1#!./perl 2# 3# This is a home for regular expression tests that do not fit into 4# the format supported by re/regexp.t. If you want to add a test 5# that does fit that format, add it to re/re_tests, not here. 6 7BEGIN { 8 chdir 't' if -d 't'; 9 require './test.pl'; 10 require './charset_tools.pl'; 11 set_up_inc(qw '../lib .'); 12 skip_all_if_miniperl("miniperl can't load Tie::Hash::NamedCapture, need for %+ and %-"); 13} 14 15use strict; 16use warnings; 17use 5.010; 18our ($REGMARK, $REGERROR); 19 20sub run_tests; 21 22$| = 1; 23 24run_tests() unless caller; 25 26# 27# Tests start here. 28# 29sub run_tests { 30 31 { 32 # Japhy -- added 03/03/2001 33 () = (my $str = "abc") =~ /(...)/; 34 $str = "def"; 35 is($1, "abc", 'Changing subject does not modify $1'); 36 } 37 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. 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]/'; 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. 54 unlike("\x8e", qr/[i-j]/, '"\x8e" !~ /[i-j]/'); 55 unlike("\xce", qr/[I-J]/, '"\xce" !~ /[I-J]/'); 56 unlike("\xd0", qr/[I-J]/, '"\xd0" !~ /[I-J]/'); 57 } 58 59 { 60 like "\x{ab}", qr/\x{ab}/, '"\x{ab}" =~ /\x{ab}/ '; 61 like "\x{abcd}", qr/\x{abcd}/, '"\x{abcd}" =~ /\x{abcd}/'; 62 } 63 64 { 65 my $message = 'bug id 20001008.001 (#4407)'; 66 67 my $strasse = "stra" . uni_to_native("\337") . "e"; 68 my @x = ("$strasse 138", "$strasse 138"); 69 for (@x) { 70 ok(s/(\d+)\s*([\w\-]+)/$1 . uc $2/e, $message); 71 ok(my ($latin) = /^(.+)(?:\s+\d)/, $message); 72 is($latin, $strasse, $message); 73 ok($latin =~ s/$strasse/straße/, $message); 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 83 { 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); 93 } 94 95 { 96 my $message = 'Match code points > 255'; 97 $_ = "abc\x{100}\x{200}\x{300}\x{380}\x{400}defg"; 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); 103 } 104 105 { 106 my $x = "\x{10FFFD}"; 107 $x =~ s/(.)/$1/g; 108 ok ord($x) == 0x10FFFD && length($x) == 1, "From Robin Houston"; 109 } 110 111 { 112 my %d = ( 113 "7f" => [0, 0, 0], 114 "80" => [1, 1, 0], 115 "ff" => [1, 1, 0], 116 "100" => [0, 1, 1], 117 ); 118 119 while (my ($code, $match) = each %d) { 120 my $message = "Properties of \\x$code"; 121 my $char = eval qq ["\\x{$code}"]; 122 123 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], $message); 126 } 127 } 128 129 { 130 # From Japhy 131 foreach (qw(c g o)) { 132 warning_like(sub {'' =~ "(?$_)"}, qr/^Useless \(\?$_\)/); 133 warning_like(sub {'' =~ "(?-$_)"}, qr/^Useless \(\?-$_\)/); 134 } 135 136 # Now test multi-error regexes 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 } 147 } 148 149 { 150 my $message = "/x tests"; 151 $_ = "foo"; 152 foreach my $pat (<<" --", <<" --") { 153 /f 154 o\r 155 o 156 \$ 157 /x 158 -- 159 /f 160 o 161 o 162 \$\r 163 /x 164 -- 165 is(eval $pat, 1, $message); 166 is($@, '', $message); 167 } 168 } 169 170 { 171 my $message = "/o feature"; 172 sub test_o {$_ [0] =~ /$_[1]/o; return $1} 173 is(test_o ('abc', '(.)..'), 'a', $message); 174 is(test_o ('abc', '..(.)'), 'a', $message); 175 } 176 177 { 178 # Test basic $^N usage outside of a regex 179 my $message = '$^N usage outside of a regex'; 180 my $x = "abcdef"; 181 ok(($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);} 196 ## Test to see if $^N is automatically localized -- it should now 197 ## have the value set in the previous test. 198 is($^N, "e", '$^N is automatically localized'); 199 200 # Now test inside (?{ ... }) 201 $message = '$^N usage inside (?{ ... })'; 202 our ($y, $z); 203 ok(($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); 210 211 } 212 213 SKIP: 214 { 215 ## Should probably put in tests for all the POSIX stuff, 216 ## but not sure how to guarantee a specific locale...... 217 218 my $message = 'Test [[:cntrl:]]'; 219 my $AllBytes = join "" => map {chr} 0 .. 255; 220 (my $x = $AllBytes) =~ s/[[:cntrl:]]//g; 221 $x = join "", sort { $a cmp $b } 222 map { chr utf8::native_to_unicode(ord $_) } split "", $x; 223 is($x, join("", map {chr} 0x20 .. 0x7E, 0x80 .. 0xFF), $message); 224 225 ($x = $AllBytes) =~ s/[^[:cntrl:]]//g; 226 $x = join "", sort { $a cmp $b } 227 map { chr utf8::native_to_unicode(ord $_) } split "", $x; 228 is($x, (join "", map {chr} 0x00 .. 0x1F, 0x7F), $message); 229 } 230 231 { 232 # With /s modifier UTF8 chars were interpreted as bytes 233 my $message = "UTF-8 chars aren't bytes"; 234 my $a = "Hello \x{263A} World"; 235 my @a = ($a =~ /./gs); 236 is($#a, 12, $message); 237 } 238 239 { 240 no warnings 'digit'; 241 # Check that \x## works. 5.6.1 and 5.005_03 fail some of these. 242 my $x; 243 $x = "\x4e" . "E"; 244 like ($x, qr/^\x4EE$/, "Check only 2 bytes of hex are matched."); 245 246 $x = "\x4e" . "i"; 247 like ($x, qr/^\x4Ei$/, "Check that invalid hex digit stops it (2)"); 248 249 $x = "\x4" . "j"; 250 like ($x, qr/^\x4j$/, "Check that invalid hex digit stops it (1)"); 251 252 $x = "\x0" . "k"; 253 like ($x, qr/^\xk$/, "Check that invalid hex digit stops it (0)"); 254 255 $x = "\x0" . "x"; 256 like ($x, qr/^\xx$/, "\\xx isn't to be treated as \\0"); 257 258 $x = "\x0" . "xa"; 259 like ($x, qr/^\xxa$/, "\\xxa isn't to be treated as \\xa"); 260 261 $x = "\x9" . "_b"; 262 like ($x, qr/^\x9_b$/, "\\x9_b isn't to be treated as \\x9b"); 263 264 # and now again in [] ranges 265 266 $x = "\x4e" . "E"; 267 like ($x, qr/^[\x4EE]{2}$/, "Check only 2 bytes of hex are matched."); 268 269 $x = "\x4e" . "i"; 270 like ($x, qr/^[\x4Ei]{2}$/, "Check that invalid hex digit stops it (2)"); 271 272 $x = "\x4" . "j"; 273 like ($x, qr/^[\x4j]{2}$/, "Check that invalid hex digit stops it (1)"); 274 275 $x = "\x0" . "k"; 276 like ($x, qr/^[\xk]{2}$/, "Check that invalid hex digit stops it (0)"); 277 278 $x = "\x0" . "x"; 279 like ($x, qr/^[\xx]{2}$/, "\\xx isn't to be treated as \\0"); 280 281 $x = "\x0" . "xa"; 282 like ($x, qr/^[\xxa]{3}$/, "\\xxa isn't to be treated as \\xa"); 283 284 $x = "\x9" . "_b"; 285 like ($x, qr/^[\x9_b]{3}$/, "\\x9_b isn't to be treated as \\x9b"); 286 287 # Check that \x{##} works. 5.6.1 fails quite a few of these. 288 289 $x = "\x9b"; 290 like ($x, qr/^\x{9_b}$/, "\\x{9_b} is to be treated as \\x9b"); 291 292 $x = "\x9b" . "y"; 293 like ($x, qr/^\x{9_b}y$/, "\\x{9_b} is to be treated as \\x9b (again)"); 294 295 $x = "\x9b" . "y"; 296 like ($x, qr/^\x{9b_}y$/, "\\x{9b_} is to be treated as \\x9b"); 297 298 $x = "\x9b" . "y"; 299 like ($x, qr/^\x{9_bq}y$/, "\\x{9_bc} is to be treated as \\x9b"); 300 301 $x = "\x0" . "y"; 302 like ($x, qr/^\x{x9b}y$/, "\\x{x9b} is to be treated as \\x0"); 303 304 $x = "\x0" . "y"; 305 like ($x, qr/^\x{0x9b}y$/, "\\x{0x9b} is to be treated as \\x0"); 306 307 $x = "\x9b" . "y"; 308 like ($x, qr/^\x{09b}y$/, "\\x{09b} is to be treated as \\x9b"); 309 310 $x = "\x9b"; 311 like ($x, qr/^[\x{9_b}]$/, "\\x{9_b} is to be treated as \\x9b"); 312 313 $x = "\x9b" . "y"; 314 like ($x, qr/^[\x{9_b}y]{2}$/, 315 "\\x{9_b} is to be treated as \\x9b (again)"); 316 317 $x = "\x9b" . "y"; 318 like ($x, qr/^[\x{9b_}y]{2}$/, "\\x{9b_} is to be treated as \\x9b"); 319 320 $x = "\x9b" . "y"; 321 like ($x, qr/^[\x{9_bq}y]{2}$/, "\\x{9_bc} is to be treated as \\x9b"); 322 323 $x = "\x0" . "y"; 324 like ($x, qr/^[\x{x9b}y]{2}$/, "\\x{x9b} is to be treated as \\x0"); 325 326 $x = "\x0" . "y"; 327 like ($x, qr/^[\x{0x9b}y]{2}$/, "\\x{0x9b} is to be treated as \\x0"); 328 329 $x = "\x9b" . "y"; 330 like ($x, qr/^[\x{09b}y]{2}$/, "\\x{09b} is to be treated as \\x9b"); 331 332 } 333 334 { 335 # High bit bug -- japhy 336 my $x = "ab\200d"; 337 like $x, qr/.*?\200/, "High bit fine"; 338 } 339 340 { 341 # The basic character classes and Unicode 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/'; 345 } 346 347 { 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); 369 } 370 371 { 372 use charnames ':full'; 373 my $message = "Folding 'LATIN LETTER A WITH GRAVE'"; 374 375 my $lower = "\N{LATIN SMALL LETTER A WITH GRAVE}"; 376 my $UPPER = "\N{LATIN CAPITAL LETTER A WITH GRAVE}"; 377 378 like($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); 382 383 $message = "Folding 'GREEK LETTER ALPHA WITH VRACHY'"; 384 385 $lower = "\N{GREEK CAPITAL LETTER ALPHA WITH VRACHY}"; 386 $UPPER = "\N{GREEK SMALL LETTER ALPHA WITH VRACHY}"; 387 388 like($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); 392 393 $message = "Folding 'LATIN LETTER Y WITH DIAERESIS'"; 394 395 $lower = "\N{LATIN SMALL LETTER Y WITH DIAERESIS}"; 396 $UPPER = "\N{LATIN CAPITAL LETTER Y WITH DIAERESIS}"; 397 398 like($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); 402 } 403 404 { 405 use charnames ':full'; 406 my $message = "GREEK CAPITAL LETTER SIGMA vs " . 407 "COMBINING GREEK PERISPOMENI"; 408 409 my $SIGMA = "\N{GREEK CAPITAL LETTER SIGMA}"; 410 my $char = "\N{COMBINING GREEK PERISPOMENI}"; 411 412 warning_is(sub {unlike("_:$char:_", qr/_:$SIGMA:_/i, $message)}, undef, 413 'Did not warn [change a5961de5f4215b5c]'); 414 } 415 416 { 417 my $message = '\X'; 418 use charnames ':full'; 419 420 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}!" 427 =~ /^(\X)!