This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Merge branch 'multi-fold' into blead
[perl5.git] / t / re / pat_advanced.t
CommitLineData
e425a60b
YO
1#!./perl
2#
c5de0829 3# This is a home for regular expression tests that do not fit into
e425a60b
YO
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
3fe9640a
JH
7BEGIN {
8 chdir 't' if -d 't';
3fe9640a
JH
9 require './test.pl';
10 require './charset_tools.pl';
624c42e2 11 set_up_inc(qw '../lib .');
624c42e2 12 skip_all_if_miniperl("miniperl can't load Tie::Hash::NamedCapture, need for %+ and %-");
5b39eca1 13}
3fe9640a 14
e425a60b
YO
15use strict;
16use warnings;
17use 5.010;
fee50582 18our ($REGMARK, $REGERROR);
e425a60b 19
e425a60b
YO
20sub run_tests;
21
22$| = 1;
23
9d45b377 24run_tests() unless caller;
e425a60b
YO
25
26#
27# Tests start here.
28#
29sub run_tests {
30
e425a60b 31 {
e425a60b
YO
32 # Japhy -- added 03/03/2001
33 () = (my $str = "abc") =~ /(...)/;
34 $str = "def";
de26e0cc 35 is($1, "abc", 'Changing subject does not modify $1');
e425a60b
YO
36 }
37
e425a60b
YO
38 SKIP:
39 {
40 # The trick is that in EBCDIC the explicit numeric range should
41 # match (as also in non-EBCDIC) but the explicit alphabetic range
42 # should not match.
b7ad6c3e
KW
43 like "\x8e", qr/[\x89-\x91]/, '"\x8e" =~ /[\x89-\x91]/';
44 like "\xce", qr/[\xc9-\xd1]/, '"\xce" =~ /[\xc9-\xd1]/';
45 like "\xd0", qr/[\xc9-\xd1]/, '"\xd0" =~ /[\xc9-\xd1]/';
e425a60b
YO
46
47 skip "Not an EBCDIC platform", 2 unless ord ('i') == 0x89 &&
48 ord ('J') == 0xd1;
49
50 # In most places these tests would succeed since \x8e does not
51 # in most character sets match 'i' or 'j' nor would \xce match
52 # 'I' or 'J', but strictly speaking these tests are here for
53 # the good of EBCDIC, so let's test these only there.
b33825c4
NC
54 unlike("\x8e", qr/[i-j]/, '"\x8e" !~ /[i-j]/');
55 unlike("\xce", qr/[I-J]/, '"\xce" !~ /[I-J]/');
2ca8589c 56 unlike("\xd0", qr/[I-J]/, '"\xd0" !~ /[I-J]/');
e425a60b
YO
57 }
58
e425a60b 59 {
b7ad6c3e
KW
60 like "\x{ab}", qr/\x{ab}/, '"\x{ab}" =~ /\x{ab}/ ';
61 like "\x{abcd}", qr/\x{abcd}/, '"\x{abcd}" =~ /\x{abcd}/';
e425a60b
YO
62 }
63
e425a60b 64 {
ee95e30c 65 my $message = 'bug id 20001008.001 (#4407)';
e425a60b 66
f36cc39d
KW
67 my $strasse = "stra" . uni_to_native("\337") . "e";
68 my @x = ("$strasse 138", "$strasse 138");
e425a60b 69 for (@x) {
de946258
NC
70 ok(s/(\d+)\s*([\w\-]+)/$1 . uc $2/e, $message);
71 ok(my ($latin) = /^(.+)(?:\s+\d)/, $message);
f36cc39d
KW
72 is($latin, $strasse, $message);
73 ok($latin =~ s/$strasse/straße/, $message);
e425a60b
YO
74 #
75 # Previous code follows, but outcommented - there were no tests.
76 #
77 # $latin =~ s/stra\337e/straße/; # \303\237 after the 2nd a
78 # use utf8; # needed for the raw UTF-8
79 # $latin =~ s!(s)tr(?:aß|s+e)!$1tr.!; # \303\237 after the a
80 }
81 }
82
e425a60b 83 {
de946258
NC
84 my $message = 'Test \x escapes';
85 ok("ba\xd4c" =~ /([a\xd4]+)/ && $1 eq "a\xd4", $message);
86 ok("ba\xd4c" =~ /([a\xd4]+)/ && $1 eq "a\x{d4}", $message);
87 ok("ba\x{d4}c" =~ /([a\xd4]+)/ && $1 eq "a\x{d4}", $message);
88 ok("ba\x{d4}c" =~ /([a\xd4]+)/ && $1 eq "a\xd4", $message);
89 ok("ba\xd4c" =~ /([a\x{d4}]+)/ && $1 eq "a\xd4", $message);
90 ok("ba\xd4c" =~ /([a\x{d4}]+)/ && $1 eq "a\x{d4}", $message);
91 ok("ba\x{d4}c" =~ /([a\x{d4}]+)/ && $1 eq "a\x{d4}", $message);
92 ok("ba\x{d4}c" =~ /([a\x{d4}]+)/ && $1 eq "a\xd4", $message);
e425a60b
YO
93 }
94
e425a60b 95 {
de946258 96 my $message = 'Match code points > 255';
e425a60b 97 $_ = "abc\x{100}\x{200}\x{300}\x{380}\x{400}defg";
de946258
NC
98 ok(/(.\x{300})./, $message);
99 ok($` eq "abc\x{100}" && length ($`) == 4, $message);
100 ok($& eq "\x{200}\x{300}\x{380}" && length ($&) == 3, $message);
101 ok($' eq "\x{400}defg" && length ($') == 5, $message);
102 ok($1 eq "\x{200}\x{300}" && length ($1) == 2, $message);
e425a60b
YO
103 }
104
e425a60b
YO
105 {
106 my $x = "\x{10FFFD}";
107 $x =~ s/(.)/$1/g;
108 ok ord($x) == 0x10FFFD && length($x) == 1, "From Robin Houston";
109 }
110
e425a60b
YO
111 {
112 my %d = (
113 "7f" => [0, 0, 0],
114 "80" => [1, 1, 0],
115 "ff" => [1, 1, 0],
116 "100" => [0, 1, 1],
117 );
a2d248f9 118
e425a60b 119 while (my ($code, $match) = each %d) {
de946258 120 my $message = "Properties of \\x$code";
e425a60b 121 my $char = eval qq ["\\x{$code}"];
a2d248f9 122
de946258
NC
123 is(0 + ($char =~ /[\x80-\xff]/), $$match[0], $message);
124 is(0 + ($char =~ /[\x80-\x{100}]/), $$match[1], $message);
125 is(0 + ($char =~ /[\x{100}]/), $$match[2], $message);
e425a60b
YO
126 }
127 }
128
e425a60b
YO
129 {
130 # From Japhy
e4e5d8ba 131 foreach (qw(c g o)) {
4d18b353
NC
132 warning_like(sub {'' =~ "(?$_)"}, qr/^Useless \(\?$_\)/);
133 warning_like(sub {'' =~ "(?-$_)"}, qr/^Useless \(\?-$_\)/);
e4e5d8ba 134 }
e425a60b
YO
135
136 # Now test multi-error regexes
f4554ed5
NC
137 foreach (['(?g-o)', qr/^Useless \(\?g\)/, qr/^Useless \(\?-o\)/],
138 ['(?g-c)', qr/^Useless \(\?g\)/, qr/^Useless \(\?-c\)/],
139 # (?c) means (?g) error won't be thrown
140 ['(?o-cg)', qr/^Useless \(\?o\)/, qr/^Useless \(\?-c\)/],
141 ['(?ogc)', qr/^Useless \(\?o\)/, qr/^Useless \(\?g\)/,
142 qr/^Useless \(\?c\)/],
143 ) {
144 my ($re, @warnings) = @$_;
145 warnings_like(sub {eval "qr/$re/"}, \@warnings, "qr/$re/ warns");
146 }
e425a60b
YO
147 }
148
e425a60b 149 {
de946258 150 my $message = "/x tests";
e425a60b 151 $_ = "foo";
14358a41 152 foreach my $pat (<<" --", <<" --") {
e425a60b
YO
153 /f
154 o\r
155 o
156 \$
157 /x
158 --
e425a60b
YO
159 /f
160 o
161 o
162 \$\r
163 /x
164 --
14358a41
NC
165 is(eval $pat, 1, $message);
166 is($@, '', $message);
167 }
e425a60b
YO
168 }
169
e425a60b 170 {
de946258 171 my $message = "/o feature";
e425a60b 172 sub test_o {$_ [0] =~ /$_[1]/o; return $1}
de946258
NC
173 is(test_o ('abc', '(.)..'), 'a', $message);
174 is(test_o ('abc', '..(.)'), 'a', $message);
e425a60b
YO
175 }
176
e425a60b
YO
177 {
178 # Test basic $^N usage outside of a regex
de946258 179 my $message = '$^N usage outside of a regex';
e425a60b 180 my $x = "abcdef";
de946258
NC
181 ok(($x =~ /cde/ and !defined $^N), $message);
182 ok(($x =~ /(cde)/ and $^N eq "cde"), $message);
183 ok(($x =~ /(c)(d)(e)/ and $^N eq "e"), $message);
184 ok(($x =~ /(c(d)e)/ and $^N eq "cde"), $message);
185 ok(($x =~ /(foo)|(c(d)e)/ and $^N eq "cde"), $message);
186 ok(($x =~ /(c(d)e)|(foo)/ and $^N eq "cde"), $message);
187 ok(($x =~ /(c(d)e)|(abc)/ and $^N eq "abc"), $message);
188 ok(($x =~ /(c(d)e)|(abc)x/ and $^N eq "cde"), $message);
189 ok(($x =~ /(c(d)e)(abc)?/ and $^N eq "cde"), $message);
190 ok(($x =~ /(?:c(d)e)/ and $^N eq "d"), $message);
191 ok(($x =~ /(?:c(d)e)(?:f)/ and $^N eq "d"), $message);
192 ok(($x =~ /(?:([abc])|([def]))*/ and $^N eq "f"), $message);
193 ok(($x =~ /(?:([ace])|([bdf]))*/ and $^N eq "f"), $message);
194 ok(($x =~ /(([ace])|([bd]))*/ and $^N eq "e"), $message);
195 {ok(($x =~ /(([ace])|([bdf]))*/ and $^N eq "f"), $message);}
e425a60b
YO
196 ## Test to see if $^N is automatically localized -- it should now
197 ## have the value set in the previous test.
de26e0cc 198 is($^N, "e", '$^N is automatically localized');
e425a60b
YO
199
200 # Now test inside (?{ ... })
de946258 201 $message = '$^N usage inside (?{ ... })';
e425a60b 202 our ($y, $z);
de946258
NC
203 ok(($x =~ /a([abc])(?{$y=$^N})c/ and $y eq "b"), $message);
204 ok(($x =~ /a([abc]+)(?{$y=$^N})d/ and $y eq "bc"), $message);
205 ok(($x =~ /a([abcdefg]+)(?{$y=$^N})d/ and $y eq "bc"), $message);
206 ok(($x =~ /(a([abcdefg]+)(?{$y=$^N})d)(?{$z=$^N})e/ and $y eq "bc"
207 and $z eq "abcd"), $message);
208 ok(($x =~ /(a([abcdefg]+)(?{$y=$^N})de)(?{$z=$^N})/ and $y eq "bc"
209 and $z eq "abcde"), $message);
e425a60b
YO
210
211 }
212
e425a60b
YO
213 SKIP:
214 {
215 ## Should probably put in tests for all the POSIX stuff,
216 ## but not sure how to guarantee a specific locale......
217
de946258 218 my $message = 'Test [[:cntrl:]]';
e425a60b
YO
219 my $AllBytes = join "" => map {chr} 0 .. 255;
220 (my $x = $AllBytes) =~ s/[[:cntrl:]]//g;
628d95af
KW
221 $x = join "", sort { $a cmp $b }
222 map { chr utf8::native_to_unicode(ord $_) } split "", $x;
de946258 223 is($x, join("", map {chr} 0x20 .. 0x7E, 0x80 .. 0xFF), $message);
e425a60b
YO
224
225 ($x = $AllBytes) =~ s/[^[:cntrl:]]//g;
628d95af
KW
226 $x = join "", sort { $a cmp $b }
227 map { chr utf8::native_to_unicode(ord $_) } split "", $x;
de946258 228 is($x, (join "", map {chr} 0x00 .. 0x1F, 0x7F), $message);
e425a60b
YO
229 }
230
e425a60b
YO
231 {
232 # With /s modifier UTF8 chars were interpreted as bytes
de946258 233 my $message = "UTF-8 chars aren't bytes";
e425a60b
YO
234 my $a = "Hello \x{263A} World";
235 my @a = ($a =~ /./gs);
de946258 236 is($#a, 12, $message);
e425a60b
YO
237 }
238
e425a60b 239 {
e425a60b
YO
240 no warnings 'digit';
241 # Check that \x## works. 5.6.1 and 5.005_03 fail some of these.
242 my $x;
243 $x = "\x4e" . "E";
b7ad6c3e 244 like ($x, qr/^\x4EE$/, "Check only 2 bytes of hex are matched.");
e425a60b
YO
245
246 $x = "\x4e" . "i";
b7ad6c3e 247 like ($x, qr/^\x4Ei$/, "Check that invalid hex digit stops it (2)");
e425a60b
YO
248
249 $x = "\x4" . "j";
b7ad6c3e 250 like ($x, qr/^\x4j$/, "Check that invalid hex digit stops it (1)");
e425a60b
YO
251
252 $x = "\x0" . "k";
b7ad6c3e 253 like ($x, qr/^\xk$/, "Check that invalid hex digit stops it (0)");
e425a60b
YO
254
255 $x = "\x0" . "x";
b7ad6c3e 256 like ($x, qr/^\xx$/, "\\xx isn't to be treated as \\0");
e425a60b
YO
257
258 $x = "\x0" . "xa";
b7ad6c3e 259 like ($x, qr/^\xxa$/, "\\xxa isn't to be treated as \\xa");
e425a60b
YO
260
261 $x = "\x9" . "_b";
b7ad6c3e 262 like ($x, qr/^\x9_b$/, "\\x9_b isn't to be treated as \\x9b");
e425a60b
YO
263
264 # and now again in [] ranges
265
266 $x = "\x4e" . "E";
b7ad6c3e 267 like ($x, qr/^[\x4EE]{2}$/, "Check only 2 bytes of hex are matched.");
e425a60b
YO
268
269 $x = "\x4e" . "i";
b7ad6c3e 270 like ($x, qr/^[\x4Ei]{2}$/, "Check that invalid hex digit stops it (2)");
e425a60b
YO
271
272 $x = "\x4" . "j";
b7ad6c3e 273 like ($x, qr/^[\x4j]{2}$/, "Check that invalid hex digit stops it (1)");
e425a60b
YO
274
275 $x = "\x0" . "k";
b7ad6c3e 276 like ($x, qr/^[\xk]{2}$/, "Check that invalid hex digit stops it (0)");
e425a60b
YO
277
278 $x = "\x0" . "x";
b7ad6c3e 279 like ($x, qr/^[\xx]{2}$/, "\\xx isn't to be treated as \\0");
e425a60b
YO
280
281 $x = "\x0" . "xa";
b7ad6c3e 282 like ($x, qr/^[\xxa]{3}$/, "\\xxa isn't to be treated as \\xa");
e425a60b
YO
283
284 $x = "\x9" . "_b";
b7ad6c3e 285 like ($x, qr/^[\x9_b]{3}$/, "\\x9_b isn't to be treated as \\x9b");
e425a60b
YO
286
287 # Check that \x{##} works. 5.6.1 fails quite a few of these.
