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 @INC = qw(../lib .); 10 require './test.pl'; 11 require './charset_tools.pl'; 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 ok "\x8e" =~ /[\x89-\x91]/, '"\x8e" =~ /[\x89-\x91]/'; 44 ok "\xce" =~ /[\xc9-\xd1]/, '"\xce" =~ /[\xc9-\xd1]/'; 45 ok "\xd0" =~ /[\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 ok "\x{ab}" =~ /\x{ab}/, '"\x{ab}" =~ /\x{ab}/ '; 61 ok "\x{abcd}" =~ /\x{abcd}/, '"\x{abcd}" =~ /\x{abcd}/'; 62 } 63 64 { 65 my $message = 'bug id 20001008.001'; 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 $x = "\x4e" . "E"; 244 ok ($x =~ /^\x4EE$/, "Check only 2 bytes of hex are matched."); 245 246 $x = "\x4e" . "i"; 247 ok ($x =~ /^\x4Ei$/, "Check that invalid hex digit stops it (2)"); 248 249 $x = "\x4" . "j"; 250 ok ($x =~ /^\x4j$/, "Check that invalid hex digit stops it (1)"); 251 252 $x = "\x0" . "k"; 253 ok ($x =~ /^\xk$/, "Check that invalid hex digit stops it (0)"); 254 255 $x = "\x0" . "x"; 256 ok ($x =~ /^\xx$/, "\\xx isn't to be treated as \\0"); 257 258 $x = "\x0" . "xa"; 259 ok ($x =~ /^\xxa$/, "\\xxa isn't to be treated as \\xa"); 260 261 $x = "\x9" . "_b"; 262 ok ($x =~ /^\x9_b$/, "\\x9_b isn't to be treated as \\x9b"); 263 264 # and now again in [] ranges 265 266 $x = "\x4e" . "E"; 267 ok ($x =~ /^[\x4EE]{2}$/, "Check only 2 bytes of hex are matched."); 268 269 $x = "\x4e" . "i"; 270 ok ($x =~ /^[\x4Ei]{2}$/, "Check that invalid hex digit stops it (2)"); 271 272 $x = "\x4" . "j"; 273 ok ($x =~ /^[\x4j]{2}$/, "Check that invalid hex digit stops it (1)"); 274 275 $x = "\x0" . "k"; 276 ok ($x =~ /^[\xk]{2}$/, "Check that invalid hex digit stops it (0)"); 277 278 $x = "\x0" . "x"; 279 ok ($x =~ /^[\xx]{2}$/, "\\xx isn't to be treated as \\0"); 280 281 $x = "\x0" . "xa"; 282 ok ($x =~ /^[\xxa]{3}$/, "\\xxa isn't to be treated as \\xa"); 283 284 $x = "\x9" . "_b"; 285 ok ($x =~ /^[\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 $x = "\x9b"; 290 ok ($x =~ /^\x{9_b}$/, "\\x{9_b} is to be treated as \\x9b"); 291 292 $x = "\x9b" . "y"; 293 ok ($x =~ /^\x{9_b}y$/, "\\x{9_b} is to be treated as \\x9b (again)"); 294 295 $x = "\x9b" . "y"; 296 ok ($x =~ /^\x{9b_}y$/, "\\x{9b_} is to be treated as \\x9b"); 297 298 $x = "\x9b" . "y"; 299 ok ($x =~ /^\x{9_bq}y$/, "\\x{9_bc} is to be treated as \\x9b"); 300 301 $x = "\x0" . "y"; 302 ok ($x =~ /^\x{x9b}y$/, "\\x{x9b} is to be treated as \\x0"); 303 304 $x = "\x0" . "y"; 305 ok ($x =~ /^\x{0x9b}y$/, "\\x{0x9b} is to be treated as \\x0"); 306 307 $x = "\x9b" . "y"; 308 ok ($x =~ /^\x{09b}y$/, "\\x{09b} is to be treated as \\x9b"); 309 310 $x = "\x9b"; 311 ok ($x =~ /^[\x{9_b}]$/, "\\x{9_b} is to be treated as \\x9b"); 312 313 $x = "\x9b" . "y"; 314 ok ($x =~ /^[\x{9_b}y]{2}$/, 315 "\\x{9_b} is to be treated as \\x9b (again)"); 316 317 $x = "\x9b" . "y"; 318 ok ($x =~ /^[\x{9b_}y]{2}$/, "\\x{9b_} is to be treated as \\x9b"); 319 320 $x = "\x9b" . "y"; 321 ok ($x =~ /^[\x{9_bq}y]{2}$/, "\\x{9_bc} is to be treated as \\x9b"); 322 323 $x = "\x0" . "y"; 324 ok ($x =~ /^[\x{x9b}y]{2}$/, "\\x{x9b} is to be treated as \\x0"); 325 326 $x = "\x0" . "y"; 327 ok ($x =~ /^[\x{0x9b}y]{2}$/, "\\x{0x9b} is to be treated as \\x0"); 328 329 $x = "\x9b" . "y"; 330 ok ($x =~ /^[\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 ok $x =~ /.*?\200/, "High bit fine"; 338 } 339 340 { 341 # The basic character classes and Unicode 342 ok "\x{0100}" =~ /\w/, 'LATIN CAPITAL LETTER A WITH MACRON in /\w/'; 343 ok "\x{0660}" =~ /\d/, 'ARABIC-INDIC DIGIT ZERO in /\d/'; 344 ok "\x{1680}" =~ /\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}/i, $message); 539 like("\N{LATIN SMALL LETTER SHARP S}", 540 qr/[\N{LATIN SMALL LETTER SHARP S}]/, $message); 541 like("\N{LATIN SMALL LETTER SHARP S}", 542 qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message); 543 544 like("ss", qr /\N{LATIN SMALL LETTER SHARP S}/i, $message); 545 like("SS", qr /\N{LATIN SMALL LETTER SHARP S}/i, $message); 546 like("ss", qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message); 547 like("SS", qr/[\N{LATIN SMALL LETTER SHARP S}]/i, $message); 548 549 like("\N{LATIN SMALL LETTER SHARP S}", qr/ss/i, $message); 550 like("\N{LATIN SMALL LETTER SHARP S}", qr/SS/i, $message); 551 552 $message = "Unoptimized named sequence in class"; 553 like("ss", qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message); 554 like("SS", qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message); 555 like("\N{LATIN SMALL LETTER SHARP S}", 556 qr/[\N{LATIN SMALL LETTER SHARP S}x]/, $message); 557 like("\N{LATIN SMALL LETTER SHARP S}", 558 qr/[\N{LATIN SMALL LETTER SHARP S}x]/i, $message); 559 } 560 561 { 562 # More whitespace: U+0085, U+2028, U+2029\n"; 563 564 # U+0085, U+00A0 need to be forced to be Unicode, the \x{100} does that. 565 ok "<\x{100}" . uni_to_native("\x{0085}") . ">" =~ /<\x{100}\s>/, '\x{0085} in \s'; 566 ok "<" . uni_to_native("\x{0085}") . ">" =~ /<\v>/, '\x{0085} in \v'; 567 ok "<\x{100}" . uni_to_native("\x{00A0}") . ">" =~ /<\x{100}\s>/, '\x{00A0} in \s'; 568 ok "<" . uni_to_native("\x{00A0}") . ">" =~ /<\h>/, '\x{00A0} in \h'; 569 my @h = map {sprintf "%05x" => $_} 0x01680, 0x02000 .. 0x0200A, 570 0x0202F, 0x0205F, 0x03000; 571 my @v = map {sprintf "%05x" => $_} 0x02028, 0x02029; 572 573 my @H = map {sprintf "%05x" => $_} 0x01361, 0x0200B, 0x02408, 0x02420, 574 0x0303F, 0xE0020, 0x180E; 575 my @V = map {sprintf "%05x" => $_} 0x0008A .. 0x0008D, 0x00348, 0x10100, 576 0xE005F, 0xE007C, 0x180E; 577 578 for my $hex (@h) { 579 my $str = eval qq ["<\\x{$hex}>"]; 580 ok $str =~ /<\s>/, "\\x{$hex} in \\s"; 581 ok $str =~ /<\h>/, "\\x{$hex} in \\h"; 582 ok $str !~ /<\v>/, "\\x{$hex} not in \\v"; 583 } 584 585 for my $hex (@v) { 586 my $str = eval qq ["<\\x{$hex}>"]; 587 ok $str =~ /<\s>/, "\\x{$hex} in \\s"; 588 ok $str =~ /<\v>/, "\\x{$hex} in \\v"; 589 ok $str !~ /<\h>/, "\\x{$hex} not in \\h"; 590 } 591 592 for my $hex (@H) { 593 my $str = eval qq ["<\\x{$hex}>"]; 594 ok $str =~ /<\S>/, "\\x{$hex} in \\S"; 595 ok $str =~ /<\H>/, "\\x{$hex} in \\H"; 596 } 597 598 for my $hex (@V) { 599 my $str = eval qq ["<\\x{$hex}>"]; 600 ok $str =~ /<\S>/, "\\x{$hex} in \\S"; 601 ok $str =~ /<\V>/, "\\x{$hex} in \\V"; 602 } 603 } 604 605 { 606 # . with /s should work on characters, as opposed to bytes 607 my $message = ". with /s works on characters, not bytes"; 608 609 my $s = "\x{e4}\x{100}"; 610 # This is not expected to match: the point is that 611 # neither should we get "Malformed UTF-8" warnings. 612 warning_is(sub {$s =~ /\G(.+?)\n/gcs}, undef, 613 "No 'Malformed UTF-8' warning"); 614 615 my @c; 616 push @c => $1 while $s =~ /\G(.)/gs; 617 618 local $" = ""; 619 is("@c", $s, $message); 620 621 # Test only chars < 256 622 my $t1 = "Q003\n\n\x{e4}\x{f6}\n\nQ004\n\n\x{e7}"; 623 my $r1 = ""; 624 while ($t1 =~ / \G ( .+? ) \n\s+ ( .+? ) ( $ | \n\s+ ) /xgcs) { 625 $r1 .= $1 . $2; 626 } 627 628 my $t2 = $t1 . "\x{100}"; # Repeat with a larger char 629 my $r2 = ""; 630 while ($t2 =~ / \G ( .+? ) \n\s+ ( .+? ) ( $ | \n\s+ ) /xgcs) { 631 $r2 .= $1 . $2; 632 } 633 $r2 =~ s/\x{100}//; 634 635 is($r1, $r2, $message); 636 } 637 638 { 639 my $message = "Unicode lookbehind"; 640 like("A\x{100}B" , qr/(?<=A.)B/, $message); 641 like("A\x{200}\x{300}B", qr/(?<=A..)B/, $message); 642 like("\x{400}AB" , qr/(?<=\x{400}.)B/, $message); 643 like("\x{500}\x{600}B" , qr/(?<=\x{500}.)B/, $message); 644 645 # Original code also contained: 646 # ok "\x{500\x{600}}B" =~ /(?<=\x{500}.)B/; 647 # but that looks like a typo. 648 } 649 650 { 651 my $message = 'UTF-8 hash keys and /$/'; 652 # http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters 653 # /2002-01/msg01327.html 654 655 my $u = "a\x{100}"; 656 my $v = substr ($u, 0, 1); 657 my $w = substr ($u, 1, 1); 658 my %u = ($u => $u, $v => $v, $w => $w); 659 for (keys %u) { 660 my $m1 = /^\w*$/ ? 1 : 0; 661 my $m2 = $u {$_} =~ /^\w*$/ ? 1 : 0; 662 is($m1, $m2, $message); 663 } 664 } 665 666 { 667 my $message = "No SEGV in s/// and UTF-8"; 668 my $s = "s#\x{100}" x 4; 669 ok($s =~ s/[^\w]/ /g, $message); 670 if ( 1 or $ENV{PERL_TEST_LEGACY_POSIX_CC} ) { 671 is($s, "s \x{100}" x 4, $message); 672 } 673 else { 674 is($s, "s " x 4, $message); 675 } 676 } 677 678 { 679 my $message = "UTF-8 bug (maybe already known?)"; 680 my $u = "foo"; 681 $u =~ s/./\x{100}/g; 682 is($u, "\x{100}\x{100}\x{100}", $message); 683 684 $u = "foobar"; 685 $u =~ s/[ao]/\x{100}/g; 686 is($u, "f\x{100}\x{100}b\x{100}r", $message); 687 688 $u =~ s/\x{100}/e/g; 689 is($u, "feeber", $message); 690 } 691 692 { 693 my $message = "UTF-8 bug with s///"; 694 # check utf8/non-utf8 mixtures 695 # try to force all float/anchored check combinations 696 697 my $c = "\x{100}"; 698 my $subst; 699 for my $re ("xx.