$p?\$/");
1574            like($utf8_char, qr/^$char?$/, "my \$c = \"$display\"; utf8::upgrade(\$c); \"\$c\" =~ /^$display?\$/");
1575            like($utf8_char, qr/^$utf8_char?$/, "my \$c = \"$display\"; utf8::upgrade(\$c); my \$p = \"$display\"; utf8::upgrade(\$p); \"\$c\" =~ /^\$p?\$/");
1576        }
1577    }
1578
1579    {
1580	# #116148: Pattern utf8ness sticks around globally
1581	# the utf8 in the first match was sticking around for the second
1582	# match
1583
1584	use feature 'unicode_strings';
1585
1586	my $x = "\x{263a}";
1587	$x =~ /$x/;
1588
1589	my $text = "Perl";
1590	ok("Perl" =~ /P.*$/i, '#116148');
1591    }
1592
1593    { # 118297: Mixing up- and down-graded strings in regex
1594        utf8::upgrade(my $u = "\x{e5}");
1595        utf8::downgrade(my $d = "\x{e5}");
1596        my $warned;
1597        local $SIG{__WARN__} = sub { $warned++ if $_[0] =~ /\AMalformed UTF-8/ };
1598        my $re = qr/$u$d/;
1599        ok(!$warned, "no warnings when interpolating mixed up-/downgraded strings in pattern");
1600        my $c = "\x{e5}\x{e5}";
1601        utf8::downgrade($c);
1602        like($c, $re, "mixed up-/downgraded pattern matches downgraded string");
1603        utf8::upgrade($c);
1604        like($c, $re, "mixed up-/downgraded pattern matches upgraded string");
1605    }
1606
1607    {
1608        # if we have 87 capture buffers defined then \87 should refer to the 87th.
1609        # test that this is true for 1..100
1610        # Note that this test causes the engine to recurse at runtime, and
1611        # hence use a lot of C stack.
1612
1613        # Compiling for all 100 nested captures blows the stack under
1614        # clang and ASan; reduce.
1615        my $max_captures = $Config{ccflags} =~ /sanitize/ ? 20 : 100;
1616
1617        for my $i (1..100) {
1618            if ($i > $max_captures) {
1619                pass("skipping $i buffers under ASan aa");
1620                pass("skipping $i buffers under ASan aba");
1621                next;
1622            }
1623            my $capture= "a";
1624            $capture= "($capture)" for 1 .. $i;
1625            for my $mid ("","b") {
1626                my $str= "a${mid}a";
1627                my $backref= "\\$i";
1628                eval {
1629                    ok($str=~/$capture$mid$backref/,"\\$i works with $i buffers '$str'=~/...$mid$backref/");
1630                    1;
1631                } or do {
1632                    is("$@","","\\$i works with $i buffers works with $i buffers '$str'=~/...$mid$backref/");
1633                };
1634            }
1635        }
1636    }
1637
1638    # this mixture of readonly (not COWable) and COWable strings
1639    # messed up the capture buffers under COW. The actual test results
1640    # are incidental; the issue is was an AddressSanitizer failure
1641    {
1642	my $c ='AB';
1643	my $res = '';
1644	for ($c, 'C', $c, 'DE') {
1645	    ok(/(.)/, "COWable match");
1646	    $res .= $1;
1647	}
1648	is($res, "ACAD");
1649    }
1650
1651
1652    {
1653	# RT #45667
1654	# /[#$x]/x didn't interpolate the var $x.
1655	my $b = 'cd';
1656	my $s = 'abcd$%#&';
1657	$s =~ s/[a#$b%]/X/g;
1658	is ($s, 'XbXX$XX&', 'RT #45667 without /x');
1659	$s = 'abcd$%#&';
1660	$s =~ s/[a#$b%]/X/gx;
1661	is ($s, 'XbXX$XX&', 'RT #45667 with /x');
1662    }
1663
1664    {
1665	no warnings "uninitialized";
1666	my @a;
1667	$a[1]++;
1668	/@a/;
1669	pass('no crash with /@a/ when array has nonexistent elems');
1670    }
1671
1672    {
1673	is runperl(prog => 'delete $::{qq-\cR-}; //; print qq-ok\n-'),
1674	   "ok\n",
1675	   'deleting *^R does not result in crashes';
1676	no warnings 'once';
1677	*^R = *caretRglobwithnoscalar;
1678	"" =~ /(?{42})/;
1679	is $^R, 42, 'assigning to *^R does not result in a crash';
1680	is runperl(
1681	     stderr => 1,
1682	     prog => 'eval q|'
1683	            .' q-..- =~ /(??{undef *^R;q--})(?{42})/; '
1684                    .' print qq-$^R\n-'
1685	            .'|'
1686	   ),
1687	   "42\n",
1688	   'undefining *^R within (??{}) does not result in a crash';
1689    }
1690
1691    SKIP: {   # Test literal range end point special handling