288
289 $x = "\x9b";
b7ad6c3e 290 like ($x, qr/^\x{9_b}$/, "\\x{9_b} is to be treated as \\x9b");
e425a60b
YO
291
292 $x = "\x9b" . "y";
b7ad6c3e 293 like ($x, qr/^\x{9_b}y$/, "\\x{9_b} is to be treated as \\x9b (again)");
e425a60b
YO
294
295 $x = "\x9b" . "y";
b7ad6c3e 296 like ($x, qr/^\x{9b_}y$/, "\\x{9b_} is to be treated as \\x9b");
e425a60b
YO
297
298 $x = "\x9b" . "y";
b7ad6c3e 299 like ($x, qr/^\x{9_bq}y$/, "\\x{9_bc} is to be treated as \\x9b");
e425a60b
YO
300
301 $x = "\x0" . "y";
b7ad6c3e 302 like ($x, qr/^\x{x9b}y$/, "\\x{x9b} is to be treated as \\x0");
e425a60b
YO
303
304 $x = "\x0" . "y";
b7ad6c3e 305 like ($x, qr/^\x{0x9b}y$/, "\\x{0x9b} is to be treated as \\x0");
e425a60b
YO
306
307 $x = "\x9b" . "y";
b7ad6c3e 308 like ($x, qr/^\x{09b}y$/, "\\x{09b} is to be treated as \\x9b");
e425a60b
YO
309
310 $x = "\x9b";
b7ad6c3e 311 like ($x, qr/^[\x{9_b}]$/, "\\x{9_b} is to be treated as \\x9b");
e425a60b
YO
312
313 $x = "\x9b" . "y";
b7ad6c3e 314 like ($x, qr/^[\x{9_b}y]{2}$/,
e425a60b
YO
315 "\\x{9_b} is to be treated as \\x9b (again)");
316
317 $x = "\x9b" . "y";
b7ad6c3e 318 like ($x, qr/^[\x{9b_}y]{2}$/, "\\x{9b_} is to be treated as \\x9b");
e425a60b
YO
319
320 $x = "\x9b" . "y";
b7ad6c3e 321 like ($x, qr/^[\x{9_bq}y]{2}$/, "\\x{9_bc} is to be treated as \\x9b");
e425a60b
YO
322
323 $x = "\x0" . "y";
b7ad6c3e 324 like ($x, qr/^[\x{x9b}y]{2}$/, "\\x{x9b} is to be treated as \\x0");
e425a60b
YO
325
326 $x = "\x0" . "y";
b7ad6c3e 327 like ($x, qr/^[\x{0x9b}y]{2}$/, "\\x{0x9b} is to be treated as \\x0");
e425a60b
YO
328
329 $x = "\x9b" . "y";
b7ad6c3e 330 like ($x, qr/^[\x{09b}y]{2}$/, "\\x{09b} is to be treated as \\x9b");
e425a60b
YO
331
332 }
333
e425a60b
YO
334 {
335 # High bit bug -- japhy
336 my $x = "ab\200d";
b7ad6c3e 337 like $x, qr/.*?\200/, "High bit fine";
e425a60b
YO
338 }
339
e425a60b
YO
340 {
341 # The basic character classes and Unicode
b7ad6c3e
KW
342 like "\x{0100}", qr/\w/, 'LATIN CAPITAL LETTER A WITH MACRON in /\w/';
343 like "\x{0660}", qr/\d/, 'ARABIC-INDIC DIGIT ZERO in /\d/';
344 like "\x{1680}", qr/\s/, 'OGHAM SPACE MARK in /\s/';
e425a60b
YO
345 }
346
e425a60b 347 {
de946258
NC
348 my $message = "Folding matches and Unicode";
349 like("a\x{100}", qr/A/i, $message);
350 like("A\x{100}", qr/a/i, $message);
351 like("a\x{100}", qr/a/i, $message);
352 like("A\x{100}", qr/A/i, $message);
353 like("\x{101}a", qr/\x{100}/i, $message);
354 like("\x{100}a", qr/\x{100}/i, $message);
355 like("\x{101}a", qr/\x{101}/i, $message);
356 like("\x{100}a", qr/\x{101}/i, $message);
357 like("a\x{100}", qr/A\x{100}/i, $message);
358 like("A\x{100}", qr/a\x{100}/i, $message);
359 like("a\x{100}", qr/a\x{100}/i, $message);
360 like("A\x{100}", qr/A\x{100}/i, $message);
361 like("a\x{100}", qr/[A]/i, $message);
362 like("A\x{100}", qr/[a]/i, $message);
363 like("a\x{100}", qr/[a]/i, $message);
364 like("A\x{100}", qr/[A]/i, $message);
365 like("\x{101}a", qr/[\x{100}]/i, $message);
366 like("\x{100}a", qr/[\x{100}]/i, $message);
367 like("\x{101}a", qr/[\x{101}]/i, $message);
368 like("\x{100}a", qr/[\x{101}]/i, $message);
e425a60b
YO
369 }
370
e425a60b
YO
371 {
372 use charnames ':full';
de946258 373 my $message = "Folding 'LATIN LETTER A WITH GRAVE'";
e425a60b
YO
374
375 my $lower = "\N{LATIN SMALL LETTER A WITH GRAVE}";
376 my $UPPER = "\N{LATIN CAPITAL LETTER A WITH GRAVE}";
0f289c68 377
de946258
NC
378 like($lower, qr/$UPPER/i, $message);
379 like($UPPER, qr/$lower/i, $message);
380 like($lower, qr/[$UPPER]/i, $message);
381 like($UPPER, qr/[$lower]/i, $message);
e425a60b 382
de946258 383 $message = "Folding 'GREEK LETTER ALPHA WITH VRACHY'";
e425a60b
YO
384
385 $lower = "\N{GREEK CAPITAL LETTER ALPHA WITH VRACHY}";
386 $UPPER = "\N{GREEK SMALL LETTER ALPHA WITH VRACHY}";
387
de946258
NC
388 like($lower, qr/$UPPER/i, $message);
389 like($UPPER, qr/$lower/i, $message);
390 like($lower, qr/[$UPPER]/i, $message);
391 like($UPPER, qr/[$lower]/i, $message);
e425a60b 392
de946258 393 $message = "Folding 'LATIN LETTER Y WITH DIAERESIS'";
e425a60b
YO
394
395 $lower = "\N{LATIN SMALL LETTER Y WITH DIAERESIS}";
396 $UPPER = "\N{LATIN CAPITAL LETTER Y WITH DIAERESIS}";
397
de946258
NC
398 like($lower, qr/$UPPER/i, $message);
399 like($UPPER, qr/$lower/i, $message);
400 like($lower, qr/[$UPPER]/i, $message);
401 like($UPPER, qr/[$lower]/i, $message);
e425a60b
YO
402 }
403
e425a60b
YO
404 {
405 use charnames ':full';
de946258 406 my $message = "GREEK CAPITAL LETTER SIGMA vs " .
e425a60b
YO
407 "COMBINING GREEK PERISPOMENI";
408
409 my $SIGMA = "\N{GREEK CAPITAL LETTER SIGMA}";
410 my $char = "\N{COMBINING GREEK PERISPOMENI}";
411
c11a8df3
NC
412 warning_is(sub {unlike("_:$char:_", qr/_:$SIGMA:_/i, $message)}, undef,
413 'Did not warn [change a5961de5f4215b5c]');
e425a60b
YO
414 }
415
e425a60b 416 {
de946258 417 my $message = '\X';
e425a60b
YO
418 use charnames ':full';
419
de946258
NC
420 ok("a!" =~ /^(\X)!/ && $1 eq "a", $message);
421 ok("\xDF!" =~ /^(\X)!/ && $1 eq "\xDF", $message);
422 ok("\x{100}!" =~ /^(\X)!/ && $1 eq "\x{100}", $message);
423 ok("\x{100}\x{300}!" =~ /^(\X)!/ && $1 eq "\x{100}\x{300}", $message);
424 ok("\N{LATIN CAPITAL LETTER E}!" =~ /^(\X)!/ &&
425 $1 eq "\N{LATIN CAPITAL LETTER E}", $message);
426 ok("\N{LATIN CAPITAL LETTER E}\N{COMBINING GRAVE ACCENT}!"
e425a60b 427 =~ /^(\X)!/ &&
de946258 428 $1 eq "\N{LATIN CAPITAL LETTER E}\N{COMBINING GRAVE ACCENT}", $message);
e425a60b 429
e425a60b
YO
430 }
431
e425a60b 432 {
de946258 433 my $message = "Final Sigma";
e425a60b
YO
434
435 my $SIGMA = "\x{03A3}"; # CAPITAL
436 my $Sigma = "\x{03C2}"; # SMALL FINAL
437 my $sigma = "\x{03C3}"; # SMALL
438
de946258
NC
439 like($SIGMA, qr/$SIGMA/i, $message);
440 like($SIGMA, qr/$Sigma/i, $message);
441 like($SIGMA, qr/$sigma/i, $message);
e425a60b 442
de946258
NC
443 like($Sigma, qr/$SIGMA/i, $message);
444 like($Sigma, qr/$Sigma/i, $message);
445 like($Sigma, qr/$sigma/i, $message);
e425a60b 446
de946258
NC
447 like($sigma, qr/$SIGMA/i, $message);
448 like($sigma, qr/$Sigma/i, $message);
449 like($sigma, qr/$sigma/i, $message);
0f289c68 450
de946258
NC
451 like($SIGMA, qr/[$SIGMA]/i, $message);
452 like($SIGMA, qr/[$Sigma]/i, $message);
453 like($SIGMA, qr/[$sigma]/i, $message);
e425a60b 454
de946258
NC
455 like($Sigma, qr/[$SIGMA]/i, $message);
456 like($Sigma, qr/[$Sigma]/i, $message);
457 like($Sigma, qr/[$sigma]/i, $message);
e425a60b 458
de946258
NC
459 like($sigma, qr/[$SIGMA]/i, $message);
460 like($sigma, qr/[$Sigma]/i, $message);
461 like($sigma, qr/[$sigma]/i, $message);
e425a60b 462
de946258 463 $message = "More final Sigma";
e425a60b
YO
464
465 my $S3 = "$SIGMA$Sigma$sigma";
466
de946258
NC
467 ok(":$S3:" =~ /:(($SIGMA)+):/i && $1 eq $S3 && $2 eq $sigma, $message);
468 ok(":$S3:" =~ /:(($Sigma)+):/i && $1 eq $S3 && $2 eq $sigma, $message);
469 ok(":$S3:" =~ /:(($sigma)+):/i && $1 eq $S3 && $2 eq $sigma, $message);
e425a60b 470
de946258
NC
471 ok(":$S3:" =~ /:(([$SIGMA])+):/i && $1 eq $S3 && $2 eq $sigma, $message);
472 ok(":$S3:" =~ /:(([$Sigma])+):/i && $1 eq $S3 && $2 eq $sigma, $message);
473 ok(":$S3:" =~ /:(([$sigma])+):/i && $1 eq $S3 && $2 eq $sigma, $message);
e425a60b
YO
474 }
475
e425a60b
YO
476 {
477 use charnames ':full';
de946258 478 my $message = "Parlez-Vous " .
e425a60b
YO
479 "Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais?";
480
de946258
NC
481 ok("Fran\N{LATIN SMALL LETTER C}ais" =~ /Fran.ais/ &&
482 $& eq "Francais", $message);
483 ok("Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~ /Fran.ais/ &&
484 $& eq "Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais", $message);
de946258
NC
485 ok("Fran\N{LATIN SMALL LETTER C}ais" =~ /Fran\Xais/ &&
486 $& eq "Francais", $message);
487 ok("Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~ /Fran\Xais/ &&
488 $& eq "Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais", $message);
489 ok("Franc\N{COMBINING CEDILLA}ais" =~ /Fran\Xais/ &&
490 $& eq "Franc\N{COMBINING CEDILLA}ais", $message);
491 ok("Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~
e425a60b 492 /Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais/ &&
de946258
NC
493 $& eq "Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais", $message);
494 ok("Franc\N{COMBINING CEDILLA}ais" =~ /Franc\N{COMBINING CEDILLA}ais/ &&
495 $& eq "Franc\N{COMBINING CEDILLA}ais", $message);
e425a60b
YO
496
497 my @f = (
498 ["Fran\N{LATIN SMALL LETTER C}ais", "Francais"],
499 ["Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais",
500 "Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais"],
501 ["Franc\N{COMBINING CEDILLA}ais", "Franc\N{COMBINING CEDILLA}ais"],
502 );
503 foreach my $entry (@f) {
504 my ($subject, $match) = @$entry;
de946258 505 ok($subject =~ /Fran(?:c\N{COMBINING CEDILLA}?|
e425a60b 506 \N{LATIN SMALL LETTER C WITH CEDILLA})ais/x &&
de946258 507 $& eq $match, $message);
e425a60b
YO
508 }
509 }
510
e425a60b 511 {
de946258 512 my $message = "Lingering (and useless) UTF8 flag doesn't mess up /i";
e425a60b
YO
513 my $pat = "ABcde";
514 my $str = "abcDE\x{100}";
515 chop $str;
de946258 516 like($str, qr/$pat/i, $message);
e425a60b
YO
517
518 $pat = "ABcde\x{100}";
519 $str = "abcDE";
520 chop $pat;
de946258 521 like($str, qr/$pat/i, $message);
e425a60b
YO
522
523 $pat = "ABcde\x{100}";
524 $str = "abcDE\x{100}";
525 chop $pat;
526 chop $str;
de946258 527 like($str, qr/$pat/i, $message);
e425a60b
YO
528 }
529
e425a60b
YO
530 {
531 use charnames ':full';
de946258 532 my $message = "LATIN SMALL LETTER SHARP S " .
e425a60b
YO
533 "(\N{LATIN SMALL LETTER SHARP S})";
534
de946258
NC
535 like("\N{LATIN SMALL LETTER SHARP S}",
536 qr/\N{LATIN SMALL LETTER SHARP S}/, $message);
537 like("\N{LATIN SMALL LETTER SHARP S}",
945961fd
KW
538 qr'\N{LATIN SMALL LETTER SHARP S}', $message);
539 like("\N{LATIN SMALL LETTER SHARP S}",
de946258
NC
540 qr/\N{LATIN SMALL LETTER SHARP S}/i, $message);
541 like("\N{LATIN SMALL LETTER SHARP S}",
945961fd
KW
542 qr'\N{LATIN SMALL LETTER SHARP S}'i, $message);
543 like("\N{LATIN SMALL LETTER SHARP S}",
de946258
NC
544 qr/[\N{LATIN SMALL LETTER SHARP S}]/, $message);
545 like("\N{LATIN SMALL LETTER SHARP S}",
945961fd
KW
546 qr'[\N{LATIN SMALL LETTER SHARP S}]', $message);
547 like("\N{LATIN SMALL LETTER SHARP S}",
de946258 548 qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message);
945961fd
KW
549 like("\N{LATIN SMALL LETTER SHARP S}",
550 qr'[\N{LATIN SMALL LETTER SHARP S}]'i, $message);
e425a60b 551
de946258 552 like("ss", qr /\N{LATIN SMALL LETTER SHARP S}/i, $message);
945961fd 553 like("ss", qr '\N{LATIN SMALL LETTER SHARP S}'i, $message);
de946258 554 like("SS", qr /\N{LATIN SMALL LETTER SHARP S}/i, $message);
945961fd 555 like("SS", qr '\N{LATIN SMALL LETTER SHARP S}'i, $message);
de946258 556 like("ss", qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message);
945961fd 557 like("ss", qr'[\N{LATIN SMALL LETTER SHARP S}]'i, $message);
de946258 558 like("SS", qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message);
945961fd 559 like("SS", qr'[\N{LATIN SMALL LETTER SHARP S}]'i, $message);
e425a60b 560
de946258
NC
561 like("\N{LATIN SMALL LETTER SHARP S}", qr/ss/i, $message);
562 like("\N{LATIN SMALL LETTER SHARP S}", qr/SS/i, $message);
0f289c68 563
de946258
NC
564 $message = "Unoptimized named sequence in class";
565 like("ss", qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message);
945961fd 566 like("ss", qr'[\N{LATIN SMALL LETTER SHARP S}x]'i, $message);
de946258 567 like("SS", qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message);
945961fd 568 like("SS", qr'[\N{LATIN SMALL LETTER SHARP S}x]'i, $message);
de946258
NC
569 like("\N{LATIN SMALL LETTER SHARP S}",
570 qr/[\N{LATIN SMALL LETTER SHARP S}x]/, $message);
571 like("\N{LATIN SMALL LETTER SHARP S}",
945961fd
KW
572 qr'[\N{LATIN SMALL LETTER SHARP S}x]', $message);
573 like("\N{LATIN SMALL LETTER SHARP S}",
de946258 574 qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message);
945961fd
KW
575 like("\N{LATIN SMALL LETTER SHARP S}",
576 qr'[\N{LATIN SMALL LETTER SHARP S}x]'i, $message);
e425a60b
YO
577 }
578
e425a60b
YO
579 {
580 # More whitespace: U+0085, U+2028, U+2029\n";
581
582 # U+0085, U+00A0 need to be forced to be Unicode, the \x{100} does that.