*$c", "x.*$c$c", "$c.*xx", "$c$c.*x", 700 "xx.*(?=$c)", "(?=$c).*xx",) { 701 unlike("xxx", qr/$re/, $message); 702 ok(+($subst = "xxx") !~ s/$re//, $message); 703 } 704 for my $re ("xx.*$c*", "$c*.*xx") { 705 like("xxx", qr/$re/, $message); 706 ok(+($subst = "xxx") =~ s/$re//, $message); 707 is($subst, "", $message); 708 } 709 for my $re ("xxy*", "y*xx") { 710 like("xx$c", qr/$re/, $message); 711 ok(+($subst = "xx$c") =~ s/$re//, $message); 712 is($subst, $c, $message); 713 unlike("xy$c", qr/$re/, $message); 714 ok(+($subst = "xy$c") !~ s/$re//, $message); 715 } 716 for my $re ("xy$c*z", "x$c*yz") { 717 like("xyz", qr/$re/, $message); 718 ok(+($subst = "xyz") =~ s/$re//, $message); 719 is($subst, "", $message); 720 } 721 } 722 723 { 724 # The second half of RT #114808 725 warning_is(sub {'aa' =~ /.+\x{100}/}, undef, 726 'utf8-only floating substr, non-utf8 target, no warning'); 727 } 728 729 { 730 my $message = "qr /.../x"; 731 my $R = qr / A B C # D E/x; 732 ok("ABCDE" =~ $R && $& eq "ABC", $message); 733 ok("ABCDE" =~ /$R/ && $& eq "ABC", $message); 734 ok("ABCDE" =~ m/$R/ && $& eq "ABC", $message); 735 ok("ABCDE" =~ /($R)/ && $1 eq "ABC", $message); 736 ok("ABCDE" =~ m/($R)/ && $1 eq "ABC", $message); 737 } 738 739 { 740 local $\; 741 $_ = 'aaaaaaaaaa'; 742 utf8::upgrade($_); chop $_; $\="\n"; 743 ok /[^\s]+/, 'm/[^\s]/ utf8'; 744 ok /[^\d]+/, 'm/[^\d]/ utf8'; 745 ok +($a = $_, $_ =~ s/[^\s]+/./g), 's/[^\s]/ utf8'; 746 ok +($a = $_, $a =~ s/[^\d]+/./g), 's/[^\s]/ utf8'; 747 } 748 749 { 750 # Subject: Odd regexp behavior 751 # From: Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk> 752 # Date: Wed, 26 Feb 2003 16:53:12 +0000 753 # Message-Id: <E18o4nw-0008Ly-00@wisbech.cl.cam.ac.uk> 754 # To: perl-unicode@perl.org 755 756 my $message = 'Markus Kuhn 2003-02-26'; 757 758 my $x = "\x{2019}\nk"; 759 ok($x =~ s/(\S)\n(\S)/$1 $2/sg, $message); 760 is($x, "\x{2019} k", $message); 761 762 $x = "b\nk"; 763 ok($x =~ s/(\S)\n(\S)/$1 $2/sg, $message); 764 is($x, "b k", $message); 765 766 like("\x{2019}", qr/\S/, $message); 767 } 768 769 { 770 ok "\x{100}\n" =~ /\x{100}\n$/, "UTF-8 length cache and fbm_compile"; 771 } 772 773 { 774 package Str; 775 use overload q /""/ => sub {${$_ [0]};}; 776 sub new {my ($c, $v) = @_; bless \$v, $c;} 777 778 package main; 779 $_ = Str -> new ("a\x{100}/\x{100}b"); 780 ok join (":", /\b(.)\x{100}/g) eq "a:/", "re_intuit_start and PL_bostr"; 781 } 782 783 { 784 my $re = qq /^([^X]*)X/; 785 utf8::upgrade ($re); 786 ok "\x{100}X" =~ /$re/, "S_cl_and ANYOF_UNICODE & ANYOF_INVERTED"; 787 my $loc_re = qq /(?l:^([^X]*)X)/; 788 utf8::upgrade ($loc_re); 789 no warnings 'locale'; 790 ok "\x{100}X" =~ /$loc_re/, "locale, S_cl_and ANYOF_UNICODE & ANYOF_INVERTED"; 791 } 792 793 { 794 ok "123\x{100}" =~ /^.*1.*23\x{100}$/, 795 'UTF-8 + multiple floating substr'; 796 } 797 798 { 799 my $message = '<20030808193656.5109.1@llama.ni-s.u-net.com>'; 800 801 # LATIN SMALL/CAPITAL LETTER A WITH MACRON 802 like(" \x{101}", qr/\x{100}/i, $message); 803 804 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW 805 like(" \x{1E01}", qr/\x{1E00}/i, $message); 806 807 # DESERET SMALL/CAPITAL LETTER LONG I 808 like(" \x{10428}", qr/\x{10400}/i, $message); 809 810 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X' 811 like(" \x{1E01}x", qr/\x{1E00}X/i, $message); 812 } 813 814 { 815 for (120 .. 130, 240 .. 260) { 816 my $head = 'x' x $_; 817 my $message = q [Don't misparse \x{...} in regexp ] . 818 q [near EXACT char count limit]; 819 for my $tail ('\x{0061}', '\x{1234}', '\x61') { 820 eval qq{like("$head$tail", qr/$head$tail/, \$message)}; 821 is($@, '', $message); 822 } 823 $message = q [Don't misparse \N{...} in regexp ] . 824 q [near EXACT char count limit]; 825 for my $tail ('\N{SNOWFLAKE}') { 826 eval qq {use charnames ':full'; 827 like("$head$tail", qr/$head$tail/, \$message)}; 828 is($@, '', $message); 829 } 830 } 831 } 832 833 { # TRIE related 834 our @got = (); 835 "words" =~ /(word|word|word)(?{push @got, $1})s$/; 836 is(@got, 1, "TRIE optimisation"); 837 838 @got = (); 839 "words" =~ /(word|word|word)(?{push @got,$1})s$/i; 840 is(@got, 1,"TRIEF optimisation"); 841 842 my @nums = map {int rand 1000} 1 .. 100; 843 my $re = "(" . (join "|", @nums) . ")"; 844 $re = qr/\b$re\b/; 845 846 foreach (@nums) { 847 ok $_ =~ /$re/, "Trie nums"; 848 } 849 850 $_ = join " ", @nums; 851 @got = (); 852 push @got, $1 while /$re/g; 853 854 my %count; 855 $count {$_} ++ for @got; 856 my $ok = 1; 857 for (@nums) { 858 $ok = 0 if --$count {$_} < 0; 859 } 860 ok $ok, "Trie min count matches"; 861 } 862 863 { 864 # TRIE related 865 # LATIN SMALL/CAPITAL LETTER A WITH MACRON 866 ok "foba \x{101}foo" =~ qr/(foo|\x{100}foo|bar)/i && 867 $1 eq "\x{101}foo", 868 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH MACRON"; 869 870 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW 871 ok "foba \x{1E01}foo" =~ qr/(foo|\x{1E00}foo|bar)/i && 872 $1 eq "\x{1E01}foo", 873 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW"; 874 875 # DESERET SMALL/CAPITAL LETTER LONG I 876 ok "foba \x{10428}foo" =~ qr/(foo|\x{10400}foo|bar)/i && 877 $1 eq "\x{10428}foo", 878 "TRIEF + DESERET SMALL/CAPITAL LETTER LONG I"; 879 880 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X' 881 ok "foba \x{1E01}xfoo" =~ qr/(foo|\x{1E00}Xfoo|bar)/i && 882 $1 eq "\x{1E01}xfoo", 883 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'"; 884 885 use charnames ':full'; 886 887 my $s = "\N{LATIN SMALL LETTER SHARP S}"; 888 ok "foba ba$s" =~ qr/(foo|Ba$s|bar)/i && $1 eq "ba$s", 889 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss"; 890 ok "foba ba$s" =~ qr/(Ba$s|foo|bar)/i && $1 eq "ba$s", 891 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss"; 892 ok "foba ba$s" =~ qr/(foo|bar|Ba$s)/i && $1 eq "ba$s", 893 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss"; 894 895 ok "foba ba$s" =~ qr/(foo|Bass|bar)/i && $1 eq "ba$s", 896 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss"; 897 898 ok "foba ba$s" =~ qr/(foo|BaSS|bar)/i && $1 eq "ba$s", 899 "TRIEF + LATIN SMALL LETTER SHARP S =~ SS"; 900 901 ok "foba ba${s}pxySS$s$s" =~ qr/(b(?:a${s}t|a${s}f|a${s}p)[xy]+$s*)/i 902 && $1 eq "ba${s}pxySS$s$s", 903 "COMMON PREFIX TRIEF + LATIN SMALL LETTER SHARP S"; 904 } 905 906 { 907 BEGIN { 908 unshift @INC, 'lib'; 909 } 910 use Cname; 911 912 ok 'fooB' =~ /\N{foo}[\N{B}\N{b}]/, "Passthrough charname"; 913 my $name = "foo\xDF"; 914 my $result = eval "'A${name}B' =~ /^A\\N{$name}B\$/"; 915 ok !$@ && $result, "Passthrough charname of non-ASCII, Latin1"; 916 # 917 # Why doesn't must_warn work here? 918 # 919 my $w; 920 local $SIG {__WARN__} = sub {$w .= "@_"}; 921 $result = eval 'q(WARN) =~ /^[\N{WARN}]$/'; 922 ok !$@ && $result && ! $w, '\N{} returning multi-char works'; 923 924 undef $w; 925 eval q [ok "\0" !~ /[\N{EMPTY-STR}XY]/, 926 "Zerolength charname in charclass doesn't match \\\\0"]; 927 ok $w && $w =~ /Ignoring zero length/, 928 'Ignoring zero length \N{} in character class warning'; 929 undef $w; 930 eval q [ok 'xy' =~ /x[\N{EMPTY-STR} y]/x, 931 'Empty string charname in [] is ignored; finds a following character']; 932 ok $w && $w =~ /Ignoring zero length/, 933 'Ignoring zero length \N{} in character class warning'; 934 undef $w; 935 eval q [ok 'x ' =~ /x[\N{EMPTY-STR} y]/, 936 'Empty string charname in [] is ignored; finds a following blank under /x']; 937 ok $w && $w =~ /Ignoring zero length/, 938 'Ignoring zero length \N{} in character class warning'; 939 940 ok 'AB' =~ /(\N{EVIL})/ && $1 eq 'A', 'Charname caching $1'; 941 ok 'ABC' =~ /(\N{EVIL})/, 'Charname caching $1'; 942 ok 'xy' =~ /x\N{EMPTY-STR}y/, 943 'Empty string charname produces NOTHING node'; 944 ok '' =~ /\N{EMPTY-STR}/, 945 'Empty string charname produces NOTHING node'; 946 ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/, 'Verify that long string works'; 947 ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/i, 'Verify under folding that long string works'; 948 949 eval '/(?