1692        unless ($::IS_EBCDIC) {
1693            skip "Valid only for EBCDIC", 24;
1694        }
1695
1696        like("\x89", qr/[i-j]/, '"\x89" should match [i-j]');
1697        unlike("\x8A", qr/[i-j]/, '"\x8A" shouldnt match [i-j]');
1698        unlike("\x90", qr/[i-j]/, '"\x90" shouldnt match [i-j]');
1699        like("\x91", qr/[i-j]/, '"\x91" should match [i-j]');
1700
1701        like("\x89", qr/[i-\N{LATIN SMALL LETTER J}]/, '"\x89" should match [i-\N{LATIN SMALL LETTER J}]');
1702        unlike("\x8A", qr/[i-\N{LATIN SMALL LETTER J}]/, '"\x8A" shouldnt match [i-\N{LATIN SMALL LETTER J}]');
1703        unlike("\x90", qr/[i-\N{LATIN SMALL LETTER J}]/, '"\x90" shouldnt match [i-\N{LATIN SMALL LETTER J}]');
1704        like("\x91", qr/[i-\N{LATIN SMALL LETTER J}]/, '"\x91" should match [i-\N{LATIN SMALL LETTER J}]');
1705
1706        like("\x89", qr/[i-\N{U+6A}]/, '"\x89" should match [i-\N{U+6A}]');
1707        unlike("\x8A", qr/[i-\N{U+6A}]/, '"\x8A" shouldnt match [i-\N{U+6A}]');
1708        unlike("\x90", qr/[i-\N{U+6A}]/, '"\x90" shouldnt match [i-\N{U+6A}]');
1709        like("\x91", qr/[i-\N{U+6A}]/, '"\x91" should match [i-\N{U+6A}]');
1710
1711        like("\x89", qr/[\N{U+69}-\N{U+6A}]/, '"\x89" should match [\N{U+69}-\N{U+6A}]');
1712        unlike("\x8A", qr/[\N{U+69}-\N{U+6A}]/, '"\x8A" shouldnt match [\N{U+69}-\N{U+6A}]');
1713        unlike("\x90", qr/[\N{U+69}-\N{U+6A}]/, '"\x90" shouldnt match [\N{U+69}-\N{U+6A}]');
1714        like("\x91", qr/[\N{U+69}-\N{U+6A}]/, '"\x91" should match [\N{U+69}-\N{U+6A}]');
1715
1716        like("\x89", qr/[i-\x{91}]/, '"\x89" should match [i-\x{91}]');
1717        like("\x8A", qr/[i-\x{91}]/, '"\x8A" should match [i-\x{91}]');
1718        like("\x90", qr/[i-\x{91}]/, '"\x90" should match [i-\x{91}]');
1719        like("\x91", qr/[i-\x{91}]/, '"\x91" should match [i-\x{91}]');
1720
1721        # Need to use eval, because tries to compile on ASCII platforms even
1722        # though the tests are skipped, and fails because 0x89-j is an illegal
1723        # range there.
1724        like("\x89", eval 'qr/[\x{89}-j]/', '"\x89" should match [\x{89}-j]');
1725        like("\x8A", eval 'qr/[\x{89}-j]/', '"\x8A" should match [\x{89}-j]');
1726        like("\x90", eval 'qr/[\x{89}-j]/', '"\x90" should match [\x{89}-j]');
1727        like("\x91", eval 'qr/[\x{89}-j]/', '"\x91" should match [\x{89}-j]');
1728    }
1729
1730    # These are based on looking at the code in regcomp.c
1731    # We don't look for specific code, just the existence of an SSC
1732    foreach my $re (qw(     qr/a?c/
1733                            qr/a?c/i
1734                            qr/[ab]?c/
1735                            qr/\R?c/
1736                            qr/\d?c/d
1737                            qr/\w?c/l
1738                            qr/\s?c/a
1739                            qr/[[:lower:]]?c/u
1740    )) {
1741      SKIP: {
1742        skip "no re-debug under miniperl" if is_miniperl;
1743        my $prog = <<"EOP";
1744use re qw(Debug COMPILE);
1745$re;
1746EOP
1747        fresh_perl_like($prog, qr/synthetic stclass/, { stderr=>1 }, "$re generates a synthetic start class");
1748      }
1749    }
1750
1751    {
1752        like "\x{AA}", qr/a?[\W_]/d, "\\W with /d synthetic start class works";
1753    }
1754
1755    SKIP: {
1756        skip("Tests are ASCII-centric, some would fail on EBCDIC", 12) if $::IS_EBCDIC;
1757
1758        # Verify that the very last Latin-1 U+00FF
1759        # (LATIN SMALL LETTER Y WITH DIAERESIS)
1760        # and its UPPER counterpart (U+0178 which is pure Unicode),
1761        # and likewise for the very first pure Unicode
1762        # (LATIN CAPITAL LETTER A WITH MACRON) fold-match properly,
1763        # and there are no off-by-one logic errors in the transition zone.
1764
1765        ok("\xFF" =~ /\xFF/i, "Y WITH DIAERESIS l =~ l");
1766        ok("\xFF" =~ /\x{178}/i, "Y WITH DIAERESIS l =~ u");
1767        ok("\x{178}" =~ /\xFF/i, "Y WITH DIAERESIS u =~ l");
1768        ok("\x{178}" =~ /\x{178}/i, "Y WITH DIAERESIS u =~ u");
1769
1770        # U+00FF with U+05D0 (non-casing Hebrew letter).