/ && 428 $1 eq "\N{LATIN CAPITAL LETTER E}\N{COMBINING GRAVE ACCENT}", $message); 429 430 } 431 432 { 433 my $message = "Final Sigma"; 434 435 my $SIGMA = "\x{03A3}"; # CAPITAL 436 my $Sigma = "\x{03C2}"; # SMALL FINAL 437 my $sigma = "\x{03C3}"; # SMALL 438 439 like($SIGMA, qr/$SIGMA/i, $message); 440 like($SIGMA, qr/$Sigma/i, $message); 441 like($SIGMA, qr/$sigma/i, $message); 442 443 like($Sigma, qr/$SIGMA/i, $message); 444 like($Sigma, qr/$Sigma/i, $message); 445 like($Sigma, qr/$sigma/i, $message); 446 447 like($sigma, qr/$SIGMA/i, $message); 448 like($sigma, qr/$Sigma/i, $message); 449 like($sigma, qr/$sigma/i, $message); 450 451 like($SIGMA, qr/[$SIGMA]/i, $message); 452 like($SIGMA, qr/[$Sigma]/i, $message); 453 like($SIGMA, qr/[$sigma]/i, $message); 454 455 like($Sigma, qr/[$SIGMA]/i, $message); 456 like($Sigma, qr/[$Sigma]/i, $message); 457 like($Sigma, qr/[$sigma]/i, $message); 458 459 like($sigma, qr/[$SIGMA]/i, $message); 460 like($sigma, qr/[$Sigma]/i, $message); 461 like($sigma, qr/[$sigma]/i, $message); 462 463 $message = "More final Sigma"; 464 465 my $S3 = "$SIGMA$Sigma$sigma"; 466 467 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, $message); 470 471 ok(":$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); 474 } 475 476 { 477 use charnames ':full'; 478 my $message = "Parlez-Vous " . 479 "Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais?"; 480 481 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); 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" =~ 492 /Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais/ && 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); 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; 505 ok($subject =~ /Fran(?:c\N{COMBINING CEDILLA}?| 506 \N{LATIN SMALL LETTER C WITH CEDILLA})ais/x && 507 $& eq $match, $message); 508 } 509 } 510 511 { 512 my $message = "Lingering (and useless) UTF8 flag doesn't mess up /i"; 513 my $pat = "ABcde"; 514 my $str = "abcDE\x{100}"; 515 chop $str; 516 like($str, qr/$pat/i, $message); 517 518 $pat = "ABcde\x{100}"; 519 $str = "abcDE"; 520 chop $pat; 521 like($str, qr/$pat/i, $message); 522 523 $pat = "ABcde\x{100}"; 524 $str = "abcDE\x{100}"; 525 chop $pat; 526 chop $str; 527 like($str, qr/$pat/i, $message); 528 } 529 530 { 531 use charnames ':full'; 532 my $message = "LATIN SMALL LETTER SHARP S " . 533 "(\N{LATIN SMALL LETTER SHARP S})"; 534 535 like("\N{LATIN SMALL LETTER SHARP S}", 536 qr/\N{LATIN SMALL LETTER SHARP S}/, $message); 537 like("\N{LATIN SMALL LETTER SHARP S}", 538 qr'\N{LATIN SMALL LETTER SHARP S}', $message); 539 like("\N{LATIN SMALL LETTER SHARP S}", 540 qr/\N{LATIN SMALL LETTER SHARP S}/i, $message); 541 like("\N{LATIN SMALL LETTER SHARP S}", 542 qr'\N{LATIN SMALL LETTER SHARP S}'i, $message); 543 like("\N{LATIN SMALL LETTER SHARP S}", 544 qr/[\N{LATIN SMALL LETTER SHARP S}]/, $message); 545 like("\N{LATIN SMALL LETTER SHARP S}", 546 qr'[\N{LATIN SMALL LETTER SHARP S}]', $message); 547 like("\N{LATIN SMALL LETTER SHARP S}", 548 qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message); 549 like("\N{LATIN SMALL LETTER SHARP S}", 550 qr'[\N{LATIN SMALL LETTER SHARP S}]'i, $message); 551 552 like("ss", qr /\N{LATIN SMALL LETTER SHARP S}/i, $message); 553 like("ss", qr '\N{LATIN SMALL LETTER SHARP S}'i, $message); 554 like("SS", qr /\N{LATIN SMALL LETTER SHARP S}/i, $message); 555 like("SS", qr '\N{LATIN SMALL LETTER SHARP S}'i, $message); 556 like("ss", qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message); 557 like("ss", qr'[\N{LATIN SMALL LETTER SHARP S}]'i, $message); 558 like("SS", qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message); 559 like("SS", qr'[\N{LATIN SMALL LETTER SHARP S}]'i, $message); 560 561 like("\N{LATIN SMALL LETTER SHARP S}", qr/ss/i, $message); 562 like("\N{LATIN SMALL LETTER SHARP S}", qr/SS/i, $message); 563 564 $message = "Unoptimized named sequence in class"; 565 like("ss", qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message); 566 like("ss", qr'[\N{LATIN SMALL LETTER SHARP S}x]'i, $message); 567 like("SS", qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message); 568 like("SS", qr'[\N{LATIN SMALL LETTER SHARP S}x]'i, $message); 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}", 572 qr'[\N{LATIN SMALL LETTER SHARP S}x]', $message); 573 like("\N{LATIN SMALL LETTER SHARP S}", 574 qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message); 575 like("\N{LATIN SMALL LETTER SHARP S}", 576 qr'[\N{LATIN SMALL LETTER SHARP S}x]'i, $message); 577 } 578 579 { 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. 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'; 587 my @h = map {sprintf "%05x" => $_} 0x01680, 0x02000 .. 0x0200A, 588 0x0202F, 0x0205F, 0x03000; 589 my @v = map {sprintf "%05x" => $_} 0x02028, 0x02029; 590 591 my @H = map {sprintf "%05x" => $_} 0x01361, 0x0200B, 0x02408, 0x02420, 592 0x0303F, 0xE0020, 0x180E; 593 my @V = map {sprintf "%05x" => $_} 0x0008A .. 0x0008D, 0x00348, 0x10100, 594 0xE005F, 0xE007C, 0x180E; 595 596 for my $hex (@h) { 597 my $str = eval qq ["<\\x{$hex}>"]; 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"; 601 } 602 603 for my $hex (@v) { 604 my $str = eval qq ["<\\x{$hex}>"]; 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"; 608 } 609 610 for my $hex (@H) { 611 my $str = eval qq ["<\\x{$hex}>"]; 612 like $str, qr/<\S>/, "\\x{$hex} in \\S"; 613 like $str, qr/<\H>/, "\\x{$hex} in \\H"; 614 } 615 616 for my $hex (@V) { 617 my $str = eval qq ["<\\x{$hex}>"]; 618 like $str, qr/<\S>/, "\\x{$hex} in \\S"; 619 like $str, qr/<\V>/, "\\x{$hex} in \\V"; 620 } 621 } 622 623 { 624 # . with /s should work on characters, as opposed to bytes 625 my $message = ". with /s works on characters, not bytes"; 626 627 my $s = "\x{e4}\x{100}"; 628 # This is not expected to match: the point is that 629 # neither should we get "Malformed UTF-8" warnings. 630 warning_is(sub {$s =~ /\G(.+?)\n/gcs}, undef, 631 "No 'Malformed UTF-8' warning"); 632 633 my @c; 634 push @c => $1 while $s =~ /\G(.)/gs; 635 636 local $" = ""; 637 is("@c", $s, $message); 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) { 643 $r1 .= $1 . $2; 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) { 649 $r2 .= $1 . $2; 650 } 651 $r2 =~ s/\x{100}//; 652 653 is($r1, $r2, $message); 654 } 655 656 { 657 my $message = "Unicode lookbehind"; 658 like("A\x{100}B", qr/(?<=A.)B/, $message); 659 like("A\x{200}\x{300}B", qr/(?<=A..)B/, $message); 660 like("\x{400}AB", qr/(?<=\x{400}.)B/, $message); 661 like("\x{500}\x{600}B", qr/(?<=\x{500}.)B/, $message); 662 663 # Original code also contained: 664 # ok "\x{500\x{600}}B" =~ /(?<=\x{500}.)B/; 665 # but that looks like a typo. 666 } 667 668 { 669 my $message = 'UTF-8 hash keys and /$/'; 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*$/ ? 1 : 0; 680 is($m1, $m2, $message); 681 } 682 } 683 684 { 685 my $message = "No SEGV in s/// and UTF-8"; 686 my $s = "s#\x{100}" x 4; 687 ok($s =~ s/[^\w]/ /g, $message); 688 if ( 1 or $ENV{PERL_TEST_LEGACY_POSIX_CC} ) { 689 is($s, "s \x{100}" x 4, $message); 690 } 691 else { 692 is($s, "s " x 4, $message); 693 } 694 } 695 696 { 697 my $message = "UTF-8 bug (maybe already known?)"; 698 my $u = "foo"; 699 $u =~ s/./\x{100}/g; 700 is($u, "\x{100}\x{100}\x{100}", $message); 701 702 $u = "foobar"; 703 $u =~ s/[ao]/\x{100}/g; 704 is($u, "f\x{100}\x{100}b\x{100}r", $message); 705 706 $u =~ s/\x{100}/e/g; 707 is($u, "feeber", $message); 708 } 709 710 { 711 my $message = "UTF-8 bug with s///"; 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)", "(?=$c).*xx",) { 719 unlike("xxx", qr/$re/, $message); 720 ok(+($subst = "xxx") !~ s/$re//, $message); 721 } 722 for my $re ("xx.*$c*", "$c*.*xx") { 723 like("xxx", qr/$re/, $message); 724 ok(+($subst = "xxx") =~ s/$re//, $message); 725 is($subst, "", $message); 726 } 727 for my $re ("xxy*", "y*xx") { 728 like("xx$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); 733 } 734 for my $re ("xy$c*z", "x$c*yz") { 735 like("xyz", qr/$re/, $message); 736 ok(+($subst = "xyz") =~ s/$re//, $message); 737 is($subst, "", $message); 738 } 739 } 740 741 { 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 { 748 my $message = "qr /.../x"; 749 my $R = qr / A B C # D E/x; 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); 755 } 756 757 { 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 } 766 767 { 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 773 774 my $message = 'Markus Kuhn 2003-02-26'; 775 776 my $x = "\x{2019}\nk"; 777 ok($x =~ s/(\S)\n(\S)/$1 $2/sg, $message); 778 is($x, "\x{2019} k", $message); 779 780 $x = "b\nk"; 781 ok($x =~ s/(\S)\n(\S)/$1 $2/sg, $message); 782 is($x, "b k", $message); 783 784 like("\x{2019}", qr/\S/, $message); 785 } 786 787 { 788 like "\x{100}\n", qr/\x{100}\n$/, "UTF-8 length cache and fbm_compile"; 789 } 790 791 { 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 } 800 801 { 802 my $re = qq /^([^X]*)X/; 803 utf8::upgrade ($re); 804 like "\x{100}X", qr/$re/, "S_cl_and ANYOF_UNICODE & ANYOF_INVERTED"; 805 my $loc_re = qq /(?l:^([^X]*)X)/; 806 utf8::upgrade ($loc_re); 807 no warnings 'locale'; 808 like "\x{100}X", qr/$loc_re/, "locale, S_cl_and ANYOF_UNICODE & ANYOF_INVERTED"; 809 } 810 811 { 812 like "123\x{100}", qr/^.*1.*23\x{100}$/, 813 'UTF-8 + multiple floating substr'; 814 } 815 816 { 817 my $message = '<20030808193656.5109.1@llama.ni-s.u-net.com>'; 818 819 # LATIN SMALL/CAPITAL LETTER A WITH MACRON 820 like(" \x{101}", qr/\x{100}/i, $message); 821 822 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW 823 like(" \x{1E01}", qr/\x{1E00}/i, $message); 824 825 # DESERET SMALL/CAPITAL LETTER LONG I 826 like(" \x{10428}", qr/\x{10400}/i, $message); 827 828 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X' 829 like(" \x{1E01}x", qr/\x{1E00}X/i, $message); 830 } 831 832 { 833 for (120 .. 130, 240 .. 260) { 834 my $head = 'x' x $_; 835 my $message = q [Don't misparse \x{...} in regexp ] . 