[[\N{EMPTY-STR}]])/'; 950 ok $@ && $@ =~ /Zero length \\N\{}/; 951 952 undef $w; 953 { 954 () = eval q ["\N{TOO MANY SPACES}"]; 955 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "Multiple spaces in a row in a charnames alias is fatal"); 956 eval q [use utf8; () = "\N{TOO MANY SPACES}"]; 957 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "... same under utf8"); 958 } 959 960 undef $w; 961 { 962 () = eval q ["\N{TRAILING SPACE }"]; 963 like ($@, qr/charnames alias definitions may not contain trailing white-space/, "Trailing white-space in a charnames alias is fatal"); 964 eval q [use utf8; () = "\N{TRAILING SPACE }"]; 965 like ($@, qr/charnames alias definitions may not contain trailing white-space/, "... same under utf8"); 966 } 967 968 undef $w; 969 my $Cedilla_Latin1 = "GAR" 970 . uni_to_native("\xC7") 971 . "ON"; 972 my $Cedilla_utf8 = $Cedilla_Latin1; 973 utf8::upgrade($Cedilla_utf8); 974 eval qq[is("\\N{$Cedilla_Latin1}", "$Cedilla_Latin1", "A cedilla in character name works")]; 975 undef $w; 976 { 977 use feature 'unicode_eval'; 978 eval qq[use utf8; is("\\N{$Cedilla_utf8}", "$Cedilla_utf8", "... same under 'use utf8': they work")]; 979 } 980 981 undef $w; 982 my $NBSP_Latin1 = "NBSP" 983 . uni_to_native("\xA0") 984 . "SEPARATED" 985 . uni_to_native("\xA0") 986 . "SPACE"; 987 my $NBSP_utf8 = $NBSP_Latin1; 988 utf8::upgrade($NBSP_utf8); 989 eval qq[is("\\N{$NBSP_Latin1}", "$NBSP_Latin1", "An NBSP in character name works")]; 990 like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... but returns a deprecation warning"); 991 undef $w; 992 { 993 use feature 'unicode_eval'; 994 eval qq[use utf8; is("\\N{$NBSP_utf8}", "$NBSP_utf8", "Same under 'use utf8': they work")]; 995 like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... but return a deprecation warning"); 996 } 997 { 998 # disable lexical warnings 999 BEGIN { ${^WARNING_BITS} = undef; $^W = 0 } 1000 undef $w; 1001 () = eval qq["\\N{$NBSP_Latin1}"]; 1002 like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "And returns a deprecation warning outside of lexical warnings"); 1003 undef $w; 1004 use feature 'unicode_eval'; 1005 eval qq[use utf8; () = "\\N{$NBSP_utf8}"]; 1006 like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... same under utf8"); 1007 } 1008 { 1009 no warnings 'deprecated'; 1010 undef $w; 1011 eval qq["\\N{$NBSP_Latin1}"]; 1012 ok (! defined $w, "... and no warning if warnings are off"); 1013 use feature 'unicode_eval'; 1014 eval qq[use utf8; "\\N{$NBSP_utf8}"]; 1015 ok (! defined $w, "... same under 'use utf8'"); 1016 } 1017 { 1018 use warnings FATAL=>'deprecated'; 1019 () = eval qq["\\N{$NBSP_Latin1}"]; 1020 like ($@, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... the warning can be fatal"); 1021 use feature 'unicode_eval'; 1022 eval qq[use utf8; () = "\\N{$NBSP_utf8}"]; 1023 like ($@, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... same under utf8"); 1024 } 1025 1026 { 1027 BEGIN { no strict; *CnameTest:: = *{"_charnames\0A::" } } 1028 package CnameTest { sub translator { pop } } 1029 BEGIN { $^H{charnames} = \&CnameTest::translator } 1030 undef $w; 1031 () = eval q ["\N{TOO MANY SPACES}"]; 1032 like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, 1033 'translators in _charnames\0* packages get validated'); 1034 } 1035 1036 # If remove the limitation in regcomp code these should work 1037 # differently 1038 undef $w; 1039 eval q [ok "\N{TOO-LONG-STR}" =~ /^\N{TOO-LONG-STR}$/, 'Verify that what once was too long a string works']; 1040 eval 'q() =~ /\N{4F}/'; 1041 ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in name gives error'; 1042 eval 'q() =~ /\N{COM,MA}/'; 1043 ok $@ && $@ =~ /Invalid character/, 'Verify that comma in name gives error'; 1044 $name = "A" . uni_to_native("\x{D7}") . "O"; 1045 eval "q(W) =~ /\\N{$name}/"; 1046 ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in name gives error'; 1047 my $utf8_name = "7 CITIES OF GOLD"; 1048 utf8::upgrade($utf8_name); 1049 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1050 ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in utf8 name gives error'; 1051 $utf8_name = "SHARP #"; 1052 utf8::upgrade($utf8_name); 1053 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1054 ok $@ && $@ =~ /Invalid character/, 'Verify that ASCII symbol in utf8 name gives error'; 1055 $utf8_name = "A HOUSE " . uni_to_native("\xF7") . " AGAINST ITSELF"; 1056 utf8::upgrade($utf8_name); 1057 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1058 ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in utf8 name gives error'; 1059 $utf8_name = "\x{664} HORSEMEN}"; 1060 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1061 ok $@ && $@ =~ /Invalid character/, 'Verify that leading above Latin1 digit in utf8 name gives error'; 1062 $utf8_name = "A \x{1F4A9} WOULD SMELL AS SWEET}"; 1063 eval "use utf8; q(W) =~ /\\N{$utf8_name}/"; 1064 ok $@ && $@ =~ /Invalid character/, 'Verify that above Latin1 symbol in utf8 name gives error'; 1065 1066 undef $w; 1067 $name = "A" . uni_to_native("\x{D1}") . "O"; 1068 eval "q(W) =~ /\\N{$name}/"; 1069 ok ! $w, 'Verify that latin1 letter in name doesnt give warning'; 1070 1071 # This tests the code path that restarts the parse when the recursive 1072 # call to S_reg() from within S_grok_bslash_N() discovers that the 1073 # pattern needs to be recalculated as UTF-8. use eval to avoid 1074 # needing literal Unicode in this source file: 1075 my $r = eval "qr/\\N{\x{100}\x{100}}/"; 1076 isnt $r, undef, "Generated regex for multi-char UTF-8 charname" 1077 or diag($@); 1078 ok "\x{100}\x{100}" =~ $r, "which matches"; 1079 } 1080 1081 { 1082 use charnames ':full'; 1083 1084 ok 'aabc' !~ /a\N{PLUS SIGN}b/, '/a\N{PLUS SIGN}b/ against aabc'; 1085 ok 'a+bc' =~ /a\N{PLUS SIGN}b/, '/a\N{PLUS SIGN}b/ against a+bc'; 1086 1087 ok ' A B' =~ /\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/, 1088 'Intermixed named and unicode escapes'; 1089 ok "\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}" =~ 1090 /\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/, 1091 'Intermixed named and unicode escapes'; 1092 ok "\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}" =~ 1093 /[\N{SPACE}\N{U+0041}][\N{SPACE}\N{U+0042}]/, 1094 'Intermixed named and unicode escapes'; 1095 ok "\0" =~ /^\N{NULL}$/, 'Verify that \N{NULL} works; is not confused with an error'; 1096 } 1097 1098 { 1099 our $brackets; 1100 $brackets = qr{ 1101 { (?> [^{}]+ | (??{ $brackets }) )* } 1102 }x; 1103 1104 ok "{b{c}d" !~ m/^((??{ $brackets }))/, "Bracket mismatch"; 1105 1106 SKIP: { 1107 our @stack = (); 1108 my @expect = qw( 1109 stuff1 1110 stuff2 1111 <stuff1>and<stuff2> 1112 right 1113 <right> 1114 <<right>> 1115 <<<right>>> 1116 <<stuff1>and<stuff2>><<<<right>>>> 1117 ); 1118 1119 local $_ = '<<<stuff1>and<stuff2>><<<<right>>>>>'; 1120 ok /^(<((?:(?>[^<>]+)|(?1))*)>(?{push @stack, $2 }))$/, 1121 "Recursion matches"; 1122 is(@stack, @expect, "Right amount of matches") 1123 or skip "Won't test individual results as count isn't equal", 1124 0 + @expect; 1125 my $idx = 0; 1126 foreach my $expect (@expect) { 1127 is($stack [$idx], $expect, 1128 "Expecting '$expect' at stack pos #$idx"); 1129 $idx ++; 1130 } 1131 } 1132 } 1133 1134 { 1135 my $s = '123453456'; 1136 $s =~ s/(?<digits>\d+)\k<digits>/$+{digits}/; 1137 ok $s eq '123456', 'Named capture (angle brackets) s///'; 1138 $s = '123453456'; 1139 $s =~ s/(?'digits'\d+)\k'digits'/$+{digits}/; 1140 ok $s eq '123456', 'Named capture (single quotes) s///'; 1141 } 1142 1143 { 1144 my @ary = ( 1145 pack('U', 0x00F1), # n-tilde 1146 '_'.pack('U', 0x00F1), # _ + n-tilde 1147 'c'.pack('U', 0x0327), # c + cedilla 1148 pack('U*', 0x00F1, 0x0327),# n-tilde + cedilla 1149 pack('U', 0x0391), # ALPHA 1150 pack('U', 0x0391).'2', # ALPHA + 2 1151 pack('U', 0x0391).'_', # ALPHA + _ 1152 ); 1153 1154 for my $uni (@ary) { 1155 my ($r1, $c1, $r2, $c2) = eval qq { 1156 use utf8; 1157 scalar ("..foo foo.." =~ /(?'${uni}'foo) \\k'${uni}'/), 1158 \$+{${uni}}, 1159 scalar ("..bar bar.." =~ /(?<${uni}>bar) \\k<${uni}>/), 1160 \$+{${uni}}; 1161 }; 1162 ok $r1, "Named capture UTF (?'')"; 1163 ok defined $c1 && $c1 eq 'foo', "Named capture UTF \%+"; 1164 ok $r2, "Named capture UTF (?<>)"; 1165 ok defined $c2 && $c2 eq 'bar', "Named capture UTF \%+"; 1166 } 1167 } 1168 1169 { 1170 my $s = 'foo bar baz'; 1171 my @res; 1172 if ('1234' =~ /(?<A>1)(?<B>2)(?<A>3)(?<B>4)/) { 1173 foreach my $name (sort keys(%-)) { 1174 my $ary = $- {$name}; 1175 foreach my $idx (0 .. $#$ary) { 1176 push @res, "$name:$idx:$ary->[$idx]"; 1177 } 1178 } 1179 } 1180 my @expect = qw (A:0:1 A:1:3 B:0:2 B:1:4); 1181 is("@res", "@expect", "Check %-"); 1182 eval' 1183 no warnings "uninitialized"; 1184 print for $- {this_key_doesnt_exist}; 1185 '; 1186 ok !