1771        ok("\xFF\x{5D0}" =~ /\xFF\x{5D0}/i, "Y WITH DIAERESIS l =~ l");
1772        ok("\xFF\x{5D0}" =~ /\x{178}\x{5D0}/i, "Y WITH DIAERESIS l =~ u");
1773        ok("\x{178}\x{5D0}" =~ /\xFF\x{5D0}/i, "Y WITH DIAERESIS u =~ l");
1774        ok("\x{178}\x{5D0}" =~ /\x{178}\x{5D0}/i, "Y WITH DIAERESIS u =~ u");
1775
1776        # U+0100.
1777        ok("\x{100}" =~ /\x{100}/i, "A WITH MACRON u =~ u");
1778        ok("\x{100}" =~ /\x{101}/i, "A WITH MACRON u =~ l");
1779        ok("\x{101}" =~ /\x{100}/i, "A WITH MACRON l =~ u");
1780        ok("\x{101}" =~ /\x{101}/i, "A WITH MACRON l =~ l");
1781    }
1782
1783    {
1784        use utf8;
1785        ok("abc" =~ /a…b…c/x, "NEL is white-space under /x");
1786    }
1787
1788    {
1789        ok('a(b)c' =~ qr(a\(b\)c), "'\\(' is a literal in qr(...)");
1790        ok('a[b]c' =~ qr[a\[b\]c], "'\\[' is a literal in qr[...]");
1791        ok('a{3}c' =~ qr{a\{3\}c},  # Only failed when { could be a meta
1792              "'\\{' is a literal in qr{...}, where it could be a quantifier");
1793
1794        # This one is for completeness
1795        ok('a<b>c' =~ qr<a\<b\>c>, "'\\<' is a literal in qr<...>)");
1796    }
1797
1798    {   # Was getting optimized into EXACT (non-folding node)
1799        my $x = qr/[x]/i;
1800        utf8::upgrade($x);
1801        like("X", qr/$x/, "UTF-8 of /[x]/i matches upper case");
1802    }
1803
1804    {   # Special handling of literal-ended ranges in [...] was breaking this
1805        use utf8;
1806        like("ÿ", qr/[ÿ-ÿ]/, "\"ÿ\" should match [ÿ-ÿ]");
1807    }
1808
1809    {	# [perl #123539]
1810        like("TffffffffffffTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT5TTTTTTTTTTTTTTTTTTTTTTTTT3TTgTTTTTTTTTTTTTTTTTTTTT2TTTTTTTTTTTTTTTTTTTTTTTHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHiHHHHHHHfffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff&ffff", qr/TffffffffffffTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT5TTTTTTTTTTTTTTTTTTTTTTTTT3TTgTTTTTTTTTTTTTTTTTTTTT2TTTTTTTTTTTTTTTTTTTTTTTHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHiHHHHHHHfffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff&ffff/il, "");
1811        like("TffffffffffffT\x{100}TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT5TTTTTTTTTTTTTTTTTTTTTTTTT3TTgTTTTTTTTTTTTTTTTTTTTT2TTTTTTTTTTTTTTTTTTTTTTTHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHiHHHHHHHfffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff&ffff", qr/TffffffffffffT\x{100}TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT5TTTTTTTTTTTTTTTTTTTTTTTTT3TTgTTTTTTTTTTTTTTTTTTTTT2TTTTTTTTTTTTTTTTTTTTTTTHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHiHHHHHHHfffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff&ffff/il, "");
1812    }
1813
1814	{	# [perl #123604]
1815		my($s, $x, @x) = ('abc', 'a', 'd');
1816		my $long = 'b' x 2000;
1817		my $eval = q{$s =~ m{$x[bbb]c} ? 1 : 0};
1818		$eval =~ s{bbb}{$long};
1819		my $match = eval $eval;
1820		ok(1, "did not crash");
1821		ok($match, "[bbb...] resolved as character class, not subscript");
1822	}
1823
1824	{	# [perl #123755]
1825		for my $pat ('(??', '(?P', '(?i-') {
1826			eval qq{ qr/$pat/ };
1827			ok(1, "qr/$pat/ did not crash");
1828			eval qq{ qr/${pat}\x{123}/ };
1829			my $e = $@;
1830			like($e, qr{\x{123}},
1831				"qr/${pat}x/ shows x in error even if it's a wide character");
1832		}
1833	}
1834
1835	{
1836		# Expect one of these sizes to cause overflow and wrap to negative
1837		for my $bits (32, 64) {
1838			my $wrapneg = 2 ** ($bits - 2) * 3;
1839			for my $sign ('', '-') {
1840				my $pat = sprintf "qr/(?%s%u)/", $sign, $wrapneg;
1841				eval $pat;
1842				ok(1, "big backref $pat did not crash");
1843			}
1844		}
1845	}
1846        {
1847            # Test that we handle qr/\8888888/ and variants without an infinite loop,
1848            # we use a test within a test so we can todo it, and make sure we don't
1849            # infinite loop our tests.
1850            # NOTE - Do not put quotes in the code!