836 q [near EXACT char count limit]; 837 for my $tail ('\x{0061}', '\x{1234}', '\x61') { 838 eval qq{like("$head$tail", qr/$head$tail/, \$message)}; 839 is($@, '', $message); 840 } 841 $message = q [Don't misparse \N{...} in regexp ] . 842 q [near EXACT char count limit]; 843 for my $tail ('\N{SNOWFLAKE}') { 844 eval qq {use charnames ':full'; 845 like("$head$tail", qr/$head$tail/, \$message)}; 846 eval qq {use charnames ':full'; 847 like("$head$tail", qr'$head$tail', \$message)}; 848 is($@, '', $message); 849 } 850 } 851 } 852 853 { # TRIE related 854 our @got = (); 855 "words" =~ /(word|word|word)(?{push @got, $1})s$/; 856 is(@got, 1, "TRIE optimisation"); 857 858 @got = (); 859 "words" =~ /(word|word|word)(?{push @got,$1})s$/i; 860 is(@got, 1,"TRIEF optimisation"); 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) { 867 like $_, qr/$re/, "Trie nums"; 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 } 882 883 { 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 } 925 926 { 927 BEGIN { 928 unshift @INC, 'lib'; 929 } 930 use Cname; # Our custom charname plugin, currently found in 931 # t/lib/Cname.pm 932 933 like 'fooB', qr/\N{foo}[\N{B}\N{b}]/, "Passthrough charname"; 934 my $name = "foo\xDF"; 935 my $result = eval "'A${name}B' =~ /^A\\N{$name}B\$/"; 936 ok !$@ && $result, "Passthrough charname of non-ASCII, Latin1"; 937 eval "qr/\\p{name=foo}/"; 938 like($@, qr/Can't find Unicode property definition "name=foo"/, 939 '\p{name=} doesn\'t see a cumstom charnames translator'); 940 # 941 # Why doesn't must_warn work here? 942 # 943 my $w; 944 local $SIG {__WARN__} = sub {$w .= "@_"}; 945 $result = eval 'q(WARN) =~ /^[\N{WARN}]$/'; 946 ok !$@ && $result && ! $w, '\N{} returning multi-char works'; 947 948 undef $w; 949 eval q [unlike "\0", qr/[\N{EMPTY-STR}XY]/, 950 "Zerolength charname in charclass doesn't match \\\\0"]; 951 ok $w && $w =~ /Ignoring zero length/, 952 'Ignoring zero length \N{} in character class warning'; 953 undef $w; 954 eval q [like 'xy', qr/x[\N{EMPTY-STR} y]/x, 955 'Empty string charname in [] is ignored; finds a following character']; 956 ok $w && $w =~ /Ignoring zero length/, 957 'Ignoring zero length \N{} in character class warning'; 958 undef $w; 959 eval q [like 'x ', qr/x[\N{EMPTY-STR} y]/, 960 'Empty string charname in [] is ignored; finds a following blank under /x']; 961 like $w, qr/Ignoring zero length/, 962 'Ignoring zero length \N{} in character class warning'; 963 964 # EVIL keeps track of its calls, and appends a new character each 965 # time: A AB ABC ABCD ... 966 ok 'AB' =~ /(\N{EVIL})/ && $1 eq 'A', 'Charname caching $1'; 967 like 'ABC', qr/(\N{EVIL})/, 'Charname caching $1'; 968 ok 'ABCD' =~ m'(\N{EVIL})' && $1 eq 'ABC', 'Charname caching $1'; 969 ok 'ABCDE' =~ m'(\N{EVIL})', 'Charname caching $1'; 970 like 'xy', qr/x\N{EMPTY-STR}y/, 971 'Empty string charname produces NOTHING node'; 972 ok 'xy' =~ 'x\N{EMPTY-STR}y', 973 'Empty string charname produces NOTHING node'; 974 like '', qr/\N{EMPTY-STR}/, 975 'Empty string charname produces NOTHING node'; 976 like "\N{LONG-STR}", qr/^\N{LONG-STR}$/, 'Verify that long string works'; 977 like "\N{LONG-STR}", qr/^\N{LONG-STR}$/i, 'Verify under folding that long string works'; 978 979 # perlhacktips points out that these work on both ASCII and EBCDIC 980 like "\xfc", qr/\N{EMPTY-STR}\xdc/i, 'Empty \N{} should change /d to /u'; 981 like "\xfc", qr'\N{EMPTY-STR}\xdc'i, 'Empty \N{} should change /d to /u'; 982 983 eval '/(?[[\N{EMPTY-STR}]])/'; 984 like $@, qr/Zero length \\N\{\}/, 'Verify zero-length return from \N{} correctly fails'; 985 ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/, 'Verify that long string works'; 986 ok "\N{LONG-STR}" =~ '^\N{LONG-STR}$', 'Verify that long string works'; 987 ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/i, 'Verify under folding that long string works'; 988 ok "\N{LONG-STR}" =~ m'^\N{LONG-STR}$'i, 'Verify under folding that long string works'; 989 990 undef $w; 991 { 992 () = eval q ["\N{TOO MANY SPACES}"]; 993 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "Multiple spaces in a row in a charnames alias is fatal"); 994 eval q [use utf8; () = "\N{TOO MANY SPACES}"]; 995 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "... same under utf8"); 996 } 997 998 undef $w; 999 { 1000 () = eval q ["\N{TRAILING SPACE }"]; 1001 like ($@, qr/charnames alias definitions may not contain trailing white-space/, "Trailing white-space in a charnames alias is fatal"); 1002 eval q [use utf8; () = "\N{TRAILING SPACE }"]; 1003 like ($@, qr/charnames alias definitions may not contain trailing white-space/, "... same under utf8"); 1004 } 1005 1006 undef $w; 1007 my $Cedilla_Latin1 = "GAR" 1008 . uni_to_native("\xC7") 1009 . "ON"; 1010 my $Cedilla_utf8 = $Cedilla_Latin1; 1011 utf8::upgrade($Cedilla_utf8); 1012 eval qq[is("\\N{$Cedilla_Latin1}", "$Cedilla_Latin1", "A cedilla in character name works")]; 1013 undef $w; 1014 { 1015 use feature 'unicode_eval'; 1016 eval qq[use utf8; is("\\N{$Cedilla_utf8}", "$Cedilla_utf8", "... same under 'use utf8': they work")]; 1017 } 1018 1019 undef $w; 1020 my $NBSP_Latin1 = "NBSP" 1021 . uni_to_native("\xA0") 1022 . "SEPARATED" 1023 . uni_to_native("\xA0") 1024 . "SPACE"; 1025 my $NBSP_utf8 = $NBSP_Latin1; 1026 utf8::upgrade($NBSP_utf8); 1027 () = eval qq[is("\\N{$NBSP_Latin1}", "$NBSP_Latin1"]; 1028 like ($@, qr/Invalid character in \\N\{...}/, "A NO-BREAK SPACE in a charnames alias is fatal"); 1029 undef $w; 1030 { 1031 use feature 'unicode_eval'; 1032 eval qq[use utf8; is("\\N{$NBSP_utf8}"]; 1033 like ($@, qr/Invalid character in \\N\{...}/, "A NO-BREAK SPACE in a charnames alias is fatal"); 1034 } 1035 1036 { 1037 BEGIN { no strict; *CnameTest:: = *{"_charnames\0A::" } } 1038 package CnameTest { sub translator { pop } } 1039 BEGIN { $^H{charnames} = \&CnameTest::translator } 1040 undef $w; 1041 () = eval q ["\N{TOO MANY SPACES}"]; 1042 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, 1043 'translators in _charnames\0* packages get validated'); 1044 } 1045 1046 # If remove the limitation in regcomp code these should work 1047 # differently 1048 undef $w; 1049 eval q [like "\N{TOO-LONG-STR}" =~ /^\N{TOO-LONG-STR}$/, 'Verify that what once was too long a string works']; 1050 eval 'q() =~ /\N{4F}/'; 1051 ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in name gives error'; 1052 eval 'q() =~ /\N{COM,MA}/'; 1053 ok $@ && $@ =~ /Invalid character/, 'Verify that comma in name gives error'; 1054 $name = "A" . uni_to_native("\x{D7}") . "O"; 1055 eval "q(W) =~ /\\N{$name}/"; 1056 ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in name gives error'; 1057 my $utf8_name = "7 CITIES OF GOLD"; 1058 utf8::upgrade($utf8_name); 1059 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1060 ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in utf8 name gives error'; 1061 $utf8_name = "SHARP #"; 1062 utf8::upgrade($utf8_name); 1063 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1064 ok $@ && $@ =~ /Invalid character/, 'Verify that ASCII symbol in utf8 name gives error'; 1065 $utf8_name = "A HOUSE " . uni_to_native("\xF7") . " AGAINST ITSELF"; 1066 utf8::upgrade($utf8_name); 1067 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1068 ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in utf8 name gives error'; 1069 $utf8_name = "\x{664} HORSEMEN}"; 1070 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1071 ok $@ && $@ =~ /Invalid character/, 'Verify that leading above Latin1 digit in utf8 name gives error'; 1072 $utf8_name = "A \x{1F4A9} WOULD SMELL AS SWEET}"; 1073 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1074 ok $@ && $@ =~ /Invalid character/, 'Verify that above Latin1 symbol in utf8 name gives error'; 1075 1076 undef $w; 1077 $name = "A" . uni_to_native("\x{D1}") . "O"; 1078 eval "q(W) =~ /\\N{$name}/"; 1079 ok ! $w, 'Verify that latin1 letter in name doesnt give warning'; 1080 1081 # This tests the code path that restarts the parse when the recursive 1082 # call to S_reg() from within S_grok_bslash_N() discovers that the 1083 # pattern needs to be recalculated as UTF-8. use eval to avoid 1084 # needing literal Unicode in this source file: 1085 my $r = eval "qr/\\N{\x{100}\x{100}}/"; 1086 isnt $r, undef, "Generated regex for multi-char UTF-8 charname" 1087 or diag($@); 1088 like "\x{100}\x{100}", $r, "which matches"; 1089 } 1090 1091 { 1092 use charnames ':full'; 1093 1094 unlike 'aabc', qr/a\N{PLUS SIGN}b/, '/a\N{PLUS SIGN}b/ against aabc'; 1095 like 'a+bc', qr/a\N{PLUS SIGN}b/, '/a\N{PLUS SIGN}b/ against a+bc'; 1096 1097 like ' A B', qr/\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/, 1098 'Intermixed named and unicode escapes'; 1099 like "\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}", 1100 qr/\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/, 1101 'Intermixed named and unicode escapes'; 1102 like "\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}", 1103 qr/[\N{SPACE}\N{U+0041}][\N{SPACE}\N{U+0042}]/, 1104 'Intermixed named and unicode escapes'; 1105 like "\0", qr/^\N{NULL}$/, 'Verify that \N{NULL} works; is not confused with an error'; 1106 } 1107 1108 { 1109 our $brackets; 1110 $brackets = qr{ 1111 { (?> [^{}]+ | (??{ $brackets }) )* } 1112 }x; 1113 1114 unlike "{b{c}d", qr/^((??{ $brackets }))/, "Bracket mismatch"; 1115 1116 SKIP: { 1117 our @stack = (); 1118 my @expect = qw( 1119 stuff1 1120 stuff2 1121 <stuff1>and<stuff2> 1122 right 1123 <right> 1124 <<right>> 1125 <<<right>>> 1126 <<stuff1>and<stuff2>><<<<right>>>> 1127 ); 1128 1129 local $_ = '<<<stuff1>and<stuff2>><<<<right>>>>>'; 1130 ok /^(<((?:(?