3 # This is a home for regular expression tests that don't fit into
4 # the format supported by op/regexp.t. If you want to add a test
5 # that does fit that format, add it to op/re_tests, not here.
9 # Test counter output is generated by a BEGIN block at bottom of file
15 our $Message = "Noname test";
17 eval 'use Config'; # Defaults assumed if this fails
19 run_tests() unless caller;
25 if ($x =~ /^abc/) {print "ok 1\n";} else {print "not ok 1\n";}
26 if ($x !~ /^def/) {print "ok 2\n";} else {print "not ok 2\n";}
28 # used to be a test for $*
29 if ($x =~ /^def/m) {print "ok 3\n";} else {print "not ok 3\n";}
32 if (/^([0-9][0-9]*)/) {print "ok 4\n";} else {print "not ok 4\n";}
34 if ($x =~ /^xxx/) {print "not ok 5\n";} else {print "ok 5\n";}
35 if ($x !~ /^abc/) {print "not ok 6\n";} else {print "ok 6\n";}
37 if ($x =~ /def/) {print "ok 7\n";} else {print "not ok 7\n";}
38 if ($x !~ /def/) {print "not ok 8\n";} else {print "ok 8\n";}
40 if ($x !~ /.def/) {print "ok 9\n";} else {print "not ok 9\n";}
41 if ($x =~ /.def/) {print "not ok 10\n";} else {print "ok 10\n";}
43 if ($x =~ /\ndef/) {print "ok 11\n";} else {print "not ok 11\n";}
44 if ($x !~ /\ndef/) {print "not ok 12\n";} else {print "ok 12\n";}
47 if (/(a*b*)(c*)/ && $1 eq 'aaabbb' && $2 eq 'ccc') {
52 if (/(a+b+c+)/ && $1 eq 'aaabbbccc') {
58 if (/a+b?c+/) {print "not ok 15\n";} else {print "ok 15\n";}
61 if (/a+b?c+/) {print "ok 16\n";} else {print "not ok 16\n";}
62 if (/a*b+c*/) {print "ok 17\n";} else {print "not ok 17\n";}
65 if (/a*b?c*/) {print "ok 18\n";} else {print "not ok 18\n";}
66 if (/a*b+c*/) {print "not ok 19\n";} else {print "ok 19\n";}
69 if (/bcd|xyz/) {print "ok 20\n";} else {print "not ok 20\n";}
70 if (/xyz|bcd/) {print "ok 21\n";} else {print "not ok 21\n";}
72 if (m|bc/*d|) {print "ok 22\n";} else {print "not ok 22\n";}
74 if (/^$_$/) {print "ok 23\n";} else {print "not ok 23\n";}
76 # used to be a test for $*
77 if ("ab\ncd\n" =~ /^cd/m) {print "ok 24\n";} else {print "not ok 24\n";}
83 @XXX = ('ok 25','not ok 25', 'ok 26','not ok 26','not ok 27');
84 while ($_ = shift(@XXX)) {
85 ?(.*)? && (print $1,"\n");
98 while (($key,$val) = each(%XXX)) {
108 if ($& eq 'xyz') {print "ok 28\n";} else {print "not ok 28\n";}
113 if ($& eq 'xyz') {print "ok 29\n";} else {print "not ok 29\n";}
118 if ($& eq 'xyz') {print "ok 30\n";} else {print "not ok 30\n";}
121 /def/; # optimized up to cmd
122 if ("$`:$&:$'" eq 'abc:def:ghi') {print "ok 31\n";} else {print "not ok 31\n";}
124 /cde/ + 0; # optimized only to spat
125 if ("$`:$&:$'" eq 'ab:cde:fghi') {print "ok 32\n";} else {print "not ok 32\n";}
127 /[d][e][f]/; # not optimized
128 if ("$`:$&:$'" eq 'abc:def:ghi') {print "ok 33\n";} else {print "not ok 33\n";}
130 $_ = 'now is the {time for all} good men to come to.';
132 if ($1 eq 'time for all') {print "ok 34\n";} else {print "not ok 34 $1\n";}
134 $_ = 'xxx {3,4} yyy zzz';
135 print /( {3,4})/ ? "ok 35\n" : "not ok 35\n";
136 print $1 eq ' ' ? "ok 36\n" : "not ok 36\n";
137 print /( {4,})/ ? "not ok 37\n" : "ok 37\n";
138 print /( {2,3}.)/ ? "ok 38\n" : "not ok 38\n";
139 print $1 eq ' y' ? "ok 39\n" : "not ok 39\n";
140 print /(y{2,3}.)/ ? "ok 40\n" : "not ok 40\n";
141 print $1 eq 'yyy ' ? "ok 41\n" : "not ok 41\n";
142 print /x {3,4}/ ? "not ok 42\n" : "ok 42\n";
143 print /^xxx {3,4}/ ? "not ok 43\n" : "ok 43\n";
145 $_ = "now is the time for all good men to come to.";
147 print join(':',@words) eq "now:is:the:time:for:all:good:men:to:come:to"
155 print join(':',@words) eq "now:is:the:time:for:all:good:men:to:come:to"
164 print join(':',@words) eq "to:to"
166 : "not ok 46 `@words'\n";
170 print join(':',@words) eq "to:to"
172 : "not ok 47 `@words'\n";
186 $t1=$t2=$t3=$t4=$t5=$t6=$t7=$t8=$t9=0;
200 $x = "$t1$t2$t3$t4$t5$t6$t7$t8$t9";
201 print $x eq '505550555' ? "ok 48\n" : "not ok 48 $x\n";
204 print "abc" =~ /^abc$|$xyz/ ? "ok 49\n" : "not ok 49\n";
206 # perl 4.009 says "unmatched ()"
207 eval '"abc" =~ /a(bc$)|$xyz/; $result = "$&:$1"';
208 print $@ eq "" ? "ok 50\n" : "not ok 50\n";
209 print $result eq "abc:bc" ? "ok 51\n" : "not ok 51\n";
214 print $` eq "" ? "ok 52\n" : "not ok 52\n" if $x;
216 print $` eq "abcfoo" ? "ok 53\n" : "not ok 53\n" if $x;
218 print $x == 0 ? "ok 54\n" : "not ok 54\n";
221 print $` eq "" ? "ok 55\n" : "not ok 55\n" if $x;
223 print $` eq "abcfoo" ? "ok 56\n" : "not ok 56\n" if $x;
225 print $x == 0 ? "ok 57\n" : "not ok 57\n";
228 print $' eq "fooabcbar" ? "ok 58\n" : "not ok 58\n" if $x;
230 print $' eq "bar" ? "ok 59\n" : "not ok 59\n" if $x;
233 print scalar @x == 2 ? "ok 60\n" : "not ok 60\n";
238 print "not " if (pos $_) != 2;
241 print "not " if defined pos $_;
245 'abc' =~ m'a(?{ $out = 2 })b';
246 print "not " if $out != 2;
250 'abc' =~ m'a(?{ $out = 3 })c';
251 print "not " if $out != 1;
254 $_ = 'foobar1 bar2 foobar3 barfoobar5 foobar6';
255 @out = /(?<!foo)bar./g;
256 print "not " if "@out" ne 'bar2 barf';
259 # Tests which depend on REG_INFTY
260 $reg_infty = defined $Config{reg_infty} ? $Config{reg_infty} : 32767;
261 $reg_infty_m = $reg_infty - 1; $reg_infty_p = $reg_infty + 1;
263 # As well as failing if the pattern matches do unexpected things, the
264 # next three tests will fail if you should have picked up a lower-than-
265 # default value for $reg_infty from Config.pm, but have not.
268 print "not " if eval q(('aaa' =~ /(a{1,$reg_infty_m})/)[0] ne 'aaa') || $@;
272 print "not " if eval q(('a' x $reg_infty_m) !~ /a{$reg_infty_m}/) || $@;
276 print "not " if eval q(('a' x ($reg_infty_m - 1)) =~ /a{$reg_infty_m}/) || $@;
280 eval "'aaa' =~ /a{1,$reg_infty}/";
281 print "not " if $@ !~ m%^\QQuantifier in {,} bigger than%;
284 eval "'aaa' =~ /a{1,$reg_infty_p}/";
286 if $@ !~ m%^\QQuantifier in {,} bigger than%;
290 # Poke a couple more parse failures
292 $context = 'x' x 256;
293 eval qq("${context}y" =~ /(?<=$context)y/);
294 print "not " if $@ !~ m%^\QLookbehind longer than 255 not%;
302 for $l (125, 140, 250, 270, 300000, 30) { # Ordered to free memory
304 print "# length=$l\nnot " unless "ba$a=" =~ /a$a=/;
308 print "not " if "b$a=" =~ /a$a=/;
313 # 20000 nodes, each taking 3 words per string, and 1 per branch
314 $long_constant_len = join '|', 12120 .. 32645;
315 $long_var_len = join '|', 8120 .. 28645;
316 %ans = ( 'ax13876y25677lbc' => 1,
317 'ax13876y25677mcb' => 0, # not b.
318 'ax13876y35677nbc' => 0, # Num too big
319 'ax13876y25677y21378obc' => 1,
320 'ax13876y25677y21378zbc' => 0, # Not followed by [k-o]
321 'ax13876y25677y21378y21378kbc' => 1,
322 'ax13876y25677y21378y21378kcb' => 0, # Not b.