1851            # NOTE - We have to triple escape the backref in the pattern below.
1852            my $code='
1853                BEGIN{require q(./test.pl);}
1854                watchdog(3);
1855                for my $len (1 .. 20) {
1856                    my $eights= q(8) x $len;
1857                    eval qq{ qr/\\\\$eights/ };
1858                }
1859                print q(No infinite loop here!);
1860            ';
1861            fresh_perl_is($code, "No infinite loop here!", {},
1862                "test that we handle things like m/\\888888888/ without infinite loops" );
1863        }
1864
1865        SKIP:
1866        {   # Test that we handle some malformed UTF-8 without looping [perl
1867            # #123562]
1868            skip "no Encode", 1 if is_miniperl;
1869            my $code='
1870                BEGIN{require q(./test.pl);}
1871                use Encode qw(_utf8_on);
1872                # \x80 and \x41 are continuation bytes in their respective
1873                # character sets
1874                my $malformed = (ord("A") == 65) ? "a\x80\n" : "a\x41\n";
1875                utf8::downgrade($malformed);
1876                _utf8_on($malformed);
1877                watchdog(3);
1878                $malformed =~ /(\n\r|\r)$/;
1879                print q(No infinite loop here!);
1880            ';
1881            fresh_perl_like($code, qr/Malformed UTF-8 character/, {},
1882                "test that we handle some UTF-8 malformations without looping" );
1883        }
1884
1885	{
1886		# [perl #123843] hits SEGV trying to compile this pattern
1887		my $match;
1888		eval q{ ($match) = ("xxyxxyxy" =~ m{(x+(y(?1))*)}) };
1889		ok(1, "compiled GOSUB in CURLYM ok");
1890		is($match, 'xxyxxyx', "matched GOSUB in CURLYM");
1891	}
1892
1893	{
1894		# [perl #123852] doesn't avoid all the capture-related work with
1895		# //n, leading to possible memory corruption
1896		eval q{ qr{()(?1)}n };
1897		my $error = $@;
1898		ok(1, "qr{()(?1)}n didn't crash");
1899		like($error, qr{Reference to nonexistent group},
1900				'gave appropriate error for qr{()(?1)}n');
1901	}
1902
1903	{
1904            # [perl #126406] panic with unmatchable quantifier
1905            my $code='
1906                no warnings "regexp";
1907                "" =~ m/(.0\N{6,0}0\N{6,0}000000000000000000000000000000000)/;
1908            ';
1909            fresh_perl_is($code, "", {},
1910                            "perl [#126406] panic");
1911	}
1912        {
1913            my $bug="[perl #126182]"; # test for infinite pattern recursion
1914            for my $tuple (
1915                    [ 'q(a)=~/(.(?2))((?<=(?=(?1)).))/', "died", "look ahead left recursion fails fast" ],
1916                    [ 'q(aa)=~/(?R)a/', "died", "left-recursion fails fast", ],
1917                    [ 'q(bbaa)=~/(?&x)(?(DEFINE)(?<x>(?&y)*a)(?<y>(?&x)*b))/',
1918                        "died", "inter-cyclic optional left recursion dies" ],
1919                    [ 'q(abc) =~ /a((?1)?)c/', "died", "optional left recursion dies" ],
1920                    [ 'q(abc) =~ /a((?1)??)c/', "died", "min mod left recursion dies" ],
1921                    [ 'q(abc) =~ /a((?1)*)c/', "died", "* left recursion dies" ],
1922                    [ 'q(abc) =~ /a((?1)+)c/', "died", "+ left recursion dies" ],
1923                    [ 'q(abc) =~ /a((?1){0,3})c/', "died", "{0,3} left recursion fails fast" ],
1924
1925                    [ 'q(aaabbb)=~/a(?R)?b/', "matched", "optional self recursion works" ],
1926                    [ '"((5maa-maa)(maa-3maa))" =~ /(\\\\((?:[^()]++|(?0))*+\\\\))/', "matched",
1927                        "recursion and possessive captures", "((5maa-maa)(maa-3maa))"],
1928                    [ '"((5maa-maa)(maa-3maa))" =~ /(\\\\((?:[^()]++|(?1))*+\\\\))/', "matched",
1929                        "recursion and possessive captures", "((5maa-maa)(maa-3maa))"],
1930                    [ '"((5maa-maa)(maa-3maa))" =~ /(\\\\((?:[^()]+|(?0))*\\\\))/', "matched",
1931                        "recursion and possessive captures", "((5maa-maa)(maa-3maa))"],
1932                    [ '"((5maa-maa)(maa-3maa))" =~ /(\\\\((?:[^()]+|(?1))*\\\\))/', "matched",
1933                        "recursion and possessive captures", "((5maa-maa)(maa-3maa))"],
1934            ) {
1935                my ($expr, $expect, $test_name, $cap1)= @$tuple;
1936                # avoid quotes in this code!
1937                my $code='
1938                    BEGIN{require q(./test.pl);}
1939                    watchdog(3);
1940                    my $status= eval(q{ !(' . $expr . ') ? q(failed) : ' .
1941                        ($cap1 ? '($1 ne q['.$cap1.']) ? qq(badmatch:$1) : ' : '') .