>[^<>]+)|(?1))*)>(?{push @stack, $2 }))$/, 1131 "Recursion matches"; 1132 is(@stack, @expect, "Right amount of matches") 1133 or skip "Won't test individual results as count isn't equal", 1134 0 + @expect; 1135 my $idx = 0; 1136 foreach my $expect (@expect) { 1137 is($stack [$idx], $expect, 1138 "Expecting '$expect' at stack pos #$idx"); 1139 $idx ++; 1140 } 1141 } 1142 } 1143 1144 { 1145 my $s = '123453456'; 1146 $s =~ s/(?<digits>\d+)\k<digits>/$+{digits}/; 1147 ok $s eq '123456', 'Named capture (angle brackets) s///'; 1148 $s = '123453456'; 1149 $s =~ s/(?'digits'\d+)\k'digits'/$+{digits}/; 1150 ok $s eq '123456', 'Named capture (single quotes) s///'; 1151 } 1152 1153 { 1154 my @ary = ( 1155 pack('U', 0x00F1), # n-tilde 1156 '_'.pack('U', 0x00F1), # _ + n-tilde 1157 'c'.pack('U', 0x0327), # c + cedilla 1158 pack('U*', 0x00F1, 0x0327),# n-tilde + cedilla 1159 pack('U', 0x0391), # ALPHA 1160 pack('U', 0x0391).'2', # ALPHA + 2 1161 pack('U', 0x0391).'_', # ALPHA + _ 1162 ); 1163 1164 for my $uni (@ary) { 1165 my ($r1, $c1, $r2, $c2) = eval qq { 1166 use utf8; 1167 scalar ("..foo foo.." =~ /(?'${uni}'foo) \\k'${uni}'/), 1168 \$+{${uni}}, 1169 scalar ("..bar bar.." =~ /(?<${uni}>bar) \\k<${uni}>/), 1170 \$+{${uni}}; 1171 }; 1172 ok $r1, "Named capture UTF (?'')"; 1173 ok defined $c1 && $c1 eq 'foo', "Named capture UTF \%+"; 1174 ok $r2, "Named capture UTF (?<>)"; 1175 ok defined $c2 && $c2 eq 'bar', "Named capture UTF \%+"; 1176 } 1177 } 1178 1179 { 1180 my $s = 'foo bar baz'; 1181 my @res; 1182 if ('1234' =~ /(?<A>1)(?<B>2)(?<A>3)(?<B>4)/) { 1183 foreach my $name (sort keys(%-)) { 1184 my $ary = $- {$name}; 1185 foreach my $idx (0 .. $#$ary) { 1186 push @res, "$name:$idx:$ary->[$idx]"; 1187 } 1188 } 1189 } 1190 my @expect = qw (A:0:1 A:1:3 B:0:2 B:1:4); 1191 is("@res", "@expect", "Check %-"); 1192 eval' 1193 no warnings "uninitialized"; 1194 print for $- {this_key_doesnt_exist}; 1195 '; 1196 ok !$@,'lvalue $- {...} should not throw an exception'; 1197 } 1198 1199 { 1200 # \c\ followed by _ 1201 unlike "x\c_y", qr/x\c\_y/, '\_ in a pattern'; 1202 like "x\c\_y", qr/x\c\_y/, '\_ in a pattern'; 1203 1204 # \c\ followed by other characters 1205 for my $c ("z", "\0", "!", chr(254), chr(256)) { 1206 my $targ = "a" . uni_to_native("\034") . "$c"; 1207 my $reg = "a\\c\\$c"; 1208 ok eval ("qq/$targ/ =~ /$reg/"), "\\c\\ in pattern"; 1209 } 1210 } 1211 1212 { # Test the (*PRUNE) pattern 1213 our $count = 0; 1214 'aaab' =~ /a+b?(?{$count++})(*FAIL)/; 1215 is($count, 9, "Expect 9 for no (*PRUNE)"); 1216 $count = 0; 1217 'aaab' =~ /a+b?(*PRUNE)(?{$count++})(*FAIL)/; 1218 is($count, 3, "Expect 3 with (*PRUNE)"); 1219 local $_ = 'aaab'; 1220 $count = 0; 1221 1 while /.(*PRUNE)(?{$count++})(*FAIL)/g; 1222 is($count, 4, "/.(*PRUNE)/"); 1223 $count = 0; 1224 'aaab' =~ /a+b?(??{'(*PRUNE)'})(?{$count++})(*FAIL)/; 1225 is($count, 3, "Expect 3 with (*PRUNE)"); 1226 local $_ = 'aaab'; 1227 $count = 0; 1228 1 while /.(??{'(*PRUNE)'})(?{$count++})(*FAIL)/g; 1229 is($count, 4, "/.(*PRUNE)/"); 1230 } 1231 1232 { # Test the (*SKIP) pattern 1233 our $count = 0; 1234 'aaab' =~ /a+b?(*SKIP)(?{$count++})(*FAIL)/; 1235 is($count, 1, "Expect 1 with (*SKIP)"); 1236 local $_ = 'aaab'; 1237 $count = 0; 1238 1 while /.(*SKIP)(?{$count++})(*FAIL)/g; 1239 is($count, 4, "/.(*SKIP)/"); 1240 $_ = 'aaabaaab'; 1241 $count = 0; 1242 our @res = (); 1243 1 while /(a+b?)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g; 1244 is($count, 2, "Expect 2 with (*SKIP)"); 1245 is("@res", "aaab aaab", "Adjacent (*SKIP) works as expected"); 1246 } 1247 1248 { # Test the (*SKIP) pattern 1249 our $count = 0; 1250 'aaab' =~ /a+b?(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/; 1251 is($count, 1, "Expect 1 with (*SKIP)"); 1252 local $_ = 'aaab'; 1253 $count = 0; 1254 1 while /.(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/g; 1255 is($count, 4, "/.(*SKIP)/"); 1256 $_ = 'aaabaaab'; 1257 $count = 0; 1258 our @res = (); 1259 1 while /(a+b?)(*MARK:foo)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g; 1260 is($count, 2, "Expect 2 with (*SKIP)"); 1261 is("@res", "aaab aaab", "Adjacent (*SKIP) works as expected"); 1262 } 1263 1264 { # Test the (*SKIP) pattern 1265 our $count = 0; 1266 'aaab' =~ /a*(*MARK:a)b?(*MARK:b)(*SKIP:a)(?{$count++})(*FAIL)/; 1267 is($count, 3, "Expect 3 with *MARK:a)b?(*MARK:b)(*SKIP:a)"); 1268 local $_ = 'aaabaaab'; 1269 $count = 0; 1270 our @res = (); 1271 1 while 1272 /(a*(*MARK:a)b?)(*MARK:x)(*SKIP:a)(?{$count++; push @res,$1})(*FAIL)/g; 1273 is($count, 5, "Expect 5 with (*MARK:a)b?)(*MARK:x)(*SKIP:a)"); 1274 is("@res", "aaab b aaab b ", 1275 "Adjacent (*MARK:a)b?)(*MARK:x)(*SKIP:a) works as expected"); 1276 } 1277 1278 { # Test the (*COMMIT) pattern 1279 our $count = 0; 1280 'aaabaaab' =~ /a+b?(*COMMIT)(?{$count++})(*FAIL)/; 1281 is($count, 1, "Expect 1 with (*COMMIT)"); 1282 local $_ = 'aaab'; 1283 $count = 0; 1284 1 while /.(*COMMIT)(?{$count++})(*FAIL)/g; 1285 is($count, 1, "/.(*COMMIT)/"); 1286 $_ = 'aaabaaab'; 1287 $count = 0; 1288 our @res = (); 1289 1 while /(a+b?)(*COMMIT)(?{$count++; push @res,$1})(*FAIL)/g; 1290 is($count, 1, "Expect 1 with (*COMMIT)"); 1291 is("@res", "aaab", "Adjacent (*COMMIT) works as expected"); 1292 1293 unlike("1\n2a\n", qr/^\d+(*COMMIT)\w+/m, "COMMIT and anchors"); 1294 } 1295 1296 { 1297 # Test named commits and the $REGERROR var 1298 local $REGERROR; 1299 for my $name ('', ':foo') { 1300 for my $pat ("(*PRUNE$name)", 1301 ($name ? "(*MARK$name)" : "") . "(*SKIP$name)", 1302 "(*COMMIT$name)") { 1303 for my $suffix ('(*FAIL)', '') { 1304 'aaaab' =~ /a+b$pat$suffix/; 1305 is($REGERROR, 1306 ($suffix ? ($name ? 'foo' : "1") : ""), 1307 "Test $pat and \$REGERROR $suffix"); 1308 } 1309 } 1310 } 1311 } 1312 1313 { 1314 # Test named commits and the $REGERROR var 1315 package Fnorble; 1316 our $REGERROR; 1317 local $REGERROR; 1318 for my $name ('', ':foo') { 1319 for my $pat ("(*PRUNE$name)", 1320 ($name ? "(*MARK$name)" : "") . "(*SKIP$name)", 1321 "(*COMMIT$name)") { 1322 for my $suffix ('(*FAIL)','') { 1323 'aaaab' =~ /a+b$pat$suffix/; 1324 ::is($REGERROR, 1325 ($suffix ? ($name ? 'foo' : "1") : ""), 1326 "Test $pat and \$REGERROR $suffix"); 1327 } 1328 } 1329 } 1330 } 1331 1332 { 1333 # Test named commits and the $REGERROR var 1334 my $message = '$REGERROR'; 1335 local $REGERROR; 1336 for my $word (qw (bar baz bop)) { 1337 $REGERROR = ""; 1338 "aaaaa$word" =~ 1339 /a+(?:bar(*COMMIT:bar)|baz(*COMMIT:baz)|bop(*COMMIT:bop))(*FAIL)/; 1340 is($REGERROR, $word, $message); 1341 } 1342 } 1343 1344 { 1345 #Mindnumbingly simple test of (*THEN) 1346 for ("ABC","BAX") { 1347 ok /A (*THEN) X | B (*THEN) C/x, "Simple (*THEN) test"; 1348 } 1349 } 1350 1351 { 1352 my $message = "Relative Recursion"; 1353 my $parens = qr/(\((?:[^()]++|(?-1))*+\))/; 1354 local $_ = 'foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))'; 1355 my ($all, $one, $two) = ('', '', ''); 1356 ok(m/foo $parens \s* \+ \s* bar $parens/x, $message); 1357 is($1, '((2*3)+4-3)', $message); 1358 is($2, '(2*(3+4)-1*(2-3))', $message); 1359 is($&, 'foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))', $message); 1360 is($&, $_, $message); 1361 } 1362 1363 { 1364 my $spaces=" "; 1365 local $_ = join 'bar', $spaces, $spaces; 1366 our $count = 0; 1367 s/(?>\s+bar)(?{$count++})//g; 1368 is($_, $spaces, "SUSPEND final string"); 1369 is($count, 1, "Optimiser should have prevented more than one match"); 1370 } 1371 1372 { 1373 # From Message-ID: <877ixs6oa6.fsf@k75.linux.bogus> 1374 my $dow_name = "nada"; 1375 my $parser = "(\$dow_name) = \$time_string =~ /(D\x{e9}\\ " . 1376 "C\x{e9}adaoin|D\x{e9}\\ Sathairn|\\w+|\x{100})/"; 1377 my $time_string = "D\x{e9} C\x{e9}adaoin"; 1378 eval $parser; 1379 ok !$@, "Test Eval worked"; 1380 is($dow_name, $time_string, "UTF-8 trie common prefix extraction"); 1381 } 1382 1383 { 1384 my $v; 1385 ($v = 'bar') =~ /(\w+)/g; 1386 $v = 'foo'; 1387 is("$1", 'bar', 1388 '$1 is safe after /g - may fail due to specialized config in pp_hot.c'); 1389 } 1390 1391 { 1392 my $message = "http://nntp.perl.org/group/perl.perl5.porters/118663"; 1393 my $qr_barR1 = qr/(bar)\g-1/; 1394 like("foobarbarxyz", $qr_barR1, $message); 1395 like("foobarbarxyz", qr/foo${qr_barR1}xyz/, $message); 1396 like("foobarbarxyz", qr/(foo)${qr_barR1}xyz/, $message); 1397 like("foobarbarxyz", qr/(foo)(bar)\g{-1}xyz/, $message); 1398 like("foobarbarxyz", qr/(foo${qr_barR1})xyz/, $message); 1399 like("foobarbarxyz", qr/(foo(bar)\g{-1})xyz/, $message); 1400 } 1401 1402 { 1403 my $message = '$REGMARK'; 1404 our @r = (); 1405 local $REGMARK; 1406 local $REGERROR; 1407 like('foofoo', qr/foo (*MARK:foo) (?{push @r,$REGMARK}) /x, $message); 1408 is("@r","foo", $message); 1409 is($REGMARK, "foo", $message); 1410 unlike('foofoo', qr/foo (*MARK:foo) (*FAIL) /x, $message); 1411 is($REGMARK, '', $message); 1412 is($REGERROR, 'foo', $message); 1413 } 1414 1415 { 1416 my $message = '\K test'; 1417 my $x; 1418 $x = "abc.def.ghi.jkl"; 1419 $x =~ s/.*\K\..*//; 1420 is($x, "abc.def.ghi", $message); 1421 1422 $x = "one two three four"; 1423 $x =~ s/o+ \Kthree//g; 1424 is($x, "one two four", $message); 1425 1426 $x = "abcde"; 1427 $x =~ s/(.)\K/$1/g; 1428 is($x, "aabbccddee", $message); 1429 } 1430 1431 { 1432 sub kt { 1433 return '4' if $_[0] eq '09028623'; 1434 } 1435 # Nested EVAL using PL_curpm (via $1 or friends) 1436 my $re; 1437 our $grabit = qr/ ([0-6][0-9]{7}) (??{ kt $1 }) [890] /x; 1438 $re = qr/^ ( (??