323 'ax13876y25677y21378y21378y21378kbc' => 0, # 5 runs
327 print "# const-len `$_' not => $ans{$_}\nnot "
328 if $ans{$_} xor /a(?=([yx]($long_constant_len)){2,4}[k-o]).*b./o;
331 print "# var-len `$_' not => $ans{$_}\nnot "
332 if $ans{$_} xor /a(?=([yx]($long_var_len)){2,4}[k-o]).*b./o;
337 $_ = " a (bla()) and x(y b((l)u((e))) and b(l(e)e)e";
338 $expect = "(bla()) ((l)u((e))) (l(e)e)";
344 (?{ $c = 1 }) # Initialize
346 (?(?{ $c == 0 }) # PREVIOUS iteration was OK, stop the loop
348 ) # Fail: will unwind one iteration back
351 [^()]+ # Match a big chunk
354 ) # Do not try to match subchunks
362 )+ # This may not match with different subblocks
367 ) # Otherwise the chunk 1 may succeed with $c>0
372 push @ans, $res while $res = matchit;
374 print "# ans='@ans'\n# expect='$expect'\nnot " if "@ans" ne "1 1 1";
380 print "# ans='@ans'\n# expect='$expect'\nnot " if "@ans" ne $expect;
384 print "not " unless "abc" =~ /^(??{"a"})b/;
389 $matched = qr/\((?:(?>[^()]+)|(??{$matched}))*\)/;
392 push(@ans, $res), push(@ans1, $&) while $res = m/$matched/g;
394 print "# ans='@ans'\n# expect='$expect'\nnot " if "@ans" ne "1 1 1";
398 print "# ans1='@ans1'\n# expect='$expect'\nnot " if "@ans1" ne $expect;
404 print "# ans='@ans'\n# expect='$expect'\nnot " if "@ans" ne $expect;
408 @ans = ('a/b' =~ m%(.*/)?(.*)%); # Stack may be bad
409 print "not " if "@ans" ne 'a/ b';
413 $code = '{$blah = 45}';
416 print "not " unless $@ and $@ =~ /not allowed at runtime/ and $blah == 12;
420 for $code ('{$blah = 45}','=xx') {
422 $res = eval { "xx" =~ /(?$code)/o };
423 if ($code eq '=xx') {
424 print "#'$@','$res','$blah'\nnot " unless not $@ and $res;
426 print "#'$@','$res','$blah'\nnot " unless $@ and $@ =~ /not allowed at runtime/ and $blah == 12;
432 $code = '{$blah = 45}';
435 print "not " if $blah != 45;
441 print "not " if $blah != 45;
447 print "not " unless pos($x) == 2;
452 print "not " unless pos($x) == 2;
462 print "not " unless f(pos($x)) == 4;
467 'foot' =~ /foo(?{$x = 12; 75})[t]/;
468 print "not " unless $^R eq '75';
473 'foot' =~ /foo(?{$x = 12; 75})[xy]/;
474 print "not " unless $^R eq '67' and $x eq '12';
479 'foot' =~ /foo(?{ $^R + 12 })((?{ $x = 12; $^R + 17 })[xy])?/;
480 print "not " unless $^R eq '79' and $x eq '12';
484 print "not " unless qr/\b\v$/i eq '(?i-xsm:\b\v$)';
488 print "not " unless qr/\b\v$/s eq '(?s-xim:\b\v$)';
492 print "not " unless qr/\b\v$/m eq '(?m-xis:\b\v$)';
496 print "not " unless qr/\b\v$/x eq '(?x-ism:\b\v$)';
500 print "not " unless qr/\b\v$/xism eq '(?msix:\b\v$)';
504 print "not " unless qr/\b\v$/ eq '(?-xism:\b\v$)';
509 foreach $ans ('', 'c') {
510 /(?<=(?=a)..)((?=c)|.)/g;
511 print "# \$1 ='$1'\n# \$ans='$ans'\nnot " unless $1 eq $ans;
517 foreach $ans ('', 'a', '') {
519 print "# \$& ='$&'\n# \$ans='$ans'\nnot " unless $& eq $ans;
525 my($v,$a,$b,$res) = @_;
527 print "not " unless $res eq $v;
531 prefixify('/a/b/lib/arch', "/a/b/lib", 'X/lib', 'X/lib/arch');
532 prefixify('/a/b/man/arch', "/a/b/man", 'X/man', 'X/man/arch');
536 print "not " unless $1 and /$1/;
540 if ($::running_as_thread) {
541 print "not ok $test # TODO & SKIP: croaks in 5.10 when threaded\n";
547 print "not " unless $b eq '9';
553 local $TODO = $::running_as_thread;
563 local $TODO = $::running_as_thread;
571 my $lex_res = ($lex_b =~ qr/$lex_b(?{ $lex_c = $lex_a++ })/);
572 print "not " unless $lex_res eq '1';
575 print "not " unless $lex_a eq '44';
578 print "not " unless $lex_c eq '43';
584 $match = eval { /$a$c$a/ };
585 # FIXME - split this one. That would require removing a lot of hard coded
587 local $TODO = $::running_as_thread;
588 ok($b eq '14' and $@ =~ /Eval-group not allowed/ and not $match);
596 my $lex_res = ($lex_b =~ qr/17(?{ $lex_c = $lex_a++ })/);
597 print "not " unless $lex_res eq '1';
600 print "not " unless $lex_a eq '44';
603 print "not " unless $lex_c eq '43';
612 '' =~ /(?{ $c = 4 })/;
613 print "not " unless $c == 4;
617 print "not " unless $c == 3;
622 my $warn_pat = shift;
623 return sub { print "not " unless $_[0] =~ /$warn_pat/ }
627 my ($warn_pat, $code) = @_;
629 eval 'BEGIN { use warnings; $SIG{__WARN__} = $warn_pat };' . $code;
636 my $warn_pat = shift;
637 return sub { must_warn(must_warn_pat($warn_pat)) }
640 my $for_future = make_must_warn('reserved for future extensions');
642 &$for_future('q(a:[b]:) =~ /[x[:foo:]]/');
644 #&$for_future('q(a=[b]=) =~ /[x[=foo=]]/');
645 print "ok $test\n"; $test++; # now a fatal croak
647 #&$for_future('q(a.[b].) =~ /[x[.foo.]]/');
648 print "ok $test\n"; $test++; # now a fatal croak
650 # test if failure of patterns returns empty list
673 print "not " if $#+ != 0 or $#- != 0;
677 print "not " if $+[0] != 2 or $-[0] != 1;
682 if defined $+[1] or defined $-[1] or defined $+[2] or defined $-[2];
687 print "not " if $#+ != 2 or $#- != 2;
691 print "not " if $+[0] != 3 or $-[0] != 0;
695 print "not " if $+[1] != 2 or $-[1] != 1;
699 print "not " if $+[2] != 3 or $-[2] != 2;
704 if defined $+[3] or defined $-[3] or defined $+[4] or defined $-[4];
709 print "not " if $#+ != 3 or $#- != 3;
713 print "not " if $+[0] != 3 or $-[0] != 0;
717 print "not " if $+[1] != 2 or $-[1] != 1;
721 print "not " if $+[3] != 3 or $-[3] != 2;
726 if defined $+[2] or defined $-[2] or defined $+[4] or defined $-[4];
731 print "not " if $#+ != 1 or $#- != 1;
735 print "not " if $+[0] != 2 or $-[0] != 0;
739 print "not " if $+[1] != 2 or $-[1] != 1;
744 if defined $+[2] or defined $-[2] or defined $+[3] or defined $-[3];
748 eval { $+[0] = 13; };
750 if $@ !~ /^Modification of a read-only value attempted/;
754 eval { $-[0] = 13; };
756 if $@ !~ /^Modification of a read-only value attempted/;
760 eval { @+ = (7, 6, 5); };
762 if $@ !~ /^Modification of a read-only value attempted/;
766 eval { @- = qw(foo bar); };
768 if $@ !~ /^Modification of a read-only value attempted/;
773 print "#$#-..$#+\nnot " if $#+ != 2 or $#- != 1;
780 print "not " unless "@a" eq "a a";
787 print "not " if $str =~ /^\G/;
791 print "not " if $str =~ /^.\G/;
795 print "not " unless $str =~ /^..\G/;
799 print "not " if $str =~ /^...\G/;
804 local $TODO = $::running_as_thread;
805 ok($str =~ /.\G./ and $& eq 'bc');
808 print "not " unless $str =~ /\G../ and $& eq 'cd';
812 undef $foo; undef $bar;
813 print "#'$str','$foo','$bar'\nnot "
814 unless $str =~ /b(?{$foo = $_; $bar = pos})c/
815 and $foo eq 'abcde' and $bar eq 2;
819 undef $foo; undef $bar;
821 print "#'$str','$foo','$bar'\nnot "
822 unless $str =~ /b(?{$foo = $_; $bar = pos})c/g
823 and $foo eq 'abcde' and $bar eq 2 and pos $str eq 3;
829 undef $foo; undef $bar;
830 print "#'$str','$foo','$bar'\nnot "
831 unless /b(?{$foo = $_; $bar = pos})c/
832 and $foo eq 'abcde' and $bar eq 2;
836 undef $foo; undef $bar;
837 print "#'$str','$foo','$bar'\nnot "
838 unless /b(?{$foo = $_; $bar = pos})c/g
839 and $foo eq 'abcde' and $bar eq 2 and pos eq 3;
843 undef $foo; undef $bar;
845 1 while /b(?{$foo = $_; $bar = pos})c/g;
846 print "#'$str','$foo','$bar'\nnot "
847 unless $foo eq 'abcde' and $bar eq 2 and not defined pos;
851 undef $foo; undef $bar;
853 print "#'$str','$foo','$bar','$_'\nnot "
854 unless s/b(?{$foo = $_; $bar = pos})c/x/g and $foo eq 'abcde|abcde'
855 and $bar eq 8 and $_ eq 'axde|axde';
862 @dummy = /([ace]).(?{push @res, $1,$2})([ce])(?{push @res, $1,$2})/g;
863 @res = map {defined $_ ? "'$_'" : 'undef'} @res;
865 print "#'@res' '$_'\nnot "
866 unless "@res" eq "'a' undef 'a' 'c' 'e' undef 'a' undef 'a' 'c'";
871 @dummy = /([ace]).(?{push @res, $`,$&,$'})([ce])(?{push @res, $`,$&,$'})/g;
872 @res = map {defined $_ ? "'$_'" : 'undef'} @res;
874 print "#'@res' '$_'\nnot "
876 "'' 'ab' 'cde|abcde' " .