$@,'lvalue $- {...} should not throw an exception'; 1187 } 1188 1189 { 1190 # \, breaks {3,4} 1191 no warnings qw{deprecated regexp}; 1192 ok "xaaay" !~ /xa{3\,4}y/, '\, in a pattern'; 1193 ok "xa{3,4}y" =~ /xa{3\,4}y/, '\, in a pattern'; 1194 1195 # \c\ followed by _ 1196 ok "x\c_y" !~ /x\c\_y/, '\_ in a pattern'; 1197 ok "x\c\_y" =~ /x\c\_y/, '\_ in a pattern'; 1198 1199 # \c\ followed by other characters 1200 for my $c ("z", "\0", "!", chr(254), chr(256)) { 1201 my $targ = "a" . uni_to_native("\034") . "$c"; 1202 my $reg = "a\\c\\$c"; 1203 ok eval ("qq/$targ/ =~ /$reg/"), "\\c\\ in pattern"; 1204 } 1205 } 1206 1207 { # Test the (*PRUNE) pattern 1208 our $count = 0; 1209 'aaab' =~ /a+b?(?{$count++})(*FAIL)/; 1210 is($count, 9, "Expect 9 for no (*PRUNE)"); 1211 $count = 0; 1212 'aaab' =~ /a+b?(*PRUNE)(?{$count++})(*FAIL)/; 1213 is($count, 3, "Expect 3 with (*PRUNE)"); 1214 local $_ = 'aaab'; 1215 $count = 0; 1216 1 while /.(*PRUNE)(?{$count++})(*FAIL)/g; 1217 is($count, 4, "/.(*PRUNE)/"); 1218 $count = 0; 1219 'aaab' =~ /a+b?(??{'(*PRUNE)'})(?{$count++})(*FAIL)/; 1220 is($count, 3, "Expect 3 with (*PRUNE)"); 1221 local $_ = 'aaab'; 1222 $count = 0; 1223 1 while /.(??{'(*PRUNE)'})(?{$count++})(*FAIL)/g; 1224 is($count, 4, "/.(*PRUNE)/"); 1225 } 1226 1227 { # Test the (*SKIP) pattern 1228 our $count = 0; 1229 'aaab' =~ /a+b?(*SKIP)(?{$count++})(*FAIL)/; 1230 is($count, 1, "Expect 1 with (*SKIP)"); 1231 local $_ = 'aaab'; 1232 $count = 0; 1233 1 while /.(*SKIP)(?{$count++})(*FAIL)/g; 1234 is($count, 4, "/.(*SKIP)/"); 1235 $_ = 'aaabaaab'; 1236 $count = 0; 1237 our @res = (); 1238 1 while /(a+b?)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g; 1239 is($count, 2, "Expect 2 with (*SKIP)"); 1240 is("@res", "aaab aaab", "Adjacent (*SKIP) works as expected"); 1241 } 1242 1243 { # Test the (*SKIP) pattern 1244 our $count = 0; 1245 'aaab' =~ /a+b?(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/; 1246 is($count, 1, "Expect 1 with (*SKIP)"); 1247 local $_ = 'aaab'; 1248 $count = 0; 1249 1 while /.(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/g; 1250 is($count, 4, "/.(*SKIP)/"); 1251 $_ = 'aaabaaab'; 1252 $count = 0; 1253 our @res = (); 1254 1 while /(a+b?)(*MARK:foo)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g; 1255 is($count, 2, "Expect 2 with (*SKIP)"); 1256 is("@res", "aaab aaab", "Adjacent (*SKIP) works as expected"); 1257 } 1258 1259 { # Test the (*SKIP) pattern 1260 our $count = 0; 1261 'aaab' =~ /a*(*MARK:a)b?(*MARK:b)(*SKIP:a)(?{$count++})(*FAIL)/; 1262 is($count, 3, "Expect 3 with *MARK:a)b?(*MARK:b)(*SKIP:a)"); 1263 local $_ = 'aaabaaab'; 1264 $count = 0; 1265 our @res = (); 1266 1 while 1267 /(a*(*MARK:a)b?)(*MARK:x)(*SKIP:a)(?{$count++; push @res,$1})(*FAIL)/g; 1268 is($count, 5, "Expect 5 with (*MARK:a)b?)(*MARK:x)(*SKIP:a)"); 1269 is("@res", "aaab b aaab b ", 1270 "Adjacent (*MARK:a)b?)(*MARK:x)(*SKIP:a) works as expected"); 1271 } 1272 1273 { # Test the (*COMMIT) pattern 1274 our $count = 0; 1275 'aaabaaab' =~ /a+b?(*COMMIT)(?{$count++})(*FAIL)/; 1276 is($count, 1, "Expect 1 with (*COMMIT)"); 1277 local $_ = 'aaab'; 1278 $count = 0; 1279 1 while /.(*COMMIT)(?{$count++})(*FAIL)/g; 1280 is($count, 1, "/.(*COMMIT)/"); 1281 $_ = 'aaabaaab'; 1282 $count = 0; 1283 our @res = (); 1284 1 while /(a+b?)(*COMMIT)(?{$count++; push @res,$1})(*FAIL)/g; 1285 is($count, 1, "Expect 1 with (*COMMIT)"); 1286 is("@res", "aaab", "Adjacent (*COMMIT) works as expected"); 1287 1288 ok("1\n2a\n" !~ /^\d+(*COMMIT)\w+/m, "COMMIT and anchors"); 1289 } 1290 1291 { 1292 # Test named commits and the $REGERROR var 1293 local $REGERROR; 1294 for my $name ('', ':foo') { 1295 for my $pat ("(*PRUNE$name)", 1296 ($name ? "(*MARK$name)" : "") . "(*SKIP$name)", 1297 "(*COMMIT$name)") { 1298 for my $suffix ('(*FAIL)', '') { 1299 'aaaab' =~ /a+b$pat$suffix/; 1300 is($REGERROR, 1301 ($suffix ? ($name ? 'foo' : "1") : ""), 1302 "Test $pat and \$REGERROR $suffix"); 1303 } 1304 } 1305 } 1306 } 1307 1308 { 1309 # Test named commits and the $REGERROR var 1310 package Fnorble; 1311 our $REGERROR; 1312 local $REGERROR; 1313 for my $name ('', ':foo') { 1314 for my $pat ("(*PRUNE$name)", 1315 ($name ? "(*MARK$name)" : "") . "(*SKIP$name)", 1316 "(*COMMIT$name)") { 1317 for my $suffix ('(*FAIL)','') { 1318 'aaaab' =~ /a+b$pat$suffix/; 1319 ::is($REGERROR, 1320 ($suffix ? ($name ? 'foo' : "1") : ""), 1321 "Test $pat and \$REGERROR $suffix"); 1322 } 1323 } 1324 } 1325 } 1326 1327 { 1328 # Test named commits and the $REGERROR var 1329 my $message = '$REGERROR'; 1330 local $REGERROR; 1331 for my $word (qw (bar baz bop)) { 1332 $REGERROR = ""; 1333 "aaaaa$word" =~ 1334 /a+(?:bar(*COMMIT:bar)|baz(*COMMIT:baz)|bop(*COMMIT:bop))(*FAIL)/; 1335 is($REGERROR, $word, $message); 1336 } 1337 } 1338 1339 { 1340 #Mindnumbingly simple test of (*THEN) 1341 for ("ABC","BAX") { 1342 ok /A (*THEN) X | B (*THEN) C/x, "Simple (*THEN) test"; 1343 } 1344 } 1345 1346 { 1347 my $message = "Relative Recursion"; 1348 my $parens = qr/(\((?:[^()]++|(?-1))*+\))/; 1349 local $_ = 'foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))'; 1350 my ($all, $one, $two) = ('', '', ''); 1351 ok(m/foo $parens \s* \+ \s* bar $parens/x, $message); 1352 is($1, '((2*3)+4-3)', $message); 1353 is($2, '(2*(3+4)-1*(2-3))', $message); 1354 is($&, 'foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))', $message); 1355 is($&, $_, $message); 1356 } 1357 1358 { 1359 my $spaces=" "; 1360 local $_ = join 'bar', $spaces, $spaces; 1361 our $count = 0; 1362 s/(?>\s+bar)(?{$count++})//g; 1363 is($_, $spaces, "SUSPEND final string"); 1364 is($count, 1, "Optimiser should have prevented more than one match"); 1365 } 1366 1367 { 1368 # From Message-ID: <877ixs6oa6.fsf@k75.linux.bogus> 1369 my $dow_name = "nada"; 1370 my $parser = "(\$dow_name) = \$time_string =~ /(D\x{e9}\\ " . 1371 "C\x{e9}adaoin|D\x{e9}\\ Sathairn|\\w+|\x{100})/"; 1372 my $time_string = "D\x{e9} C\x{e9}adaoin"; 1373 eval $parser; 1374 ok !$@, "Test Eval worked"; 1375 is($dow_name, $time_string, "UTF-8 trie common prefix extraction"); 1376 } 1377 1378 { 1379 my $v; 1380 ($v = 'bar') =~ /(\w+)/g; 1381 $v = 'foo'; 1382 is("$1", 'bar', 1383 '$1 is safe after /g - may fail due to specialized config in pp_hot.c'); 1384 } 1385 1386 { 1387 my $message = "http://nntp.perl.org/group/perl.perl5.porters/118663"; 1388 my $qr_barR1 = qr/(bar)\g-1/; 1389 like("foobarbarxyz", $qr_barR1, $message); 1390 like("foobarbarxyz", qr/foo${qr_barR1}xyz/, $message); 1391 like("foobarbarxyz", qr/(foo)${qr_barR1}xyz/, $message); 1392 like("foobarbarxyz", qr/(foo)(bar)\g{-1}xyz/, $message); 1393 like("foobarbarxyz", qr/(foo${qr_barR1})xyz/, $message); 1394 like("foobarbarxyz", qr/(foo(bar)\g{-1})xyz/, $message); 1395 } 1396 1397 { 1398 my $message = '$REGMARK'; 1399 our @r = (); 1400 local $REGMARK; 1401 local $REGERROR; 1402 like('foofoo', qr/foo (*MARK:foo) (?{push @r,$REGMARK}) /x, $message); 1403 is("@r","foo", $message); 1404 is($REGMARK, "foo", $message); 1405 unlike('foofoo', qr/foo (*MARK:foo) (*FAIL) /x, $message); 1406 is($REGMARK, '', $message); 1407 is($REGERROR, 'foo', $message); 1408 } 1409 1410 { 1411 my $message = '\K test'; 1412 my $x; 1413 $x = "abc.def.ghi.jkl"; 1414 $x =~ s/.*\K\..*//; 1415 is($x, "abc.def.ghi", $message); 1416 1417 $x = "one two three four"; 1418 $x =~ s/o+ \Kthree//g; 1419 is($x, "one two four", $message); 1420 1421 $x = "abcde"; 1422 $x =~ s/(.)\K/$1/g; 1423 is($x, "aabbccddee", $message); 1424 } 1425 1426 { 1427 sub kt { 1428 return '4' if $_[0] eq '09028623'; 1429 } 1430 # Nested EVAL using PL_curpm (via $1 or friends) 1431 my $re; 1432 our $grabit = qr/ ([0-6][0-9]{7}) (??{ kt $1 }) [890] /x; 1433 $re = qr/^ ( (??{ $grabit }) ) $ /x; 1434 my @res = '0902862349' =~ $re; 1435 is(join ("-", @res), "0902862349", 1436 'PL_curpm is set properly on nested eval'); 1437 1438 our $qr = qr/ (o) (??{ $1 }) /x; 1439 ok 'boob'=~/( b (??{ $qr }) b )/x && 1, "PL_curpm, nested eval"; 1440 } 1441 1442 { 1443 use charnames ":full"; 1444 ok "\N{ROMAN NUMERAL ONE}" =~ /\p{Alphabetic}/, "I =~ Alphabetic"; 1445 ok "\N{ROMAN NUMERAL ONE}" =~ /\p{Uppercase}/, "I =~ Uppercase"; 1446 ok "\N{ROMAN NUMERAL ONE}" !~ /\p{Lowercase}/, "I !