{ $grabit }) ) $ /x; 1439 my @res = '0902862349' =~ $re; 1440 is(join ("-", @res), "0902862349", 1441 'PL_curpm is set properly on nested eval'); 1442 1443 our $qr = qr/ (o) (??{ $1 }) /x; 1444 ok 'boob'=~/( b (??{ $qr }) b )/x && 1, "PL_curpm, nested eval"; 1445 } 1446 1447 { 1448 use charnames ":full"; 1449 like "\N{ROMAN NUMERAL ONE}", qr/\p{Alphabetic}/, "I =~ Alphabetic"; 1450 like "\N{ROMAN NUMERAL ONE}", qr/\p{Uppercase}/, "I =~ Uppercase"; 1451 unlike "\N{ROMAN NUMERAL ONE}", qr/\p{Lowercase}/, "I !~ Lowercase"; 1452 like "\N{ROMAN NUMERAL ONE}", qr/\p{IDStart}/, "I =~ ID_Start"; 1453 like "\N{ROMAN NUMERAL ONE}", qr/\p{IDContinue}/, "I =~ ID_Continue"; 1454 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Alphabetic}/, "i =~ Alphabetic"; 1455 unlike "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Uppercase}/, "i !~ Uppercase"; 1456 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Uppercase}/i, "i =~ Uppercase under /i"; 1457 unlike "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Titlecase}/, "i !~ Titlecase"; 1458 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Titlecase}/i, "i =~ Titlecase under /i"; 1459 like "\N{ROMAN NUMERAL ONE}", qr/\p{Lowercase}/i, "I =~ Lowercase under /i"; 1460 1461 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{Lowercase}/, "i =~ Lowercase"; 1462 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{IDStart}/, "i =~ ID_Start"; 1463 like "\N{SMALL ROMAN NUMERAL ONE}", qr/\p{IDContinue}/, "i =~ ID_Continue" 1464 } 1465 1466 { # More checking that /i works on the few properties that it makes a 1467 # difference. Uppercase, Lowercase, and Titlecase were done in the 1468 # block above 1469 like "A", qr/\p{PosixUpper}/, "A =~ PosixUpper"; 1470 like "A", qr/\p{PosixUpper}/i, "A =~ PosixUpper under /i"; 1471 unlike "A", qr/\p{PosixLower}/, "A !~ PosixLower"; 1472 like "A", qr/\p{PosixLower}/i, "A =~ PosixLower under /i"; 1473 unlike "a", qr/\p{PosixUpper}/, "a !~ PosixUpper"; 1474 like "a", qr/\p{PosixUpper}/i, "a =~ PosixUpper under /i"; 1475 like "a", qr/\p{PosixLower}/, "a =~ PosixLower"; 1476 like "a", qr/\p{PosixLower}/i, "a =~ PosixLower under /i"; 1477 1478 like uni_to_native("\xC0"), qr/\p{XPosixUpper}/, "\\xC0 =~ XPosixUpper"; 1479 like uni_to_native("\xC0"), qr/\p{XPosixUpper}/i, "\\xC0 =~ XPosixUpper under /i"; 1480 unlike uni_to_native("\xC0"), qr/\p{XPosixLower}/, "\\xC0 !~ XPosixLower"; 1481 like uni_to_native("\xC0"), qr/\p{XPosixLower}/i, "\\xC0 =~ XPosixLower under /i"; 1482 unlike uni_to_native("\xE0"), qr/\p{XPosixUpper}/, "\\xE0 !~ XPosixUpper"; 1483 like uni_to_native("\xE0"), qr/\p{XPosixUpper}/i, "\\xE0 =~ XPosixUpper under /i"; 1484 like uni_to_native("\xE0"), qr/\p{XPosixLower}/, "\\xE0 =~ XPosixLower"; 1485 like uni_to_native("\xE0"), qr/\p{XPosixLower}/i, "\\xE0 =~ XPosixLower under /i"; 1486 1487 like uni_to_native("\xC0"), qr/\p{UppercaseLetter}/, "\\xC0 =~ UppercaseLetter"; 1488 like uni_to_native("\xC0"), qr/\p{UppercaseLetter}/i, "\\xC0 =~ UppercaseLetter under /i"; 1489 unlike uni_to_native("\xC0"), qr/\p{LowercaseLetter}/, "\\xC0 !~ LowercaseLetter"; 1490 like uni_to_native("\xC0"), qr/\p{LowercaseLetter}/i, "\\xC0 =~ LowercaseLetter under /i"; 1491 unlike uni_to_native("\xC0"), qr/\p{TitlecaseLetter}/, "\\xC0 !~ TitlecaseLetter"; 1492 like uni_to_native("\xC0"), qr/\p{TitlecaseLetter}/i, "\\xC0 =~ TitlecaseLetter under /i"; 1493 unlike uni_to_native("\xE0"), qr/\p{UppercaseLetter}/, "\\xE0 !~ UppercaseLetter"; 1494 like uni_to_native("\xE0"), qr/\p{UppercaseLetter}/i, "\\xE0 =~ UppercaseLetter under /i"; 1495 like uni_to_native("\xE0"), qr/\p{LowercaseLetter}/, "\\xE0 =~ LowercaseLetter"; 1496 like uni_to_native("\xE0"), qr/\p{LowercaseLetter}/i, "\\xE0 =~ LowercaseLetter under /i"; 1497 unlike uni_to_native("\xE0"), qr/\p{TitlecaseLetter}/, "\\xE0 !~ TitlecaseLetter"; 1498 like uni_to_native("\xE0"), qr/\p{TitlecaseLetter}/i, "\\xE0 =~ TitlecaseLetter under /i"; 1499 unlike "\x{1C5}", qr/\p{UppercaseLetter}/, "\\x{1C5} !~ UppercaseLetter"; 1500 like "\x{1C5}", qr/\p{UppercaseLetter}/i, "\\x{1C5} =~ UppercaseLetter under /i"; 1501 unlike "\x{1C5}", qr/\p{LowercaseLetter}/, "\\x{1C5} !~ LowercaseLetter"; 1502 like "\x{1C5}", qr/\p{LowercaseLetter}/i, "\\x{1C5} =~ LowercaseLetter under /i"; 1503 like "\x{1C5}", qr/\p{TitlecaseLetter}/, "\\x{1C5} =~ TitlecaseLetter"; 1504 like "\x{1C5}", qr/\p{TitlecaseLetter}/i, "\\x{1C5} =~ TitlecaseLetter under /i"; 1505 } 1506 1507 { 1508 # requirement of Unicode Technical Standard #18, 1.7 Code Points 1509 # cf. http://www.unicode.org/reports/tr18/#Supplementary_Characters 1510 for my $u (0x7FF, 0x800, 0xFFFF, 0x10000) { 1511 no warnings 'utf8'; # oops 1512 my $c = chr $u; 1513 my $x = sprintf '%04X', $u; 1514 like "A${c}B", qr/A[\0-\x{10000}]B/, "Unicode range - $x"; 1515 } 1516 } 1517 1518 { 1519 my $res=""; 1520 1521 if ('1' =~ /(?|(?<digit>1)|(?<digit>2))/) { 1522 $res = "@{$- {digit}}"; 1523 } 1524 is($res, "1", 1525 "Check that (?|...) doesnt cause dupe entries in the names array"); 1526 1527 $res = ""; 1528 if ('11' =~ /(?|(?<digit>1)|(?<digit>2))(?&digit)/) { 1529 $res = "@{$- {digit}}"; 1530 } 1531 is($res, "1", 1532 "Check that (?&..) to a buffer inside a (?|...) goes to the leftmost"); 1533 } 1534 1535 { 1536 use warnings; 1537 my $message = "ASCII pattern that really is UTF-8"; 1538 my @w; 1539 local $SIG {__WARN__} = sub {push @w, "@_"}; 1540 my $c = qq (\x{DF}); 1541 like($c, qr/${c}|\x{100}/, $message); 1542 is("@w", '', $message); 1543 } 1544 1545 { 1546 my $message = "Corruption of match results of qr// across scopes"; 1547 my $qr = qr/(fo+)(ba+r)/; 1548 'foobar' =~ /$qr/; 1549 is("$1$2", "foobar", $message); 1550 { 1551 'foooooobaaaaar' =~ /$qr/; 1552 is("$1$2", 'foooooobaaaaar', $message); 1553 } 1554 is("$1$2", "foobar", $message); 1555 } 1556 1557 { 1558 my $message = "HORIZWS"; 1559 local $_ = "\t \r\n \n \t".chr(11)."\n"; 1560 s/\H/H/g; 1561 s/\h/h/g; 1562 is($_, "hhHHhHhhHH", $message); 1563 $_ = "\t \r\n \n \t" . chr (11) . "\n"; 1564 utf8::upgrade ($_); 1565 s/\H/H/g; 1566 s/\h/h/g; 1567 is($_, "hhHHhHhhHH", $message); 1568 } 1569 1570 { 1571 # Various whitespace special patterns 1572 my @h = map {chr utf8::unicode_to_native($_) } 1573 0x09, 0x20, 0xa0, 0x1680, 0x2000, 1574 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, 0x2006, 1575 0x2007, 0x2008, 0x2009, 0x200a, 0x202f, 0x205f, 1576 0x3000; 1577 my @v = map {chr utf8::unicode_to_native($_) } 1578 0x0a, 0x0b, 0x0c, 0x0d, 0x85, 0x2028, 1579 0x2029; 1580 my @lb = (uni_to_native("\x0D\x0A"), 1581 map {chr utf8::unicode_to_native($_) } 1582 0x0A .. 0x0D, 0x85, 0x2028, 0x2029); 1583 foreach my $t ([\@h, qr/\h/, qr/\h+/], 1584 [\@v, qr/\v/, qr/\v+/], 1585 [\@lb, qr/\R/, qr/\R+/],) { 1586 my $ary = shift @$t; 1587 foreach my $pat (@$t) { 1588 foreach my $str (@$ary) { 1589 my $temp_str = $str; 1590 $temp_str = display($temp_str); 1591 ok $str =~ /($pat)/, $temp_str . " =~ /($pat)"; 1592 my $temp_1 = $1; 1593 is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "' after ($pat)"); 1594 utf8::upgrade ($str); 1595 ok $str =~ /($pat)/, "Upgraded " . $temp_str . " =~ /($pat)/"; 1596 is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "'(upgraded) after ($pat)"); 1597 } 1598 } 1599 } 1600 } 1601 1602 { 1603 # Check that \\xDF match properly in its various forms 1604 # Test that \xDF matches properly. this is pretty hacky stuff, 1605 # but its actually needed. The malarky with '-' is to prevent 1606 # compilation caching from playing any role in the test. 1607 my @df = (chr utf8::unicode_to_native(0xDF), '-', chr utf8::unicode_to_native(0xDF)); 1608 utf8::upgrade ($df [2]); 1609 my @strs = ('ss', 'sS', 'Ss', 'SS', chr utf8::unicode_to_native(0xDF)); 1610 my @ss = map {("$_", "$_")} @strs; 1611 utf8::upgrade ($ss [$_ * 2 + 1]) for 0 .. $#strs; 1612 1613 for my $ssi (0 .. $#ss) { 1614 for my $dfi (0 .. $#df) { 1615 my $pat = $df [$dfi]; 1616 my $str = $ss [$ssi]; 1617 my $utf_df = ($dfi > 1) ? 'utf8' : ''; 1618 my $utf_ss = ($ssi % 2) ? 'utf8' : ''; 1619 my $sstr; # We hard-code the ebcdic value below to avoid 1620 # perturbing the test 1621 ($sstr = $str) =~ s/\xDF/\\xDF/ if $::IS_ASCII; 1622 ($sstr = $str) =~ s/\x59/\\x59/ if $::IS_EBCDIC; 1623 1624 if ($utf_df || $utf_ss || length ($ss [$ssi]) == 1) { 1625 my $ret = $str =~ /$pat/i; 1626 next if $pat eq '-'; 1627 if ($::IS_ASCII) { 1628 ok $ret, "\"$sstr\" =~ /\\xDF/i " . 1629 "(str is @{[$utf_ss||'latin']}, pat is " . 1630 "@{[$utf_df||'latin']})"; 1631 } 1632 else { 1633 ok $ret, "\"$sstr\" =~ /\\x59/i " . 1634 "(str is @{[$utf_ss||'latin']}, pat is " . 1635 "@{[$utf_df||'latin']})"; 1636 } 1637 } 1638 else { 1639 my $ret = $str !~ /$pat/i; 1640 next if $pat eq '-'; 1641 if ($::IS_EBCDIC) { 1642 ok $ret, "\"$sstr\" !~ /\\x59/i " . 1643 "(str is @{[$utf_ss||'latin']}, pat is " . 1644 "@{[$utf_df||'latin']})"; 1645 } 1646 else { 1647 ok $ret, "\"$sstr\" !~ /\\xDF/i " . 1648 "(str is @{[$utf_ss||'latin']}, pat is " . 1649 "@{[$utf_df||'latin']})"; 1650 } 1651 } 1652 } 1653 } 1654 } 1655 1656 { 1657 my $message = "BBC(Bleadperl Breaks CPAN) Today: String::Multibyte"; 1658 my $re = qr/(?:[\x00-\xFF]{4})/; 1659 my $hyp = "\0\0\0-"; 1660 my $esc = "\0\0\0\\"; 1661 1662 my $str = "$esc$hyp$hyp$esc$esc"; 1663 my @a = ($str =~ /\G(?:\Q$esc$esc\E|\Q$esc$hyp\E|$re)/g); 1664 1665 is(@a,3, $message); 1666 local $" = "="; 1667 is("@a","$esc$hyp=$hyp=$esc$esc", $message); 1668 } 1669 1670 { 1671 # Test for keys in %+ and %- 1672 my $message = 'Test keys in %+ and %-'; 1673 no warnings 'uninitialized'; 1674 local $_ = "abcdef"; 1675 /(?<foo>a)|(?