877 "'' 'abc' 'de|abcde' " .
878 "'abcd' 'e|' 'abcde' " .
879 "'abcde|' 'ab' 'cde' " .
880 "'abcde|' 'abc' 'de'" ;
884 #Some more \G anchor checks
885 $foo='aabbccddeeffgg';
891 local $TODO = $::running_as_thread;
898 local $TODO = $::running_as_thread;
905 local $TODO = $::running_as_thread;
910 local $TODO = $::running_as_thread;
917 print "not " unless($1 eq 'aa');
922 print "not " unless($1 eq 'bb');
928 print "not " unless($1 eq 'cd');
934 print "not " unless('123||x|123|' eq join '|', @res);
938 # see if matching against temporaries (created via pp_helem()) is safe
939 { foo => "ok $test\n".$^X }->{foo} =~ /^(.*)\n/g;
943 # See if $i work inside (?{}) in the presense of saved substrings and
947 s/(\w)(?{push @b, $1})/,$1,/g for @a;
949 print "# \@b='@b', expect 'f o o b a r'\nnot " unless("@b" eq "f o o b a r");
953 print "not " unless("@a" eq ",f,,o,,o, ,b,,a,,r,");
958 { (?> [^{}]+ | (??{ $brackets }) )* }
962 print "ok $test\n"; # Did we survive?
965 "something { long { and } hairy" =~ $brackets;
966 print "ok $test\n"; # Did we survive?
969 "something { long { and } hairy" =~ m/((??{ $brackets }))/;
970 print "not " unless $1 eq "{ and }";
976 m/^-.*bb/mg and print "not ";
982 $text =~ /\GXb*X/g and print 'not ';
986 $text = "xA\n" x 500;
987 $text =~ /^\s*A/m and print 'not ';
992 @res = ($text =~ /.*?(b).*?\b/g);
993 "@res" eq 'b b' or print 'not ';
1000 @c = grep(/[^\s]/,@a);
1001 print "not " if "@b" ne "@c";
1006 @c = grep(/[\S]/,@a);
1007 print "not " if "@b" ne "@c";
1012 @c = grep(/[^\S]/,@a);
1013 print "not " if "@b" ne "@c";
1018 @c = grep(/[\s]/,@a);
1019 print "not " if "@b" ne "@c";
1024 @c = grep(/[^\d]/,@a);
1025 print "not " if "@b" ne "@c";
1030 @c = grep(/[\D]/,@a);
1031 print "not " if "@b" ne "@c";
1036 @c = grep(/[^\D]/,@a);
1037 print "not " if "@b" ne "@c";
1042 @c = grep(/[\d]/,@a);
1043 print "not " if "@b" ne "@c";
1048 @c = grep(/[^\w]/,@a);
1049 print "not " if "@b" ne "@c";
1054 @c = grep(/[\W]/,@a);
1055 print "not " if "@b" ne "@c";
1060 @c = grep(/[^\W]/,@a);
1061 print "not " if "@b" ne "@c";
1066 @c = grep(/[\w]/,@a);
1069 # see if backtracking optimization works correctly
1070 "\n\n" =~ /\n $ \n/x or print "not ";
1074 "\n\n" =~ /\n* $ \n/x or print "not ";
1078 "\n\n" =~ /\n+ $ \n/x or print "not ";
1082 [] =~ /^ARRAY/ or print "# [] \nnot ";
1089 use overload '""' => sub { 'Object S' };
1090 sub new { bless [] }
1095 $a and $a =~ /^Object\sS/ or print "# '$a' \nnot ";
1099 # test result of match used as match (!)
1100 'a1b' =~ ('xyz' =~ /y/) and $` eq 'a' or print "not ";
1104 'a1b' =~ ('xyz' =~ /t/) and $` eq 'a' or print "not ";
1110 local $SIG{__WARN__} = sub { $w = 1 };
1112 $w = 1 if ("1\n" x 102) =~ /^\s*\n/m;
1114 print $w ? "not " : "", "ok $test\n";
1117 my %space = ( spc => " ",
1122 # There's no \v but the vertical tabulator seems miraculously
1123 # be 11 both in ASCII and EBCDIC.
1127 my @space0 = sort grep { $space{$_} =~ /\s/ } keys %space;
1128 my @space1 = sort grep { $space{$_} =~ /[[:space:]]/ } keys %space;
1129 my @space2 = sort grep { $space{$_} =~ /[[:blank:]]/ } keys %space;
1131 print "not " unless "@space0" eq "cr ff lf spc tab";
1132 print "ok $test # @space0\n";
1135 print "not " unless "@space1" eq "cr ff lf spc tab vt";
1136 print "ok $test # @space1\n";
1139 print "not " unless "@space2" eq "spc tab";
1140 print "ok $test # @space2\n";
1143 # bugid 20001021.005 - this caused a SEGV
1144 print "not " unless undef =~ /^([^\/]*)(.*)$/;
1148 # bugid 20000731.001
1150 print "not " unless "A \x{263a} B z C" =~ /A . B (??{ "z" }) C/;
1154 my $ordA = ord('A');
1157 if (/(.)(\C)(\C)(.)/) {
1162 print "not ok 233\n";
1164 if ($ordA == 65) { # ASCII (or equivalent), should be UTF-8
1168 print "not ok 234\n";
1173 print "not ok 235\n";
1175 } elsif ($ordA == 193) { # EBCDIC (or equivalent), should be UTF-EBCDIC
1179 print "not ok 234\n";
1184 print "not ok 235\n";
1188 print "not ok $_ # ord('A') == $ordA\n";
1194 print "not ok 236\n";
1198 print "not ok $_\n";
1204 # currently \C are still tagged as UTF-8
1209 print "not ok 238\n";
1211 } elsif ($ordA == 193) {
1215 print "not ok 238\n";
1218 print "not ok 238 # ord('A') == $ordA\n";
1222 print "not ok $_\n";
1227 # currently \C are still tagged as UTF-8
1232 print "not ok 240\n";
1234 } elsif ($ordA == 193) {
1238 print "not ok 240\n";
1241 print "not ok 240 # ord('A') == $ordA\n";
1245 print "not ok $_\n";
1250 # japhy -- added 03/03/2001
1251 () = (my $str = "abc") =~ /(...)/;
1253 print "not " if $1 ne "abc";
1257 # The 242 and 243 go with the 244 and 245.
1258 # The trick is that in EBCDIC the explicit numeric range should match
1259 # (as also in non-EBCDIC) but the explicit alphabetic range should not match.
1261 if ("\x8e" =~ /[\x89-\x91]/) {
1264 print "not ok 242\n";
1267 if ("\xce" =~ /[\xc9-\xd1]/) {
1270 print "not ok 243\n";
1273 # In most places these tests would succeed since \x8e does not
1274 # in most character sets match 'i' or 'j' nor would \xce match
1275 # 'I' or 'J', but strictly speaking these tests are here for
1276 # the good of EBCDIC, so let's test these only there.
1277 if (ord('i') == 0x89 && ord('J') == 0xd1) { # EBCDIC
1278 if ("\x8e" !~ /[i-j]/) {
1281 print "not ok 244\n";
1283 if ("\xce" !~ /[I-J]/) {
1286 print "not ok 245\n";
1290 print "ok $_ # Skip: only in EBCDIC\n";
1294 print "not " unless "\x{ab}" =~ /\x{ab}/;
1297 print "not " unless "\x{abcd}" =~ /\x{abcd}/;
1301 # bug id 20001008.001
1304 my @x = ("stra\337e 138","stra\337e 138");
1306 s/(\d+)\s*([\w\-]+)/$1 . uc $2/e;
1307 my($latin) = /^(.+)(?:\s+\d)/;
1308 print $latin eq "stra\337e" ? "ok $test\n" : # 248,249
1309 "#latin[$latin]\nnot ok $test\n";
1311 $latin =~ s/stra\337e/straße/; # \303\237 after the 2nd a
1312 use utf8; # needed for the raw UTF-8
1313 $latin =~ s!(s)tr(?:aß|s+e)!$1tr.!; # \303\237 after the a
1318 print "not " unless "ba\xd4c" =~ /([a\xd4]+)/ && $1 eq "a\xd4";
1321 print "not " unless "ba\xd4c" =~ /([a\xd4]+)/ && $1 eq "a\x{d4}";
1324 print "not " unless "ba\x{d4}c" =~ /([a\xd4]+)/ && $1 eq "a\x{d4}";
1327 print "not " unless "ba\x{d4}c" =~ /([a\xd4]+)/ && $1 eq "a\xd4";
1330 print "not " unless "ba\xd4c" =~ /([a\x{d4}]+)/ && $1 eq "a\xd4";
1333 print "not " unless "ba\xd4c" =~ /([a\x{d4}]+)/ && $1 eq "a\x{d4}";
1336 print "not " unless "ba\x{d4}c" =~ /([a\x{d4}]+)/ && $1 eq "a\x{d4}";
1339 print "not " unless "ba\x{d4}c" =~ /([a\x{d4}]+)/ && $1 eq "a\xd4";
1344 # the first half of 20001028.003
1347 my ($Y) = $X =~ /(.*)/;
1348 print "not " unless $Y eq v1448 && length($Y) == 1;
1355 my $X = "Szab\x{f3},Bal\x{e1}zs";
1357 $Y =~ s/(B)/$1/ for 0..3;
1358 print "not " unless $Y eq $X && $X eq "Szab\x{f3},Bal\x{e1}zs";
1363 # the second half of 20001028.003
1366 $X =~ s/^/chr(1488)/e;
1367 print "not " unless length $X == 1 && ord($X) == 1488;
1378 print "not " unless $x eq "\x{100}B" && length($x) == 2;
1383 # bug id 20001230.002
1385 print "not " unless "École" =~ /^\C\C(.)/ && $1 eq 'c';
1388 print "not " unless "École" =~ /^\C\C(c)/;
1393 $test = 264; # till 575
1395 use charnames ":full";
1397 # This is far from complete testing, there are dozens of character
1398 # classes in Unicode. The mixing of literals and \N{...} is
1399 # intentional so that in non-Latin-1 places we test the native
1400 # characters, not the Unicode code points.