b7ad6c3e
KW
583 like "<\x{100}" . uni_to_native("\x{0085}") . ">", qr/<\x{100}\s>/, '\x{0085} in \s';
584 like "<" . uni_to_native("\x{0085}") . ">", qr/<\v>/, '\x{0085} in \v';
585 like "<\x{100}" . uni_to_native("\x{00A0}") . ">", qr/<\x{100}\s>/, '\x{00A0} in \s';
586 like "<" . uni_to_native("\x{00A0}") . ">", qr/<\h>/, '\x{00A0} in \h';
a9c9e371 587 my @h = map {sprintf "%05x" => $_} 0x01680, 0x02000 .. 0x0200A,
e425a60b
YO
588 0x0202F, 0x0205F, 0x03000;
589 my @v = map {sprintf "%05x" => $_} 0x02028, 0x02029;
590
591 my @H = map {sprintf "%05x" => $_} 0x01361, 0x0200B, 0x02408, 0x02420,
a9c9e371 592 0x0303F, 0xE0020, 0x180E;
e425a60b 593 my @V = map {sprintf "%05x" => $_} 0x0008A .. 0x0008D, 0x00348, 0x10100,
a9c9e371 594 0xE005F, 0xE007C, 0x180E;
e425a60b
YO
595
596 for my $hex (@h) {
597 my $str = eval qq ["<\\x{$hex}>"];
b7ad6c3e
KW
598 like $str, qr/<\s>/, "\\x{$hex} in \\s";
599 like $str, qr/<\h>/, "\\x{$hex} in \\h";
600 unlike $str, qr/<\v>/, "\\x{$hex} not in \\v";
e425a60b
YO
601 }
602
603 for my $hex (@v) {
604 my $str = eval qq ["<\\x{$hex}>"];
b7ad6c3e
KW
605 like $str, qr/<\s>/, "\\x{$hex} in \\s";
606 like $str, qr/<\v>/, "\\x{$hex} in \\v";
607 unlike $str, qr/<\h>/, "\\x{$hex} not in \\h";
e425a60b
YO
608 }
609
610 for my $hex (@H) {
611 my $str = eval qq ["<\\x{$hex}>"];
b7ad6c3e
KW
612 like $str, qr/<\S>/, "\\x{$hex} in \\S";
613 like $str, qr/<\H>/, "\\x{$hex} in \\H";
e425a60b
YO
614 }
615
616 for my $hex (@V) {
617 my $str = eval qq ["<\\x{$hex}>"];
b7ad6c3e
KW
618 like $str, qr/<\S>/, "\\x{$hex} in \\S";
619 like $str, qr/<\V>/, "\\x{$hex} in \\V";
e425a60b
YO
620 }
621 }
622
e425a60b
YO
623 {
624 # . with /s should work on characters, as opposed to bytes
de946258 625 my $message = ". with /s works on characters, not bytes";
e425a60b
YO
626
627 my $s = "\x{e4}\x{100}";
628 # This is not expected to match: the point is that
629 # neither should we get "Malformed UTF-8" warnings.
c11a8df3
NC
630 warning_is(sub {$s =~ /\G(.+?)\n/gcs}, undef,
631 "No 'Malformed UTF-8' warning");
e425a60b
YO
632
633 my @c;
634 push @c => $1 while $s =~ /\G(.)/gs;
635
636 local $" = "";
de946258 637 is("@c", $s, $message);
e425a60b
YO
638
639 # Test only chars < 256
640 my $t1 = "Q003\n\n\x{e4}\x{f6}\n\nQ004\n\n\x{e7}";
641 my $r1 = "";
642 while ($t1 =~ / \G ( .+? ) \n\s+ ( .+? ) ( $ | \n\s+ ) /xgcs) {
0f289c68 643 $r1 .= $1 . $2;
e425a60b
YO
644 }
645
646 my $t2 = $t1 . "\x{100}"; # Repeat with a larger char
647 my $r2 = "";
648 while ($t2 =~ / \G ( .+? ) \n\s+ ( .+? ) ( $ | \n\s+ ) /xgcs) {
0f289c68 649 $r2 .= $1 . $2;
e425a60b
YO
650 }
651 $r2 =~ s/\x{100}//;
652
de946258 653 is($r1, $r2, $message);
e425a60b
YO
654 }
655
e425a60b 656 {
de946258 657 my $message = "Unicode lookbehind";
b7ad6c3e 658 like("A\x{100}B", qr/(?<=A.)B/, $message);
de946258 659 like("A\x{200}\x{300}B", qr/(?<=A..)B/, $message);
b7ad6c3e
KW
660 like("\x{400}AB", qr/(?<=\x{400}.)B/, $message);
661 like("\x{500}\x{600}B", qr/(?<=\x{500}.)B/, $message);
e425a60b
YO
662
663 # Original code also contained:
664 # ok "\x{500\x{600}}B" =~ /(?<=\x{500}.)B/;
665 # but that looks like a typo.
666 }
667
e425a60b 668 {
de946258 669 my $message = 'UTF-8 hash keys and /$/';
e425a60b
YO
670 # http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters
671 # /2002-01/msg01327.html
672
673 my $u = "a\x{100}";
674 my $v = substr ($u, 0, 1);
675 my $w = substr ($u, 1, 1);
676 my %u = ($u => $u, $v => $v, $w => $w);
677 for (keys %u) {
678 my $m1 = /^\w*$/ ? 1 : 0;
679 my $m2 = $u {$_} =~ /^\w*$/ ? 1 : 0;
de946258 680 is($m1, $m2, $message);
e425a60b
YO
681 }
682 }
683
e425a60b 684 {
de946258 685 my $message = "No SEGV in s/// and UTF-8";
e425a60b 686 my $s = "s#\x{100}" x 4;
de946258 687 ok($s =~ s/[^\w]/ /g, $message);
2e84be61 688 if ( 1 or $ENV{PERL_TEST_LEGACY_POSIX_CC} ) {
de946258 689 is($s, "s \x{100}" x 4, $message);
e425a60b
YO
690 }
691 else {
de946258 692 is($s, "s " x 4, $message);
e425a60b
YO
693 }
694 }
695
e425a60b 696 {
de946258 697 my $message = "UTF-8 bug (maybe already known?)";
e425a60b
YO
698 my $u = "foo";
699 $u =~ s/./\x{100}/g;
de946258 700 is($u, "\x{100}\x{100}\x{100}", $message);
e425a60b
YO
701
702 $u = "foobar";
703 $u =~ s/[ao]/\x{100}/g;
de946258 704 is($u, "f\x{100}\x{100}b\x{100}r", $message);
e425a60b
YO
705
706 $u =~ s/\x{100}/e/g;
de946258 707 is($u, "feeber", $message);
e425a60b
YO
708 }
709
e425a60b 710 {
de946258 711 my $message = "UTF-8 bug with s///";
e425a60b
YO
712 # check utf8/non-utf8 mixtures
713 # try to force all float/anchored check combinations
714
715 my $c = "\x{100}";
716 my $subst;
717 for my $re ("xx.*$c", "x.*$c$c", "$c.*xx", "$c$c.*x",
718 "xx.*(?=$c)", "(?=$c).*xx",) {
de946258
NC
719 unlike("xxx", qr/$re/, $message);
720 ok(+($subst = "xxx") !~ s/$re//, $message);
e425a60b
YO
721 }
722 for my $re ("xx.*$c*", "$c*.*xx") {
de946258
NC
723 like("xxx", qr/$re/, $message);
724 ok(+($subst = "xxx") =~ s/$re//, $message);
725 is($subst, "", $message);
e425a60b
YO
726 }
727 for my $re ("xxy*", "y*xx") {
de946258
NC
728 like("xx$c", qr/$re/, $message);
729 ok(+($subst = "xx$c") =~ s/$re//, $message);
730 is($subst, $c, $message);
731 unlike("xy$c", qr/$re/, $message);
732 ok(+($subst = "xy$c") !~ s/$re//, $message);
e425a60b
YO
733 }
734 for my $re ("xy$c*z", "x$c*yz") {
de946258
NC
735 like("xyz", qr/$re/, $message);
736 ok(+($subst = "xyz") =~ s/$re//, $message);
737 is($subst, "", $message);
e425a60b
YO
738 }
739 }
740
e425a60b 741 {
c72077c4
AC
742 # The second half of RT #114808
743 warning_is(sub {'aa' =~ /.+\x{100}/}, undef,
744 'utf8-only floating substr, non-utf8 target, no warning');
745 }
746
747 {
de946258 748 my $message = "qr /.../x";
e425a60b 749 my $R = qr / A B C # D E/x;
de946258
NC
750 ok("ABCDE" =~ $R && $& eq "ABC", $message);
751 ok("ABCDE" =~ /$R/ && $& eq "ABC", $message);
752 ok("ABCDE" =~ m/$R/ && $& eq "ABC", $message);
753 ok("ABCDE" =~ /($R)/ && $1 eq "ABC", $message);
754 ok("ABCDE" =~ m/($R)/ && $1 eq "ABC", $message);
e425a60b
YO
755 }
756
e425a60b
YO
757 {
758 local $\;
759 $_ = 'aaaaaaaaaa';
760 utf8::upgrade($_); chop $_; $\="\n";
761 ok /[^\s]+/, 'm/[^\s]/ utf8';
762 ok /[^\d]+/, 'm/[^\d]/ utf8';
763 ok +($a = $_, $_ =~ s/[^\s]+/./g), 's/[^\s]/ utf8';
764 ok +($a = $_, $a =~ s/[^\d]+/./g), 's/[^\s]/ utf8';
765 }
766
e425a60b
YO
767 {
768 # Subject: Odd regexp behavior
769 # From: Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>
770 # Date: Wed, 26 Feb 2003 16:53:12 +0000
771 # Message-Id: <E18o4nw-0008Ly-00@wisbech.cl.cam.ac.uk>
772 # To: perl-unicode@perl.org
773
de946258 774 my $message = 'Markus Kuhn 2003-02-26';
0f289c68 775
e425a60b 776 my $x = "\x{2019}\nk";
de946258
NC
777 ok($x =~ s/(\S)\n(\S)/$1 $2/sg, $message);
778 is($x, "\x{2019} k", $message);
e425a60b
YO
779
780 $x = "b\nk";
de946258
NC
781 ok($x =~ s/(\S)\n(\S)/$1 $2/sg, $message);
782 is($x, "b k", $message);
e425a60b 783
de946258 784 like("\x{2019}", qr/\S/, $message);
e425a60b
YO
785 }
786
e425a60b 787 {
b7ad6c3e 788 like "\x{100}\n", qr/\x{100}\n$/, "UTF-8 length cache and fbm_compile";
e425a60b
YO
789 }
790
e425a60b
YO
791 {
792 package Str;
793 use overload q /""/ => sub {${$_ [0]};};
794 sub new {my ($c, $v) = @_; bless \$v, $c;}
795
796 package main;
797 $_ = Str -> new ("a\x{100}/\x{100}b");
798 ok join (":", /\b(.)\x{100}/g) eq "a:/", "re_intuit_start and PL_bostr";
799 }
800
e425a60b
YO
801 {
802 my $re = qq /^([^X]*)X/;
803 utf8::upgrade ($re);
8f46dbea 804 like "\x{100}X", qr/$re/, "S_cl_and ANYOF_UNICODE & ANYOF_INVERTED";
0ab1ff8e
KW
805 my $loc_re = qq /(?l:^([^X]*)X)/;
806 utf8::upgrade ($loc_re);
613abc6d 807 no warnings 'locale';
8f46dbea 808 like "\x{100}X", qr/$loc_re/, "locale, S_cl_and ANYOF_UNICODE & ANYOF_INVERTED";
e425a60b
YO
809 }
810
e425a60b 811 {
b7ad6c3e 812 like "123\x{100}", qr/^.*1.*23\x{100}$/,
e425a60b
YO
813 'UTF-8 + multiple floating substr';
814 }
815
e425a60b 816 {
de946258 817 my $message = '<20030808193656.5109.1@llama.ni-s.u-net.com>';
e425a60b
YO
818
819 # LATIN SMALL/CAPITAL LETTER A WITH MACRON
de946258 820 like(" \x{101}", qr/\x{100}/i, $message);
e425a60b
YO
821
822 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW
de946258 823 like(" \x{1E01}", qr/\x{1E00}/i, $message);
e425a60b
YO
824
825 # DESERET SMALL/CAPITAL LETTER LONG I
de946258 826 like(" \x{10428}", qr/\x{10400}/i, $message);
e425a60b
YO
827
828 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'
de946258 829 like(" \x{1E01}x", qr/\x{1E00}X/i, $message);
e425a60b
YO
830 }
831
e425a60b 832 {
26faadbd 833 for (120 .. 130, 240 .. 260) {
e425a60b 834 my $head = 'x' x $_;
de946258 835 my $message = q [Don't misparse \x{...} in regexp ] .
26faadbd 836 q [near EXACT char count limit];
e425a60b 837 for my $tail ('\x{0061}', '\x{1234}', '\x61') {
14358a41
NC
838 eval qq{like("$head$tail", qr/$head$tail/, \$message)};
839 is($@, '', $message);
e425a60b 840 }
de946258 841 $message = q [Don't misparse \N{...} in regexp ] .
26faadbd 842 q [near EXACT char count limit];
e425a60b 843 for my $tail ('\N{SNOWFLAKE}') {
14358a41
NC
844 eval qq {use charnames ':full';
845 like("$head$tail", qr/$head$tail/, \$message)};
945961fd
KW
846 eval qq {use charnames ':full';
847 like("$head$tail", qr'$head$tail', \$message)};
14358a41 848 is($@, '', $message);
e425a60b
YO
849 }
850 }
851 }
852
e425a60b
YO
853 { # TRIE related
854 our @got = ();
855 "words" =~ /(word|word|word)(?{push @got, $1})s$/;
de26e0cc 856 is(@got, 1, "TRIE optimisation");
e425a60b
YO
857
858 @got = ();
859 "words" =~ /(word|word|word)(?{push @got,$1})s$/i;
de26e0cc 860 is(@got, 1,"TRIEF optimisation");
e425a60b
YO
861
862 my @nums = map {int rand 1000} 1 .. 100;
863 my $re = "(" . (join "|", @nums) . ")";
864 $re = qr/\b$re\b/;
865
866 foreach (@nums) {
b7ad6c3e 867 like $_, qr/$re/, "Trie nums";
e425a60b
YO
868 }
869
870 $_ = join " ", @nums;
871 @got = ();
872 push @got, $1 while /$re/g;
873
874 my %count;
875 $count {$_} ++ for @got;
876 my $ok = 1;
877 for (@nums) {
878 $ok = 0 if --$count {$_} < 0;
879 }
880 ok $ok, "Trie min count matches";
881 }
882
e425a60b
YO
883 {
884 # TRIE related
885 # LATIN SMALL/CAPITAL LETTER A WITH MACRON
886 ok "foba \x{101}foo" =~ qr/(foo|\x{100}foo|bar)/i &&
887 $1 eq "\x{101}foo",
888 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH MACRON";
889
890 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW
891 ok "foba \x{1E01}foo" =~ qr/(foo|\x{1E00}foo|bar)/i &&
892 $1 eq "\x{1E01}foo",
893 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW";
894
895 # DESERET SMALL/CAPITAL LETTER LONG I
896 ok "foba \x{10428}foo" =~ qr/(foo|\x{10400}foo|bar)/i &&
897 $1 eq "\x{10428}foo",
898 "TRIEF + DESERET SMALL/CAPITAL LETTER LONG I";
899
900 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'
901 ok "foba \x{1E01}xfoo" =~ qr/(foo|\x{1E00}Xfoo|bar)/i &&
902 $1 eq "\x{1E01}xfoo",
903 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'";
904
905 use charnames ':full';
906
907 my $s = "\N{LATIN SMALL LETTER SHARP S}";
908 ok "foba ba$s" =~ qr/(foo|Ba$s|bar)/i && $1 eq "ba$s",
909 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss";
910 ok "foba ba$s" =~ qr/(Ba$s|foo|bar)/i && $1 eq "ba$s",
911 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss";
912 ok "foba ba$s" =~ qr/(foo|bar|Ba$s)/i && $1 eq "ba$s",
913 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss";
914
915 ok "foba ba$s" =~ qr/(foo|Bass|bar)/i && $1 eq "ba$s",
916 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss";
917
918 ok "foba ba$s" =~ qr/(foo|BaSS|bar)/i && $1 eq "ba$s",
919 "TRIEF + LATIN SMALL LETTER SHARP S =~ SS";
920
921 ok "foba ba${s}pxySS$s$s" =~ qr/(b(?:a${s}t|a${s}f|a${s}p)[xy]+$s*)/i
922 && $1 eq "ba${s}pxySS$s$s",
923 "COMMON PREFIX TRIEF + LATIN SMALL LETTER SHARP S";
924 }
925
e425a60b 926 {
3fe9640a
JH
927 BEGIN {
928 unshift @INC, 'lib';
929 }
b99b8a84
KW
930 use Cname; # Our custom charname plugin, currently found in
931 # t/lib/Cname.pm
0f289c68 932
b7ad6c3e 933 like 'fooB', qr/\N{foo}[\N{B}\N{b}]/, "Passthrough charname";
94ca1619
KW
934 my $name = "foo\xDF";
935 my $result = eval "'A${name}B' =~ /^A\\N{$name}B\$/";
936 ok !$@ && $result, "Passthrough charname of non-ASCII, Latin1";
e425a60b
YO
937 #
938 # Why doesn't must_warn work here?