<foo>b)/; 1676 is((join ",", sort keys %+), "foo", $message); 1677 is((join ",", sort keys %-), "foo", $message); 1678 is((join ",", sort values %+), "a", $message); 1679 is((join ",", sort map "@$_", values %-), "a ", $message); 1680 /(?<bar>a)(?<bar>b)(?<quux>.)/; 1681 is((join ",", sort keys %+), "bar,quux", $message); 1682 is((join ",", sort keys %-), "bar,quux", $message); 1683 is((join ",", sort values %+), "a,c", $message); # leftmost 1684 is((join ",", sort map "@$_", values %-), "a b,c", $message); 1685 /(?<un>a)(?<deux>c)?/; # second buffer won't capture 1686 is((join ",", sort keys %+), "un", $message); 1687 is((join ",", sort keys %-), "deux,un", $message); 1688 is((join ",", sort values %+), "a", $message); 1689 is((join ",", sort map "@$_", values %-), ",a", $message); 1690 } 1691 1692 { 1693 # length() on captures, the numbered ones end up in Perl_magic_len 1694 local $_ = "aoeu " . uni_to_native("\xe6") . "var ook"; 1695 /^ \w+ \s (?<eek>\S+)/x; 1696 1697 is(length $`, 0, q[length $`]); 1698 is(length $', 4, q[length $']); 1699 is(length $&, 9, q[length $&]); 1700 is(length $1, 4, q[length $1]); 1701 is(length $+{eek}, 4, q[length $+{eek} == length $1]); 1702 } 1703 1704 { 1705 my $ok = -1; 1706 1707 $ok = exists ($-{x}) ? 1 : 0 if 'bar' =~ /(?<x>foo)|bar/; 1708 is($ok, 1, '$-{x} exists after "bar"=~/(?<x>foo)|bar/'); 1709 is(scalar (%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/'); 1710 is(scalar (%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/'); 1711 1712 $ok = -1; 1713 $ok = exists ($+{x}) ? 1 : 0 if 'bar' =~ /(?<x>foo)|bar/; 1714 is($ok, 0, '$+{x} not exists after "bar"=~/(?<x>foo)|bar/'); 1715 is(scalar (%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/'); 1716 is(scalar (%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/'); 1717 1718 $ok = -1; 1719 $ok = exists ($-{x}) ? 1 : 0 if 'foo' =~ /(?<x>foo)|bar/; 1720 is($ok, 1, '$-{x} exists after "foo"=~/(?<x>foo)|bar/'); 1721 is(scalar (%+), 1, 'scalar %+ == 1 after "foo"=~/(?<x>foo)|bar/'); 1722 is(scalar (%-), 1, 'scalar %- == 1 after "foo"=~/(?<x>foo)|bar/'); 1723 1724 $ok = -1; 1725 $ok = exists ($+{x}) ? 1 : 0 if 'foo'=~/(?<x>foo)|bar/; 1726 is($ok, 1, '$+{x} exists after "foo"=~/(?<x>foo)|bar/'); 1727 } 1728 1729 { 1730 local $_; 1731 ($_ = 'abc') =~ /(abc)/g; 1732 $_ = '123'; 1733 is("$1", 'abc', "/g leads to unsafe match vars: $1"); 1734 1735 fresh_perl_is(<<'EOP', ">abc<\n", {}, 'mention $&'); 1736$&; 1737my $x; 1738($x='abc')=~/(abc)/g; 1739$x='123'; 1740print ">$1<\n"; 1741EOP 1742 1743 fresh_perl_is(<<'EOP', ">abc<\n", {}, 'no mention of $&'); 1744my $x; 1745($x='abc')=~/(abc)/g; 1746$x='123'; 1747print ">$1<\n"; 1748EOP 1749 } 1750 1751 { 1752 # Message-ID: <20070818091501.7eff4831@r2d2> 1753 my $str = ""; 1754 for (0 .. 5) { 1755 my @x; 1756 $str .= "@x"; # this should ALWAYS be the empty string 1757 'a' =~ /(a|)/; 1758 push @x, 1; 1759 } 1760 is(length $str, 0, "Trie scope error, string should be empty"); 1761 $str = ""; 1762 my @foo = ('a') x 5; 1763 for (@foo) { 1764 my @bar; 1765 $str .= "@bar"; 1766 s/a|/push @bar, 1/e; 1767 } 1768 is(length $str, 0, "Trie scope error, string should be empty"); 1769 } 1770 1771 { 1772# more TRIE/AHOCORASICK problems with mixed utf8 / latin-1 and case folding 1773 for my $ord (160 .. 255) { 1774 my $chr = utf8::unicode_to_native($ord); 1775 my $chr_byte = chr($chr); 1776 my $chr_utf8 = chr($chr); utf8::upgrade($chr_utf8); 1777 my $rx = qr{$chr_byte|X}i; 1778 like($chr_utf8, $rx, "utf8/latin, codepoint $chr"); 1779 } 1780 } 1781 1782 { 1783 our $a = 3; "" =~ /(??{ $a })/; 1784 our $b = $a; 1785 is($b, $a, "Copy of scalar used for postponed subexpression"); 1786 } 1787 1788 { 1789 our @ctl_n = (); 1790 our @plus = (); 1791 our $nested_tags; 1792 $nested_tags = qr{ 1793 < 1794 (\w+) 1795 (?{ 1796 push @ctl_n,$^N; 1797 push @plus,$+; 1798 }) 1799 > 1800 (??{$nested_tags})* 1801 </\s* \w+ \s*> 1802 }x; 1803 1804 my $match = '<bla><blubb></blubb></bla>' =~ m/^$nested_tags$/; 1805 ok $match, 'nested construct matches'; 1806 is("@ctl_n", "bla blubb", '$^N inside of (?{}) works as expected'); 1807 is("@plus", "bla blubb", '$+ inside of (?{}) works as expected'); 1808 } 1809 1810 SKIP: { 1811 # XXX: This set of tests is essentially broken, POSIX character classes 1812 # should not have differing definitions under Unicode. 1813 # There are property names for that. 1814 skip "Tests assume ASCII", 4 unless $::IS_ASCII; 1815 1816 my @notIsPunct = grep {/[[:punct:]]/ and not /\p{IsPunct}/} 1817 map {chr} 0x20 .. 0x7f; 1818 is(join ('', @notIsPunct), '$+<=>^`|~', 1819 '[:punct:] disagrees with IsPunct on Symbols'); 1820 1821 my @isPrint = grep {not /[[:print:]]/ and /\p{IsPrint}/} 1822 map {chr} 0 .. 0x1f, 0x7f .. 0x9f; 1823 is(join ('', @isPrint), "", 1824 'IsPrint agrees with [:print:] on control characters'); 1825 1826 my @isPunct = grep {/[[:punct:]]/ != /\p{IsPunct}/} 1827 map {chr} 0x80 .. 0xff; 1828 is(join ('', @isPunct), "\xa1\xa7\xab\xb6\xb7\xbb\xbf", # ¡ « · » ¿ 1829 'IsPunct disagrees with [:punct:] outside ASCII'); 1830 1831 my @isPunctLatin1 = eval q { 1832 grep {/[[:punct:]]/u != /\p{IsPunct}/} map {chr} 0x80 .. 0xff; 1833 }; 1834 skip "Eval failed ($@)", 1 if $@; 1835 skip "PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS set to 0", 1 1836 if !$ENV{PERL_TEST_LEGACY_POSIX_CC}; 1837 is(join ('', @isPunctLatin1), '', 1838 'IsPunct agrees with [:punct:] with explicit Latin1'); 1839 } 1840 1841 { 1842 # Tests for [#perl 71942] 1843 our $count_a; 1844 our $count_b; 1845 1846 my $c = 0; 1847 for my $re ( 1848# [ 1849# should match?, 1850# input string, 1851# re 1, 1852# re 2, 1853# expected values of count_a and count_b, 1854# ] 1855 [ 1856 0, 1857 "xababz", 1858 qr/a+(?{$count_a++})b?(*COMMIT)(*FAIL)/, 1859 qr/a+(?{$count_b++})b?(*COMMIT)z/, 1860 1, 1861 ], 1862 [ 1863 0, 1864 "xababz", 1865 qr/a+(?{$count_a++})b?(*COMMIT)\s*(*FAIL)/, 1866 qr/a+(?{$count_b++})b?(*COMMIT)\s*z/, 1867 1, 1868 ], 1869 [ 1870 0, 1871 "xababz", 1872 qr/a+(?{$count_a++})(?:b|)?(*COMMIT)(*FAIL)/, 1873 qr/a+(?{$count_b++})(?:b|)?(*COMMIT)z/, 1874 1, 1875 ], 1876 [ 1877 0, 1878 "xababz", 1879 qr/a+(?{$count_a++})b{0,6}(*COMMIT)(*FAIL)/, 1880 qr/a+(?{$count_b++})b{0,6}(*COMMIT)z/, 1881 1, 1882 ], 1883 [ 1884 0, 1885 "xabcabcz", 1886 qr/a+(?{$count_a++})(bc){0,6}(*COMMIT)(*FAIL)/, 1887 qr/a+(?{$count_b++})(bc){0,6}(*COMMIT)z/, 1888 1, 1889 ], 1890 [ 1891 0, 1892 "xabcabcz", 1893 qr/a+(?{$count_a++})(bc*){0,6}(*COMMIT)(*FAIL)/, 1894 qr/a+(?{$count_b++})(bc*){0,6}(*COMMIT)z/, 1895 1, 1896 ], 1897 1898 1899 [ 1900 0, 1901 "aaaabtz", 1902 qr/a+(?{$count_a++})b?(*PRUNE)(*FAIL)/, 1903 qr/a+(?{$count_b++})b?(*PRUNE)z/, 1904 4, 1905 ], 1906 [ 1907 0, 1908 "aaaabtz", 1909 qr/a+(?{$count_a++})b?(*PRUNE)\s*(*FAIL)/, 1910 qr/a+(?{$count_b++})b?(*PRUNE)\s*z/, 1911 4, 1912 ], 1913 [ 1914 0, 1915 "aaaabtz", 1916 qr/a+(?{$count_a++})(?:b|)(*PRUNE)(*FAIL)/, 1917 qr/a+(?{$count_b++})(?:b|)(*PRUNE)z/, 1918 4, 1919 ], 1920 [ 1921 0, 1922 "aaaabtz", 1923 qr/a+(?{$count_a++})b{0,6}(*PRUNE)(*FAIL)/, 1924 qr/a+(?{$count_b++})b{0,6}(*PRUNE)z/, 1925 4, 1926 ], 1927 [ 1928 0, 1929 "aaaabctz", 1930 qr/a+(?{$count_a++})(bc){0,6}(*PRUNE)(*FAIL)/, 1931 qr/a+(?{$count_b++})(bc){0,6}(*PRUNE)z/, 1932 4, 1933 ], 1934 [ 1935 0, 1936 "aaaabctz", 1937 qr/a+(?{$count_a++})(bc*){0,6}(*PRUNE)(*FAIL)/, 1938 qr/a+(?{$count_b++})(bc*){0,6}(*PRUNE)z/, 1939 4, 1940 ], 1941 1942 [ 1943 0, 1944 "aaabaaab", 1945 qr/a+(?{$count_a++;})b?(*SKIP)(*FAIL)/, 1946 qr/a+(?{$count_b++;})b?(*SKIP)z/, 1947 2, 1948 ], 1949 [ 1950 0, 1951 "aaabaaab", 1952 qr/a+(?{$count_a++;})b?(*SKIP)\s*(*FAIL)/, 1953 qr/a+(?{$count_b++;})b?(*SKIP)\s*z/, 1954 2, 1955 ], 1956 [ 1957 0, 1958 "aaabaaab", 1959 qr/a+(?{$count_a++;})(?:b|)(*SKIP)(*FAIL)/, 1960 qr/a+(?{$count_b++;})(?:b|)(*SKIP)z/, 1961 2, 1962 ], 1963 [ 1964 0, 1965 "aaabaaab", 1966 qr/a+(?{$count_a++;})b{0,6}(*SKIP)(*FAIL)/, 1967 qr/a+(?{$count_b++;})b{0,6}(*SKIP)z/, 1968 2, 1969 ], 1970 [ 1971 0, 1972 "aaabcaaabc", 1973 qr/a+(?{$count_a++;})(bc){0,6}(*SKIP)(*FAIL)/, 1974 qr/a+(?{$count_b++;})(bc){0,6}(*SKIP)z/, 1975 2, 1976 ], 1977 [ 1978 0, 1979 "aaabcaaabc", 1980 qr/a+(?{$count_a++;})(bc*){0,6}(*SKIP)(*FAIL)/, 1981 qr/a+(?