1404 "\N{CYRILLIC SMALL LETTER A}" => 'Ll',
1406 "\N{GREEK CAPITAL LETTER ALPHA}" => 'Lu',
1407 "\N{HIRAGANA LETTER SMALL A}" => 'Lo',
1408 "\N{COMBINING GRAVE ACCENT}" => 'Mn',
1410 "\N{ARABIC-INDIC DIGIT ZERO}" => 'Nd',
1417 for my $char (map { s/^\S+ //; $_ }
1418 sort map { sprintf("%06x", ord($_))." $_" } keys %s) {
1419 my $class = $s{$char};
1420 my $code = sprintf("%06x", ord($char));
1421 printf "#\n# 0x$code\n#\n";
1422 print "# IsAlpha\n";
1423 if ($class =~ /^[LM]/) {
1424 print "not " unless $char =~ /\p{IsAlpha}/;
1425 print "ok $test\n"; $test++;
1426 print "not " if $char =~ /\P{IsAlpha}/;
1427 print "ok $test\n"; $test++;
1429 print "not " if $char =~ /\p{IsAlpha}/;
1430 print "ok $test\n"; $test++;
1431 print "not " unless $char =~ /\P{IsAlpha}/;
1432 print "ok $test\n"; $test++;
1434 print "# IsAlnum\n";
1435 if ($class =~ /^[LMN]/ && $char ne "_") {
1436 print "not " unless $char =~ /\p{IsAlnum}/;
1437 print "ok $test\n"; $test++;
1438 print "not " if $char =~ /\P{IsAlnum}/;
1439 print "ok $test\n"; $test++;
1441 print "not " if $char =~ /\p{IsAlnum}/;
1442 print "ok $test\n"; $test++;
1443 print "not " unless $char =~ /\P{IsAlnum}/;
1444 print "ok $test\n"; $test++;
1446 print "# IsASCII\n";
1447 if (ord("A") == 193) {
1448 print "ok $test # Skip: in EBCDIC\n"; $test++;
1449 print "ok $test # Skip: in EBCDIC\n"; $test++;
1451 if ($code le '00007f') {
1452 print "not " unless $char =~ /\p{IsASCII}/;
1453 print "ok $test\n"; $test++;
1454 print "not " if $char =~ /\P{IsASCII}/;
1455 print "ok $test\n"; $test++;
1457 print "not " if $char =~ /\p{IsASCII}/;
1458 print "ok $test\n"; $test++;
1459 print "not " unless $char =~ /\P{IsASCII}/;
1460 print "ok $test\n"; $test++;
1463 print "# IsCntrl\n";
1464 if ($class =~ /^C/) {
1465 print "not " unless $char =~ /\p{IsCntrl}/;
1466 print "ok $test\n"; $test++;
1467 print "not " if $char =~ /\P{IsCntrl}/;
1468 print "ok $test\n"; $test++;
1470 print "not " if $char =~ /\p{IsCntrl}/;
1471 print "ok $test\n"; $test++;
1472 print "not " unless $char =~ /\P{IsCntrl}/;
1473 print "ok $test\n"; $test++;
1475 print "# IsBlank\n";
1476 if ($class =~ /^Z[lp]/ || $char eq " ") {
1477 print "not " unless $char =~ /\p{IsBlank}/;
1478 print "ok $test\n"; $test++;
1479 print "not " if $char =~ /\P{IsBlank}/;
1480 print "ok $test\n"; $test++;
1482 print "not " if $char =~ /\p{IsBlank}/;
1483 print "ok $test\n"; $test++;
1484 print "not " unless $char =~ /\P{IsBlank}/;
1485 print "ok $test\n"; $test++;
1487 print "# IsDigit\n";
1488 if ($class =~ /^Nd$/) {
1489 print "not " unless $char =~ /\p{IsDigit}/;
1490 print "ok $test\n"; $test++;
1491 print "not " if $char =~ /\P{IsDigit}/;
1492 print "ok $test\n"; $test++;
1494 print "not " if $char =~ /\p{IsDigit}/;
1495 print "ok $test\n"; $test++;
1496 print "not " unless $char =~ /\P{IsDigit}/;
1497 print "ok $test\n"; $test++;
1499 print "# IsGraph\n";
1500 if ($class =~ /^([LMNPS])|Co/) {
1501 print "not " unless $char =~ /\p{IsGraph}/;
1502 print "ok $test\n"; $test++;
1503 print "not " if $char =~ /\P{IsGraph}/;
1504 print "ok $test\n"; $test++;
1506 print "not " if $char =~ /\p{IsGraph}/;
1507 print "ok $test\n"; $test++;
1508 print "not " unless $char =~ /\P{IsGraph}/;
1509 print "ok $test\n"; $test++;
1511 print "# IsLower\n";
1512 if ($class =~ /^Ll$/) {
1513 print "not " unless $char =~ /\p{IsLower}/;
1514 print "ok $test\n"; $test++;
1515 print "not " if $char =~ /\P{IsLower}/;
1516 print "ok $test\n"; $test++;
1518 print "not " if $char =~ /\p{IsLower}/;
1519 print "ok $test\n"; $test++;
1520 print "not " unless $char =~ /\P{IsLower}/;
1521 print "ok $test\n"; $test++;
1523 print "# IsPrint\n";
1524 if ($class =~ /^([LMNPS])|Co|Zs/) {
1525 print "not " unless $char =~ /\p{IsPrint}/;
1526 print "ok $test\n"; $test++;
1527 print "not " if $char =~ /\P{IsPrint}/;
1528 print "ok $test\n"; $test++;
1530 print "not " if $char =~ /\p{IsPrint}/;
1531 print "ok $test\n"; $test++;
1532 print "not " unless $char =~ /\P{IsPrint}/;
1533 print "ok $test\n"; $test++;
1535 print "# IsPunct\n";
1536 if ($class =~ /^P/ || $char eq "_") {
1537 print "not " unless $char =~ /\p{IsPunct}/;
1538 print "ok $test\n"; $test++;
1539 print "not " if $char =~ /\P{IsPunct}/;
1540 print "ok $test\n"; $test++;
1542 print "not " if $char =~ /\p{IsPunct}/;
1543 print "ok $test\n"; $test++;
1544 print "not " unless $char =~ /\P{IsPunct}/;
1545 print "ok $test\n"; $test++;
1547 print "# IsSpace\n";
1548 if ($class =~ /^Z/ || ($code =~ /^(0009|000A|000B|000C|000D)$/)) {
1549 print "not " unless $char =~ /\p{IsSpace}/;
1550 print "ok $test\n"; $test++;
1551 print "not " if $char =~ /\P{IsSpace}/;
1552 print "ok $test\n"; $test++;
1554 print "not " if $char =~ /\p{IsSpace}/;
1555 print "ok $test\n"; $test++;
1556 print "not " unless $char =~ /\P{IsSpace}/;
1557 print "ok $test\n"; $test++;
1559 print "# IsUpper\n";
1560 if ($class =~ /^L[ut]/) {
1561 print "not " unless $char =~ /\p{IsUpper}/;
1562 print "ok $test\n"; $test++;
1563 print "not " if $char =~ /\P{IsUpper}/;
1564 print "ok $test\n"; $test++;
1566 print "not " if $char =~ /\p{IsUpper}/;
1567 print "ok $test\n"; $test++;
1568 print "not " unless $char =~ /\P{IsUpper}/;
1569 print "ok $test\n"; $test++;
1572 if ($class =~ /^[LMN]/ || $char eq "_") {
1573 print "not " unless $char =~ /\p{IsWord}/;
1574 print "ok $test\n"; $test++;
1575 print "not " if $char =~ /\P{IsWord}/;
1576 print "ok $test\n"; $test++;
1578 print "not " if $char =~ /\p{IsWord}/;
1579 print "ok $test\n"; $test++;
1580 print "not " unless $char =~ /\P{IsWord}/;
1581 print "ok $test\n"; $test++;
1587 $_ = "abc\x{100}\x{200}\x{300}\x{380}\x{400}defg";
1589 if (/(.\x{300})./) {
1592 print "not " unless $` eq "abc\x{100}" && length($`) == 4;
1595 print "not " unless $& eq "\x{200}\x{300}\x{380}" && length($&) == 3;
1598 print "not " unless $' eq "\x{400}defg" && length($') == 5;
1601 print "not " unless $1 eq "\x{200}\x{300}" && length($1) == 2;
1604 for (576..580) { print "not ok $_\n" }
1609 # bug id 20010306.008
1612 # The original bug report had 'no utf8' here but that was irrelevant.