1942                        ' q(matched) })
1943                                || ( ( $@ =~ /Infinite recursion/ ) ? qq(died) : q(strange-death) );
1944                    print $status;
1945                ';
1946                fresh_perl_is($code, $expect, {}, "$bug - $test_name" );
1947            }
1948        }
1949        {
1950            my $is_cygwin = $^O eq "cygwin";
1951            local $::TODO = "this flaps on github cygwin vm, but not on cygwin iron #18129"
1952              if $is_cygwin;
1953            my $expected = "Timeout";
1954            my $code = '
1955                BEGIN{require q(test.pl);}
1956                watchdog(3);
1957                $SIG{ALRM} = sub {print "'.$expected.'\n"; exit(1)};
1958                alarm 1;
1959                $_ = "a" x 1000 . "b" x 1000 . "c" x 1000;
1960                /.*a.*b.*c.*[de]/;
1961                print "increase the multipliers in the regex above to run the regex longer";
1962            ';
1963            # this flaps on github cygwin vm, but not on cygwin iron #18129
1964            # so on cygwin it's run for 50 seconds to see if it fails eventually
1965            my $max = $is_cygwin ? 50 : 1;
1966            my ($iter, $result, $status);
1967            for my $i (1..$max) {
1968                $iter = $i;
1969                $result = fresh_perl($code,{});
1970                $status = $?;
1971                last if $result ne $expected;
1972            }
1973            is($result, $expected, "Test Perl 73464")
1974              or diag "PROG:", $code, "STATUS:", $status, "failed on iteration: $iter";
1975        }
1976
1977        {   # [perl #128686], crashed the the interpreter
1978            my $AE = chr utf8::unicode_to_native(0xC6);
1979            my $ae = chr utf8::unicode_to_native(0xE6);
1980            my $re = qr/[$ae\s]/i;
1981            ok($AE !~ $re, '/[\xE6\s]/i doesn\'t match \xC6 when not in UTF-8');
1982            utf8::upgrade $AE;
1983            ok($AE =~ $re, '/[\xE6\s]/i matches \xC6 when in UTF-8');
1984        }
1985
1986        {
1987            is(0+("\n" =~ m'\n'), 1, q|m'\n' should interpolate escapes|);
1988        }
1989
1990        {
1991            my $str = "a\xB6";
1992            ok( $str =~ m{^(a|a\x{b6})$}, "fix [perl #129950] - latin1 case" );
1993            utf8::upgrade($str);
1994            ok( $str =~ m{^(a|a\x{b6})$}, "fix [perl #129950] - utf8 case" );
1995        }
1996        {
1997            my $got= run_perl( switches => [ '-l' ], prog => <<'EOF_CODE' );
1998            my $died= !eval {
1999                $_=qq(ab);
2000                print;
2001                my $p=qr/(?{ s!!x! })/;
2002                /$p/;
2003                print;
2004                /a/;
2005                /$p/;
2006                print;
2007                /b/;
2008                /$p/;
2009                print;
2010                //;
2011                1;
2012            };
2013            $error = $died ? ($@ || qq(Zombie)) : qq(none);
2014            print $died ? qq(died) : qq(lived);
2015            print qq(Error: $@);
2016EOF_CODE
2017            my @got= split /\n/, $got;
2018            is($got[0],"ab","empty pattern in regex codeblock: got expected start string");
2019            is($got[1],"xab",
2020                "empty pattern in regex codeblock: first subst with no last-match worked right");
2021            is($got[2],"xxb","empty pattern in regex codeblock: second subst worked right");
2022            is($got[3],"xxx","empty pattern in regex codeblock: third subst worked right");
2023            is($got[4],"died","empty pattern in regex codeblock: died as expected");
2024            like($got[5],qr/Error: Infinite recursion via empty pattern/,
2025           "empty pattern in regex codeblock: produced the right exception message" );
2026        }
2027
2028    # This test is based on the one directly above, which happened to
2029    # leak. Repeat the test, but stripped down to the bare essentials
2030    # of the leak, which is to die while executing a regex which is
2031    # already the current regex, thus causing the saved outer set of
2032    # capture offsets to leak. The test itself doesn't do anything
2033    # except sit around hoping not to be triggered by ASan
2034    {
2035        eval {
2036            my $s = "abcd";
2037            $s =~ m{([abcd]) (?{ die if $1 eq 'd'; })}gx;
2038            $s =~ //g;
2039            $s =~ //g;
2040            $s =~ //g;
2041        };
2042        pass("call to current regex doesn't leak");
2043    }
2044
2045
2046
2047    {
2048        # [perl #130495] /x comment skipping stopped a byte short, leading
2049        # to assertion failure or 'malformed utf-8 character" warning
2050        fresh_perl_is(
2051            "use utf8; m{a#\x{124}}x", '', {wide_chars => 1},
2052            '[perl #130495] utf-8 character at end of /x comment should not misparse',
2053        );
2054    }
2055    {
2056        # [perl #130522] causes out-of-bounds read detected by clang with
2057        # address=sanitized when length of the STCLASS string is greater than
2058        # length of target string.