939 #
940 my $w;
941 local $SIG {__WARN__} = sub {$w .= "@_"};
a78e2a97 942 $result = eval 'q(WARN) =~ /^[\N{WARN}]$/';
8f0cd35a 943 ok !$@ && $result && ! $w, '\N{} returning multi-char works';
e425a60b
YO
944
945 undef $w;
b7ad6c3e 946 eval q [unlike "\0", qr/[\N{EMPTY-STR}XY]/,
5895685f 947 "Zerolength charname in charclass doesn't match \\\\0"];
ff3f963a
KW
948 ok $w && $w =~ /Ignoring zero length/,
949 'Ignoring zero length \N{} in character class warning';
67048c1b 950 undef $w;
b7ad6c3e 951 eval q [like 'xy', qr/x[\N{EMPTY-STR} y]/x,
67048c1b
KW
952 'Empty string charname in [] is ignored; finds a following character'];
953 ok $w && $w =~ /Ignoring zero length/,
954 'Ignoring zero length \N{} in character class warning';
955 undef $w;
b7ad6c3e 956 eval q [like 'x ', qr/x[\N{EMPTY-STR} y]/,
67048c1b 957 'Empty string charname in [] is ignored; finds a following blank under /x'];
b7ad6c3e 958 like $w, qr/Ignoring zero length/,
67048c1b 959 'Ignoring zero length \N{} in character class warning';
e425a60b 960
b99b8a84
KW
961 # EVIL keeps track of its calls, and appends a new character each
962 # time: A AB ABC ABCD ...
e425a60b 963 ok 'AB' =~ /(\N{EVIL})/ && $1 eq 'A', 'Charname caching $1';
b7ad6c3e 964 like 'ABC', qr/(\N{EVIL})/, 'Charname caching $1';
945961fd
KW
965 ok 'ABCD' =~ m'(\N{EVIL})' && $1 eq 'ABC', 'Charname caching $1';
966 ok 'ABCDE' =~ m'(\N{EVIL})', 'Charname caching $1';
b7ad6c3e 967 like 'xy', qr/x\N{EMPTY-STR}y/,
e425a60b 968 'Empty string charname produces NOTHING node';
945961fd
KW
969 ok 'xy' =~ 'x\N{EMPTY-STR}y',
970 'Empty string charname produces NOTHING node';
b7ad6c3e 971 like '', qr/\N{EMPTY-STR}/,
e425a60b 972 'Empty string charname produces NOTHING node';
b7ad6c3e
KW
973 like "\N{LONG-STR}", qr/^\N{LONG-STR}$/, 'Verify that long string works';
974 like "\N{LONG-STR}", qr/^\N{LONG-STR}$/i, 'Verify under folding that long string works';
1e629c22
KW
975
976 # perlhacktips points out that these work on both ASCII and EBCDIC
977 like "\xfc", qr/\N{EMPTY-STR}\xdc/i, 'Empty \N{} should change /d to /u';
945961fd 978 like "\xfc", qr'\N{EMPTY-STR}\xdc'i, 'Empty \N{} should change /d to /u';
ff3f963a 979
902994e4 980 eval '/(?[[\N{EMPTY-STR}]])/';
b7ad6c3e 981 like $@, qr/Zero length \\N\{\}/, 'Verify zero-length return from \N{} correctly fails';
945961fd
KW
982 ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/, 'Verify that long string works';
983 ok "\N{LONG-STR}" =~ '^\N{LONG-STR}$', 'Verify that long string works';
984 ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/i, 'Verify under folding that long string works';
985 ok "\N{LONG-STR}" =~ m'^\N{LONG-STR}$'i, 'Verify under folding that long string works';
902994e4 986
bd299e29 987 undef $w;
bd299e29 988 {
3036c853 989 () = eval q ["\N{TOO MANY SPACES}"];
2d8eb851 990 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "Multiple spaces in a row in a charnames alias is fatal");
3036c853 991 eval q [use utf8; () = "\N{TOO MANY SPACES}"];
2d8eb851 992 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "... same under utf8");
94ec3a20 993 }
bd299e29
KW
994
995 undef $w;
bd299e29 996 {
94ec3a20 997 () = eval q ["\N{TRAILING SPACE }"];
2d8eb851 998 like ($@, qr/charnames alias definitions may not contain trailing white-space/, "Trailing white-space in a charnames alias is fatal");
94ec3a20 999 eval q [use utf8; () = "\N{TRAILING SPACE }"];
2d8eb851 1000 like ($@, qr/charnames alias definitions may not contain trailing white-space/, "... same under utf8");
94ec3a20 1001 }
bd299e29 1002
2d8eb851 1003 undef $w;
7fc82458 1004 my $Cedilla_Latin1 = "GAR"
d1cef54a 1005 . uni_to_native("\xC7")
7fc82458
KW
1006 . "ON";
1007 my $Cedilla_utf8 = $Cedilla_Latin1;
1008 utf8::upgrade($Cedilla_utf8);
1009 eval qq[is("\\N{$Cedilla_Latin1}", "$Cedilla_Latin1", "A cedilla in character name works")];
1010 undef $w;
1011 {
1012 use feature 'unicode_eval';
1013 eval qq[use utf8; is("\\N{$Cedilla_utf8}", "$Cedilla_utf8", "... same under 'use utf8': they work")];
1014 }
1015
1016 undef $w;
df758df2 1017 my $NBSP_Latin1 = "NBSP"
d1cef54a 1018 . uni_to_native("\xA0")
df758df2 1019 . "SEPARATED"
d1cef54a 1020 . uni_to_native("\xA0")
df758df2
KW
1021 . "SPACE";
1022 my $NBSP_utf8 = $NBSP_Latin1;
1023 utf8::upgrade($NBSP_utf8);
61463a3e
KW
1024 () = eval qq[is("\\N{$NBSP_Latin1}", "$NBSP_Latin1"];
1025 like ($@, qr/Invalid character in \\N\{...}/, "A NO-BREAK SPACE in a charnames alias is fatal");
df758df2
KW
1026 undef $w;
1027 {
1028 use feature 'unicode_eval';
61463a3e
KW
1029 eval qq[use utf8; is("\\N{$NBSP_utf8}"];
1030 like ($@, qr/Invalid character in \\N\{...}/, "A NO-BREAK SPACE in a charnames alias is fatal");
df758df2
KW
1031 }
1032
6a642c21
FC
1033 {
1034 BEGIN { no strict; *CnameTest:: = *{"_charnames\0A::" } }
1035 package CnameTest { sub translator { pop } }
1036 BEGIN { $^H{charnames} = \&CnameTest::translator }
1037 undef $w;
1038 () = eval q ["\N{TOO MANY SPACES}"];
2d8eb851 1039 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/,
6a642c21
FC
1040 'translators in _charnames\0* packages get validated');
1041 }
1042
ff3f963a
KW
1043 # If remove the limitation in regcomp code these should work
1044 # differently
1045 undef $w;
b7ad6c3e 1046 eval q [like "\N{TOO-LONG-STR}" =~ /^\N{TOO-LONG-STR}$/, 'Verify that what once was too long a string works'];
cb233ae3 1047 eval 'q() =~ /\N{4F}/';
4d7cd482 1048 ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in name gives error';
cb233ae3 1049 eval 'q() =~ /\N{COM,MA}/';
4d7cd482 1050 ok $@ && $@ =~ /Invalid character/, 'Verify that comma in name gives error';
f36cc39d 1051 $name = "A" . uni_to_native("\x{D7}") . "O";
cb233ae3 1052 eval "q(W) =~ /\\N{$name}/";
4d7cd482 1053 ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in name gives error';
bde9e88d
KW
1054 my $utf8_name = "7 CITIES OF GOLD";
1055 utf8::upgrade($utf8_name);
1056 eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
1057 ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in utf8 name gives error';
1058 $utf8_name = "SHARP #";
1059 utf8::upgrade($utf8_name);
1060 eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
1061 ok $@ && $@ =~ /Invalid character/, 'Verify that ASCII symbol in utf8 name gives error';
f36cc39d 1062 $utf8_name = "A HOUSE " . uni_to_native("\xF7") . " AGAINST ITSELF";
bde9e88d
KW
1063 utf8::upgrade($utf8_name);
1064 eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
1065 ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in utf8 name gives error';
1066 $utf8_name = "\x{664} HORSEMEN}";
1067 eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
1068 ok $@ && $@ =~ /Invalid character/, 'Verify that leading above Latin1 digit in utf8 name gives error';
1069 $utf8_name = "A \x{1F4A9} WOULD SMELL AS SWEET}";
1070 eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
1071 ok $@ && $@ =~ /Invalid character/, 'Verify that above Latin1 symbol in utf8 name gives error';
1072
cb233ae3 1073 undef $w;
f36cc39d 1074 $name = "A" . uni_to_native("\x{D1}") . "O";
cb233ae3
KW
1075 eval "q(W) =~ /\\N{$name}/";
1076 ok ! $w, 'Verify that latin1 letter in name doesnt give warning';
0f289c68 1077
ba7b73c5
NC
1078 # This tests the code path that restarts the parse when the recursive
1079 # call to S_reg() from within S_grok_bslash_N() discovers that the
1080 # pattern needs to be recalculated as UTF-8. use eval to avoid
1081 # needing literal Unicode in this source file:
1082 my $r = eval "qr/\\N{\x{100}\x{100}}/";
1083 isnt $r, undef, "Generated regex for multi-char UTF-8 charname"
1084 or diag($@);
b7ad6c3e 1085 like "\x{100}\x{100}", $r, "which matches";
e425a60b
YO
1086 }
1087
e425a60b
YO
1088 {
1089 use charnames ':full';
1090
b7ad6c3e
KW
1091 unlike 'aabc', qr/a\N{PLUS SIGN}b/, '/a\N{PLUS SIGN}b/ against aabc';
1092 like 'a+bc', qr/a\N{PLUS SIGN}b/, '/a\N{PLUS SIGN}b/ against a+bc';
e425a60b 1093
b7ad6c3e 1094 like ' A B', qr/\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/,
e425a60b 1095 'Intermixed named and unicode escapes';
b7ad6c3e
KW
1096 like "\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}",
1097 qr/\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/,
e425a60b 1098 'Intermixed named and unicode escapes';
b7ad6c3e
KW
1099 like "\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}",
1100 qr/[\N{SPACE}\N{U+0041}][\N{SPACE}\N{U+0042}]/,
0f289c68 1101 'Intermixed named and unicode escapes';
b7ad6c3e 1102 like "\0", qr/^\N{NULL}$/, 'Verify that \N{NULL} works; is not confused with an error';
e425a60b
YO
1103 }
1104
e425a60b
YO
1105 {
1106 our $brackets;
1107 $brackets = qr{
1108 { (?> [^{}]+ | (??{ $brackets }) )* }
1109 }x;
1110
b7ad6c3e 1111 unlike "{b{c}d", qr/^((??{ $brackets }))/, "Bracket mismatch";
e425a60b
YO
1112
1113 SKIP: {
1114 our @stack = ();
1115 my @expect = qw(
1116 stuff1
1117 stuff2
1118 <stuff1>and<stuff2>
1119 right
1120 <right>
1121 <<right>>
1122 <<<right>>>
1123 <<stuff1>and<stuff2>><<<<right>>>>
1124 );
1125
1126 local $_ = '<<<stuff1>and<stuff2>><<<<right>>>>>';
1127 ok /^(<((?:(?>[^<>]+)|(?1))*)>(?{push @stack, $2 }))$/,
1128 "Recursion matches";
de26e0cc 1129 is(@stack, @expect, "Right amount of matches")
e425a60b
YO
1130 or skip "Won't test individual results as count isn't equal",
1131 0 + @expect;
1132 my $idx = 0;
1133 foreach my $expect (@expect) {
de26e0cc
NC
1134 is($stack [$idx], $expect,
1135 "Expecting '$expect' at stack pos #$idx");
e425a60b
YO
1136 $idx ++;
1137 }
1138 }
1139 }
1140
e425a60b
YO
1141 {
1142 my $s = '123453456';
1143 $s =~ s/(?<digits>\d+)\k<digits>/$+{digits}/;
1144 ok $s eq '123456', 'Named capture (angle brackets) s///';
1145 $s = '123453456';
1146 $s =~ s/(?'digits'\d+)\k'digits'/$+{digits}/;
0f289c68 1147 ok $s eq '123456', 'Named capture (single quotes) s///';
e425a60b
YO
1148 }
1149
e425a60b
YO
1150 {
1151 my @ary = (
f36cc39d
KW
1152 pack('U', 0x00F1), # n-tilde
1153 '_'.pack('U', 0x00F1), # _ + n-tilde
e425a60b 1154 'c'.pack('U', 0x0327), # c + cedilla
f36cc39d 1155 pack('U*', 0x00F1, 0x0327),# n-tilde + cedilla
e425a60b
YO
1156 pack('U', 0x0391), # ALPHA
1157 pack('U', 0x0391).'2', # ALPHA + 2
1158 pack('U', 0x0391).'_', # ALPHA + _
1159 );
1160
1161 for my $uni (@ary) {
1162 my ($r1, $c1, $r2, $c2) = eval qq {
1163 use utf8;
1164 scalar ("..foo foo.." =~ /(?'${uni}'foo) \\k'${uni}'/),
1165 \$+{${uni}},
1166 scalar ("..bar bar.." =~ /(?<${uni}>bar) \\k<${uni}>/),
1167 \$+{${uni}};
1168 };
1169 ok $r1, "Named capture UTF (?'')";
1170 ok defined $c1 && $c1 eq 'foo', "Named capture UTF \%+";
1171 ok $r2, "Named capture UTF (?<>)";
1172 ok defined $c2 && $c2 eq 'bar', "Named capture UTF \%+";
1173 }
1174 }
1175
e425a60b
YO
1176 {
1177 my $s = 'foo bar baz';
1178 my @res;
1179 if ('1234' =~ /(?<A>1)(?<B>2)(?<A>3)(?<B>4)/) {
1180 foreach my $name (sort keys(%-)) {
1181 my $ary = $- {$name};
1182 foreach my $idx (0 .. $#$ary) {
1183 push @res, "$name:$idx:$ary->[$idx]";
1184 }
1185 }
1186 }
1187 my @expect = qw (A:0:1 A:1:3 B:0:2 B:1:4);
de26e0cc 1188 is("@res", "@expect", "Check %-");
e425a60b
YO
1189 eval'
1190 no warnings "uninitialized";
1191 print for $- {this_key_doesnt_exist};
1192 ';
1193 ok !$@,'lvalue $- {...} should not throw an exception';
1194 }
1195
e425a60b 1196 {
e425a60b 1197 # \c\ followed by _
b7ad6c3e
KW
1198 unlike "x\c_y", qr/x\c\_y/, '\_ in a pattern';
1199 like "x\c\_y", qr/x\c\_y/, '\_ in a pattern';
e425a60b
YO
1200
1201 # \c\ followed by other characters
1202 for my $c ("z", "\0", "!", chr(254), chr(256)) {
f36cc39d 1203 my $targ = "a" . uni_to_native("\034") . "$c";
e425a60b
YO
1204 my $reg = "a\\c\\$c";
1205 ok eval ("qq/$targ/ =~ /$reg/"), "\\c\\ in pattern";
1206 }
1207 }
1208
e425a60b
YO
1209 { # Test the (*PRUNE) pattern
1210 our $count = 0;