1613 $a =~ m/\w/; # used to core dump
1621 # bugid 20010410.006
1623 '/(.*?)\{(.*?)\}/csg',
1624 '/(.*?)\{(.*?)\}/cg',
1625 '/(.*?)\{(.*?)\}/sg',
1626 '/(.*?)\{(.*?)\}/g',
1627 '/(.+?)\{(.+?)\}/csg',
1633 $input = "a{b}c{d}";
1635 while (eval \$input =~ $rx) {
1636 print "# \\\$1 = '\$1' \\\$2 = '\$2'\n";
1640 print "not " unless $i == 2;
1641 print "ok " . $test++ . "\n";
1646 # from Robin Houston
1648 my $x = "\x{10FFFD}";
1650 print "not " unless ord($x) == 0x10FFFD && length($x) == 1;
1657 print "not " if $x =~ /[\x80-\xff]/;
1660 print "not " if $x =~ /[\x80-\x{100}]/;
1663 print "not " if $x =~ /[\x{100}]/;
1666 print "not " if $x =~ /\p{InLatin1Supplement}/;
1669 print "not " unless $x =~ /\P{InLatin1Supplement}/;
1672 print "not " if $x =~ /\p{InLatinExtendedA}/;
1675 print "not " unless $x =~ /\P{InLatinExtendedA}/;
1682 print "not " unless $x =~ /[\x80-\xff]/;
1685 print "not " unless $x =~ /[\x80-\x{100}]/;
1688 print "not " if $x =~ /[\x{100}]/;
1691 print "not " unless $x =~ /\p{InLatin1Supplement}/;
1694 print "not " if $x =~ /\P{InLatin1Supplement}/;
1697 print "not " if $x =~ /\p{InLatinExtendedA}/;
1700 print "not " unless $x =~ /\P{InLatinExtendedA}/;
1707 print "not " unless $x =~ /[\x80-\xff]/;
1710 print "not " unless $x =~ /[\x80-\x{100}]/;
1713 print "not " if $x =~ /[\x{100}]/;
1716 # the next two tests must be ignored on EBCDIC
1717 print "not " unless $x =~ /\p{InLatin1Supplement}/ or ord("A") == 193;
1720 print "not " if $x =~ /\P{InLatin1Supplement}/ and ord("A") != 193;
1723 print "not " if $x =~ /\p{InLatinExtendedA}/;
1726 print "not " unless $x =~ /\P{InLatinExtendedA}/;
1733 print "not " if $x =~ /[\x80-\xff]/;
1736 print "not " unless $x =~ /[\x80-\x{100}]/;
1739 print "not " unless $x =~ /[\x{100}]/;
1742 print "not " if $x =~ /\p{InLatin1Supplement}/;
1745 print "not " unless $x =~ /\P{InLatin1Supplement}/;
1748 print "not " unless $x =~ /\p{InLatinExtendedA}/;
1751 print "not " if $x =~ /\P{InLatinExtendedA}/;
1759 local $SIG{__WARN__} = sub { $w .= shift };
1763 print "not " if $w !~ /^Useless \(\?c\)/;
1768 print "not " if $w !~ /^Useless \(\?-c\)/;
1773 print "not " if $w !~ /^Useless \(\?g\)/;
1778 print "not " if $w !~ /^Useless \(\?-g\)/;
1783 print "not " if $w !~ /^Useless \(\?o\)/;
1788 print "not " if $w !~ /^Useless \(\?-o\)/;
1791 # now test multi-error regexes
1795 print "not " if $w !~ /^Useless \(\?g\).*\nUseless \(\?-o\)/;
1800 print "not " if $w !~ /^Useless \(\?g\).*\nUseless \(\?-c\)/;
1804 eval 'qr/(?o-cg)/'; # (?c) means (?g) error won't be thrown
1805 print "not " if $w !~ /^Useless \(\?o\).*\nUseless \(\?-c\)/;
1810 print "not " if $w !~ /^Useless \(\?o\).*\nUseless \(\?g\).*\nUseless \(\?c\)/;
1814 # More Unicode "class" tests
1817 use charnames ':full';
1819 print "not " unless "\N{LATIN CAPITAL LETTER A}" =~ /\p{InBasicLatin}/;
1822 print "not " unless "\N{LATIN CAPITAL LETTER A WITH GRAVE}" =~ /\p{InLatin1Supplement}/;
1825 print "not " unless "\N{LATIN CAPITAL LETTER A WITH MACRON}" =~ /\p{InLatinExtendedA}/;
1828 print "not " unless "\N{LATIN SMALL LETTER B WITH STROKE}" =~ /\p{InLatinExtendedB}/;
1831 print "not " unless "\N{KATAKANA LETTER SMALL A}" =~ /\p{InKatakana}/;
1837 eval <<"EOT"; die if $@;
1842 /x && print "ok 631\n";
1845 eval <<"EOT"; die if $@;
1850 /x && print "ok 632\n";
1854 sub test_o { $_[0] =~/$_[1]/o; return $1}
1855 if(test_o('abc','(.)..') eq 'a') {
1858 print "not ok 633\n";
1860 if(test_o('abc','..(.)') eq 'a') {
1863 print "not ok 634\n";
1866 # 635..639: ID 20010619.003 (only the space character is
1867 # supposed to be [:print:], not the whole isprint()).
1869 print "not " if "\n" =~ /[[:print:]]/;
1872 print "not " if "\t" =~ /[[:print:]]/;
1875 # Amazingly vertical tabulator is the same in ASCII and EBCDIC.
1876 print "not " if "\014" =~ /[[:print:]]/;
1879 print "not " if "\r" =~ /[[:print:]]/;
1882 print "not " unless " " =~ /[[:print:]]/;
1886 ## Test basic $^N usage outside of a regex
1889 $T="ok 640\n";if ($x =~ /cde/ and not defined $^N) {print $T} else {print "not $T"};
1890 $T="ok 641\n";if ($x =~ /(cde)/ and $^N eq "cde") {print $T} else {print "not $T"};
1891 $T="ok 642\n";if ($x =~ /(c)(d)(e)/ and $^N eq "e") {print $T} else {print "not $T"};
1892 $T="ok 643\n";if ($x =~ /(c(d)e)/ and $^N eq "cde") {print $T} else {print "not $T"};
1893 $T="ok 644\n";if ($x =~ /(foo)|(c(d)e)/ and $^N eq "cde") {print $T} else {print "not $T"};
1894 $T="ok 645\n";if ($x =~ /(c(d)e)|(foo)/ and $^N eq "cde") {print $T} else {print "not $T"};
1895 $T="ok 646\n";if ($x =~ /(c(d)e)|(abc)/ and $^N eq "abc") {print $T} else {print "not $T"};
1896 $T="ok 647\n";if ($x =~ /(c(d)e)|(abc)x/ and $^N eq "cde") {print $T} else {print "not $T"};
1897 $T="ok 648\n";if ($x =~ /(c(d)e)(abc)?/ and $^N eq "cde") {print $T} else {print "not $T"};
1898 $T="ok 649\n";if ($x =~ /(?:c(d)e)/ and $^N eq "d" ) {print $T} else {print "not $T"};
1899 $T="ok 650\n";if ($x =~ /(?:c(d)e)(?:f)/ and $^N eq "d" ) {print $T} else {print "not $T"};
1900 $T="ok 651\n";if ($x =~ /(?:([abc])|([def]))*/ and $^N eq "f" ){print $T} else {print "not $T"};
1901 $T="ok 652\n";if ($x =~ /(?:([ace])|([bdf]))*/ and $^N eq "f" ){print $T} else {print "not $T"};
1902 $T="ok 653\n";if ($x =~ /(([ace])|([bd]))*/ and $^N eq "e" ){print $T} else {print "not $T"};
1904 $T="ok 654\n";if($x =~ /(([ace])|([bdf]))*/ and $^N eq "f" ){print $T} else {print "not $T"};
1906 ## test to see if $^N is automatically localized -- it should now
1907 ## have the value set in test 653
1908 $T="ok 655\n";if ($^N eq "e" ){print $T} else {print "not $T"};
1911 ## Now test inside (?{...})
1913 $T="ok 656\n";if ($x =~ /a([abc])(?{$y=$^N})c/ and $y eq "b" ){print $T} else {print "not $T"};
1914 $T="ok 657\n";if ($x =~ /a([abc]+)(?{$y=$^N})d/ and $y eq "bc"){print $T} else {print "not $T"};
1915 $T="ok 658\n";if ($x =~ /a([abcdefg]+)(?{$y=$^N})d/ and $y eq "bc"){print $T} else {print "not $T"};
1916 $T="ok 659\n";if ($x =~ /(a([abcdefg]+)(?{$y=$^N})d)(?{$z=$^N})e/ and $y eq "bc" and $z eq "abcd")
1917 {print $T} else {print "not $T"};
1918 $T="ok 660\n";if ($x =~ /(a([abcdefg]+)(?{$y=$^N})de)(?{$z=$^N})/ and $y eq "bc" and $z eq "abcde")
1919 {print $T} else {print "not $T"};
1921 # Test the Unicode script classes
1923 print "not " unless chr(0x100) =~ /\p{IsLatin}/; # outside Latin-1
1926 print "not " unless chr(0x212b) =~ /\p{IsLatin}/; # Angstrom sign, very outside
1929 print "not " unless chr(0x5d0) =~ /\p{IsHebrew}/; # inside InHebrew
1932 print "not " unless chr(0xfb4f) =~ /\p{IsHebrew}/; # outside InHebrew
1935 # # singleton (not in a range, this test must be ignored on EBCDIC)
1936 # print "not " unless chr(0xb5) =~ /\p{IsGreek}/ or ord("A") == 193;
1938 print "ok 665 # 0xb5 moved from Greek to Common with Unicode 4.0.1\n";
1940 print "not " unless chr(0x37a) =~ /\p{IsGreek}/; # singleton
1943 print "not " unless chr(0x386) =~ /\p{IsGreek}/; # singleton
1946 print "not " unless chr(0x387) =~ /\P{IsGreek}/; # not there
1949 print "not " unless chr(0x388) =~ /\p{IsGreek}/; # range
1952 print "not " unless chr(0x38a) =~ /\p{IsGreek}/; # range
1955 print "not " unless chr(0x38b) =~ /\P{IsGreek}/; # not there
1958 print "not " unless chr(0x38c) =~ /\p{IsGreek}/; # singleton
1961 if (ord("A") == 65) {
1963 ## Test [:cntrl:]...
1965 ## Should probably put in tests for all the POSIX stuff, but not sure how to
1966 ## guarantee a specific locale......