2059        my $re = qr{(?=\0z)\0?z?$}i;
2060        my($yes, $no) = (1, "");
2061        for my $test (
2062            [ $no,  undef,   '<undef>' ],
2063            [ $no,  '',      '' ],
2064            [ $no,  "\0",    '\0' ],
2065            [ $yes, "\0z",   '\0z' ],
2066            [ $no,  "\0z\0", '\0z\0' ],
2067            [ $yes, "\0z\n", '\0z\n' ],
2068        ) {
2069            my($result, $target, $disp) = @$test;
2070            no warnings qw/uninitialized/;
2071            is($target =~ $re, $result, "[perl #130522] with target '$disp'");
2072        }
2073    }
2074    {
2075	# [perl #129377] backref to an unmatched capture should not cause
2076	# reading before start of string.
2077	SKIP: {
2078	    skip "no re-debug under miniperl" if is_miniperl;
2079	    my $prog = <<'EOP';
2080use re qw(Debug EXECUTE);
2081"x" =~ m{ () y | () \1 }x;
2082EOP
2083	    fresh_perl_like($prog, qr{
2084		\A (?! .* ^ \s+ - )
2085	    }msx, { stderr => 1 }, "Offsets in debug output are not negative");
2086	}
2087    }
2088    {
2089        # buffer overflow
2090
2091        # This test also used to leak - fixed by the commit which added
2092        # this line.
2093
2094        fresh_perl_is("BEGIN{\$^H=0x200000}\ns/[(?{//xx",
2095                      "Unmatched [ in regex; marked by <-- HERE in m/[ <-- HERE (?{/ at (eval 1) line 1.\n",
2096                      {}, "buffer overflow for regexp component");
2097    }
2098    {
2099        # [perl #129281] buffer write overflow, detected by ASAN, valgrind
2100        fresh_perl_is('/0(?0)|^*0(?0)|^*(^*())0|/', '', {}, "don't bump whilem_c too much");
2101    }
2102    {
2103        # RT #131893 - fails with ASAN -fsanitize=undefined
2104        fresh_perl_is('qr/0(0?(0||00*))|/', '', {}, "integer overflow during compilation");
2105    }
2106
2107    {
2108        # RT #131575 intuit skipping back from the end to find the highest
2109        # possible start point, was potentially hopping back beyond pos()
2110        # and crashing by calling fbm_instr with a negative length
2111
2112        my $text = "=t=\x{5000}";
2113        pos($text) = 3;
2114        ok(scalar($text !~ m{(~*=[a-z]=)}g), "RT #131575");
2115    }
2116    {
2117        fresh_perl_is('"AA" =~ m/AA{1,0}/','',{},"handle OPFAIL insert properly");
2118    }
2119    {
2120        fresh_perl_is('$_="0\x{1000000}";/^000?\0000/','',{},"dont throw assert errors trying to fbm past end of string");
2121    }
2122    {   # [perl $132227]
2123        fresh_perl_is("('0ba' . ('ss' x 300)) =~ m/0B\\N{U+41}" . $sharp_s x 150 . '/i and print "1\n"',  1,{},"Use of sharp s under /di that changes to /ui");
2124
2125        # A variation, but as far as khw knows not part of 132227
2126        fresh_perl_is("'0bssa' =~ m/0B" . $sharp_s . "\\N{U+41}" . '/i and print "1\n"',  1,{},"Use of sharp s under /di that changes to /ui");
2127    }
2128    {   # [perl $132164]
2129        fresh_perl_is('m m0*0+\Rm', "",{},"Undefined behavior in address sanitizer");
2130    }
2131    {   # [perl #133642]
2132        fresh_perl_is('no warnings "experimental::vlb";
2133                      m/((?<=(0?)))/', "",{},"Was getting 'Double free'");
2134    }
2135    {   # [perl #133782]
2136        # this would panic on DEBUGGING builds
2137        fresh_perl_is(<<'CODE', "ok\nok\n",{}, 'Bad length magic was left on $^R');
2138while( "\N{U+100}bc" =~ /(..?)(?{$^N})/g ) {
2139  print "ok\n" if length($^R)==length("$^R");
2140}
2141CODE
2142    }
2143    {   # [perl #133871], ASAN/valgrind out-of-bounds access
2144        fresh_perl_like('qr/(?|(())|())|//', qr/syntax error/, {}, "[perl #133871]");
2145    }
2146    {   # [perl #133871], ASAN/valgrind out-of-bounds access
2147        fresh_perl_like('qr/\p{nv:NAnq}/', qr/Can't find Unicode property definition/, {}, "GH #17367");