{$count_b++;})(bc*){0,6}(*SKIP)z/, 1982 2, 1983 ], 1984 1985 1986 [ 1987 0, 1988 "aaddbdaabyzc", 1989 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) (*FAIL) \s* c \1 /x, 1990 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) z \s* c \1 /x, 1991 4, 1992 ], 1993 [ 1994 0, 1995 "aaddbdaabyzc", 1996 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) \s* (*FAIL) \s* c \1 /x, 1997 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) \s* z \s* c \1 /x, 1998 4, 1999 ], 2000 [ 2001 0, 2002 "aaddbdaabyzc", 2003 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (?:b|) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2004 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (?:b|) (*SKIP:T1) z \s* c \1 /x, 2005 4, 2006 ], 2007 [ 2008 0, 2009 "aaddbdaabyzc", 2010 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b{0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x, 2011 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b{0,6} (*SKIP:T1) z \s* c \1 /x, 2012 4, 2013 ], 2014 [ 2015 0, 2016 "aaddbcdaabcyzc", 2017 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (bc){0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x, 2018 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (bc){0,6} (*SKIP:T1) z \s* c \1 /x, 2019 4, 2020 ], 2021 [ 2022 0, 2023 "aaddbcdaabcyzc", 2024 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (bc*){0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x, 2025 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (bc*){0,6} (*SKIP:T1) z \s* c \1 /x, 2026 4, 2027 ], 2028 2029 2030 [ 2031 0, 2032 "aaaaddbdaabyzc", 2033 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2034 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2035 2, 2036 ], 2037 [ 2038 0, 2039 "aaaaddbdaabyzc", 2040 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) \s* (*FAIL) \s* c \1 /x, 2041 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) \s* z \s* c \1 /x, 2042 2, 2043 ], 2044 [ 2045 0, 2046 "aaaaddbdaabyzc", 2047 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (?:b|) (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2048 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (?:b|) (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2049 2, 2050 ], 2051 [ 2052 0, 2053 "aaaaddbdaabyzc", 2054 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b{0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2055 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b{0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2056 2, 2057 ], 2058 [ 2059 0, 2060 "aaaaddbcdaabcyzc", 2061 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (bc){0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2062 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (bc){0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2063 2, 2064 ], 2065 [ 2066 0, 2067 "aaaaddbcdaabcyzc", 2068 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (bc*){0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2069 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (bc*){0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2070 2, 2071 ], 2072 2073 2074 [ 2075 0, 2076 "AbcdCBefgBhiBqz", 2077 qr/(A (.*) (?{ $count_a++ }) C? (*THEN) | A D) (*FAIL)/x, 2078 qr/(A (.*) (?{ $count_b++ }) C? (*THEN) | A D) z/x, 2079 1, 2080 ], 2081 [ 2082 0, 2083 "AbcdCBefgBhiBqz", 2084 qr/(A (.*) (?{ $count_a++ }) C? (*THEN) | A D) \s* (*FAIL)/x, 2085 qr/(A (.*) (?{ $count_b++ }) C? (*THEN) | A D) \s* z/x, 2086 1, 2087 ], 2088 [ 2089 0, 2090 "AbcdCBefgBhiBqz", 2091 qr/(A (.*) (?{ $count_a++ }) (?:C|) (*THEN) | A D) (*FAIL)/x, 2092 qr/(A (.*) (?{ $count_b++ }) (?:C|) (*THEN) | A D) z/x, 2093 1, 2094 ], 2095 [ 2096 0, 2097 "AbcdCBefgBhiBqz", 2098 qr/(A (.*) (?{ $count_a++ }) C{0,6} (*THEN) | A D) (*FAIL)/x, 2099 qr/(A (.*) (?{ $count_b++ }) C{0,6} (*THEN) | A D) z/x, 2100 1, 2101 ], 2102 [ 2103 0, 2104 "AbcdCEBefgBhiBqz", 2105 qr/(A (.*) (?{ $count_a++ }) (CE){0,6} (*THEN) | A D) (*FAIL)/x, 2106 qr/(A (.*) (?{ $count_b++ }) (CE){0,6} (*THEN) | A D) z/x, 2107 1, 2108 ], 2109 [ 2110 0, 2111 "AbcdCBefgBhiBqz", 2112 qr/(A (.*) (?{ $count_a++ }) (CE*){0,6} (*THEN) | A D) (*FAIL)/x, 2113 qr/(A (.*) (?{ $count_b++ }) (CE*){0,6} (*THEN) | A D) z/x, 2114 1, 2115 ], 2116 ) { 2117 $c++; 2118 $count_a = 0; 2119 $count_b = 0; 2120 2121 my $match_a = ($re->[1] =~ $re->[2]) || 0; 2122 my $match_b = ($re->[1] =~ $re->[3]) || 0; 2123 2124 is($match_a, $re->[0], "match a " . ($re->[0] ? "succeeded" : "failed") . " ($c)"); 2125 is($match_b, $re->[0], "match b " . ($re->[0] ? "succeeded" : "failed") . " ($c)"); 2126 is($count_a, $re->[4], "count a ($c)"); 2127 is($count_b, $re->[4], "count b ($c)"); 2128 } 2129 } 2130 2131 { # Bleadperl v5.13.8-292-gf56b639 breaks NEZUMI/Unicode-LineBreak-1.011 2132 # \xdf in lookbehind failed to compile as is multi-char fold 2133 my $message = "Lookbehind with \\xdf matchable compiles"; 2134 my $r = eval 'qr{ 2135 (?u: (?<=^url:) | 2136 (?<=[/]) (?=[^/]) | 2137 (?<=[^-.]) (?=[-~.,_?\#%=&]) | 2138 (?<=[=&]) (?=.) 2139 )}iox'; 2140 is($@, '', $message); 2141 object_ok($r, 'Regexp', $message); 2142 } 2143 2144 # RT #82610 2145 like 'foo/file.fob', qr,^(?=[^\.])[^/]*/(?=[^\.])[^/]*\.fo[^/]$,; 2146 2147 { # This was failing unless an explicit /d was added 2148 my $E0 = uni_to_native("\xE0"); 2149 my $p = qr/[_$E0]/i; 2150 utf8::upgrade($p); 2151 like(uni_to_native("\xC0"), qr/$p/, "Verify \"\\xC0\" =~ /[\\xE0_]/i; pattern in utf8"); 2152 } 2153 2154 like "x", qr/\A(?>(?:(?:)A|B|C?x))\z/, 2155 "Check TRIE does not overwrite EXACT following NOTHING at start - RT #111842"; 2156 2157 { 2158 my $single = "z"; 2159 my $upper = "\x{390}"; # Fold is 3 chars. 2160 my $multi = CORE::fc($upper); 2161 2162 my $failed = 0; 2163 2164 # Try forcing a node to be split, with a multi-char fold at the 2165 # boundary 2166 for my $repeat (1 .. 300) { 2167 my $string = $single x $repeat; 2168 my $lhs = $string . $upper; 2169 if ($lhs !~ m/$string$multi/i) { 2170 $failed = $repeat; 2171 last; 2172 } 2173 } 2174 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed"); 2175 2176 $failed = 0; 2177 for my $repeat (1 .. 300) { 2178 my $string = $single x $repeat; 2179 my $lhs = $string . "\N{LATIN SMALL LIGATURE FFI}"; 2180 if ($lhs !~ m/${string}ff\N{LATIN SMALL LETTER I}/i) { 2181 $failed = $repeat; 2182 last; 2183 } 2184 } 2185 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed"); 2186 2187 $failed = 0; 2188 for my $repeat (1 .. 300) { 2189 my $string = $single x $repeat; 2190 my $lhs = $string . "\N{LATIN SMALL LIGATURE FFL}"; 2191 if ($lhs !~ m/${string}ff\N{U+6c}/i) { 2192 $failed = $repeat; 2193 last; 2194 } 2195 } 2196 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed"); 2197 2198 # This tests that under /d matching that an 'ss' split across two 2199 # parts of a node doesn't end up turning into something that matches 2200 # \xDF unless it is in utf8. 2201 $failed = 0; 2202 $single = 'a'; # Is non-terminal multi-char fold char 2203 for my $repeat (1 .. 300) { 2204 my $string = $single x $repeat; 2205 my $lhs = "$string\N{LATIN SMALL LETTER SHARP S}"; 2206 utf8::downgrade($lhs); 2207 $string .= "s"; 2208 if ($lhs =~ m/${string}s/di) { 2209 $failed = $repeat; 2210 last; 2211 } 2212 } 2213 ok(! $failed, "Matched multi-char fold 'ss' across EXACTF node boundaries; if failed, was at count $failed"); 2214 2215 for my $non_finals ("t", "ft", "ift", "sift") { 2216 my $base_pat = $non_finals . "enKalt"; # (The tail is taken from 2217 # the trouble ticket, is 2218 # arbitrary) 2219 for my $utf8 ("non-UTF-8", "UTF-8") { 2220 2221 # Try at different lengths to be sure to get a node boundary 2222 for my $repeat (120 .. 270) { # [perl #133756] 2223 my $head = ("b" x $repeat) . "\xDC"; 2224 my $pat = $base_pat; 2225 utf8::upgrade($pat) if $utf8 eq "UTF-8"; 2226 $pat = $head . $pat; 2227 my $text = $head . $base_pat; 2228 2229 if ($text !~ /$pat/i) { 2230 $failed = $repeat; 2231 last; 2232 } 2233 } 2234 2235 ok(! $failed, "A non-final fold character " 2236 . (length($non_finals) - 1) 2237 . " characters from the end of an EXACTFish" 2238 . " $utf8 pattern works; if failed, was at count $failed"); 2239 } 2240 } 2241 } 2242 2243 { 2244 fresh_perl_is('print eval "\"\x{101}\" =~ /[[:lower:]]/", "\n"; print eval "\"\x{100}\" =~ /[[:lower:]]/i", "\n";', 2245 "1\n1", # Both re's should match 2246 {}, 2247 "get [:lower:] swash in first eval; test under /i in second"); 2248 } 2249 2250 { 2251 fresh_perl_is(<<'EOF', 2252 my $s = "\x{41c}"; 2253 $s =~ /(.*)/ or die; 2254 $ls = lc $1; 2255 print $ls eq lc $s ? "good\n" : "bad: [$ls]\n"; 2256EOF 2257 "good\n", 2258 {}, 2259 "swash triggered by lc() doesn't corrupt \$1" 2260 ); 2261 } 2262 2263 { 2264 #' RT #119075 2265 no warnings 'regexp'; # Silence "has useless greediness modifier" 2266 local $@; 2267 eval { /a{0}?/; }; 2268 ok(! $@, 2269 "PCRE regression test: No 'Quantifier follows nothing in regex' warning"); 2270 2271 } 2272 2273 { 2274 unlike("\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2275 like("\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2276 unlike("\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2277 like("\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2278 unlike("\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2279 like("\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2280 2281 unlike("_\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2282 like("_\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2283 unlike("_\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2284 like("_\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2285 unlike("_\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2286 like("_\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2287 } 2288 2289 # These are defined later, so won't be known at regex compile time above 2290 sub IsMyRuntimeProperty { 2291 return "B6\n"; 2292 } 2293 2294 sub IsntMyRuntimeProperty { 2295 return "!B6\n"; 2296 } 2297 2298 { # [perl 121777] 2299 my $regex; 2300 { package Some; 2301 # define a Unicode propertyIs_q 2302 sub Is_q 2303 { 2304 sprintf '%x', ord 'q' 2305 } 2306 $regex = qr/\p{Is_q}/; 2307 2308 # If we uncomment the following line, prior to the patch that 2309 # fixed this, everything would work because we would have expanded 2310 # the property by the time the regex in the 'like' below got 2311 # compiled. 2312 #'q' =~ $regex; 2313 } 2314 2315 like('q', $regex, 'User-defined property matches outside package'); 2316 2317 package Some { 2318 main::like('abcq', qr/abc$regex/, 'Run-time compiled in-package user-defined property matches'); 2319 } 2320 } 2321 2322 { # From Lingua::Stem::UniNE; no ticket filed but related to #121778 2323 use utf8; 2324 my $word = 'рабта'; 2325 $word =~ s{ (?: 2326 ия # definite articles for nouns: 2327 | ът # ∙ masculine 2328 | та # ∙ feminine 2329 | то # ∙ neutral 2330 | те # ∙ plural 2331 ) $ }{}x; 2332 is($word, 'раб', "Handles UTF8 trie correctly"); 2333 } 2334 2335 { # [perl #122460] 2336 my $a = "rdvark"; 2337 $a =~ /(?