~ Lowercase"; 1447 ok "\N{ROMAN NUMERAL ONE}" =~ /\p{IDStart}/, "I =~ ID_Start"; 1448 ok "\N{ROMAN NUMERAL ONE}" =~ /\p{IDContinue}/, "I =~ ID_Continue"; 1449 ok "\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{Alphabetic}/, "i =~ Alphabetic"; 1450 ok "\N{SMALL ROMAN NUMERAL ONE}" !~ /\p{Uppercase}/, "i !~ Uppercase"; 1451 ok "\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{Uppercase}/i, "i =~ Uppercase under /i"; 1452 ok "\N{SMALL ROMAN NUMERAL ONE}" !~ /\p{Titlecase}/, "i !~ Titlecase"; 1453 ok "\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{Titlecase}/i, "i =~ Titlecase under /i"; 1454 ok "\N{ROMAN NUMERAL ONE}" =~ /\p{Lowercase}/i, "I =~ Lowercase under /i"; 1455 1456 ok "\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{Lowercase}/, "i =~ Lowercase"; 1457 ok "\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{IDStart}/, "i =~ ID_Start"; 1458 ok "\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{IDContinue}/, "i =~ ID_Continue" 1459 } 1460 1461 { # More checking that /i works on the few properties that it makes a 1462 # difference. Uppercase, Lowercase, and Titlecase were done in the 1463 # block above 1464 ok "A" =~ /\p{PosixUpper}/, "A =~ PosixUpper"; 1465 ok "A" =~ /\p{PosixUpper}/i, "A =~ PosixUpper under /i"; 1466 ok "A" !~ /\p{PosixLower}/, "A !~ PosixLower"; 1467 ok "A" =~ /\p{PosixLower}/i, "A =~ PosixLower under /i"; 1468 ok "a" !~ /\p{PosixUpper}/, "a !~ PosixUpper"; 1469 ok "a" =~ /\p{PosixUpper}/i, "a =~ PosixUpper under /i"; 1470 ok "a" =~ /\p{PosixLower}/, "a =~ PosixLower"; 1471 ok "a" =~ /\p{PosixLower}/i, "a =~ PosixLower under /i"; 1472 1473 ok uni_to_native("\xC0") =~ /\p{XPosixUpper}/, "\\xC0 =~ XPosixUpper"; 1474 ok uni_to_native("\xC0") =~ /\p{XPosixUpper}/i, "\\xC0 =~ XPosixUpper under /i"; 1475 ok uni_to_native("\xC0") !~ /\p{XPosixLower}/, "\\xC0 !~ XPosixLower"; 1476 ok uni_to_native("\xC0") =~ /\p{XPosixLower}/i, "\\xC0 =~ XPosixLower under /i"; 1477 ok uni_to_native("\xE0") !~ /\p{XPosixUpper}/, "\\xE0 !~ XPosixUpper"; 1478 ok uni_to_native("\xE0") =~ /\p{XPosixUpper}/i, "\\xE0 =~ XPosixUpper under /i"; 1479 ok uni_to_native("\xE0") =~ /\p{XPosixLower}/, "\\xE0 =~ XPosixLower"; 1480 ok uni_to_native("\xE0") =~ /\p{XPosixLower}/i, "\\xE0 =~ XPosixLower under /i"; 1481 1482 ok uni_to_native("\xC0") =~ /\p{UppercaseLetter}/, "\\xC0 =~ UppercaseLetter"; 1483 ok uni_to_native("\xC0") =~ /\p{UppercaseLetter}/i, "\\xC0 =~ UppercaseLetter under /i"; 1484 ok uni_to_native("\xC0") !~ /\p{LowercaseLetter}/, "\\xC0 !~ LowercaseLetter"; 1485 ok uni_to_native("\xC0") =~ /\p{LowercaseLetter}/i, "\\xC0 =~ LowercaseLetter under /i"; 1486 ok uni_to_native("\xC0") !~ /\p{TitlecaseLetter}/, "\\xC0 !~ TitlecaseLetter"; 1487 ok uni_to_native("\xC0") =~ /\p{TitlecaseLetter}/i, "\\xC0 =~ TitlecaseLetter under /i"; 1488 ok uni_to_native("\xE0") !~ /\p{UppercaseLetter}/, "\\xE0 !~ UppercaseLetter"; 1489 ok uni_to_native("\xE0") =~ /\p{UppercaseLetter}/i, "\\xE0 =~ UppercaseLetter under /i"; 1490 ok uni_to_native("\xE0") =~ /\p{LowercaseLetter}/, "\\xE0 =~ LowercaseLetter"; 1491 ok uni_to_native("\xE0") =~ /\p{LowercaseLetter}/i, "\\xE0 =~ LowercaseLetter under /i"; 1492 ok uni_to_native("\xE0") !~ /\p{TitlecaseLetter}/, "\\xE0 !~ TitlecaseLetter"; 1493 ok uni_to_native("\xE0") =~ /\p{TitlecaseLetter}/i, "\\xE0 =~ TitlecaseLetter under /i"; 1494 ok "\x{1C5}" !~ /\p{UppercaseLetter}/, "\\x{1C5} !~ UppercaseLetter"; 1495 ok "\x{1C5}" =~ /\p{UppercaseLetter}/i, "\\x{1C5} =~ UppercaseLetter under /i"; 1496 ok "\x{1C5}" !~ /\p{LowercaseLetter}/, "\\x{1C5} !~ LowercaseLetter"; 1497 ok "\x{1C5}" =~ /\p{LowercaseLetter}/i, "\\x{1C5} =~ LowercaseLetter under /i"; 1498 ok "\x{1C5}" =~ /\p{TitlecaseLetter}/, "\\x{1C5} =~ TitlecaseLetter"; 1499 ok "\x{1C5}" =~ /\p{TitlecaseLetter}/i, "\\x{1C5} =~ TitlecaseLetter under /i"; 1500 } 1501 1502 { 1503 # requirement of Unicode Technical Standard #18, 1.7 Code Points 1504 # cf. http://www.unicode.org/reports/tr18/#Supplementary_Characters 1505 for my $u (0x7FF, 0x800, 0xFFFF, 0x10000) { 1506 no warnings 'utf8'; # oops 1507 my $c = chr $u; 1508 my $x = sprintf '%04X', $u; 1509 ok "A${c}B" =~ /A[\0-\x{10000}]B/, "Unicode range - $x"; 1510 } 1511 } 1512 1513 { 1514 my $res=""; 1515 1516 if ('1' =~ /(?|(?<digit>1)|(?<digit>2))/) { 1517 $res = "@{$- {digit}}"; 1518 } 1519 is($res, "1", 1520 "Check that (?|...) doesnt cause dupe entries in the names array"); 1521 1522 $res = ""; 1523 if ('11' =~ /(?|(?<digit>1)|(?<digit>2))(?&digit)/) { 1524 $res = "@{$- {digit}}"; 1525 } 1526 is($res, "1", 1527 "Check that (?&..) to a buffer inside a (?|...) goes to the leftmost"); 1528 } 1529 1530 { 1531 use warnings; 1532 my $message = "ASCII pattern that really is UTF-8"; 1533 my @w; 1534 local $SIG {__WARN__} = sub {push @w, "@_"}; 1535 my $c = qq (\x{DF}); 1536 like($c, qr/${c}|\x{100}/, $message); 1537 is("@w", '', $message); 1538 } 1539 1540 { 1541 my $message = "Corruption of match results of qr// across scopes"; 1542 my $qr = qr/(fo+)(ba+r)/; 1543 'foobar' =~ /$qr/; 1544 is("$1$2", "foobar", $message); 1545 { 1546 'foooooobaaaaar' =~ /$qr/; 1547 is("$1$2", 'foooooobaaaaar', $message); 1548 } 1549 is("$1$2", "foobar", $message); 1550 } 1551 1552 { 1553 my $message = "HORIZWS"; 1554 local $_ = "\t \r\n \n \t".chr(11)."\n"; 1555 s/\H/H/g; 1556 s/\h/h/g; 1557 is($_, "hhHHhHhhHH", $message); 1558 $_ = "\t \r\n \n \t" . chr (11) . "\n"; 1559 utf8::upgrade ($_); 1560 s/\H/H/g; 1561 s/\h/h/g; 1562 is($_, "hhHHhHhhHH", $message); 1563 } 1564 1565 { 1566 # Various whitespace special patterns 1567 my @h = map {chr utf8::unicode_to_native($_) } 1568 0x09, 0x20, 0xa0, 0x1680, 0x2000, 1569 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, 0x2006, 1570 0x2007, 0x2008, 0x2009, 0x200a, 0x202f, 0x205f, 1571 0x3000; 1572 my @v = map {chr utf8::unicode_to_native($_) } 1573 0x0a, 0x0b, 0x0c, 0x0d, 0x85, 0x2028, 1574 0x2029; 1575 my @lb = (uni_to_native("\x0D\x0A"), 1576 map {chr utf8::unicode_to_native($_) } 1577 0x0A .. 0x0D, 0x85, 0x2028, 0x2029); 1578 foreach my $t ([\@h, qr/\h/, qr/\h+/], 1579 [\@v, qr/\v/, qr/\v+/], 1580 [\@lb, qr/\R/, qr/\R+/],) { 1581 my $ary = shift @$t; 1582 foreach my $pat (@$t) { 1583 foreach my $str (@$ary) { 1584 my $temp_str = $str; 1585 $temp_str = display($temp_str); 1586 ok $str =~ /($pat)/, $temp_str . " =~ /($pat)"; 1587 my $temp_1 = $1; 1588 is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "' after ($pat)"); 1589 utf8::upgrade ($str); 1590 ok $str =~ /($pat)/, "Upgraded " . $temp_str . " =~ /($pat)/"; 1591 is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "'(upgraded) after ($pat)"); 1592 } 1593 } 1594 } 1595 } 1596 1597 { 1598 # Check that \\xDF match properly in its various forms 1599 # Test that \xDF matches properly. this is pretty hacky stuff, 1600 # but its actually needed. The malarky with '-' is to prevent 1601 # compilation caching from playing any role in the test. 1602 my @df = (chr utf8::unicode_to_native(0xDF), '-', chr utf8::unicode_to_native(0xDF)); 1603 utf8::upgrade ($df [2]); 1604 my @strs = ('ss', 'sS', 'Ss', 'SS', chr utf8::unicode_to_native(0xDF)); 1605 my @ss = map {("$_", "$_")} @strs; 1606 utf8::upgrade ($ss [$_ * 2 + 1]) for 0 .. $#strs; 1607 1608 for my $ssi (0 .. $#ss) { 1609 for my $dfi (0 .. $#df) { 1610 my $pat = $df [$dfi]; 1611 my $str = $ss [$ssi]; 1612 my $utf_df = ($dfi > 1) ? 'utf8' : ''; 1613 my $utf_ss = ($ssi % 2) ? 'utf8' : ''; 1614 my $sstr; # We hard-code the ebcdic value below to avoid 1615 # perturbing the test 1616 ($sstr = $str) =~ s/\xDF/\\xDF/ if $::IS_ASCII; 1617 ($sstr = $str) =~ s/\x59/\\x59/ if $::IS_EBCDIC; 1618 1619 if ($utf_df || $utf_ss || length ($ss [$ssi]) == 1) { 1620 my $ret = $str =~ /$pat/i; 1621 next if $pat eq '-'; 1622 if ($::IS_ASCII) { 1623 ok $ret, "\"$sstr\" =~ /\\xDF/i " . 1624 "(str is @{[$utf_ss||'latin']}, pat is " . 1625 "@{[$utf_df||'latin']})"; 1626 } 1627 else { 1628 ok $ret, "\"$sstr\" =~ /\\x59/i " . 1629 "(str is @{[$utf_ss||'latin']}, pat is " . 1630 "@{[$utf_df||'latin']})"; 1631 } 1632 } 1633 else { 1634 my $ret = $str !~ /$pat/i; 1635 next if $pat eq '-'; 1636 if ($::IS_EBCDIC) { 1637 ok $ret, "\"$sstr\" !~ /\\x59/i " . 1638 "(str is @{[$utf_ss||'latin']}, pat is " . 1639 "@{[$utf_df||'latin']})"; 1640 } 1641 else { 1642 ok $ret, "\"$sstr\" !~ /\\xDF/i " . 1643 "(str is @{[$utf_ss||'latin']}, pat is " . 