2148    }
2149    {   # GH #17370, ASAN/valgrind out-of-bounds access
2150        fresh_perl_like('qr/\p{nv:qnan}/', qr/Can't find Unicode property definition/, {}, "GH #17370");
2151    }
2152    {   # GH #17371, segfault
2153        fresh_perl_like('qr/\p{nv=\\\\\}(?0)|\337ss|\337ss//', qr/Unicode property wildcard not terminated/, {}, "GH #17371");
2154    }
2155    {   # GH #17384, ASAN/valgrind out-of-bounds access
2156        fresh_perl_like('"q0" =~ /\p{__::Is0}/', qr/Unknown user-defined property name \\p\{__::Is0}/, {}, "GH #17384");
2157    }
2158
2159  SKIP:
2160    {   # [perl #133921], segfault
2161        skip "Not valid for EBCDIC", 5 if $::IS_EBCDIC;
2162
2163        fresh_perl_is('qr0||ß+p00000F00000ù\Q00000ÿ00000x00000x0c0e0\Qx0\Qx0\x{0c!}\;\;î0\x�ÿÿÿþ���ù\Q`\Qx`�{0c!}e;���ù\ò`\Qm`\x{0c!}\;\;îçÿ �ç���!�F��/;îçÿù\Q���xÿÿÿÿ���ù���`x{0c!}e;���ù\Q`\Qx`\x{c!}\;\;îç!}\;îçÿù\Q‡ \x�ÿÿÿÿ��>=\Qx`\Qx`��ù\ò`\Qx`\x{0c!};\;îçÿ �F�n0t0�c �€d;t ���ù �ç€��!00000000000000000000000m/0000000000000000000000000000000m/\x{){} )|i', "", {}, "[perl #133921]");
2164        fresh_perl_is('|ß+W0ü0r0\Qx0\Qx0x0c0G00000000000000000O000000000x0x0x0c!}\;îçÿù\Q0 \x�ÿÿÿÿ���ù\Q`\Qx`�{0d ;���ù\ò`\Qm`\x{0c!}\;\;îçÿ �ç���!�F��/;îçÿù\Q���xÿÿÿÿ���ù���`x{0c!};���ù\Q`\Qq`\x{c!}\;\;îç!}\;îçÿù\Q‡ \x�ÿÿÿÿ��>=\Qx`\Qx`��ù\ò`\Qx`\x{0c!};\;îçÿ �0000000F�m0t0�c �€d;t ���ù �ç€��!00000000000000000000000m/0000000000000000000000000000000m/\x{){} )|i', "", {}, "[perl #133921]");
2165
2166fresh_perl_is('s|ß+W0ü0f0\Qx0\Qx0x0c0G0xgive0000000000000O0h000x0 \xòÿÿÿ��ù\Q`\Q
2167
2168
2169
2170
2171	ç
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187x{0c!}\;\;çÿ �q0/i0/!�F��/;îçÿù\Q���xÿÿÿÿ���ù���`x{0c!}e;���ù\Q`\Qx`\x{0c!}\;ÿÿÿÿ!}\;îçÿù\Q‡\x�ÿÿÿÿ��>=\Qx`\Qx`��ù\ò`ÿ��>=\Qx`\Qx`��ù\ò`\Qx`\x{0c!};\;îçÿ �u00000F�000t0�p �€d? ���ù �ç€��!00000000000000000000000m/0000000000000000000000000000000m/0�\���} )|i', "", {}, "[perl #133921]");
2188
2189        fresh_perl_is('a aú��úv sWtrt�\ó||ß+Wüef�ù\Qx`\Qx`\x{1c!gGnuc given1111111111111O1111each111\jx` \x�òÿÿÿ���ù\Qx`\Q
2190
2191
2192
2193
2194
2195	ç
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211x{1c!}\;\;îçÿp �qr/elsif/!�eF��/;îçÿù\Q���xÿÿÿÿ���ùHQx���`Lx{1c!}e;���ù\Qx`\Qx`\x{1c!}\;ÿÿÿÿc!}\;îçÿù\Qx‡\x�ÿÿÿÿ��>=\Qx`\Qx`��ù\òx`ÿ��>=\Qx`\Qx`��ù\òx`\Qx`\x{1c!}8;\;îçÿp �unshifteF�normat0�cmp �€d?not ���ùp �ç€��!0000000000000000000000000m/00000000000000000000000000000000m/0R�\���} )|\aï||K??p€¿ÿÿfúd{\{gri{\x{1x/}� ð¹NuntiÀh', "", {}, "[perl #133921]");
2212
2213    fresh_perl_is('s|ß+W0ü0f0\Qx0\Qx0x0c0g0c 000n0000000000000O0h000x0 \xòÿÿÿ��ù\Q`\Q
2214
2215
2216
2217
2218
2219	ç
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235x{0c!}\;\;îçÿ �/0f/!�F��/;îçÿù\Q���xÿÿÿÿ���ù���`x{0c!};���ù\Q`\Qx`\x{0c!}\;ÿÿÿÿ!}\;îçÿù\Q‡\x�ÿÿÿÿ��>=\Qx`\Qx`��ù\ò`ÿ��>=\Qx`\Qx`��ù\ò`\Qx`\x{0c!};\;îçÿ �000t0F�000t0�p �€d?n ���ù �ç€��!00000000000000000000000m/0000000000000000000000000000000m/�\���} )|i', "", {}, "[perl #133933]");
2236    }
2237
2238    {   # perl #133998]
2239        fresh_perl_is('print "\x{110000}" =~ qr/(?l)|[^\S\pC\s]/', 1, {},
2240        '/[\S\s]/l works');
2241    }
2242
2243    {   # perl #133995]
2244        use utf8;
2245        fresh_perl_is('"έδωσαν ελληνικήვე" =~ m/[^0](?=0)0?/', "",
2246                      {wide_chars => 1},
2247                      '[^0] doesnt crash on UTF-8 target string');
2248    }
2249
2250    {   # [perl #133992]  This is a tokenizer bug of parsing a pattern
2251        fresh_perl_is(q:$z = do {
2252                                use utf8;
2253                                "q!тест! =~ m'"
2254                        };
2255                        $z .= 'è(?#„';
2256                        $z .= "'";
2257                        eval $z;:, "", {}, 'foo');