1211 'aaab' =~ /a+b?(?{$count++})(*FAIL)/;
de26e0cc 1212 is($count, 9, "Expect 9 for no (*PRUNE)");
e425a60b
YO
1213 $count = 0;
1214 'aaab' =~ /a+b?(*PRUNE)(?{$count++})(*FAIL)/;
de26e0cc 1215 is($count, 3, "Expect 3 with (*PRUNE)");
e425a60b
YO
1216 local $_ = 'aaab';
1217 $count = 0;
1218 1 while /.(*PRUNE)(?{$count++})(*FAIL)/g;
de26e0cc 1219 is($count, 4, "/.(*PRUNE)/");
e425a60b
YO
1220 $count = 0;
1221 'aaab' =~ /a+b?(??{'(*PRUNE)'})(?{$count++})(*FAIL)/;
de26e0cc 1222 is($count, 3, "Expect 3 with (*PRUNE)");
e425a60b
YO
1223 local $_ = 'aaab';
1224 $count = 0;
1225 1 while /.(??{'(*PRUNE)'})(?{$count++})(*FAIL)/g;
de26e0cc 1226 is($count, 4, "/.(*PRUNE)/");
e425a60b
YO
1227 }
1228
e425a60b
YO
1229 { # Test the (*SKIP) pattern
1230 our $count = 0;
1231 'aaab' =~ /a+b?(*SKIP)(?{$count++})(*FAIL)/;
de26e0cc 1232 is($count, 1, "Expect 1 with (*SKIP)");
e425a60b
YO
1233 local $_ = 'aaab';
1234 $count = 0;
1235 1 while /.(*SKIP)(?{$count++})(*FAIL)/g;
de26e0cc 1236 is($count, 4, "/.(*SKIP)/");
e425a60b
YO
1237 $_ = 'aaabaaab';
1238 $count = 0;
1239 our @res = ();
1240 1 while /(a+b?)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g;
de26e0cc
NC
1241 is($count, 2, "Expect 2 with (*SKIP)");
1242 is("@res", "aaab aaab", "Adjacent (*SKIP) works as expected");
e425a60b
YO
1243 }
1244
e425a60b
YO
1245 { # Test the (*SKIP) pattern
1246 our $count = 0;
1247 'aaab' =~ /a+b?(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/;
de26e0cc 1248 is($count, 1, "Expect 1 with (*SKIP)");
e425a60b
YO
1249 local $_ = 'aaab';
1250 $count = 0;
1251 1 while /.(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/g;
de26e0cc 1252 is($count, 4, "/.(*SKIP)/");
e425a60b
YO
1253 $_ = 'aaabaaab';
1254 $count = 0;
1255 our @res = ();
1256 1 while /(a+b?)(*MARK:foo)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g;
de26e0cc
NC
1257 is($count, 2, "Expect 2 with (*SKIP)");
1258 is("@res", "aaab aaab", "Adjacent (*SKIP) works as expected");
e425a60b
YO
1259 }
1260
e425a60b
YO
1261 { # Test the (*SKIP) pattern
1262 our $count = 0;
1263 'aaab' =~ /a*(*MARK:a)b?(*MARK:b)(*SKIP:a)(?{$count++})(*FAIL)/;
de26e0cc 1264 is($count, 3, "Expect 3 with *MARK:a)b?(*MARK:b)(*SKIP:a)");
e425a60b
YO
1265 local $_ = 'aaabaaab';
1266 $count = 0;
1267 our @res = ();
1268 1 while
1269 /(a*(*MARK:a)b?)(*MARK:x)(*SKIP:a)(?{$count++; push @res,$1})(*FAIL)/g;
de26e0cc
NC
1270 is($count, 5, "Expect 5 with (*MARK:a)b?)(*MARK:x)(*SKIP:a)");
1271 is("@res", "aaab b aaab b ",
1272 "Adjacent (*MARK:a)b?)(*MARK:x)(*SKIP:a) works as expected");
e425a60b
YO
1273 }
1274
e425a60b
YO
1275 { # Test the (*COMMIT) pattern
1276 our $count = 0;
1277 'aaabaaab' =~ /a+b?(*COMMIT)(?{$count++})(*FAIL)/;
de26e0cc 1278 is($count, 1, "Expect 1 with (*COMMIT)");
e425a60b
YO
1279 local $_ = 'aaab';
1280 $count = 0;
1281 1 while /.(*COMMIT)(?{$count++})(*FAIL)/g;
de26e0cc 1282 is($count, 1, "/.(*COMMIT)/");
e425a60b
YO
1283 $_ = 'aaabaaab';
1284 $count = 0;
1285 our @res = ();
1286 1 while /(a+b?)(*COMMIT)(?{$count++; push @res,$1})(*FAIL)/g;
de26e0cc
NC
1287 is($count, 1, "Expect 1 with (*COMMIT)");
1288 is("@res", "aaab", "Adjacent (*COMMIT) works as expected");
3542935d 1289
b7ad6c3e 1290 unlike("1\n2a\n", qr/^\d+(*COMMIT)\w+/m, "COMMIT and anchors");
e425a60b
YO
1291 }
1292
e425a60b
YO
1293 {
1294 # Test named commits and the $REGERROR var
fee50582 1295 local $REGERROR;
e425a60b
YO
1296 for my $name ('', ':foo') {
1297 for my $pat ("(*PRUNE$name)",
1298 ($name ? "(*MARK$name)" : "") . "(*SKIP$name)",
0f289c68 1299 "(*COMMIT$name)") {
e425a60b
YO
1300 for my $suffix ('(*FAIL)', '') {
1301 'aaaab' =~ /a+b$pat$suffix/;
de26e0cc 1302 is($REGERROR,
e425a60b 1303 ($suffix ? ($name ? 'foo' : "1") : ""),
de26e0cc 1304 "Test $pat and \$REGERROR $suffix");
e425a60b
YO
1305 }
1306 }
1307 }
1308 }
1309
e425a60b
YO
1310 {
1311 # Test named commits and the $REGERROR var
1312 package Fnorble;
1313 our $REGERROR;
fee50582 1314 local $REGERROR;
e425a60b
YO
1315 for my $name ('', ':foo') {
1316 for my $pat ("(*PRUNE$name)",
1317 ($name ? "(*MARK$name)" : "") . "(*SKIP$name)",
0f289c68 1318 "(*COMMIT$name)") {
e425a60b
YO
1319 for my $suffix ('(*FAIL)','') {
1320 'aaaab' =~ /a+b$pat$suffix/;
de26e0cc 1321 ::is($REGERROR,
e425a60b 1322 ($suffix ? ($name ? 'foo' : "1") : ""),
de26e0cc 1323 "Test $pat and \$REGERROR $suffix");
e425a60b
YO
1324 }
1325 }
0f289c68
YO
1326 }
1327 }
e425a60b 1328
e425a60b
YO
1329 {
1330 # Test named commits and the $REGERROR var
de946258 1331 my $message = '$REGERROR';
fee50582 1332 local $REGERROR;
e425a60b
YO
1333 for my $word (qw (bar baz bop)) {
1334 $REGERROR = "";
1335 "aaaaa$word" =~
1336 /a+(?:bar(*COMMIT:bar)|baz(*COMMIT:baz)|bop(*COMMIT:bop))(*FAIL)/;
de946258 1337 is($REGERROR, $word, $message);
0f289c68 1338 }
e425a60b
YO
1339 }
1340
e425a60b
YO
1341 {
1342 #Mindnumbingly simple test of (*THEN)
1343 for ("ABC","BAX") {
1344 ok /A (*THEN) X | B (*THEN) C/x, "Simple (*THEN) test";
1345 }
1346 }
1347
e425a60b 1348 {
de946258 1349 my $message = "Relative Recursion";
e425a60b
YO
1350 my $parens = qr/(\((?:[^()]++|(?-1))*+\))/;
1351 local $_ = 'foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))';
1352 my ($all, $one, $two) = ('', '', '');
de946258
NC
1353 ok(m/foo $parens \s* \+ \s* bar $parens/x, $message);
1354 is($1, '((2*3)+4-3)', $message);
1355 is($2, '(2*(3+4)-1*(2-3))', $message);
1356 is($&, 'foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))', $message);
1357 is($&, $_, $message);
e425a60b
YO
1358 }
1359
1360 {
1361 my $spaces=" ";
1362 local $_ = join 'bar', $spaces, $spaces;
1363 our $count = 0;
1364 s/(?>\s+bar)(?{$count++})//g;
de26e0cc
NC
1365 is($_, $spaces, "SUSPEND final string");
1366 is($count, 1, "Optimiser should have prevented more than one match");
e425a60b
YO
1367 }
1368
e425a60b
YO
1369 {
1370 # From Message-ID: <877ixs6oa6.fsf@k75.linux.bogus>
1371 my $dow_name = "nada";
1372 my $parser = "(\$dow_name) = \$time_string =~ /(D\x{e9}\\ " .
1373 "C\x{e9}adaoin|D\x{e9}\\ Sathairn|\\w+|\x{100})/";
1374 my $time_string = "D\x{e9} C\x{e9}adaoin";
1375 eval $parser;
1376 ok !$@, "Test Eval worked";
de26e0cc 1377 is($dow_name, $time_string, "UTF-8 trie common prefix extraction");
e425a60b
YO
1378 }
1379
e425a60b
YO
1380 {
1381 my $v;
1382 ($v = 'bar') =~ /(\w+)/g;
1383 $v = 'foo';
de26e0cc
NC
1384 is("$1", 'bar',
1385 '$1 is safe after /g - may fail due to specialized config in pp_hot.c');
e425a60b
YO
1386 }
1387
e425a60b 1388 {
de946258 1389 my $message = "http://nntp.perl.org/group/perl.perl5.porters/118663";
e425a60b 1390 my $qr_barR1 = qr/(bar)\g-1/;
de946258
NC
1391 like("foobarbarxyz", $qr_barR1, $message);
1392 like("foobarbarxyz", qr/foo${qr_barR1}xyz/, $message);
1393 like("foobarbarxyz", qr/(foo)${qr_barR1}xyz/, $message);
1394 like("foobarbarxyz", qr/(foo)(bar)\g{-1}xyz/, $message);
1395 like("foobarbarxyz", qr/(foo${qr_barR1})xyz/, $message);
1396 like("foobarbarxyz", qr/(foo(bar)\g{-1})xyz/, $message);
0f289c68 1397 }
e425a60b 1398
e425a60b 1399 {
de946258 1400 my $message = '$REGMARK';
e425a60b 1401 our @r = ();
fee50582
YO
1402 local $REGMARK;
1403 local $REGERROR;
de946258
NC
1404 like('foofoo', qr/foo (*MARK:foo) (?{push @r,$REGMARK}) /x, $message);
1405 is("@r","foo", $message);
1406 is($REGMARK, "foo", $message);
1407 unlike('foofoo', qr/foo (*MARK:foo) (*FAIL) /x, $message);
1408 is($REGMARK, '', $message);
1409 is($REGERROR, 'foo', $message);
e425a60b
YO
1410 }
1411
e425a60b 1412 {
de946258 1413 my $message = '\K test';
e425a60b
YO
1414 my $x;
1415 $x = "abc.def.ghi.jkl";
1416 $x =~ s/.*\K\..*//;
de946258 1417 is($x, "abc.def.ghi", $message);
0f289c68 1418
e425a60b
YO
1419 $x = "one two three four";
1420 $x =~ s/o+ \Kthree//g;
de946258 1421 is($x, "one two four", $message);
0f289c68 1422
e425a60b
YO
1423 $x = "abcde";
1424 $x =~ s/(.)\K/$1/g;
de946258 1425 is($x, "aabbccddee", $message);
e425a60b
YO
1426 }
1427
e425a60b
YO
1428 {
1429 sub kt {
1430 return '4' if $_[0] eq '09028623';
1431 }
1432 # Nested EVAL using PL_curpm (via $1 or friends)
1433 my $re;
1434 our $grabit = qr/ ([0-6][0-9]{7}) (??{ kt $1 }) [890] /x;
1435 $re = qr/^ ( (??{ $grabit }) ) $ /x;
1436 my @res = '0902862349' =~ $re;
de26e0cc
NC
1437 is(join ("-", @res), "0902862349",
1438 'PL_curpm is set properly on nested eval');
e425a60b
YO
1439
1440 our $qr = qr/ (o) (??{ $1 }) /x;
1441 ok 'boob'=~/( b (??{ $qr }) b )/x && 1, "PL_curpm, nested eval";
1442 }
1443
e425a60b
YO
1444 {
1445 use charnames ":full";
b7ad6c3e
KW
1446 like "\N{ROMAN NUMERAL ONE}", qr/\p{Alphabetic}/, "I =~ Alphabetic";
1447 like "\N{ROMAN NUMERAL ONE}", qr/\p{Uppercase}/, "I =~ Uppercase";
1448 unlike "\N{ROMAN NUMERAL ONE}", qr/\p{Lowercase}/, "I !~ Lowercase";
1449 like "\N{ROMAN NUMERAL ONE}", qr/\p{IDStart}/, "I =~ ID_Start";
1450 like "\N{ROMAN NUMERAL ONE}", qr/\p{IDContinue}/, "I =~ ID_Continue";
1451 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Alphabetic}/, "i =~ Alphabetic";
1452 unlike "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Uppercase}/, "i !~ Uppercase";
1453 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Uppercase}/i, "i =~ Uppercase under /i";
1454 unlike "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Titlecase}/, "i !~ Titlecase";
1455 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Titlecase}/i, "i =~ Titlecase under /i";
1456 like "\N{ROMAN NUMERAL ONE}", qr/\p{Lowercase}/i, "I =~ Lowercase under /i";
1457
1458 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Lowercase}/, "i =~ Lowercase";
1459 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{IDStart}/, "i =~ ID_Start";
1460 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{IDContinue}/, "i =~ ID_Continue"
e425a60b
YO
1461 }
1462
0ab1ff8e
KW
1463 { # More checking that /i works on the few properties that it makes a
1464 # difference. Uppercase, Lowercase, and Titlecase were done in the
1465 # block above
b7ad6c3e
KW
1466 like "A", qr/\p{PosixUpper}/, "A =~ PosixUpper";
1467 like "A", qr/\p{PosixUpper}/i, "A =~ PosixUpper under /i";
1468 unlike "A", qr/\p{PosixLower}/, "A !~ PosixLower";
1469 like "A", qr/\p{PosixLower}/i, "A =~ PosixLower under /i";
1470 unlike "a", qr/\p{PosixUpper}/, "a !~ PosixUpper";
1471 like "a", qr/\p{PosixUpper}/i, "a =~ PosixUpper under /i";
1472 like "a", qr/\p{PosixLower}/, "a =~ PosixLower";
1473 like "a", qr/\p{PosixLower}/i, "a =~ PosixLower under /i";
1474
1475 like uni_to_native("\xC0"), qr/\p{XPosixUpper}/, "\\xC0 =~ XPosixUpper";
1476 like uni_to_native("\xC0"), qr/\p{XPosixUpper}/i, "\\xC0 =~ XPosixUpper under /i";
1477 unlike uni_to_native("\xC0"), qr/\p{XPosixLower}/, "\\xC0 !~ XPosixLower";
1478 like uni_to_native("\xC0"), qr/\p{XPosixLower}/i, "\\xC0 =~ XPosixLower under /i";
1479 unlike uni_to_native("\xE0"), qr/\p{XPosixUpper}/, "\\xE0 !~ XPosixUpper";
1480 like uni_to_native("\xE0"), qr/\p{XPosixUpper}/i, "\\xE0 =~ XPosixUpper under /i";
1481 like uni_to_native("\xE0"), qr/\p{XPosixLower}/, "\\xE0 =~ XPosixLower";
1482 like uni_to_native("\xE0"), qr/\p{XPosixLower}/i, "\\xE0 =~ XPosixLower under /i";
1483
1484 like uni_to_native("\xC0"), qr/\p{UppercaseLetter}/, "\\xC0 =~ UppercaseLetter";
1485 like uni_to_native("\xC0"), qr/\p{UppercaseLetter}/i, "\\xC0 =~ UppercaseLetter under /i";
1486 unlike uni_to_native("\xC0"), qr/\p{LowercaseLetter}/, "\\xC0 !~ LowercaseLetter";
1487 like uni_to_native("\xC0"), qr/\p{LowercaseLetter}/i, "\\xC0 =~ LowercaseLetter under /i";
1488 unlike uni_to_native("\xC0"), qr/\p{TitlecaseLetter}/, "\\xC0 !~ TitlecaseLetter";
1489 like uni_to_native("\xC0"), qr/\p{TitlecaseLetter}/i, "\\xC0 =~ TitlecaseLetter under /i";
1490 unlike uni_to_native("\xE0"), qr/\p{UppercaseLetter}/, "\\xE0 !~ UppercaseLetter";
1491 like uni_to_native("\xE0"), qr/\p{UppercaseLetter}/i, "\\xE0 =~ UppercaseLetter under /i";