1968 $AllBytes = join('', map { chr($_) } 0..255);
1969 ($x = $AllBytes) =~ s/[[:cntrl:]]//g;
1970 if ($x ne join('', map { chr($_) } 0x20..0x7E, 0x80..0xFF)) {
1975 ($x = $AllBytes) =~ s/[^[:cntrl:]]//g;
1976 if ($x ne join('', map { chr($_) } 0..0x1F, 0x7F)) { print "not " }
1979 print "ok $_ # Skip: EBCDIC\n" for 673..674;
1982 # With /s modifier UTF8 chars were interpreted as bytes
1984 my $a = "Hello \x{263A} World";
1986 my @a = ($a =~ /./gs);
1988 print "not " unless $#a == 12;
1992 @a = ("foo\nbar" =~ /./g);
1993 print "ok 676\n" if @a == 6 && "@a" eq "f o o b a r";
1995 @a = ("foo\nbar" =~ /./gs);
1996 print "ok 677\n" if @a == 7 && "@a" eq "f o o \n b a r";
1998 @a = ("foo\nbar" =~ /\C/g);
1999 print "ok 678\n" if @a == 7 && "@a" eq "f o o \n b a r";
2001 @a = ("foo\nbar" =~ /\C/gs);
2002 print "ok 679\n" if @a == 7 && "@a" eq "f o o \n b a r";
2004 @a = ("foo\n\x{100}bar" =~ /./g);
2005 print "ok 680\n" if @a == 7 && "@a" eq "f o o \x{100} b a r";
2007 @a = ("foo\n\x{100}bar" =~ /./gs);
2008 print "ok 681\n" if @a == 8 && "@a" eq "f o o \n \x{100} b a r";
2010 ($a, $b) = map { chr } ord('A') == 65 ? (0xc4, 0x80) : (0x8c, 0x41);
2012 @a = ("foo\n\x{100}bar" =~ /\C/g);
2013 print "ok 682\n" if @a == 9 && "@a" eq "f o o \n $a $b b a r";
2015 @a = ("foo\n\x{100}bar" =~ /\C/gs);
2016 print "ok 683\n" if @a == 9 && "@a" eq "f o o \n $a $b b a r";
2019 # [ID 20010814.004] pos() doesn't work when using =~m// in list context
2021 $a = join ":", /b./gc;
2022 $b = join ":", /a./gc;
2024 print "$a $b $c" eq 'ba:ba ad:ae 10' ? "ok 684\n" : "not ok 684\t# $a $b $c\n";
2028 # [ID 20010407.006] matching utf8 return values from functions does not work
2030 package ID_20010407_006;
2039 $x =~ /(..)/; $y = $1;
2040 print "not " unless length($y) == 2 && $y eq $x;
2043 x =~ /(..)/; $y = $1;
2044 print "not " unless length($y) == 2 && $y eq $x;
2051 # Force scalar context on the patern match
2053 my($ok, $name) = @_;
2054 my $todo = $TODO ? " # TODO $TODO" : '';
2056 printf "%sok %d - %s\n", ($ok ? "" : "not "), $test,
2057 ($name||$Message)."$todo\tLine ".((caller)[2]);
2059 printf "# Failed test at line %d\n", (caller)[2] unless $ok;
2066 # Check that \x## works. 5.6.1 and 5.005_03 fail some of these.
2068 ok ($x =~ /^\x4EE$/, "Check only 2 bytes of hex are matched.");
2071 ok ($x =~ /^\x4Ei$/, "Check that invalid hex digit stops it (2)");
2074 ok ($x =~ /^\x4j$/, "Check that invalid hex digit stops it (1)");
2077 ok ($x =~ /^\xk$/, "Check that invalid hex digit stops it (0)");
2080 ok ($x =~ /^\xx$/, "\\xx isn't to be treated as \\0");
2083 ok ($x =~ /^\xxa$/, "\\xxa isn't to be treated as \\xa");
2086 ok ($x =~ /^\x9_b$/, "\\x9_b isn't to be treated as \\x9b");
2088 print "# and now again in [] ranges\n";
2091 ok ($x =~ /^[\x4EE]{2}$/, "Check only 2 bytes of hex are matched.");
2094 ok ($x =~ /^[\x4Ei]{2}$/, "Check that invalid hex digit stops it (2)");
2097 ok ($x =~ /^[\x4j]{2}$/, "Check that invalid hex digit stops it (1)");
2100 ok ($x =~ /^[\xk]{2}$/, "Check that invalid hex digit stops it (0)");
2103 ok ($x =~ /^[\xx]{2}$/, "\\xx isn't to be treated as \\0");
2106 ok ($x =~ /^[\xxa]{3}$/, "\\xxa isn't to be treated as \\xa");
2109 ok ($x =~ /^[\x9_b]{3}$/, "\\x9_b isn't to be treated as \\x9b");
2114 # Check that \x{##} works. 5.6.1 fails quite a few of these.
2117 ok ($x =~ /^\x{9_b}$/, "\\x{9_b} is to be treated as \\x9b");
2120 ok ($x =~ /^\x{9_b}y$/, "\\x{9_b} is to be treated as \\x9b (again)");
2123 ok ($x =~ /^\x{9b_}y$/, "\\x{9b_} is to be treated as \\x9b");
2126 ok ($x =~ /^\x{9_bq}y$/, "\\x{9_bc} is to be treated as \\x9b");
2129 ok ($x =~ /^\x{x9b}y$/, "\\x{x9b} is to be treated as \\x0");
2132 ok ($x =~ /^\x{0x9b}y$/, "\\x{0x9b} is to be treated as \\x0");
2135 ok ($x =~ /^\x{09b}y$/, "\\x{09b} is to be treated as \\x9b");
2137 print "# and now again in [] ranges\n";
2140 ok ($x =~ /^[\x{9_b}]$/, "\\x{9_b} is to be treated as \\x9b");
2143 ok ($x =~ /^[\x{9_b}y]{2}$/, "\\x{9_b} is to be treated as \\x9b (again)");
2146 ok ($x =~ /^[\x{9b_}y]{2}$/, "\\x{9b_} is to be treated as \\x9b");
2149 ok ($x =~ /^[\x{9_bq}y]{2}$/, "\\x{9_bc} is to be treated as \\x9b");
2152 ok ($x =~ /^[\x{x9b}y]{2}$/, "\\x{x9b} is to be treated as \\x0");
2155 ok ($x =~ /^[\x{0x9b}y]{2}$/, "\\x{0x9b} is to be treated as \\x0");
2158 ok ($x =~ /^[\x{09b}y]{2}$/, "\\x{09b} is to be treated as \\x9b");
2162 # high bit bug -- japhy
2164 $x =~ /.*?\200/ or print "not ";
2168 print "# some Unicode properties\n";
2171 # Dashes, underbars, case.
2172 print "not " unless "\x80" =~ /\p{in-latin1_SUPPLEMENT}/;
2175 # Complement, leading and trailing whitespace.
2176 print "not " unless "\x80" =~ /\P{ ^ In Latin 1 Supplement }/;
2179 # No ^In, dashes, case, dash, any intervening (word-break) whitespace.
2180 # (well, newlines don't work...)