2258    }
2259
2260    {   # [perl #134325]
2261        my $quote="\\Q";
2262        my $back="\\\\";
2263        my $ff="\xff";
2264        my $s = sprintf "/\\1|(|%s)%s%s   /i",
2265                        $quote x 8 . $back x 69,
2266                        $quote x 5 . $back x 4,
2267                        $ff x 48;
2268        like(fresh_perl("$s", { stderr => 1, }), qr/Unmatched \(/);
2269   }
2270
2271   {    # GitHub #17196, caused assertion failure
2272        fresh_perl_like('("0" x 258) =~ /(?l)0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000/',
2273                        qr/^$/,
2274                        {},
2275                        "Assertion failure with /l exact string longer than a single node");
2276    }
2277
2278SKIP:
2279    {   # [perl #134334], Assertion failure
2280        my $utf8_locale = find_utf8_ctype_locale();
2281        skip "no UTF-8 locale available" unless $utf8_locale;
2282        fresh_perl_like("use POSIX; POSIX::setlocale(&LC_CTYPE, '$utf8_locale'); 'ssss' =~ /\xDF+?sX/il;",
2283                        qr/^$/,
2284                        {},
2285                        "Assertion failure matching /il on single char folding to multi");
2286    }
2287
2288    {   # Test ANYOFHs
2289        my $pat = qr/[\x{4000001}\x{4000003}\x{4000005}]+/;
2290        unlike("\x{4000000}", $pat, "4000000 isn't in pattern");
2291        like("\x{4000001}", $pat, "4000001 is in pattern");
2292        unlike("\x{4000002}", $pat, "4000002 isn't in pattern");
2293        like("\x{4000003}", $pat, "4000003 is in pattern");
2294        unlike("\x{4000004}", $pat, "4000004 isn't in pattern");
2295        like("\x{4000005}", $pat, "4000005 is in pattern");
2296        unlike("\x{4000006}", $pat, "4000006 isn't in pattern");
2297
2298        # gh #17319
2299        $pat = qr/[\N{U+200D}\N{U+2000}]()/;
2300        unlike("\x{1FFF}", $pat, "1FFF isn't in pattern");
2301        like("\x{2000}", $pat, "2000 is in pattern");
2302        unlike("\x{2001}", $pat, "2001 isn't in pattern");
2303        unlike("\x{200C}", $pat, "200C isn't in pattern");
2304        like("\x{200D}", $pat, "200 is in pattern");
2305        unlike("\x{200E}", $pat, "200E isn't in pattern");
2306    }
2307
2308    # gh17490: test recursion check
2309    {
2310        my $eval = '(?{1})';
2311        my $re = sprintf '(?&FOO)(?(DEFINE)(?<FOO>%sfoo))', $eval x 20;
2312        my $result = eval qq{"foo" =~ /$re/};
2313        is($@ // '', '', "many evals did not die");
2314        ok($result, "regexp correctly matched");
2315    }
2316
2317    # gh16947: test regexp corruption (GOSUB)
2318    {
2319        fresh_perl_is(q{
2320            'xy' =~ /x(?0)|x(?|y|y)/ && print 'ok'
2321        }, 'ok', {}, 'gh16947: test regexp corruption (GOSUB)');
2322    }
2323    # gh16947: test fix doesn't break SUSPEND
2324    {
2325        fresh_perl_is(q{ 'sx' =~ m{ss++}i; print 'ok' },
2326                'ok', {}, "gh16947: test fix doesn't break SUSPEND");
2327    }
2328
2329    # gh17730: should not crash
2330    {
2331        fresh_perl_is(q{
2332            "q00" =~ m{(((*ACCEPT)0)*00)?0(?1)}; print "ok"
2333        }, 'ok', {}, 'gh17730: should not crash');
2334    }
2335
2336    # gh17743: more regexp corruption via GOSUB
2337    {
2338        fresh_perl_is(q{
2339            "0" =~ /((0(?0)|000(?|0000|0000)(?0))|)/; print "ok"
2340        }, 'ok', {}, 'gh17743: test regexp corruption (1)');
2341
2342        fresh_perl_is(q{
2343            "000000000000" =~ /(0(())(0((?0)())|000(?|\x{ef}\x{bf}\x{bd}|\x{ef}\x{bf}\x{bd}))|)/;
2344            print "ok"
2345        }, 'ok', {}, 'gh17743: test regexp corruption (2)');
2346    }
2347
2348} # End of sub run_tests
2349
23501;
2351
2352#
2353# ex: set ts=8 sts=4 sw=4 et:
2354#
2355