1492 like uni_to_native("\xE0"), qr/\p{LowercaseLetter}/, "\\xE0 =~ LowercaseLetter";
1493 like uni_to_native("\xE0"), qr/\p{LowercaseLetter}/i, "\\xE0 =~ LowercaseLetter under /i";
1494 unlike uni_to_native("\xE0"), qr/\p{TitlecaseLetter}/, "\\xE0 !~ TitlecaseLetter";
1495 like uni_to_native("\xE0"), qr/\p{TitlecaseLetter}/i, "\\xE0 =~ TitlecaseLetter under /i";
1496 unlike "\x{1C5}", qr/\p{UppercaseLetter}/, "\\x{1C5} !~ UppercaseLetter";
1497 like "\x{1C5}", qr/\p{UppercaseLetter}/i, "\\x{1C5} =~ UppercaseLetter under /i";
1498 unlike "\x{1C5}", qr/\p{LowercaseLetter}/, "\\x{1C5} !~ LowercaseLetter";
1499 like "\x{1C5}", qr/\p{LowercaseLetter}/i, "\\x{1C5} =~ LowercaseLetter under /i";
1500 like "\x{1C5}", qr/\p{TitlecaseLetter}/, "\\x{1C5} =~ TitlecaseLetter";
1501 like "\x{1C5}", qr/\p{TitlecaseLetter}/i, "\\x{1C5} =~ TitlecaseLetter under /i";
0ab1ff8e
KW
1502 }
1503
e425a60b
YO
1504 {
1505 # requirement of Unicode Technical Standard #18, 1.7 Code Points
1506 # cf. http://www.unicode.org/reports/tr18/#Supplementary_Characters
1507 for my $u (0x7FF, 0x800, 0xFFFF, 0x10000) {
1508 no warnings 'utf8'; # oops
1509 my $c = chr $u;
1510 my $x = sprintf '%04X', $u;
b7ad6c3e 1511 like "A${c}B", qr/A[\0-\x{10000}]B/, "Unicode range - $x";
e425a60b
YO
1512 }
1513 }
1514
e425a60b
YO
1515 {
1516 my $res="";
1517
1518 if ('1' =~ /(?|(?<digit>1)|(?<digit>2))/) {
1519 $res = "@{$- {digit}}";
1520 }
de26e0cc
NC
1521 is($res, "1",
1522 "Check that (?|...) doesnt cause dupe entries in the names array");
0f289c68 1523
e425a60b
YO
1524 $res = "";
1525 if ('11' =~ /(?|(?<digit>1)|(?<digit>2))(?&digit)/) {
1526 $res = "@{$- {digit}}";
1527 }
de26e0cc
NC
1528 is($res, "1",
1529 "Check that (?&..) to a buffer inside a (?|...) goes to the leftmost");
e425a60b
YO
1530 }
1531
e425a60b
YO
1532 {
1533 use warnings;
de946258 1534 my $message = "ASCII pattern that really is UTF-8";
e425a60b
YO
1535 my @w;
1536 local $SIG {__WARN__} = sub {push @w, "@_"};
0f289c68 1537 my $c = qq (\x{DF});
de946258
NC
1538 like($c, qr/${c}|\x{100}/, $message);
1539 is("@w", '', $message);
0f289c68 1540 }
e425a60b 1541
e425a60b 1542 {
de946258 1543 my $message = "Corruption of match results of qr// across scopes";
e425a60b
YO
1544 my $qr = qr/(fo+)(ba+r)/;
1545 'foobar' =~ /$qr/;
de946258 1546 is("$1$2", "foobar", $message);
e425a60b
YO
1547 {
1548 'foooooobaaaaar' =~ /$qr/;
de946258 1549 is("$1$2", 'foooooobaaaaar', $message);
e425a60b 1550 }
de946258 1551 is("$1$2", "foobar", $message);
e425a60b
YO
1552 }
1553
e425a60b 1554 {
de946258 1555 my $message = "HORIZWS";
e425a60b
YO
1556 local $_ = "\t \r\n \n \t".chr(11)."\n";
1557 s/\H/H/g;
1558 s/\h/h/g;
de946258 1559 is($_, "hhHHhHhhHH", $message);
e425a60b
YO
1560 $_ = "\t \r\n \n \t" . chr (11) . "\n";
1561 utf8::upgrade ($_);
1562 s/\H/H/g;
1563 s/\h/h/g;
de946258 1564 is($_, "hhHHhHhhHH", $message);
0f289c68 1565 }
e425a60b 1566
e425a60b 1567 {
de946258 1568 # Various whitespace special patterns
f36cc39d
KW
1569 my @h = map {chr utf8::unicode_to_native($_) }
1570 0x09, 0x20, 0xa0, 0x1680, 0x2000,
e425a60b
YO
1571 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, 0x2006,
1572 0x2007, 0x2008, 0x2009, 0x200a, 0x202f, 0x205f,
1573 0x3000;
f36cc39d
KW
1574 my @v = map {chr utf8::unicode_to_native($_) }
1575 0x0a, 0x0b, 0x0c, 0x0d, 0x85, 0x2028,
e425a60b 1576 0x2029;
f36cc39d
KW
1577 my @lb = (uni_to_native("\x0D\x0A"),
1578 map {chr utf8::unicode_to_native($_) }
1579 0x0A .. 0x0D, 0x85, 0x2028, 0x2029);
e425a60b
YO
1580 foreach my $t ([\@h, qr/\h/, qr/\h+/],
1581 [\@v, qr/\v/, qr/\v+/],
1582 [\@lb, qr/\R/, qr/\R+/],) {
1583 my $ary = shift @$t;
1584 foreach my $pat (@$t) {
1585 foreach my $str (@$ary) {
df6841b6
KW
1586 my $temp_str = $str;
1587 $temp_str = display($temp_str);
1588 ok $str =~ /($pat)/, $temp_str . " =~ /($pat)";
1589 my $temp_1 = $1;
1590 is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "' after ($pat)");
e425a60b 1591 utf8::upgrade ($str);
df6841b6
KW
1592 ok $str =~ /($pat)/, "Upgraded " . $temp_str . " =~ /($pat)/";
1593 is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "'(upgraded) after ($pat)");
e425a60b
YO
1594 }
1595 }
1596 }
1597 }
1598
e425a60b 1599 {
de946258 1600 # Check that \\xDF match properly in its various forms
e425a60b
YO
1601 # Test that \xDF matches properly. this is pretty hacky stuff,
1602 # but its actually needed. The malarky with '-' is to prevent
1603 # compilation caching from playing any role in the test.
f36cc39d 1604 my @df = (chr utf8::unicode_to_native(0xDF), '-', chr utf8::unicode_to_native(0xDF));
e425a60b 1605 utf8::upgrade ($df [2]);
f36cc39d 1606 my @strs = ('ss', 'sS', 'Ss', 'SS', chr utf8::unicode_to_native(0xDF));
e425a60b
YO
1607 my @ss = map {("$_", "$_")} @strs;
1608 utf8::upgrade ($ss [$_ * 2 + 1]) for 0 .. $#strs;
1609
1610 for my $ssi (0 .. $#ss) {
1611 for my $dfi (0 .. $#df) {
1612 my $pat = $df [$dfi];
1613 my $str = $ss [$ssi];
1614 my $utf_df = ($dfi > 1) ? 'utf8' : '';
1615 my $utf_ss = ($ssi % 2) ? 'utf8' : '';
f36cc39d
KW
1616 my $sstr; # We hard-code the ebcdic value below to avoid
1617 # perturbing the test
1618 ($sstr = $str) =~ s/\xDF/\\xDF/ if $::IS_ASCII;
1619 ($sstr = $str) =~ s/\x59/\\x59/ if $::IS_EBCDIC;
e425a60b
YO
1620
1621 if ($utf_df || $utf_ss || length ($ss [$ssi]) == 1) {
1622 my $ret = $str =~ /$pat/i;
1623 next if $pat eq '-';
f36cc39d
KW
1624 if ($::IS_ASCII) {
1625 ok $ret, "\"$sstr\" =~ /\\xDF/i " .
1626 "(str is @{[$utf_ss||'latin']}, pat is " .
1627 "@{[$utf_df||'latin']})";
1628 }
1629 else {
1630 ok $ret, "\"$sstr\" =~ /\\x59/i " .
e425a60b
YO
1631 "(str is @{[$utf_ss||'latin']}, pat is " .
1632 "@{[$utf_df||'latin']})";
f36cc39d 1633 }
e425a60b
YO
1634 }
1635 else {
1636 my $ret = $str !~ /$pat/i;
1637 next if $pat eq '-';
f36cc39d
KW
1638 if ($::IS_EBCDIC) {
1639 ok $ret, "\"$sstr\" !~ /\\x59/i " .
e425a60b
YO
1640 "(str is @{[$utf_ss||'latin']}, pat is " .
1641 "@{[$utf_df||'latin']})";
f36cc39d
KW
1642 }
1643 else {
1644 ok $ret, "\"$sstr\" !~ /\\xDF/i " .
1645 "(str is @{[$utf_ss||'latin']}, pat is " .
1646 "@{[$utf_df||'latin']})";
1647 }
e425a60b
YO
1648 }
1649 }
1650 }
1651 }
1652
e425a60b 1653 {
de946258 1654 my $message = "BBC(Bleadperl Breaks CPAN) Today: String::Multibyte";
e425a60b
YO
1655 my $re = qr/(?:[\x00-\xFF]{4})/;
1656 my $hyp = "\0\0\0-";
1657 my $esc = "\0\0\0\\";
1658
1659 my $str = "$esc$hyp$hyp$esc$esc";
1660 my @a = ($str =~ /\G(?:\Q$esc$esc\E|\Q$esc$hyp\E|$re)/g);
1661
de946258 1662 is(@a,3, $message);
e425a60b 1663 local $" = "=";
de946258 1664 is("@a","$esc$hyp=$hyp=$esc$esc", $message);
e425a60b
YO
1665 }
1666
e425a60b
YO
1667 {
1668 # Test for keys in %+ and %-
de946258 1669 my $message = 'Test keys in %+ and %-';
7fba2966
FC
1670 no warnings 'uninitialized';
1671 local $_ = "abcdef";
e425a60b 1672 /(?<foo>a)|(?<foo>b)/;
de946258
NC
1673 is((join ",", sort keys %+), "foo", $message);
1674 is((join ",", sort keys %-), "foo", $message);
1675 is((join ",", sort values %+), "a", $message);
1676 is((join ",", sort map "@$_", values %-), "a ", $message);
e425a60b 1677 /(?<bar>a)(?<bar>b)(?<quux>.)/;
de946258
NC
1678 is((join ",", sort keys %+), "bar,quux", $message);
1679 is((join ",", sort keys %-), "bar,quux", $message);
1680 is((join ",", sort values %+), "a,c", $message); # leftmost
1681 is((join ",", sort map "@$_", values %-), "a b,c", $message);
e425a60b 1682 /(?<un>a)(?<deux>c)?/; # second buffer won't capture
de946258
NC
1683 is((join ",", sort keys %+), "un", $message);
1684 is((join ",", sort keys %-), "deux,un", $message);
1685 is((join ",", sort values %+), "a", $message);
1686 is((join ",", sort map "@$_", values %-), ",a", $message);
e425a60b
YO
1687 }
1688
e425a60b
YO
1689 {
1690 # length() on captures, the numbered ones end up in Perl_magic_len
7fba2966 1691 local $_ = "aoeu " . uni_to_native("\xe6") . "var ook";
e425a60b
YO
1692 /^ \w+ \s (?<eek>\S+)/x;
1693
de26e0cc
NC
1694 is(length $`, 0, q[length $`]);
1695 is(length $', 4, q[length $']);
1696 is(length $&, 9, q[length $&]);
1697 is(length $1, 4, q[length $1]);
1698 is(length $+{eek}, 4, q[length $+{eek} == length $1]);
e425a60b
YO
1699 }
1700
e425a60b
YO
1701 {
1702 my $ok = -1;
1703
1704 $ok = exists ($-{x}) ? 1 : 0 if 'bar' =~ /(?<x>foo)|bar/;
de26e0cc
NC
1705 is($ok, 1, '$-{x} exists after "bar"=~/(?<x>foo)|bar/');
1706 is(scalar (%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/');
1707 is(scalar (%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/');
e425a60b
YO
1708
1709 $ok = -1;
1710 $ok = exists ($+{x}) ? 1 : 0 if 'bar' =~ /(?<x>foo)|bar/;
de26e0cc
NC
1711 is($ok, 0, '$+{x} not exists after "bar"=~/(?<x>foo)|bar/');
1712 is(scalar (%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/');
1713 is(scalar (%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/');
e425a60b
YO
1714
1715 $ok = -1;
1716 $ok = exists ($-{x}) ? 1 : 0 if 'foo' =~ /(?<x>foo)|bar/;
de26e0cc
NC
1717 is($ok, 1, '$-{x} exists after "foo"=~/(?<x>foo)|bar/');
1718 is(scalar (%+), 1, 'scalar %+ == 1 after "foo"=~/(?<x>foo)|bar/');
1719 is(scalar (%-), 1, 'scalar %- == 1 after "foo"=~/(?<x>foo)|bar/');
e425a60b
YO
1720
1721 $ok = -1;
1722 $ok = exists ($+{x}) ? 1 : 0 if 'foo'=~/(?<x>foo)|bar/;
de26e0cc 1723 is($ok, 1, '$+{x} exists after "foo"=~/(?<x>foo)|bar/');
e425a60b
YO
1724 }
1725
e425a60b
YO
1726 {
1727 local $_;
1728 ($_ = 'abc') =~ /(abc)/g;
0f289c68 1729 $_ = '123';
de26e0cc 1730 is("$1", 'abc', "/g leads to unsafe match vars: $1");
1b474ee3
NC
1731
1732 fresh_perl_is(<<'EOP', ">abc<\n", {}, 'mention $&');
1733$&;
1734my $x;
1735($x='abc')=~/(abc)/g;
1736$x='123';
1737print ">$1<\n";
1738EOP
1739
1b474ee3
NC
1740 fresh_perl_is(<<'EOP', ">abc<\n", {}, 'no mention of $&');
1741my $x;
1742($x='abc')=~/(abc)/g;
1743$x='123';
1744print ">$1<\n";
1745EOP
e425a60b
YO
1746 }
1747
e425a60b 1748 {
de946258 1749 # Message-ID: <20070818091501.7eff4831@r2d2>
e425a60b
YO
1750 my $str = "";
1751 for (0 .. 5) {
1752 my @x;
1753 $str .= "@x"; # this should ALWAYS be the empty string
1754 'a' =~ /(a|)/;
1755 push @x, 1;
1756 }
de26e0cc 1757 is(length $str, 0, "Trie scope error, string should be empty");
e425a60b
YO
1758 $str = "";
1759 my @foo = ('a') x 5;
1760 for (@foo) {
1761 my @bar;
1762 $str .= "@bar";
1763 s/a|/push @bar, 1/e;
1764 }
de26e0cc 1765 is(length $str, 0, "Trie scope error, string should be empty");
e425a60b
YO
1766 }
1767
e425a60b 1768 {
e425a60b 1769# more TRIE/AHOCORASICK problems with mixed utf8 / latin-1 and case folding
f36cc39d
KW
1770 for my $ord (160 .. 255) {
1771 my $chr = utf8::unicode_to_native($ord);
0f289c68
YO
1772 my $chr_byte = chr($chr);
1773 my $chr_utf8 = chr($chr); utf8::upgrade($chr_utf8);
1774 my $rx = qr{$chr_byte|X}i;
b7ad6c3e 1775 like($chr_utf8, $rx, "utf8/latin, codepoint $chr");
0f289c68 1776 }
e425a60b
YO
1777 }
1778
1779 {
e425a60b
YO
1780 our $a = 3; "" =~ /(??{ $a })/;
1781 our $b = $a;
de26e0cc 1782 is($b, $a, "Copy of scalar used for postponed subexpression");
e425a60b
YO
1783 }
1784
e425a60b 1785 {
e425a60b
YO
1786 our @ctl_n = ();
1787 our @plus = ();
1788 our $nested_tags;
1789 $nested_tags = qr{
1790 <
1791 (\w+)
1792 (?{
1793 push @ctl_n,$^N;
1794 push @plus,$+;
1795 })
1796 >
1797 (??{$nested_tags})*
1798 </\s* \w+ \s*>
1799 }x;
1800
1801 my $match = '<bla><blubb></blubb></bla>' =~ m/^$nested_tags$/;
1802 ok $match, 'nested construct matches';
de26e0cc
NC
1803 is("@ctl_n", "bla blubb", '$^N inside of (?{}) works as expected');
1804 is("@plus", "bla blubb", '$+ inside of (?{}) works as expected');
e425a60b
YO
1805 }
1806
e425a60b
YO
1807 SKIP: {
1808 # XXX: This set of tests is essentially broken, POSIX character classes
0f289c68 1809 # should not have differing definitions under Unicode.