1644 "@{[$utf_df||'latin']})"; 1645 } 1646 } 1647 } 1648 } 1649 } 1650 1651 { 1652 my $message = "BBC(Bleadperl Breaks CPAN) Today: String::Multibyte"; 1653 my $re = qr/(?:[\x00-\xFF]{4})/; 1654 my $hyp = "\0\0\0-"; 1655 my $esc = "\0\0\0\\"; 1656 1657 my $str = "$esc$hyp$hyp$esc$esc"; 1658 my @a = ($str =~ /\G(?:\Q$esc$esc\E|\Q$esc$hyp\E|$re)/g); 1659 1660 is(@a,3, $message); 1661 local $" = "="; 1662 is("@a","$esc$hyp=$hyp=$esc$esc", $message); 1663 } 1664 1665 { 1666 # Test for keys in %+ and %- 1667 my $message = 'Test keys in %+ and %-'; 1668 no warnings 'uninitialized'; 1669 local $_ = "abcdef"; 1670 /(?<foo>a)|(?<foo>b)/; 1671 is((join ",", sort keys %+), "foo", $message); 1672 is((join ",", sort keys %-), "foo", $message); 1673 is((join ",", sort values %+), "a", $message); 1674 is((join ",", sort map "@$_", values %-), "a ", $message); 1675 /(?<bar>a)(?<bar>b)(?<quux>.)/; 1676 is((join ",", sort keys %+), "bar,quux", $message); 1677 is((join ",", sort keys %-), "bar,quux", $message); 1678 is((join ",", sort values %+), "a,c", $message); # leftmost 1679 is((join ",", sort map "@$_", values %-), "a b,c", $message); 1680 /(?<un>a)(?<deux>c)?/; # second buffer won't capture 1681 is((join ",", sort keys %+), "un", $message); 1682 is((join ",", sort keys %-), "deux,un", $message); 1683 is((join ",", sort values %+), "a", $message); 1684 is((join ",", sort map "@$_", values %-), ",a", $message); 1685 } 1686 1687 { 1688 # length() on captures, the numbered ones end up in Perl_magic_len 1689 local $_ = "aoeu " . uni_to_native("\xe6") . "var ook"; 1690 /^ \w+ \s (?<eek>\S+)/x; 1691 1692 is(length $`, 0, q[length $`]); 1693 is(length $', 4, q[length $']); 1694 is(length $&, 9, q[length $&]); 1695 is(length $1, 4, q[length $1]); 1696 is(length $+{eek}, 4, q[length $+{eek} == length $1]); 1697 } 1698 1699 { 1700 my $ok = -1; 1701 1702 $ok = exists ($-{x}) ? 1 : 0 if 'bar' =~ /(?<x>foo)|bar/; 1703 is($ok, 1, '$-{x} exists after "bar"=~/(?<x>foo)|bar/'); 1704 is(scalar (%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/'); 1705 is(scalar (%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/'); 1706 1707 $ok = -1; 1708 $ok = exists ($+{x}) ? 1 : 0 if 'bar' =~ /(?<x>foo)|bar/; 1709 is($ok, 0, '$+{x} not exists after "bar"=~/(?<x>foo)|bar/'); 1710 is(scalar (%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/'); 1711 is(scalar (%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/'); 1712 1713 $ok = -1; 1714 $ok = exists ($-{x}) ? 1 : 0 if 'foo' =~ /(?<x>foo)|bar/; 1715 is($ok, 1, '$-{x} exists after "foo"=~/(?<x>foo)|bar/'); 1716 is(scalar (%+), 1, 'scalar %+ == 1 after "foo"=~/(?<x>foo)|bar/'); 1717 is(scalar (%-), 1, 'scalar %- == 1 after "foo"=~/(?<x>foo)|bar/'); 1718 1719 $ok = -1; 1720 $ok = exists ($+{x}) ? 1 : 0 if 'foo'=~/(?<x>foo)|bar/; 1721 is($ok, 1, '$+{x} exists after "foo"=~/(?<x>foo)|bar/'); 1722 } 1723 1724 { 1725 local $_; 1726 ($_ = 'abc') =~ /(abc)/g; 1727 $_ = '123'; 1728 is("$1", 'abc', "/g leads to unsafe match vars: $1"); 1729 1730 fresh_perl_is(<<'EOP', ">abc<\n", {}, 'mention $&'); 1731$&; 1732my $x; 1733($x='abc')=~/(abc)/g; 1734$x='123'; 1735print ">$1<\n"; 1736EOP 1737 1738 fresh_perl_is(<<'EOP', ">abc<\n", {}, 'no mention of $&'); 1739my $x; 1740($x='abc')=~/(abc)/g; 1741$x='123'; 1742print ">$1<\n"; 1743EOP 1744 } 1745 1746 { 1747 # Message-ID: <20070818091501.7eff4831@r2d2> 1748 my $str = ""; 1749 for (0 .. 5) { 1750 my @x; 1751 $str .= "@x"; # this should ALWAYS be the empty string 1752 'a' =~ /(a|)/; 1753 push @x, 1; 1754 } 1755 is(length $str, 0, "Trie scope error, string should be empty"); 1756 $str = ""; 1757 my @foo = ('a') x 5; 1758 for (@foo) { 1759 my @bar; 1760 $str .= "@bar"; 1761 s/a|/push @bar, 1/e; 1762 } 1763 is(length $str, 0, "Trie scope error, string should be empty"); 1764 } 1765 1766 { 1767# more TRIE/AHOCORASICK problems with mixed utf8 / latin-1 and case folding 1768 for my $ord (160 .. 255) { 1769 my $chr = utf8::unicode_to_native($ord); 1770 my $chr_byte = chr($chr); 1771 my $chr_utf8 = chr($chr); utf8::upgrade($chr_utf8); 1772 my $rx = qr{$chr_byte|X}i; 1773 ok($chr_utf8 =~ $rx, "utf8/latin, codepoint $chr"); 1774 } 1775 } 1776 1777 { 1778 our $a = 3; "" =~ /(??{ $a })/; 1779 our $b = $a; 1780 is($b, $a, "Copy of scalar used for postponed subexpression"); 1781 } 1782 1783 { 1784 our @ctl_n = (); 1785 our @plus = (); 1786 our $nested_tags; 1787 $nested_tags = qr{ 1788 < 1789 (\w+) 1790 (?{ 1791 push @ctl_n,$^N; 1792 push @plus,$+; 1793 }) 1794 > 1795 (??{$nested_tags})* 1796 </\s* \w+ \s*> 1797 }x; 1798 1799 my $match = '<bla><blubb></blubb></bla>' =~ m/^$nested_tags$/; 1800 ok $match, 'nested construct matches'; 1801 is("@ctl_n", "bla blubb", '$^N inside of (?{}) works as expected'); 1802 is("@plus", "bla blubb", '$+ inside of (?{}) works as expected'); 1803 } 1804 1805 SKIP: { 1806 # XXX: This set of tests is essentially broken, POSIX character classes 1807 # should not have differing definitions under Unicode. 1808 # There are property names for that. 1809 skip "Tests assume ASCII", 4 unless $::IS_ASCII; 1810 1811 my @notIsPunct = grep {/[[:punct:]]/ and not /\p{IsPunct}/} 1812 map {chr} 0x20 .. 0x7f; 1813 is(join ('', @notIsPunct), '$+<=>^`|~', 1814 '[:punct:] disagrees with IsPunct on Symbols'); 1815 1816 my @isPrint = grep {not /[[:print:]]/ and /\p{IsPrint}/} 1817 map {chr} 0 .. 0x1f, 0x7f .. 0x9f; 1818 is(join ('', @isPrint), "", 1819 'IsPrint agrees with [:print:] on control characters'); 1820 1821 my @isPunct = grep {/[[:punct:]]/ != /\p{IsPunct}/} 1822 map {chr} 0x80 .. 0xff; 1823 is(join ('', @isPunct), "\xa1\xa7\xab\xb6\xb7\xbb\xbf", # ¡ « · » ¿ 1824 'IsPunct disagrees with [:punct:] outside ASCII'); 1825 1826 my @isPunctLatin1 = eval q { 1827 grep {/[[:punct:]]/u != /\p{IsPunct}/} map {chr} 0x80 .. 0xff; 1828 }; 1829 skip "Eval failed ($@)", 1 if $@; 1830 skip "PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS set to 0", 1 1831 if !$ENV{PERL_TEST_LEGACY_POSIX_CC}; 1832 is(join ('', @isPunctLatin1), '', 1833 'IsPunct agrees with [:punct:] with explicit Latin1'); 1834 } 1835 1836 { 1837 # Tests for [#perl 71942] 1838 our $count_a; 1839 our $count_b; 1840 1841 my $c = 0; 1842 for my $re ( 1843# [ 1844# should match?, 1845# input string, 1846# re 1, 1847# re 2, 1848# expected values of count_a and count_b, 1849# ] 1850 [ 1851 0, 1852 "xababz", 1853 qr/a+(?{$count_a++})b?(*COMMIT)(*FAIL)/, 1854 qr/a+(?{$count_b++})b?(*COMMIT)z/, 1855 1, 1856 ], 1857 [ 1858 0, 1859 "xababz", 1860 qr/a+(?{$count_a++})b?(*COMMIT)\s*(*FAIL)/, 1861 qr/a+(?{$count_b++})b?(*COMMIT)\s*z/, 1862 1, 1863 ], 1864 [ 1865 0, 1866 "xababz", 1867 qr/a+(?{$count_a++})(?:b|)?(*COMMIT)(*FAIL)/, 1868 qr/a+(?{$count_b++})(?:b|)?(*COMMIT)z/, 1869 1, 1870 ], 1871 [ 1872 0, 1873 "xababz", 1874 qr/a+(?{$count_a++})b{0,6}(*COMMIT)(*FAIL)/, 1875 qr/a+(?{$count_b++})b{0,6}(*COMMIT)z/, 1876 1, 1877 ], 1878 [ 1879 0, 1880 "xabcabcz", 1881 qr/a+(?{$count_a++})(bc){0,6}(*COMMIT)(*FAIL)/, 1882 qr/a+(?{$count_b++})(bc){0,6}(*COMMIT)z/, 1883 1, 1884 ], 1885 [ 1886 0, 1887 "xabcabcz", 1888 qr/a+(?{$count_a++})(bc*){0,6}(*COMMIT)(*FAIL)/, 1889 qr/a+(?{$count_b++})(bc*){0,6}(*COMMIT)z/, 1890 1, 1891 ], 1892 1893 1894 [ 1895 0, 1896 "aaaabtz", 1897 qr/a+(?{$count_a++})b?(*PRUNE)(*FAIL)/, 1898 qr/a+(?{$count_b++})b?(*PRUNE)z/, 1899 4, 1900 ], 1901 [ 1902 0, 1903 "aaaabtz", 1904 qr/a+(?{$count_a++})b?(*PRUNE)\s*(*FAIL)/, 1905 qr/a+(?{$count_b++})b?(*PRUNE)\s*z/, 1906 4, 1907 ], 1908 [ 1909 0, 1910 "aaaabtz", 1911 qr/a+(?{$count_a++})(?:b|)(*PRUNE)(*FAIL)/, 1912 qr/a+(?{$count_b++})(?:b|)(*PRUNE)z/, 1913 4, 1914 ], 1915 [ 1916 0, 1917 "aaaabtz", 1918 qr/a+(?{$count_a++})b{0,6}(*PRUNE)(*FAIL)/, 1919 qr/a+(?{$count_b++})b{0,6}(*PRUNE)z/, 1920 4, 1921 ], 1922 [ 1923 0, 1924 "aaaabctz", 1925 qr/a+(?{$count_a++})(bc){0,6}(*PRUNE)(*FAIL)/, 1926 qr/a+(?{$count_b++})(bc){0,6}(*PRUNE)z/, 1927 4, 1928 ], 1929 [ 1930 0, 1931 "aaaabctz", 1932 qr/a+(?{$count_a++})(bc*){0,6}(*PRUNE)(*FAIL)/, 1933 qr/a+(?{$count_b++})(bc*){0,6}(*PRUNE)z/, 1934 4, 1935 ], 1936 1937 [ 1938 0, 1939 "aaabaaab", 1940 qr/a+(?{$count_a++;})b?(*SKIP)(*FAIL)/, 1941 qr/a+(?{$count_b++;})b?(*SKIP)z/, 1942 2, 1943 ], 1944 [ 1945 0, 1946 "aaabaaab", 1947 qr/a+(?{$count_a++;})b?(*SKIP)\s*(*FAIL)/, 1948 qr/a+(?{$count_b++;})b?(*SKIP)\s*z/, 1949 2, 1950 ], 1951 [ 1952 0, 1953 "aaabaaab", 1954 qr/a+(?{$count_a++;})(?:b|)(*SKIP)(*FAIL)/, 1955 qr/a+(?{$count_b++;})(?:b|)(*SKIP)z/, 1956 2, 1957 ], 1958 [ 1959 0, 1960 "aaabaaab", 1961 qr/a+(?{$count_a++;})b{0,6}(*SKIP)(*FAIL)/, 1962 qr/a+(?{$count_b++;})b{0,6}(*SKIP)z/, 1963 2, 1964 ], 1965 [ 1966 0, 1967 "aaabcaaabc", 1968 qr/a+(?{$count_a++;})(bc){0,6}(*SKIP)(*FAIL)/, 1969 qr/a+(?