{})(?=[A-Za-z0-9_])a*?/g; 2338 is (pos $a, 0, "optimizer correctly thinks (?=...) is 0-length"); 2339 } 2340 2341 { # [perl #123417] multi-char \N{...} tripping roundly 2342 use Cname; 2343 my $qr = qr$(\N{foo})$; 2344 "afoot" =~ eval "qr/$qr/"; 2345 is "$1" || $@, "foo", 'multichar \N{...} stringified and retoked'; 2346 } 2347 2348 is (scalar split(/\b{sb}/, "Don't think twice. It's all right."), 2349 2, '\b{wb} splits sentences correctly'); 2350 2351 ok "my/dir/audio_07.mp3" =~ 2352 qr/(.*)\/(.*)\/(.*)\.(?<=(?=(?:\.(?!\d+\b)\w{1,4}$)$)\.)(.*)$()/, 2353 "[perl #133948]"; 2354 2355 2356 # !!! NOTE! Keep the following tests last -- they may crash perl 2357 2358 print "# Tests that follow may crash perl\n"; 2359 { 2360 eval '/\k/'; 2361 like $@, qr/\QSequence \k... not terminated in regex;\E/, 2362 'Lone \k not allowed'; 2363 } 2364 2365 { 2366 my $message = "Substitution with lookahead (possible segv)"; 2367 $_ = "ns1ns1ns1"; 2368 s/ns(?=\d)/ns_/g; 2369 is($_, "ns_1ns_1ns_1", $message); 2370 $_ = "ns1"; 2371 s/ns(?=\d)/ns_/; 2372 is($_, "ns_1", $message); 2373 $_ = "123"; 2374 s/(?=\d+)|(?<=\d)/!Bang!/g; 2375 is($_, "!Bang!1!Bang!2!Bang!3!Bang!", $message); 2376 } 2377 2378 { 2379 # Earlier versions of Perl said this was fatal. 2380 my $message = "U+0FFFF shouldn't crash the regex engine"; 2381 no warnings 'utf8'; 2382 my $a = eval "chr(65535)"; 2383 use warnings; 2384 my $warning_message; 2385 local $SIG{__WARN__} = sub { $warning_message = $_[0] }; 2386 eval $a =~ /[a-z]/; 2387 ok(1, $message); # If it didn't crash, it worked. 2388 } 2389 2390 TODO: { # Was looping 2391 todo_skip('Triggers thread clone SEGV. See #86550') 2392 if $::running_as_thread && $::running_as_thread; 2393 watchdog(10 * ($ENV{PERL_TEST_TIME_OUT_FACTOR} || 1)); 2394 like("\x{00DF}", qr/[\x{1E9E}_]*/i, "\"\\x{00DF}\" =~ /[\\x{1E9E}_]*/i was looping"); 2395 } 2396 2397 { # Bug #90536, caused failed assertion 2398 unlike("s\N{U+DF}", qr/^\x{00DF}/i, "\"s\\N{U+DF}\", qr/^\\x{00DF}/i"); 2399 } 2400 2401 # User-defined Unicode properties to match above-Unicode code points 2402 sub Is_31_Bit_Super { return "110000\t7FFFFFFF\n" } 2403 sub Is_Portable_Super { return '!utf8::Any' } # Matches beyond 32 bits 2404 2405 { # Assertion was failing on on 64-bit platforms; just didn't work on 32. 2406 no warnings qw(non_unicode portable); 2407 use Config; 2408 2409 # We use 'ok' instead of 'like' because the warnings are lexically 2410 # scoped, and want to turn them off, so have to do the match in this 2411 # scope. 2412 if ($Config{uvsize} < 8) { 2413 like(chr(0x7FFF_FFFE), qr/\p{Is_31_Bit_Super}/, 2414 "chr(0x7FFF_FFFE) can match a Unicode property"); 2415 like(chr(0x7FFF_FFFF), qr/\p{Is_31_Bit_Super}/, 2416 "chr(0x7FFF_FFFF) can match a Unicode property"); 2417 my $p = qr/^[\x{7FFF_FFFF}]$/; 2418 like(chr(0x7FFF_FFFF), qr/$p/, 2419 "chr(0x7FFF_FFFF) can match itself in a [class]"); 2420 like(chr(0x7FFF_FFFF), qr/$p/, # Tests any caching 2421 "chr(0x7FFF_FFFF) can match itself in a [class] subsequently"); 2422 } 2423 else { 2424 no warnings 'overflow'; 2425 like(chr(0x7FFF_FFFF_FFFF_FFFE), qr/\p{Is_Portable_Super}/, 2426 "chr(0x7FFF_FFFF_FFFF_FFFE) can match a Unicode property"); 2427 like(chr(0x7FFF_FFFF_FFFF_FFFF), qr/^\p{Is_Portable_Super}$/, 2428 "chr(0x7FFF_FFFF_FFFF_FFFF) can match a Unicode property"); 2429 2430 my $p = eval 'qr/^\x{7FFF_FFFF_FFFF_FFFF}$/'; 2431 like(chr(0x7FFF_FFFF_FFFF_FFFF), qr/$p/, 2432 "chr(0x7FFF_FFFF_FFFF_FFFF) can match itself in a [class]"); 2433 like(chr(0x7FFF_FFFF_FFFF_FFFF), qr/$p/, # Tests any caching 2434 "chr(0x7FFF_FFFF_FFFF_FFFF) can match itself in a [class] subsequently"); 2435 2436 # This test is because something was declared as 32 bits, but 2437 # should have been cast to 64; only a problem where 2438 # sizeof(STRLEN) != sizeof(UV) 2439 unlike(chr(0x7FFF_FFFF_FFFF_FFFE), qr/\p{Is_31_Bit_Super}/, 2440 "chr(0x7FFF_FFFF_FFFF_FFFE) shouldn't match a range ending in 0x7FFF_FFFF"); 2441 } 2442 } 2443 2444 { # [perl #112530], the code below caused a panic 2445 sub InFoo { "a\tb\n9\ta\n" } 2446 like(chr(0xA), qr/\p{InFoo}/, 2447 "Overlapping ranges in user-defined properties"); 2448 } 2449 2450 { # [perl #125990], the final 2 tests below each caused a panic. 2451 # The \0's are not necessary; it could be a printable character 2452 # instead, but were in the ticket, so using them. 2453 my $sharp_s = chr utf8::unicode_to_native(0xdf); 2454 my $string = ("\0" x 8) 2455 . ($sharp_s x 3) 2456 . ("\0" x 42) 2457 . "ý"; 2458 my $folded_string = ("\0" x 8) 2459 . ("ss" x 3) 2460 . ("\0" x 42) 2461 . "ý"; 2462 utf8::downgrade($string); 2463 utf8::downgrade($folded_string); 2464 2465 use Cname; 2466 like($string, qr/$string/i, "LATIN SMALL SHARP S matches itself under /id"); 2467 unlike($folded_string, qr/$string/i, "LATIN SMALL SHARP S doesn't match 'ss' under /di"); 2468 like($folded_string, qr/\N{EMPTY-STR}$string/i, "\\N{} earlier than LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'"); 2469 like($folded_string, qr/$string\N{EMPTY-STR}/i, "\\N{} after LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'"); 2470 } 2471 2472 { # [perl #126606 crashed the interpreter 2473 use Cname; 2474 like("sS", qr/\N{EMPTY-STR}Ss|/i, '\N{} with empty branch alternation works'); 2475 like("sS", qr'\N{EMPTY-STR}Ss|'i, '\N{} with empty branch alternation works'); 2476 } 2477 2478 { # Regexp:Grammars was broken: 2479 # http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2013-06/msg01290.html 2480 fresh_perl_like('use warnings; "abc" =~ qr{(?&foo){0}abc(?<foo>)}', 2481 qr/Quantifier unexpected on zero-length expression/, 2482 {}, 2483 'No segfault on qr{(?&foo){0}abc(?<foo>)}'); 2484 } 2485 2486 SKIP: 2487 { # [perl #125826] buffer overflow in TRIE_STORE_REVCHAR 2488 # (during compilation, so use a fresh perl) 2489 $Config{uvsize} == 8 2490 or skip("need large code-points for this test", 1); 2491 2492 fresh_perl_is('/\x{E000000000}|/ and print qq(ok\n)', "ok\n", {}, 2493 "buffer overflow in TRIE_STORE_REVCHAR"); 2494 } 2495 2496 { 2497 fresh_perl_like('use warnings; s�0(?(?!00000000000000000000000000·000000)\500000000�0000000000000000000000000000000000000000000000000000·00000000000000000000000000000000�0', 2498 qr/Switch \(\?\(condition\)\.\.\. not terminated/, 2499 {}, 2500 'No segfault [perl #126886]'); 2501 } 2502 2503 { 2504 # [perl 130010] Downstream application texinfo started to report panics 2505 # as of commit a5540cf. 2506 2507 runperl( prog => 'A::xx(); package A; sub InFullwidth{ return qq|\n| } sub xx { split /[^\s\p{InFullwidth}]/, q|x| }' ); 2508 ok(! $?, "User-defined pattern did not cause panic [perl 130010]"); 2509 } 2510 2511 { # [perl #133999] Previously assertion failure 2512 fresh_perl_like('0 =~ /\p{nv:(\B(*COMMIT)C+)}/', 2513 qr/No Unicode property value wildcard matches/, 2514 {}, 2515 "Assertion failure with *COMMIT and wildcard property"); 2516 } 2517 2518 { # [perl #134029] Previously assertion failure 2519 fresh_perl_like('qr/\p{upper:]}|\337(?|ss)|)(?0/', 2520 qr/Unicode property wildcard not terminated/, 2521 {}, 2522 "Assertion failure with single character wildcard"); 2523 } 2524 2525 { # [perl #134034] Previously assertion failure 2526 fresh_perl_is('use utf8; q!Ȧिम한글��ყაოსაა!=~/(?li)\b{wb}\B(*COMMIT)0/;', 2527 "", {}, "*COMMIT caused positioning beyond EOS"); 2528 } 2529 2530 { # [GH #17486] Previously assertion failure 2531 fresh_perl_is('0=~/(?iaa)ss\337(?0)|/', 2532 "", {}, "EXACTFUP node isn't changed into something else"); 2533 } 2534 2535 { # [GH #17593] 2536 fresh_perl_is('qr/((?+2147483647))/', 2537 "Invalid reference to group in regex; marked by <--" 2538 . " HERE in m/((?+2147483647) <-- HERE )/ at - line 1.", 2539 {}, "integer overflow, undefined behavior in ASAN"); 2540 fresh_perl_is('qr/((?-2147483647))/', 2541 "Reference to nonexistent group in regex; marked by <--" 2542 . " HERE in m/((?-2147483647) <-- HERE )/ at - line 1.", 2543 {}, "Large negative relative capture group"); 2544 fresh_perl_is('qr/((?+18446744073709551615))/', 2545 "Invalid reference to group in regex; marked by <--" 2546 . " HERE in m/((?+18446744073709551615 <-- HERE ))/ at -" 2547 . " line 1.", 2548 {}, "Too large relative group number"); 2549 fresh_perl_is('qr/((?-18446744073709551615))/', 2550 "Invalid reference to group in regex; marked by <--" 2551 . " HERE in m/((?-18446744073709551615 <-- HERE ))/ at -" 2552 . " line 1.", 2553 {}, "Too large negative relative group number"); 2554 } 2555 2556 { # GH #17734, ASAN use after free 2557 fresh_perl_like('no warnings "experimental::uniprop_wildcards"; 2558 my $re = q<[[\p{name=/[Y-]+Z/}]]>; 2559 eval { "\N{BYZANTINE MUSICAL SYMBOL PSILI}" 2560 =~ /$re/ }; print $@ if $@; print "Done\n";', 2561 qr/Done/, 2562 {}, "GH #17734"); 2563 } 2564 2565 { # GH $17278 assertion fails 2566 fresh_perl_is('use locale; 2567 my $A_grave = "\N{LATIN CAPITAL LETTER A WITH GRAVE}"; 2568 utf8::encode($A_grave); 2569 my $a_grave = "\N{LATIN SMALL LETTER A WITH GRAVE}"; 2570 utf8::encode($a_grave); 2571 2572 my $z="q!$a_grave! =~ m!(?^i)[$A_grave]!"; 2573 utf8::decode($z); 2574 print eval $z, "\n";', 2575 1, 2576 {}, "GH #17278"); 2577 } 2578 2579 2580 # !!! NOTE that tests that aren't at all likely to crash perl should go 2581 # a ways above, above these last ones. There's a comment there that, like 2582 # this comment, contains the word 'NOTE' 2583 2584 done_testing(); 2585} # End of sub run_tests 2586 25871; 2588