e425a60b 1810 # There are property names for that.
ef237063 1811 skip "Tests assume ASCII", 4 unless $::IS_ASCII;
e425a60b
YO
1812
1813 my @notIsPunct = grep {/[[:punct:]]/ and not /\p{IsPunct}/}
1814 map {chr} 0x20 .. 0x7f;
1a3f15c1 1815 is(join ('', @notIsPunct), '$+<=>^`|~',
de26e0cc 1816 '[:punct:] disagrees with IsPunct on Symbols');
e425a60b
YO
1817
1818 my @isPrint = grep {not /[[:print:]]/ and /\p{IsPrint}/}
1819 map {chr} 0 .. 0x1f, 0x7f .. 0x9f;
de26e0cc
NC
1820 is(join ('', @isPrint), "",
1821 'IsPrint agrees with [:print:] on control characters');
e425a60b
YO
1822
1823 my @isPunct = grep {/[[:punct:]]/ != /\p{IsPunct}/}
1824 map {chr} 0x80 .. 0xff;
7620cb10 1825 is(join ('', @isPunct), "\xa1\xa7\xab\xb6\xb7\xbb\xbf", # ¡ « · » ¿
de26e0cc 1826 'IsPunct disagrees with [:punct:] outside ASCII');
e425a60b
YO
1827
1828 my @isPunctLatin1 = eval q {
3fc5553b 1829 grep {/[[:punct:]]/u != /\p{IsPunct}/} map {chr} 0x80 .. 0xff;
e425a60b
YO
1830 };
1831 skip "Eval failed ($@)", 1 if $@;
1832 skip "PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS set to 0", 1
1833 if !$ENV{PERL_TEST_LEGACY_POSIX_CC};
de26e0cc
NC
1834 is(join ('', @isPunctLatin1), '',
1835 'IsPunct agrees with [:punct:] with explicit Latin1');
0f289c68 1836 }
d1c771f5 1837
d1c771f5
B
1838 {
1839 # Tests for [#perl 71942]
1840 our $count_a;
1841 our $count_b;
1842
1843 my $c = 0;
1844 for my $re (
1845# [
1846# should match?,
1847# input string,
1848# re 1,
1849# re 2,
1850# expected values of count_a and count_b,
1851# ]
1852 [
1853 0,
1854 "xababz",
1855 qr/a+(?{$count_a++})b?(*COMMIT)(*FAIL)/,
1856 qr/a+(?{$count_b++})b?(*COMMIT)z/,
1857 1,
1858 ],
1859 [
1860 0,
1861 "xababz",
1862 qr/a+(?{$count_a++})b?(*COMMIT)\s*(*FAIL)/,
1863 qr/a+(?{$count_b++})b?(*COMMIT)\s*z/,
1864 1,
1865 ],
1866 [
1867 0,
1868 "xababz",
1869 qr/a+(?{$count_a++})(?:b|)?(*COMMIT)(*FAIL)/,
1870 qr/a+(?{$count_b++})(?:b|)?(*COMMIT)z/,
1871 1,
1872 ],
1873 [
1874 0,
1875 "xababz",
1876 qr/a+(?{$count_a++})b{0,6}(*COMMIT)(*FAIL)/,
1877 qr/a+(?{$count_b++})b{0,6}(*COMMIT)z/,
1878 1,
1879 ],
1880 [
1881 0,
1882 "xabcabcz",
1883 qr/a+(?{$count_a++})(bc){0,6}(*COMMIT)(*FAIL)/,
1884 qr/a+(?{$count_b++})(bc){0,6}(*COMMIT)z/,
1885 1,
1886 ],
1887 [
1888 0,
1889 "xabcabcz",
1890 qr/a+(?{$count_a++})(bc*){0,6}(*COMMIT)(*FAIL)/,
1891 qr/a+(?{$count_b++})(bc*){0,6}(*COMMIT)z/,
1892 1,
1893 ],
1894
1895
1896 [
1897 0,
1898 "aaaabtz",
1899 qr/a+(?{$count_a++})b?(*PRUNE)(*FAIL)/,
1900 qr/a+(?{$count_b++})b?(*PRUNE)z/,
1901 4,
1902 ],
1903 [
1904 0,
1905 "aaaabtz",
1906 qr/a+(?{$count_a++})b?(*PRUNE)\s*(*FAIL)/,
1907 qr/a+(?{$count_b++})b?(*PRUNE)\s*z/,
1908 4,
1909 ],
1910 [
1911 0,
1912 "aaaabtz",
1913 qr/a+(?{$count_a++})(?:b|)(*PRUNE)(*FAIL)/,
1914 qr/a+(?{$count_b++})(?:b|)(*PRUNE)z/,
1915 4,
1916 ],
1917 [
1918 0,
1919 "aaaabtz",
1920 qr/a+(?{$count_a++})b{0,6}(*PRUNE)(*FAIL)/,
1921 qr/a+(?{$count_b++})b{0,6}(*PRUNE)z/,
1922 4,
1923 ],
1924 [
1925 0,
1926 "aaaabctz",
1927 qr/a+(?{$count_a++})(bc){0,6}(*PRUNE)(*FAIL)/,
1928 qr/a+(?{$count_b++})(bc){0,6}(*PRUNE)z/,
1929 4,
1930 ],
1931 [
1932 0,
1933 "aaaabctz",
1934 qr/a+(?{$count_a++})(bc*){0,6}(*PRUNE)(*FAIL)/,
1935 qr/a+(?{$count_b++})(bc*){0,6}(*PRUNE)z/,
1936 4,
1937 ],
1938
1939 [
1940 0,
1941 "aaabaaab",
1942 qr/a+(?{$count_a++;})b?(*SKIP)(*FAIL)/,
1943 qr/a+(?{$count_b++;})b?(*SKIP)z/,
1944 2,
1945 ],
1946 [
1947 0,
1948 "aaabaaab",
1949 qr/a+(?{$count_a++;})b?(*SKIP)\s*(*FAIL)/,
1950 qr/a+(?{$count_b++;})b?(*SKIP)\s*z/,
1951 2,
1952 ],
1953 [
1954 0,
1955 "aaabaaab",
1956 qr/a+(?{$count_a++;})(?:b|)(*SKIP)(*FAIL)/,
1957 qr/a+(?{$count_b++;})(?:b|)(*SKIP)z/,
1958 2,
1959 ],
1960 [
1961 0,
1962 "aaabaaab",
1963 qr/a+(?{$count_a++;})b{0,6}(*SKIP)(*FAIL)/,
1964 qr/a+(?{$count_b++;})b{0,6}(*SKIP)z/,
1965 2,
1966 ],
1967 [
1968 0,
1969 "aaabcaaabc",
1970 qr/a+(?{$count_a++;})(bc){0,6}(*SKIP)(*FAIL)/,
1971 qr/a+(?{$count_b++;})(bc){0,6}(*SKIP)z/,
1972 2,
1973 ],
1974 [
1975 0,
1976 "aaabcaaabc",
1977 qr/a+(?{$count_a++;})(bc*){0,6}(*SKIP)(*FAIL)/,
1978 qr/a+(?{$count_b++;})(bc*){0,6}(*SKIP)z/,
1979 2,
1980 ],
1981
1982
1983 [
1984 0,
1985 "aaddbdaabyzc",
1986 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) (*FAIL) \s* c \1 /x,
1987 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) z \s* c \1 /x,
1988 4,
1989 ],
1990 [
1991 0,
1992 "aaddbdaabyzc",
1993 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) \s* (*FAIL) \s* c \1 /x,
1994 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) \s* z \s* c \1 /x,
1995 4,
1996 ],
1997 [
1998 0,
1999 "aaddbdaabyzc",
2000 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (?:b|) (*SKIP:T1) (*FAIL) \s* c \1 /x,
2001 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (?:b|) (*SKIP:T1) z \s* c \1 /x,
2002 4,
2003 ],
2004 [
2005 0,
2006 "aaddbdaabyzc",
2007 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b{0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x,
2008 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b{0,6} (*SKIP:T1) z \s* c \1 /x,
2009 4,
2010 ],
2011 [
2012 0,
2013 "aaddbcdaabcyzc",
2014 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (bc){0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x,
2015 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (bc){0,6} (*SKIP:T1) z \s* c \1 /x,
2016 4,
2017 ],
2018 [
2019 0,
2020 "aaddbcdaabcyzc",
2021 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (bc*){0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x,
2022 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (bc*){0,6} (*SKIP:T1) z \s* c \1 /x,
2023 4,
2024 ],
2025
2026
2027 [
2028 0,
2029 "aaaaddbdaabyzc",
2030 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x,
2031 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) z \s* c \1 /x,
2032 2,
2033 ],
2034 [
2035 0,
2036 "aaaaddbdaabyzc",
2037 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) \s* (*FAIL) \s* c \1 /x,
2038 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) \s* z \s* c \1 /x,
2039 2,
2040 ],
2041 [
2042 0,
2043 "aaaaddbdaabyzc",
2044 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (?:b|) (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x,
2045 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (?:b|) (*MARK:T1) (*SKIP:T1) z \s* c \1 /x,
2046 2,
2047 ],
2048 [
2049 0,
2050 "aaaaddbdaabyzc",
2051 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b{0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x,
2052 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b{0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x,
2053 2,
2054 ],
2055 [
2056 0,
2057 "aaaaddbcdaabcyzc",
2058 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (bc){0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x,
2059 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (bc){0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x,
2060 2,
2061 ],
2062 [
2063 0,
2064 "aaaaddbcdaabcyzc",
2065 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (bc*){0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x,
2066 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (bc*){0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x,
2067 2,
2068 ],
2069
2070
2071 [
2072 0,
2073 "AbcdCBefgBhiBqz",
2074 qr/(A (.*) (?{ $count_a++ }) C? (*THEN) | A D) (*FAIL)/x,
2075 qr/(A (.*) (?{ $count_b++ }) C? (*THEN) | A D) z/x,
2076 1,
2077 ],
2078 [
2079 0,
2080 "AbcdCBefgBhiBqz",
2081 qr/(A (.*) (?{ $count_a++ }) C? (*THEN) | A D) \s* (*FAIL)/x,
2082 qr/(A (.*) (?{ $count_b++ }) C? (*THEN) | A D) \s* z/x,
2083 1,
2084 ],
2085 [
2086 0,
2087 "AbcdCBefgBhiBqz",
2088 qr/(A (.*) (?{ $count_a++ }) (?:C|) (*THEN) | A D) (*FAIL)/x,
2089 qr/(A (.*) (?{ $count_b++ }) (?:C|) (*THEN) | A D) z/x,
2090 1,
2091 ],
2092 [
2093 0,
2094 "AbcdCBefgBhiBqz",
2095 qr/(A (.*) (?{ $count_a++ }) C{0,6} (*THEN) | A D) (*FAIL)/x,
2096 qr/(A (.*) (?{ $count_b++ }) C{0,6} (*THEN) | A D) z/x,
2097 1,
2098 ],
2099 [
2100 0,
2101 "AbcdCEBefgBhiBqz",
2102 qr/(A (.*) (?{ $count_a++ }) (CE){0,6} (*THEN) | A D) (*FAIL)/x,
2103 qr/(A (.*) (?{ $count_b++ }) (CE){0,6} (*THEN) | A D) z/x,
2104 1,
2105 ],
2106 [
2107 0,
2108 "AbcdCBefgBhiBqz",
2109 qr/(A (.*) (?{ $count_a++ }) (CE*){0,6} (*THEN) | A D) (*FAIL)/x,
2110 qr/(A (.*) (?{ $count_b++ }) (CE*){0,6} (*THEN) | A D) z/x,
2111 1,
2112 ],
2113 ) {
2114 $c++;
2115 $count_a = 0;
2116 $count_b = 0;
2117
2118 my $match_a = ($re->[1] =~ $re->[2]) || 0;
2119 my $match_b = ($re->[1] =~ $re->[3]) || 0;
2120
de26e0cc
NC
2121 is($match_a, $re->[0], "match a " . ($re->[0] ? "succeeded" : "failed") . " ($c)");
2122 is($match_b, $re->[0], "match b " . ($re->[0] ? "succeeded" : "failed") . " ($c)");
2123 is($count_a, $re->[4], "count a ($c)");
2124 is($count_b, $re->[4], "count b ($c)");
d1c771f5
B
2125 }
2126 }
e425a60b 2127
b57e4118
KW
2128 { # Bleadperl v5.13.8-292-gf56b639 breaks NEZUMI/Unicode-LineBreak-1.011
2129 # \xdf in lookbehind failed to compile as is multi-char fold
14358a41
NC
2130 my $message = "Lookbehind with \\xdf matchable compiles";
2131 my $r = eval 'qr{
b57e4118
KW
2132 (?u: (?<=^url:) |
2133 (?<=[/]) (?=[^/]) |
2134 (?<=[^-.]) (?=[-~.,_?\#%=&]) |
2135 (?<=[=&]) (?=.)
14358a41
NC
2136 )}iox';
2137 is($@, '', $message);
bbce3ca6 2138 object_ok($r, 'Regexp', $message);
b57e4118
KW
2139 }
2140
84193928 2141 # RT #82610
8f46dbea 2142 like 'foo/file.fob', qr,^(?=[^\.])[^/]*/(?=[^\.])[^/]*\.fo[^/]$,;
84193928 2143
295c2f7d 2144 { # This was failing unless an explicit /d was added
f36cc39d
KW
2145 my $E0 = uni_to_native("\xE0");
2146 my $p = qr/[_$E0]/i;
295c2f7d 2147 utf8::upgrade($p);
f36cc39d 2148 like(uni_to_native("\xC0"), qr/$p/, "Verify \"\\xC0\" =~ /[\\xE0_]/i; pattern in utf8");
295c2f7d
KW
2149 }
2150
b7ad6c3e 2151 like "x", qr/\A(?>(?:(?:)A|B|C?x))\z/,
a40630bf 2152 "Check TRIE does not overwrite EXACT following NOTHING at start - RT #111842";
c221a219 2153
26faadbd 2154 {
9e3a0e9b 2155 my $single = "z";
26faadbd
KW
2156 my $upper = "\x{390}"; # Fold is 3 chars.
2157 my $multi = CORE::fc($upper);
2158
2159 my $failed = 0;
2160
2161 # Try forcing a node to be split, with a multi-char fold at the
2162 # boundary
2163 for my $repeat (1 .. 300) {
2164 my $string = $single x $repeat;
2165 my $lhs = $string . $upper;
2166 if ($lhs !~ m/$string$multi/i) {
2167 $failed = $repeat;
2168 last;
2169 }
2170 }
2171 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed");
2172
b9c48b5b
KW
2173 $failed = 0;
2174 for my $repeat (1 .. 300) {
2175 my $string = $single x $repeat;
2176 my $lhs = $string . "\N{LATIN SMALL LIGATURE FFI}";
2177 if ($lhs !~ m/${string}ff\N{LATIN SMALL LETTER I}/i) {
2178 $failed = $repeat;
2179 last;
2180 }
2181 }
2182 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed");
2183
2184 $failed = 0;
2185 for my $repeat (1 .. 300) {
2186 my $string = $single x $repeat;
2187 my $lhs = $string . "\N{LATIN SMALL LIGATURE FFL}";
2188 if ($lhs !~ m/${string}ff\N{U+6c}/i) {
2189 $failed = $repeat;
2190 last;
2191 }
2192 }
2193 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed");
88b3a463
KW
2194
2195 # This tests that under /d matching that an 'ss' split across two
2196 # parts of a node doesn't end up turning into something that matches
2197 # \xDF unless it is in utf8.