2181 print "not " unless "\x80" =~ /\p{latin-1 supplement}/;
2186 print "not " unless "a" =~ /\pL/;
2189 print "not " unless "a" =~ /\p{IsLl}/;
2192 print "not " if "a" =~ /\p{IsLu}/;
2195 print "not " unless "a" =~ /\p{Ll}/;
2198 print "not " if "a" =~ /\p{Lu}/;
2201 print "not " unless "A" =~ /\pL/;
2204 print "not " unless "A" =~ /\p{IsLu}/;
2207 print "not " if "A" =~ /\p{IsLl}/;
2210 print "not " unless "A" =~ /\p{Lu}/;
2213 print "not " if "A" =~ /\p{Ll}/;
2216 print "not " if "a" =~ /\PL/;
2219 print "not " if "a" =~ /\P{IsLl}/;
2222 print "not " unless "a" =~ /\P{IsLu}/;
2225 print "not " if "a" =~ /\P{Ll}/;
2228 print "not " unless "a" =~ /\P{Lu}/;
2231 print "not " if "A" =~ /\PL/;
2234 print "not " if "A" =~ /\P{IsLu}/;
2237 print "not " unless "A" =~ /\P{IsLl}/;
2240 print "not " if "A" =~ /\P{Lu}/;
2243 print "not " unless "A" =~ /\P{Ll}/;
2249 print "not " if "a" =~ /\p{Common}/;
2252 print "not " unless "1" =~ /\p{Common}/;
2257 print "not " if "a" =~ /\p{Inherited}/;
2260 print "not " unless "\x{300}" =~ /\p{Inherited}/;
2265 # L& and LC are the same
2266 print "not " unless "a" =~ /\p{LC}/ and "a" =~ /\p{L&}/;
2269 print "not " if "1" =~ /\p{LC}/ or "1" =~ /\p{L&}/;
2274 print "not " unless "a" =~ /\p{Lowercase Letter}/;
2277 print "not " if "A" =~ /\p{lowercaseletter}/;
2282 print "not " unless "\x{AC00}" =~ /\p{HangulSyllables}/;
2287 # Script=, Block=, Category=
2289 print "not " unless "\x{0100}" =~ /\p{Script=Latin}/;
2292 print "not " unless "\x{0100}" =~ /\p{Block=LatinExtendedA}/;
2295 print "not " unless "\x{0100}" =~ /\p{Category=UppercaseLetter}/;
2300 print "# the basic character classes and Unicode \n";
2302 # 0100;LATIN CAPITAL LETTER A WITH MACRON;Lu;0;L;0041 0304;;;;N;LATIN CAPITAL LETTER A MACRON;;;0101;
2303 print "not " unless "\x{0100}" =~ /\w/;
2306 # 0660;ARABIC-INDIC DIGIT ZERO;Nd;0;AN;;0;0;0;N;;;;;
2307 print "not " unless "\x{0660}" =~ /\d/;
2310 # 1680;OGHAM SPACE MARK;Zs;0;WS;;;;;N;;;;;
2311 print "not " unless "\x{1680}" =~ /\s/;
2316 print "# folding matches and Unicode\n";
2318 print "not " unless "a\x{100}" =~ /A/i;
2321 print "not " unless "A\x{100}" =~ /a/i;
2324 print "not " unless "a\x{100}" =~ /a/i;
2327 print "not " unless "A\x{100}" =~ /A/i;
2330 print "not " unless "\x{101}a" =~ /\x{100}/i;
2333 print "not " unless "\x{100}a" =~ /\x{100}/i;
2336 print "not " unless "\x{101}a" =~ /\x{101}/i;
2339 print "not " unless "\x{100}a" =~ /\x{101}/i;
2342 print "not " unless "a\x{100}" =~ /A\x{100}/i;
2345 print "not " unless "A\x{100}" =~ /a\x{100}/i;
2348 print "not " unless "a\x{100}" =~ /a\x{100}/i;
2351 print "not " unless "A\x{100}" =~ /A\x{100}/i;
2354 print "not " unless "a\x{100}" =~ /[A]/i;
2357 print "not " unless "A\x{100}" =~ /[a]/i;
2360 print "not " unless "a\x{100}" =~ /[a]/i;
2363 print "not " unless "A\x{100}" =~ /[A]/i;
2366 print "not " unless "\x{101}a" =~ /[\x{100}]/i;
2369 print "not " unless "\x{100}a" =~ /[\x{100}]/i;
2372 print "not " unless "\x{101}a" =~ /[\x{101}]/i;
2375 print "not " unless "\x{100}a" =~ /[\x{101}]/i;
2381 use charnames ':full';
2383 print "# LATIN LETTER A WITH GRAVE\n";
2384 my $lower = "\N{LATIN SMALL LETTER A WITH GRAVE}";
2385 my $UPPER = "\N{LATIN CAPITAL LETTER A WITH GRAVE}";
2387 print $lower =~ m/$UPPER/i ? "ok 774\n" : "not ok 774\n";
2388 print $UPPER =~ m/$lower/i ? "ok 775\n" : "not ok 775\n";
2389 print $lower =~ m/[$UPPER]/i ? "ok 776\n" : "not ok 776\n";
2390 print $UPPER =~ m/[$lower]/i ? "ok 777\n" : "not ok 777\n";
2392 print "# GREEK LETTER ALPHA WITH VRACHY\n";
2394 $lower = "\N{GREEK CAPITAL LETTER ALPHA WITH VRACHY}";
2395 $UPPER = "\N{GREEK SMALL LETTER ALPHA WITH VRACHY}";
2397 print $lower =~ m/$UPPER/i ? "ok 778\n" : "not ok 778\n";
2398 print $UPPER =~ m/$lower/i ? "ok 779\n" : "not ok 779\n";
2399 print $lower =~ m/[$UPPER]/i ? "ok 780\n" : "not ok 780\n";
2400 print $UPPER =~ m/[$lower]/i ? "ok 781\n" : "not ok 781\n";
2402 print "# LATIN LETTER Y WITH DIAERESIS\n";
2404 $lower = "\N{LATIN CAPITAL LETTER Y WITH DIAERESIS}";
2405 $UPPER = "\N{LATIN SMALL LETTER Y WITH DIAERESIS}";
2406 print $lower =~ m/$UPPER/i ? "ok 782\n" : "not ok 782\n";
2407 print $UPPER =~ m/$lower/i ? "ok 783\n" : "not ok 783\n";
2408 print $lower =~ m/[$UPPER]/i ? "ok 784\n" : "not ok 784\n";
2409 print $UPPER =~ m/[$lower]/i ? "ok 785\n" : "not ok 785\n";
2414 use charnames ':full';
2416 print "# GREEK CAPITAL LETTER SIGMA vs COMBINING GREEK PERISPOMENI\n";
2418 my $SIGMA = "\N{GREEK CAPITAL LETTER SIGMA}";
2419 my $char = "\N{COMBINING GREEK PERISPOMENI}";
2421 # Before #13843 this was failing by matching falsely.
2422 print "_:$char:_" =~ m/_:$SIGMA:_/i ? "not ok 786\n" : "ok 786\n";
2428 use charnames ':full';
2430 print "a!" =~ /^(\X)!/ && $1 eq "a" ?
2431 "ok 787\n" : "not ok 787 # $1\n";
2432 print "\xDF!" =~ /^(\X)!/ && $1 eq "\xDF" ?
2433 "ok 788\n" : "not ok 788 # $1\n";
2434 print "\x{100}!" =~ /^(\X)!/ && $1 eq "\x{100}" ?
2435 "ok 789\n" : "not ok 789 # $1\n";
2436 print "\x{100}\x{300}!" =~ /^(\X)!/ && $1 eq "\x{100}\x{300}" ?
2437 "ok 790\n" : "not ok 790 # $1\n";
2438 print "\N{LATIN CAPITAL LETTER E}!" =~ /^(\X)!/ &&
2439 $1 eq "\N{LATIN CAPITAL LETTER E}" ?
2440 "ok 791\n" : "not ok 791 # $1\n";
2441 print "\N{LATIN CAPITAL LETTER E}\N{COMBINING GRAVE ACCENT}!" =~
2443 $1 eq "\N{LATIN CAPITAL LETTER E}\N{COMBINING GRAVE ACCENT}" ?
2444 "ok 792\n" : "not ok 792 # $1\n";
2448 print "#\\C and \\X\n";
2450 print "!abc!" =~ /a\Cc/ ? "ok 793\n" : "not ok 793\n";
2451 print "!abc!" =~ /a\Xc/ ? "ok 794\n" : "not ok 794\n";
2455 print "# FINAL SIGMA\n";
2457 my $SIGMA = "\x{03A3}"; # CAPITAL
2458 my $Sigma = "\x{03C2}"; # SMALL FINAL
2459 my $sigma = "\x{03C3}"; # SMALL
2461 print $SIGMA =~ /$SIGMA/i ? "ok 795\n" : "not ok 795\n";
2462 print $SIGMA =~ /$Sigma/i ? "ok 796\n" : "not ok 796\n";
2463 print $SIGMA =~ /$sigma/i ? "ok 797\n" : "not ok 797\n";
2465 print $Sigma =~ /$SIGMA/i ? "ok 798\n" : "not ok 798\n";
2466 print $Sigma =~ /$Sigma/i ? "ok 799\n" : "not ok 799\n";
2467 print $Sigma =~ /$sigma/i ? "ok 800\n" : "not ok 800\n";
2469 print $sigma =~ /$SIGMA/i ? "ok 801\n" : "not ok 801\n";
2470 print $sigma =~ /$Sigma/i ? "ok 802\n" : "not ok 802\n";
2471 print $sigma =~ /$sigma/i ? "ok 803\n" : "not ok 803\n";
2473 print $SIGMA =~ /[$SIGMA]/i ? "ok 804\n" : "not ok 804\n";
2474 print $SIGMA =~ /[$Sigma]/i ? "ok 805\n" : "not ok 805\n";
2475 print $SIGMA =~ /[$sigma]/i ? "ok 806\n" : "not ok 806\n";
2477 print $Sigma =~ /[$SIGMA]/i ? "ok 807\n" : "not ok 807\n";
2478 print $Sigma =~ /[$Sigma]/i ? "ok 808\n" : "not ok 808\n";
2479 print $Sigma =~ /[$sigma]/i ? "ok 809\n" : "not ok 809\n";
2481 print $sigma =~ /[$SIGMA]/i ? "ok 810\n" : "not ok 810\n";
2482 print $sigma =~ /[$Sigma]/i ? "ok 811\n" : "not ok 811\n";
2483 print $sigma =~ /[$sigma]/i ? "ok 812\n" : "not ok 812\n";
2487 print "# parlez-vous?\n";
2489 use charnames ':full';
2491 print "fran\N{LATIN SMALL LETTER C}ais" =~
2494 "ok 813\n" : "not ok 813\n";
2496 print "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~
2498 $& eq "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" ?
2499 "ok 814\n" : "not ok 814\n";
2501 print "fran\N{LATIN SMALL LETTER C}ais" =~
2504 "ok 815\n" : "not ok 815\n";
2506 print "franc\N{COMBINING CEDILLA}ais" =~
2507 /franc\C\Cais/ ? # COMBINING CEDILLA is two bytes when encoded
2508 "ok 816\n" : "not ok 816\n";
2510 print "fran\N{LATIN SMALL LETTER C}ais" =~
2513 "ok 817\n" : "not ok 817\n";
2515 print "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~
2517 $& eq "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" ?
2518 "ok 818\n" : "not ok 818\n";
2520 print "franc\N{COMBINING CEDILLA}ais" =~
2522 $& eq "franc\N{COMBINING CEDILLA}ais" ?
2523 "ok 819\n" : "not ok 819\n";
2525 print "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~
2526 /fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais/ &&
2527 $& eq "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" ?
2528 "ok 820\n" : "not ok 820\n";
2530 print "franc\N{COMBINING CEDILLA}ais" =~
2531 /franc\N{COMBINING CEDILLA}ais/ &&
2532 $& eq "franc\N{COMBINING CEDILLA}ais" ?
2533 "ok 821\n" : "not ok 821\n";
2535 print "fran\N{LATIN SMALL LETTER C}ais" =~
2536 /fran(?:c\N{COMBINING CEDILLA}?|\N{LATIN SMALL LETTER C WITH CEDILLA})ais/ &&
2538 "ok 822\n" : "not ok 822\n";
2540 print "fran\N{LATIN SMALL LETTER C}ais" =~
2541 /fran(?:c\N{COMBINING CEDILLA}?|\N{LATIN SMALL LETTER C WITH CEDILLA})ais/ &&
2543 "ok 823\n" : "not ok 823\n";
2545 print "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~
2546 /fran(?:c\N{COMBINING CEDILLA}?|\N{LATIN SMALL LETTER C WITH CEDILLA})ais/ &&
2547 $& eq "fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" ?