{$count_b++;})(bc){0,6}(*SKIP)z/, 1970 2, 1971 ], 1972 [ 1973 0, 1974 "aaabcaaabc", 1975 qr/a+(?{$count_a++;})(bc*){0,6}(*SKIP)(*FAIL)/, 1976 qr/a+(?{$count_b++;})(bc*){0,6}(*SKIP)z/, 1977 2, 1978 ], 1979 1980 1981 [ 1982 0, 1983 "aaddbdaabyzc", 1984 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) (*FAIL) \s* c \1 /x, 1985 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) z \s* c \1 /x, 1986 4, 1987 ], 1988 [ 1989 0, 1990 "aaddbdaabyzc", 1991 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) \s* (*FAIL) \s* c \1 /x, 1992 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b? (*SKIP:T1) \s* z \s* c \1 /x, 1993 4, 1994 ], 1995 [ 1996 0, 1997 "aaddbdaabyzc", 1998 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (?:b|) (*SKIP:T1) (*FAIL) \s* c \1 /x, 1999 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (?:b|) (*SKIP:T1) z \s* c \1 /x, 2000 4, 2001 ], 2002 [ 2003 0, 2004 "aaddbdaabyzc", 2005 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? b{0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x, 2006 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? b{0,6} (*SKIP:T1) z \s* c \1 /x, 2007 4, 2008 ], 2009 [ 2010 0, 2011 "aaddbcdaabcyzc", 2012 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (bc){0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x, 2013 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (bc){0,6} (*SKIP:T1) z \s* c \1 /x, 2014 4, 2015 ], 2016 [ 2017 0, 2018 "aaddbcdaabcyzc", 2019 qr/a (?{$count_a++;}) (*MARK:T1) (a*) .*? (bc*){0,6} (*SKIP:T1) (*FAIL) \s* c \1 /x, 2020 qr/a (?{$count_b++;}) (*MARK:T1) (a*) .*? (bc*){0,6} (*SKIP:T1) z \s* c \1 /x, 2021 4, 2022 ], 2023 2024 2025 [ 2026 0, 2027 "aaaaddbdaabyzc", 2028 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2029 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2030 2, 2031 ], 2032 [ 2033 0, 2034 "aaaaddbdaabyzc", 2035 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) \s* (*FAIL) \s* c \1 /x, 2036 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b? (*MARK:T1) (*SKIP:T1) \s* z \s* c \1 /x, 2037 2, 2038 ], 2039 [ 2040 0, 2041 "aaaaddbdaabyzc", 2042 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (?:b|) (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2043 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (?:b|) (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2044 2, 2045 ], 2046 [ 2047 0, 2048 "aaaaddbdaabyzc", 2049 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? b{0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2050 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? b{0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2051 2, 2052 ], 2053 [ 2054 0, 2055 "aaaaddbcdaabcyzc", 2056 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (bc){0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2057 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (bc){0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2058 2, 2059 ], 2060 [ 2061 0, 2062 "aaaaddbcdaabcyzc", 2063 qr/a (?{$count_a++;}) (a?) (*MARK:T1) (a*) .*? (bc*){0,6} (*MARK:T1) (*SKIP:T1) (*FAIL) \s* c \1 /x, 2064 qr/a (?{$count_b++;}) (a?) (*MARK:T1) (a*) .*? (bc*){0,6} (*MARK:T1) (*SKIP:T1) z \s* c \1 /x, 2065 2, 2066 ], 2067 2068 2069 [ 2070 0, 2071 "AbcdCBefgBhiBqz", 2072 qr/(A (.*) (?{ $count_a++ }) C? (*THEN) | A D) (*FAIL)/x, 2073 qr/(A (.*) (?{ $count_b++ }) C? (*THEN) | A D) z/x, 2074 1, 2075 ], 2076 [ 2077 0, 2078 "AbcdCBefgBhiBqz", 2079 qr/(A (.*) (?{ $count_a++ }) C? (*THEN) | A D) \s* (*FAIL)/x, 2080 qr/(A (.*) (?{ $count_b++ }) C? (*THEN) | A D) \s* z/x, 2081 1, 2082 ], 2083 [ 2084 0, 2085 "AbcdCBefgBhiBqz", 2086 qr/(A (.*) (?{ $count_a++ }) (?:C|) (*THEN) | A D) (*FAIL)/x, 2087 qr/(A (.*) (?{ $count_b++ }) (?:C|) (*THEN) | A D) z/x, 2088 1, 2089 ], 2090 [ 2091 0, 2092 "AbcdCBefgBhiBqz", 2093 qr/(A (.*) (?{ $count_a++ }) C{0,6} (*THEN) | A D) (*FAIL)/x, 2094 qr/(A (.*) (?{ $count_b++ }) C{0,6} (*THEN) | A D) z/x, 2095 1, 2096 ], 2097 [ 2098 0, 2099 "AbcdCEBefgBhiBqz", 2100 qr/(A (.*) (?{ $count_a++ }) (CE){0,6} (*THEN) | A D) (*FAIL)/x, 2101 qr/(A (.*) (?{ $count_b++ }) (CE){0,6} (*THEN) | A D) z/x, 2102 1, 2103 ], 2104 [ 2105 0, 2106 "AbcdCBefgBhiBqz", 2107 qr/(A (.*) (?{ $count_a++ }) (CE*){0,6} (*THEN) | A D) (*FAIL)/x, 2108 qr/(A (.*) (?{ $count_b++ }) (CE*){0,6} (*THEN) | A D) z/x, 2109 1, 2110 ], 2111 ) { 2112 $c++; 2113 $count_a = 0; 2114 $count_b = 0; 2115 2116 my $match_a = ($re->[1] =~ $re->[2]) || 0; 2117 my $match_b = ($re->[1] =~ $re->[3]) || 0; 2118 2119 is($match_a, $re->[0], "match a " . ($re->[0] ? "succeeded" : "failed") . " ($c)"); 2120 is($match_b, $re->[0], "match b " . ($re->[0] ? "succeeded" : "failed") . " ($c)"); 2121 is($count_a, $re->[4], "count a ($c)"); 2122 is($count_b, $re->[4], "count b ($c)"); 2123 } 2124 } 2125 2126 { # Bleadperl v5.13.8-292-gf56b639 breaks NEZUMI/Unicode-LineBreak-1.011 2127 # \xdf in lookbehind failed to compile as is multi-char fold 2128 my $message = "Lookbehind with \\xdf matchable compiles"; 2129 my $r = eval 'qr{ 2130 (?u: (?<=^url:) | 2131 (?<=[/]) (?=[^/]) | 2132 (?<=[^-.]) (?=[-~.,_?\#%=&]) | 2133 (?<=[=&]) (?=.) 2134 )}iox'; 2135 is($@, '', $message); 2136 object_ok($r, 'Regexp', $message); 2137 } 2138 2139 # RT #82610 2140 ok 'foo/file.fob' =~ m,^(?=[^\.])[^/]*/(?=[^\.])[^/]*\.fo[^/]$,; 2141 2142 { # This was failing unless an explicit /d was added 2143 my $E0 = uni_to_native("\xE0"); 2144 my $p = qr/[_$E0]/i; 2145 utf8::upgrade($p); 2146 like(uni_to_native("\xC0"), qr/$p/, "Verify \"\\xC0\" =~ /[\\xE0_]/i; pattern in utf8"); 2147 } 2148 2149 ok "x" =~ /\A(?>(?:(?:)A|B|C?x))\z/, 2150 "Check TRIE does not overwrite EXACT following NOTHING at start - RT #111842"; 2151 2152 { 2153 my $single = ":"; 2154 my $upper = "\x{390}"; # Fold is 3 chars. 2155 my $multi = CORE::fc($upper); 2156 2157 my $failed = 0; 2158 2159 # Try forcing a node to be split, with a multi-char fold at the 2160 # boundary 2161 for my $repeat (1 .. 300) { 2162 my $string = $single x $repeat; 2163 my $lhs = $string . $upper; 2164 if ($lhs !~ m/$string$multi/i) { 2165 $failed = $repeat; 2166 last; 2167 } 2168 } 2169 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed"); 2170 2171 $failed = 0; 2172 for my $repeat (1 .. 300) { 2173 my $string = $single x $repeat; 2174 my $lhs = $string . "\N{LATIN SMALL LIGATURE FFI}"; 2175 if ($lhs !~ m/${string}ff\N{LATIN SMALL LETTER I}/i) { 2176 $failed = $repeat; 2177 last; 2178 } 2179 } 2180 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed"); 2181 2182 $failed = 0; 2183 for my $repeat (1 .. 300) { 2184 my $string = $single x $repeat; 2185 my $lhs = $string . "\N{LATIN SMALL LIGATURE FFL}"; 2186 if ($lhs !~ m/${string}ff\N{U+6c}/i) { 2187 $failed = $repeat; 2188 last; 2189 } 2190 } 2191 ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed"); 2192 2193 # This tests that under /d matching that an 'ss' split across two 2194 # parts of a node doesn't end up turning into something that matches 2195 # \xDF unless it is in utf8. 2196 $failed = 0; 2197 $single = 'a'; # Is non-terminal multi-char fold char 2198 for my $repeat (1 .. 300) { 2199 my $string = $single x $repeat; 2200 my $lhs = "$string\N{LATIN SMALL LETTER SHARP S}"; 2201 utf8::downgrade($lhs); 2202 $string .= "s"; 2203 if ($lhs =~ m/${string}s/di) { 2204 $failed = $repeat; 2205 last; 2206 } 2207 } 2208 ok(! $failed, "Matched multi-char fold 'ss' across EXACTF node boundaries; if failed, was at count $failed"); 2209 } 2210 2211 { 2212 fresh_perl_is('print eval "\"\x{101}\" =~ /[[:lower:]]/", "\n"; print eval "\"\x{100}\" =~ /[[:lower:]]/i", "\n";', 2213 "1\n1", # Both re's should match 2214 {}, 2215 "get [:lower:] swash in first eval; test under /i in second"); 2216 } 2217 2218 { 2219 fresh_perl_is(<<'EOF', 2220 my $s = "\x{41c}"; 2221 $s =~ /(.*)/ or die; 2222 $ls = lc $1; 2223 print $ls eq lc $s ? "good\n" : "bad: [$ls]\n"; 2224EOF 2225 "good\n", 2226 {}, 2227 "swash triggered by lc() doesn't corrupt \$1" 2228 ); 2229 } 2230 2231 { 2232 #' RT #119075 2233 no warnings 'regexp'; # Silence "has useless greediness modifier" 2234 local $@; 2235 eval { /a{0}?