2198 $failed = 0;
2199 $single = 'a'; # Is non-terminal multi-char fold char
2200 for my $repeat (1 .. 300) {
2201 my $string = $single x $repeat;
2202 my $lhs = "$string\N{LATIN SMALL LETTER SHARP S}";
2203 utf8::downgrade($lhs);
2204 $string .= "s";
2205 if ($lhs =~ m/${string}s/di) {
2206 $failed = $repeat;
2207 last;
2208 }
2209 }
2210 ok(! $failed, "Matched multi-char fold 'ss' across EXACTF node boundaries; if failed, was at count $failed");
73976260
KW
2211
2212 for my $non_finals ("t", "ft", "ift", "sift") {
2213 my $base_pat = $non_finals . "enKalt"; # (The tail is taken from
2214 # the trouble ticket, is
2215 # arbitrary)
2216 for my $utf8 ("non-UTF-8", "UTF-8") {
2217
2218 # Try at different lengths to be sure to get a node boundary
2219 for my $repeat (120 .. 270) { # [perl #133756]
2220 my $head = ("b" x $repeat) . "\xDC";
2221 my $pat = $base_pat;
2222 utf8::upgrade($pat) if $utf8 eq "UTF-8";
2223 $pat = $head . $pat;
2224 my $text = $head . $base_pat;
2225
2226 if ($text !~ /$pat/i) {
2227 $failed = $repeat;
2228 last;
2229 }
2230 }
2231
2232 ok(! $failed, "A non-final fold character "
2233 . (length($non_finals) - 1)
2234 . " characters from the end of an EXACTFish"
2235 . " $utf8 pattern works; if failed, was at count $failed");
2236 }
2237 }
26faadbd
KW
2238 }
2239
2b233a8f 2240 {
2b233a8f
KW
2241 fresh_perl_is('print eval "\"\x{101}\" =~ /[[:lower:]]/", "\n"; print eval "\"\x{100}\" =~ /[[:lower:]]/i", "\n";',
2242 "1\n1", # Both re's should match
20e5bab4 2243 {},
2b233a8f
KW
2244 "get [:lower:] swash in first eval; test under /i in second");
2245 }
2246
c5de0829 2247 {
3553f4fa
DM
2248 fresh_perl_is(<<'EOF',
2249 my $s = "\x{41c}";
2250 $s =~ /(.*)/ or die;
2251 $ls = lc $1;
2252 print $ls eq lc $s ? "good\n" : "bad: [$ls]\n";
2253EOF
2254 "good\n",
2255 {},
2256 "swash triggered by lc() doesn't corrupt \$1"
2257 );
2258 }
2259
2260 {
c5de0829 2261 #' RT #119075
fd2268bc 2262 no warnings 'regexp'; # Silence "has useless greediness modifier"
c5de0829
JK
2263 local $@;
2264 eval { /a{0}?/; };
2265 ok(! $@,
2266 "PCRE regression test: No 'Quantifier follows nothing in regex' warning");
2267
2268 }
2269
709d08c2
KW
2270 {
2271 unlike("\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
2272 like("\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
2273 unlike("\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
2274 like("\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
2275 unlike("\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
2276 like("\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
2277
2278 unlike("_\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
2279 like("_\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
2280 unlike("_\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
2281 like("_\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
2282 unlike("_\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
2283 like("_\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
2284 }
2285
2286 # These are defined later, so won't be known at regex compile time above
2287 sub IsMyRuntimeProperty {
2288 return "B6\n";
2289 }
2290
2291 sub IsntMyRuntimeProperty {
2292 return "!B6\n";
2293 }
2294
80836a6e
KW
2295 { # [perl 121777]
2296 my $regex;
2297 { package Some;
2298 # define a Unicode propertyIs_q
2299 sub Is_q
2300 {
2301 sprintf '%x', ord 'q'
2302 }
2303 $regex = qr/\p{Is_q}/;
2304
2305 # If we uncomment the following line, prior to the patch that
de68ac66
KW
2306 # fixed this, everything would work because we would have expanded
2307 # the property by the time the regex in the 'like' below got
2308 # compiled.
80836a6e
KW
2309 #'q' =~ $regex;
2310 }
2311
2312 like('q', $regex, 'User-defined property matches outside package');
c0611711
KW
2313
2314 package Some {
2315 main::like('abcq', qr/abc$regex/, 'Run-time compiled in-package user-defined property matches');
2316 }
80836a6e
KW
2317 }
2318
bc031a7d
KW
2319 { # From Lingua::Stem::UniNE; no ticket filed but related to #121778
2320 use utf8;
2321 my $word = 'рабта';
2322 $word =~ s{ (?:
2323 ия # definite articles for nouns:
2324 | ът # ∙ masculine
2325 | та # ∙ feminine
2326 | то # ∙ neutral
2327 | те # ∙ plural
2328 ) $ }{}x;
2329 is($word, 'раб', "Handles UTF8 trie correctly");
2330 }
2331
842a9d21
KW
2332 { # [perl #122460]
2333 my $a = "rdvark";
2334 $a =~ /(?{})(?=[A-Za-z0-9_])a*?/g;
2335 is (pos $a, 0, "optimizer correctly thinks (?=...) is 0-length");
2336 }
2337
4cbd7e22
FC
2338 { # [perl #123417] multi-char \N{...} tripping roundly
2339 use Cname;
2340 my $qr = qr$(\N{foo})$;
2341 "afoot" =~ eval "qr/$qr/";
2342 is "$1" || $@, "foo", 'multichar \N{...} stringified and retoked';
2343 }
2344
7195e5da
KW
2345 is (scalar split(/\b{sb}/, "Don't think twice. It's all right."),
2346 2, '\b{wb} splits sentences correctly');
2347
39ce401c
KW
2348 ok "my/dir/audio_07.mp3" =~
2349 qr/(.*)\/(.*)\/(.*)\.(?<=(?=(?:\.(?!\d+\b)\w{1,4}$)$)\.)(.*)$()/,
2350 "[perl #133948]";
2351
4cbd7e22 2352
de085464
KW
2353 # !!! NOTE! Keep the following tests last -- they may crash perl
2354
e425a60b 2355 print "# Tests that follow may crash perl\n";
e425a60b
YO
2356 {
2357 eval '/\k/';
b7ad6c3e 2358 like $@, qr/\QSequence \k... not terminated in regex;\E/,
e425a60b
YO
2359 'Lone \k not allowed';
2360 }
2361
e425a60b 2362 {
de946258 2363 my $message = "Substitution with lookahead (possible segv)";
e425a60b
YO
2364 $_ = "ns1ns1ns1";
2365 s/ns(?=\d)/ns_/g;
de946258 2366 is($_, "ns_1ns_1ns_1", $message);
e425a60b
YO
2367 $_ = "ns1";
2368 s/ns(?=\d)/ns_/;
de946258 2369 is($_, "ns_1", $message);
e425a60b
YO
2370 $_ = "123";
2371 s/(?=\d+)|(?<=\d)/!Bang!/g;
de946258 2372 is($_, "!Bang!1!Bang!2!Bang!3!Bang!", $message);
e425a60b
YO
2373 }
2374
6182169b
KW
2375 {
2376 # Earlier versions of Perl said this was fatal.
de946258 2377 my $message = "U+0FFFF shouldn't crash the regex engine";
6182169b
KW
2378 no warnings 'utf8';
2379 my $a = eval "chr(65535)";
2380 use warnings;
2381 my $warning_message;
2382 local $SIG{__WARN__} = sub { $warning_message = $_[0] };
2383 eval $a =~ /[a-z]/;
de946258 2384 ok(1, $message); # If it didn't crash, it worked.
6182169b 2385 }
b57e4118 2386
06345901
NC
2387 TODO: { # Was looping
2388 todo_skip('Triggers thread clone SEGV. See #86550')
2389 if $::running_as_thread && $::running_as_thread;
478c6b74 2390 watchdog(10); # Use a bigger value for busy systems
6438af90
KW
2391 like("\x{00DF}", qr/[\x{1E9E}_]*/i, "\"\\x{00DF}\" =~ /[\\x{1E9E}_]*/i was looping");
2392 }
2393
f9126265
KW
2394 { # Bug #90536, caused failed assertion
2395 unlike("s\N{U+DF}", qr/^\x{00DF}/i, "\"s\\N{U+DF}\", qr/^\\x{00DF}/i");
2396 }
2397
361ee0fe 2398 # User-defined Unicode properties to match above-Unicode code points
56a85032 2399 sub Is_31_Bit_Super { return "110000\t7FFFFFFF\n" }
361ee0fe 2400 sub Is_Portable_Super { return '!utf8::Any' } # Matches beyond 32 bits
45d91b83
KW
2401
2402 { # Assertion was failing on on 64-bit platforms; just didn't work on 32.
2403 no warnings qw(non_unicode portable);
2404 use Config;
2405
2406 # We use 'ok' instead of 'like' because the warnings are lexically
2407 # scoped, and want to turn them off, so have to do the match in this
c0236afe 2408 # scope.
76513bdc 2409 if ($Config{uvsize} < 8) {
8f46dbea 2410 like(chr(0x7FFF_FFFE), qr/\p{Is_31_Bit_Super}/,
56a85032 2411 "chr(0x7FFF_FFFE) can match a Unicode property");
8f46dbea 2412 like(chr(0x7FFF_FFFF), qr/\p{Is_31_Bit_Super}/,
56a85032
KW
2413 "chr(0x7FFF_FFFF) can match a Unicode property");
2414 my $p = qr/^[\x{7FFF_FFFF}]$/;
8f46dbea 2415 like(chr(0x7FFF_FFFF), qr/$p/,
56a85032 2416 "chr(0x7FFF_FFFF) can match itself in a [class]");
8f46dbea 2417 like(chr(0x7FFF_FFFF), qr/$p/, # Tests any caching
56a85032 2418 "chr(0x7FFF_FFFF) can match itself in a [class] subsequently");
84ea5ef6 2419 }
76513bdc
KW
2420 else {
2421 no warnings 'overflow';
8f46dbea 2422 like(chr(0x7FFF_FFFF_FFFF_FFFE), qr/\p{Is_Portable_Super}/,
56a85032 2423 "chr(0x7FFF_FFFF_FFFF_FFFE) can match a Unicode property");
8f46dbea 2424 like(chr(0x7FFF_FFFF_FFFF_FFFF), qr/^\p{Is_Portable_Super}$/,
56a85032 2425 "chr(0x7FFF_FFFF_FFFF_FFFF) can match a Unicode property");
76513bdc 2426
56a85032 2427 my $p = qr/^[\x{7FFF_FFFF_FFFF_FFFF}]$/;
8f46dbea 2428 like(chr(0x7FFF_FFFF_FFFF_FFFF), qr/$p/,
56a85032 2429 "chr(0x7FFF_FFFF_FFFF_FFFF) can match itself in a [class]");
8f46dbea 2430 like(chr(0x7FFF_FFFF_FFFF_FFFF), qr/$p/, # Tests any caching
56a85032 2431 "chr(0x7FFF_FFFF_FFFF_FFFF) can match itself in a [class] subsequently");
76513bdc
KW
2432
2433 # This test is because something was declared as 32 bits, but
2434 # should have been cast to 64; only a problem where
2435 # sizeof(STRLEN) != sizeof(UV)
8f46dbea
KW
2436 unlike(chr(0x7FFF_FFFF_FFFF_FFFE), qr/\p{Is_31_Bit_Super}/,
2437 "chr(0x7FFF_FFFF_FFFF_FFFE) shouldn't match a range ending in 0x7FFF_FFFF");
76513bdc 2438 }
45d91b83
KW
2439 }
2440
9d501133
KW
2441 { # [perl #112530], the code below caused a panic
2442 sub InFoo { "a\tb\n9\ta\n" }
f36cc39d 2443 like(chr(0xA), qr/\p{InFoo}/,
9d501133
KW
2444 "Overlapping ranges in user-defined properties");
2445 }
2446
512c0f5a
KW
2447 { # [perl #125990], the final 2 tests below each caused a panic.
2448 # The \0's are not necessary; it could be a printable character
2449 # instead, but were in the ticket, so using them.
2450 my $sharp_s = chr utf8::unicode_to_native(0xdf);
2451 my $string = ("\0" x 8)
2452 . ($sharp_s x 3)
2453 . ("\0" x 42)
2454 . "ý";
2455 my $folded_string = ("\0" x 8)
2456 . ("ss" x 3)
2457 . ("\0" x 42)
2458 . "ý";
2459 utf8::downgrade($string);
2460 utf8::downgrade($folded_string);
2461
37ee5518 2462 use Cname;
512c0f5a
KW
2463 like($string, qr/$string/i, "LATIN SMALL SHARP S matches itself under /id");
2464 unlike($folded_string, qr/$string/i, "LATIN SMALL SHARP S doesn't match 'ss' under /di");
37ee5518
KW
2465 like($folded_string, qr/\N{EMPTY-STR}$string/i, "\\N{} earlier than LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'");
2466 like($folded_string, qr/$string\N{EMPTY-STR}/i, "\\N{} after LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'");
512c0f5a
KW
2467 }
2468
90c17285 2469 { # [perl #126606 crashed the interpreter
37ee5518
KW
2470 use Cname;
2471 like("sS", qr/\N{EMPTY-STR}Ss|/i, '\N{} with empty branch alternation works');
945961fd 2472 like("sS", qr'\N{EMPTY-STR}Ss|'i, '\N{} with empty branch alternation works');
90c17285
KW
2473 }
2474
2e3a23da
KW
2475 { # Regexp:Grammars was broken:
2476 # http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2013-06/msg01290.html
2477 fresh_perl_like('use warnings; "abc" =~ qr{(?&foo){0}abc(?<foo>)}',
aaa63dae 2478 qr/Quantifier unexpected on zero-length expression/,
20e5bab4 2479 {},
2e3a23da
KW
2480 'No segfault on qr{(?&foo){0}abc(?<foo>)}');
2481 }
2482
668fcfea
TC
2483 SKIP:
2484 { # [perl #125826] buffer overflow in TRIE_STORE_REVCHAR
2485 # (during compilation, so use a fresh perl)
2486 $Config{uvsize} == 8
2487 or skip("need large code-points for this test", 1);
760c7c2f 2488
d22ec717 2489 fresh_perl_is('/\x{E000000000}|/ and print qq(ok\n)', "ok\n", {},
668fcfea
TC
2490 "buffer overflow in TRIE_STORE_REVCHAR");
2491 }
2492
8571c4b0
KW
2493 {
2494 fresh_perl_like('use warnings; s\00(?(?!00000000000000000000000000·000000)\500000000\00000000000000000000000000000000000000000000000000000·00000000000000000000000000000000\00',
2495 qr/Switch \(\?\(condition\)\.\.\. not terminated/,
2496 {},
2497 'No segfault [perl #126886]');
2498 }
2499
553fa53c
JK
2500 {
2501 # [perl 130010] Downstream application texinfo started to report panics
2502 # as of commit a5540cf.
2503
2504 runperl( prog => 'A::xx(); package A; sub InFullwidth{ return qq|\n| } sub xx { split /[^\s\p{InFullwidth}]/, q|x| }' );
2505 ok(! $?, "User-defined pattern did not cause panic [perl 130010]");
2506 }
2507
b92b2705
KW
2508 { # [perl #133999] Previously assertion failure
2509 fresh_perl_like('0 =~ /\p{nv:(\B(*COMMIT)C+)}/',
2510 qr/No Unicode property value wildcard matches/,
2511 {},
2512 "Assertion failure with *COMMIT and wildcard property");
2513 }
2514
11b48e76
KW
2515 { # [perl #134029] Previously assertion failure
2516 fresh_perl_like('qr/\p{upper:]}|\337(?|ss)|)(?0/',
2517 qr/Unicode property wildcard not terminated/,
2518 {},
2519 "Assertion failure with single character wildcard");
2520 }
2521
8c9c2723
KW
2522 { # [perl #134034] Previously assertion failure
2523 fresh_perl_is('use utf8; q!Ȧिम한글💣΢ყაოსაა!=~/(?li)\b{wb}\B(*COMMIT)0/;',
2524 "", {}, "*COMMIT caused positioning beyond EOS");
2525 }
2526
553fa53c 2527
b57e4118 2528 # !!! NOTE that tests that aren't at all likely to crash perl should go
de085464
KW
2529 # a ways above, above these last ones. There's a comment there that, like
2530 # this comment, contains the word 'NOTE'
f4554ed5
NC
2531
2532 done_testing();
e425a60b
YO
2533} # End of sub run_tests
2534
25351;