/; }; 2236 ok(! $@, 2237 "PCRE regression test: No 'Quantifier follows nothing in regex' warning"); 2238 2239 } 2240 2241 { 2242 unlike("\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2243 like("\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2244 unlike("\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2245 like("\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2246 unlike("\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2247 like("\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2248 2249 unlike("_\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2250 like("_\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2251 unlike("_\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada"); 2252 like("_\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2253 unlike("_\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2254 like("_\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada"); 2255 } 2256 2257 # These are defined later, so won't be known at regex compile time above 2258 sub IsMyRuntimeProperty { 2259 return "B6\n"; 2260 } 2261 2262 sub IsntMyRuntimeProperty { 2263 return "!B6\n"; 2264 } 2265 2266 { # [perl 121777] 2267 my $regex; 2268 { package Some; 2269 # define a Unicode propertyIs_q 2270 sub Is_q 2271 { 2272 sprintf '%x', ord 'q' 2273 } 2274 $regex = qr/\p{Is_q}/; 2275 2276 # If we uncomment the following line, prior to the patch that 2277 # fixed this, everything would work because we would have expanded 2278 # the property by the time the regex in the 'like' below got 2279 # compiled. 2280 #'q' =~ $regex; 2281 } 2282 2283 like('q', $regex, 'User-defined property matches outside package'); 2284 2285 package Some { 2286 main::like('abcq', qr/abc$regex/, 'Run-time compiled in-package user-defined property matches'); 2287 } 2288 } 2289 2290 { # From Lingua::Stem::UniNE; no ticket filed but related to #121778 2291 use utf8; 2292 my $word = 'рабта'; 2293 $word =~ s{ (?: 2294 ия # definite articles for nouns: 2295 | ът # ∙ masculine 2296 | та # ∙ feminine 2297 | то # ∙ neutral 2298 | те # ∙ plural 2299 ) $ }{}x; 2300 is($word, 'раб', "Handles UTF8 trie correctly"); 2301 } 2302 2303 { # [perl #122460] 2304 my $a = "rdvark"; 2305 $a =~ /(?{})(?=[A-Za-z0-9_])a*?/g; 2306 is (pos $a, 0, "optimizer correctly thinks (?=...) is 0-length"); 2307 } 2308 2309 { # [perl #123417] multi-char \N{...} tripping roundly 2310 use Cname; 2311 my $qr = qr$(\N{foo})$; 2312 "afoot" =~ eval "qr/$qr/"; 2313 is "$1" || $@, "foo", 'multichar \N{...} stringified and retoked'; 2314 } 2315 { # empty \N{...} tripping roundly 2316 no warnings 'deprecated'; 2317 BEGIN { $^H{charnames} = sub { "" } } 2318 my $qr = qr$(a\N{foo}t)$; 2319 "at" =~ eval "qr/$qr/"; 2320 is "$1" || $@, "at", 'empty \N{...} stringified and retoked'; 2321 } 2322 2323 is (scalar split(/\b{sb}/, "Don't think twice. It's all right."), 2324 2, '\b{wb} splits sentences correctly'); 2325 2326 2327 # !!! NOTE! Keep the following tests last -- they may crash perl 2328 2329 print "# Tests that follow may crash perl\n"; 2330 { 2331 eval '/\k/'; 2332 ok $@ =~ /\QSequence \k... not terminated in regex;\E/, 2333 'Lone \k not allowed'; 2334 } 2335 2336 { 2337 my $message = "Substitution with lookahead (possible segv)"; 2338 $_ = "ns1ns1ns1"; 2339 s/ns(?=\d)/ns_/g; 2340 is($_, "ns_1ns_1ns_1", $message); 2341 $_ = "ns1"; 2342 s/ns(?=\d)/ns_/; 2343 is($_, "ns_1", $message); 2344 $_ = "123"; 2345 s/(?=\d+)|(?<=\d)/!Bang!/g; 2346 is($_, "!Bang!1!Bang!2!Bang!3!Bang!", $message); 2347 } 2348 2349 { 2350 # Earlier versions of Perl said this was fatal. 2351 my $message = "U+0FFFF shouldn't crash the regex engine"; 2352 no warnings 'utf8'; 2353 my $a = eval "chr(65535)"; 2354 use warnings; 2355 my $warning_message; 2356 local $SIG{__WARN__} = sub { $warning_message = $_[0] }; 2357 eval $a =~ /[a-z]/; 2358 ok(1, $message); # If it didn't crash, it worked. 2359 } 2360 2361 TODO: { # Was looping 2362 todo_skip('Triggers thread clone SEGV. See #86550') 2363 if $::running_as_thread && $::running_as_thread; 2364 watchdog(10); # Use a bigger value for busy systems 2365 like("\x{00DF}", qr/[\x{1E9E}_]*/i, "\"\\x{00DF}\" =~ /[\\x{1E9E}_]*/i was looping"); 2366 } 2367 2368 { # Bug #90536, caused failed assertion 2369 unlike("s\N{U+DF}", qr/^\x{00DF}/i, "\"s\\N{U+DF}\", qr/^\\x{00DF}/i"); 2370 } 2371 2372 # User-defined Unicode properties to match above-Unicode code points 2373 sub Is_32_Bit_Super { return "110000\tFFFFFFFF\n" } 2374 sub Is_Portable_Super { return '!utf8::Any' } # Matches beyond 32 bits 2375 2376 { # Assertion was failing on on 64-bit platforms; just didn't work on 32. 2377 no warnings qw(non_unicode portable); 2378 no warnings 'deprecated'; # These are above IV_MAX 2379 use Config; 2380 2381 # We use 'ok' instead of 'like' because the warnings are lexically 2382 # scoped, and want to turn them off, so have to do the match in this 2383 # scope. 2384 if ($Config{uvsize} < 8) { 2385 ok(chr(0xFFFF_FFFE) =~ /\p{Is_32_Bit_Super}/, 2386 "chr(0xFFFF_FFFE) can match a Unicode property"); 2387 ok(chr(0xFFFF_FFFF) =~ /\p{Is_32_Bit_Super}/, 2388 "chr(0xFFFF_FFFF) can match a Unicode property"); 2389 my $p = qr/^[\x{FFFF_FFFF}]$/; 2390 ok(chr(0xFFFF_FFFF) =~ $p, 2391 "chr(0xFFFF_FFFF) can match itself in a [class]"); 2392 ok(chr(0xFFFF_FFFF) =~ $p, # Tests any caching 2393 "chr(0xFFFF_FFFF) can match itself in a [class] subsequently"); 2394 } 2395 else { 2396 no warnings 'overflow'; 2397 ok(chr(0xFFFF_FFFF_FFFF_FFFE) =~ qr/\p{Is_Portable_Super}/, 2398 "chr(0xFFFF_FFFF_FFFF_FFFE) can match a Unicode property"); 2399 ok(chr(0xFFFF_FFFF_FFFF_FFFF) =~ qr/^\p{Is_Portable_Super}$/, 2400 "chr(0xFFFF_FFFF_FFFF_FFFF) can match a Unicode property"); 2401 2402 my $p = qr/^[\x{FFFF_FFFF_FFFF_FFFF}]$/; 2403 ok(chr(0xFFFF_FFFF_FFFF_FFFF) =~ $p, 2404 "chr(0xFFFF_FFFF_FFFF_FFFF) can match itself in a [class]"); 2405 ok(chr(0xFFFF_FFFF_FFFF_FFFF) =~ $p, # Tests any caching 2406 "chr(0xFFFF_FFFF_FFFF_FFFF) can match itself in a [class] subsequently"); 2407 2408 # This test is because something was declared as 32 bits, but 2409 # should have been cast to 64; only a problem where 2410 # sizeof(STRLEN) != sizeof(UV) 2411 ok(chr(0xFFFF_FFFF_FFFF_FFFE) !~ qr/\p{Is_32_Bit_Super}/, "chr(0xFFFF_FFFF_FFFF_FFFE) shouldn't match a range ending in 0xFFFF_FFFF"); 2412 } 2413 } 2414 2415 { # [perl #112530], the code below caused a panic 2416 sub InFoo { "a\tb\n9\ta\n" } 2417 like(chr(0xA), qr/\p{InFoo}/, 2418 "Overlapping ranges in user-defined properties"); 2419 } 2420 2421 { # [perl #125990], the final 2 tests below each caused a panic. 2422 # The \0's are not necessary; it could be a printable character 2423 # instead, but were in the ticket, so using them. 2424 my $sharp_s = chr utf8::unicode_to_native(0xdf); 2425 my $string = ("\0" x 8) 2426 . ($sharp_s x 3) 2427 . ("\0" x 42) 2428 . "ý"; 2429 my $folded_string = ("\0" x 8) 2430 . ("ss" x 3) 2431 . ("\0" x 42) 2432 . "ý"; 2433 utf8::downgrade($string); 2434 utf8::downgrade($folded_string); 2435 2436 like($string, qr/$string/i, "LATIN SMALL SHARP S matches itself under /id"); 2437 unlike($folded_string, qr/$string/i, "LATIN SMALL SHARP S doesn't match 'ss' under /di"); 2438 no warnings 'deprecated'; 2439 like($folded_string, qr/\N{}$string/i, "\\N{} earlier than LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'"); 2440 like($folded_string, qr/$string\N{}/i, "\\N{} after LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'"); 2441 } 2442 2443 { # Regexp:Grammars was broken: 2444 # http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2013-06/msg01290.html 2445 fresh_perl_like('use warnings; "abc" =~ qr{(?&foo){0}abc(?<foo>)}', 2446 qr/Quantifier unexpected on zero-length expression/, 2447 {}, 2448 'No segfault on qr{(?&foo){0}abc(?<foo>)}'); 2449 } 2450 2451 SKIP: 2452 { # [perl #125826] buffer overflow in TRIE_STORE_REVCHAR 2453 # (during compilation, so use a fresh perl) 2454 $Config{uvsize} == 8 2455 or skip("need large code-points for this test", 1); 2456 2457 # This is above IV_MAX on 32 bit machines, so turn off those warnings 2458 fresh_perl_is('no warnings "deprecated"; /\x{E000000000}|/ and print qq(ok\n)', "ok\n", {}, 2459 "buffer overflow in TRIE_STORE_REVCHAR"); 2460 } 2461 2462 { 2463 fresh_perl_like('use warnings; s�0(?(?!00000000000000000000000000·000000)\500000000�0000000000000000000000000000000000000000000000000000·00000000000000000000000000000000�0', 2464 qr/Switch \(\?\(condition\)\.\.\. not terminated/, 2465 {}, 2466 'No segfault [perl #126886]'); 2467 } 2468 2469 # !!! NOTE that tests that aren't at all likely to crash perl should go 2470 # a ways above, above these last ones. There's a comment there that, like 2471 # this comment, contains the word 'NOTE' 2472 2473 done_testing(); 2474} # End of sub run_tests 2475 24761; 2477