2548 "ok 824\n" : "not ok 824\n";
2550 print "franc\N{COMBINING CEDILLA}ais" =~
2551 /fran(?:c\N{COMBINING CEDILLA}?|\N{LATIN SMALL LETTER C WITH CEDILLA})ais/ &&
2552 $& eq "franc\N{COMBINING CEDILLA}ais" ?
2553 "ok 825\n" : "not ok 825\n";
2557 print "# Does lingering (and useless) UTF8 flag mess up /i matching?\n";
2560 my $regex = "ABcde";
2561 my $string = "abcDE\x{100}";
2563 if ($string =~ m/$regex/i) {
2566 print "not ok 826\n";
2571 my $regex = "ABcde\x{100}";
2572 my $string = "abcDE";
2574 if ($string =~ m/$regex/i) {
2577 print "not ok 827\n";
2582 my $regex = "ABcde\x{100}";
2583 my $string = "abcDE\x{100}";
2586 if ($string =~ m/$regex/i) {
2589 print "not ok 828\n";
2595 print "# more SIGMAs\n";
2597 my $SIGMA = "\x{03A3}"; # CAPITAL
2598 my $Sigma = "\x{03C2}"; # SMALL FINAL
2599 my $sigma = "\x{03C3}"; # SMALL
2601 my $S3 = "$SIGMA$Sigma$sigma";
2603 print ":$S3:" =~ /:(($SIGMA)+):/i && $1 eq $S3 && $2 eq $sigma ?
2604 "ok 829\n" : "not ok 829\n";
2605 print ":$S3:" =~ /:(($Sigma)+):/i && $1 eq $S3 && $2 eq $sigma ?
2606 "ok 830\n" : "not ok 830\n";
2607 print ":$S3:" =~ /:(($sigma)+):/i && $1 eq $S3 && $2 eq $sigma ?
2608 "ok 831\n" : "not ok 831\n";
2610 print ":$S3:" =~ /:(([$SIGMA])+):/i && $1 eq $S3 && $2 eq $sigma ?
2611 "ok 832\n" : "not ok 832\n";
2612 print ":$S3:" =~ /:(([$Sigma])+):/i && $1 eq $S3 && $2 eq $sigma ?
2613 "ok 833\n" : "not ok 833\n";
2614 print ":$S3:" =~ /:(([$sigma])+):/i && $1 eq $S3 && $2 eq $sigma ?
2615 "ok 834\n" : "not ok 834\n";
2619 print "# LATIN SMALL LETTER SHARP S\n";
2621 use charnames ':full';
2625 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /\N{LATIN SMALL LETTER SHARP S}/);
2626 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /\N{LATIN SMALL LETTER SHARP S}/i);
2628 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /[\N{LATIN SMALL LETTER SHARP S}]/);
2629 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i);
2631 ok("ss" =~ /\N{LATIN SMALL LETTER SHARP S}/i);
2632 ok("SS" =~ /\N{LATIN SMALL LETTER SHARP S}/i);
2633 ok("ss" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i);
2634 ok("SS" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i);
2636 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /ss/i);
2637 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /SS/i);
2641 print "# more whitespace: U+0085, U+2028, U+2029\n";
2643 # U+0085 needs to be forced to be Unicode, the \x{100} does that.
2645 print "<\x{100}\x{0085}>" =~ /<\x{100}e>/ ? "ok 845\n" : "not ok 845\n";
2647 print "<\x{100}\x{0085}>" =~ /<\x{100}\s>/ ? "ok 845\n" : "not ok 845\n";
2649 print "<\x{2028}>" =~ /<\s>/ ? "ok 846\n" : "not ok 846\n";
2650 print "<\x{2029}>" =~ /<\s>/ ? "ok 847\n" : "not ok 847\n";
2654 print "# . with /s should work on characters, as opposed to bytes\n";
2656 my $s = "\x{e4}\x{100}";
2658 # This is not expected to match: the point is that
2659 # neither should we get "Malformed UTF-8" warnings.
2660 print $s =~ /\G(.+?)\n/gcs ?
2661 "not ok 848\n" : "ok 848\n";
2665 while ($s =~ /\G(.)/gs) {
2669 print join("", @c) eq $s ? "ok 849\n" : "not ok 849\n";
2671 my $t1 = "Q003\n\n\x{e4}\x{f6}\n\nQ004\n\n\x{e7}"; # test only chars < 256
2673 while ($t1 =~ / \G ( .+? ) \n\s+ ( .+? ) ( $ | \n\s+ ) /xgcs) {
2677 my $t2 = $t1 . "\x{100}"; # repeat with a larger char
2679 while ($t2 =~ / \G ( .+? ) \n\s+ ( .+? ) ( $ | \n\s+ ) /xgcs) {
2683 print $r1 eq $r2 ? "ok 850\n" : "not ok 850\n";
2687 print "# Unicode lookbehind\n";
2689 print "A\x{100}B" =~ /(?<=A.)B/ ? "ok 851\n" : "not ok 851\n";
2690 print "A\x{200}\x{300}B" =~ /(?<=A..)B/ ? "ok 852\n" : "not ok 852\n";
2691 print "\x{400}AB" =~ /(?<=\x{400}.)B/ ? "ok 853\n" : "not ok 853\n";
2692 print "\x{500\x{600}}B" =~ /(?<=\x{500}.)B/ ? "ok 854\n" : "not ok 854\n";
2696 print "# UTF-8 hash keys and /\$/\n";
2697 # http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2002-01/msg01327.html
2700 my $v = substr($u,0,1);
2701 my $w = substr($u,1,1);
2702 my %u = ( $u => $u, $v => $v, $w => $w );
2705 my $m1 = /^\w*$/ ? 1 : 0;
2706 my $m2 = $u{$_}=~/^\w*$/ ? 1 : 0;
2707 print $m1 == $m2 ? "ok $i\n" : "not ok $i # $m1 $m2\n";
2713 print "# [ID 20020124.005]\n";
2716 for my $char ("a", "\x{df}", "\x{100}"){
2717 $x = "$char b $char";
2722 print substr($x,0,1) eq substr($x,-1,1) ?
2723 "ok $i\n" : "not ok $i # debug: $x\n";
2729 print "# SEGV in s/// and UTF-8\n";
2730 $s = "s#\x{100}" x 4;
2732 print $s eq "s \x{100}" x 4 ? "ok 861\n" : "not ok 861\n";
2736 print "# UTF-8 bug (maybe alreayd known?)\n";
2740 $u =~ s/./\x{100}/g;
2741 print $u eq "\x{100}\x{100}\x{100}" ? "ok 862\n" : "not ok 862\n";
2744 $u =~ s/[ao]/\x{100}/g;
2745 print $u eq "f\x{100}\x{100}b\x{100}r" ? "ok 863\n" : "not ok 863\n";
2747 $u =~ s/\x{100}/e/g;
2748 print $u eq "feeber" ? "ok 864\n" : "not ok 864\n";
2752 print "# UTF-8 bug with s///\n";
2753 # check utf8/non-utf8 mixtures
2754 # try to force all float/anchored check combinations
2759 "xx.*$c", "x.*$c$c", "$c.*xx", "$c$c.*x", "xx.*(?=$c)", "(?=$c).*xx",
2761 print "xxx" =~ /$re/ ? "not ok $test\n" : "ok $test\n";
2763 print +($subst = "xxx") =~ s/$re// ? "not ok $test\n" : "ok $test\n";
2766 for my $re ("xx.*$c*", "$c*.*xx") {
2767 print "xxx" =~ /$re/ ? "ok $test\n" : "not ok $test\n";
2769 ($subst = "xxx") =~ s/$re//;
2770 print $subst eq '' ? "ok $test\n" : "not ok $test\t# $subst\n";
2773 for my $re ("xxy*", "y*xx") {
2774 print "xx$c" =~ /$re/ ? "ok $test\n" : "not ok $test\n";
2776 ($subst = "xx$c") =~ s/$re//;
2777 print $subst eq $c ? "ok $test\n" : "not ok $test\n";
2779 print "xy$c" =~ /$re/ ? "not ok $test\n" : "ok $test\n";
2781 print +($subst = "xy$c") =~ /$re/ ? "not ok $test\n" : "ok $test\n";
2784 for my $re ("xy$c*z", "x$c*yz") {
2785 print "xyz" =~ /$re/ ? "ok $test\n" : "not ok $test\n";
2787 ($subst = "xyz") =~ s/$re//;
2788 print $subst eq '' ? "ok $test\n" : "not ok $test\n";
2794 print "# qr/.../x\n";
2797 my $R = qr/ A B C # D E/x;
2799 print eval {"ABCDE" =~ $R} ? "ok $test\n" : "not ok $test\n";
2802 print eval {"ABCDE" =~ m/$R/} ? "ok $test\n" : "not ok $test\n";
2805 print eval {"ABCDE" =~ m/($R)/} ? "ok $test\n" : "not ok $test\n";
2810 print "# illegal Unicode properties\n";
2813 print eval { "a" =~ /\pq / } ? "not ok $test\n" : "ok $test\n";
2816 print eval { "a" =~ /\p{qrst} / } ? "not ok $test\n" : "ok $test\n";
2821 print "# [ID 20020412.005] wrong pmop flags checked when empty pattern\n";
2822 # requires reuse of last successful pattern
2830 printf "not ok %s\t# 'match once' %s on %s iteration\n", $test,
2831 $match ? 'succeeded' : 'failed', $_ ? 'second' : 'first';
2836 my $result = join '', $test =~ //g;
2837 if ($result eq $test) {
2840 printf "not ok %s\t# expected '%s', got '%s'\n", $test, $test, $result;
2845 print "# user-defined character properties\n";
2879 print "\x{3040}" =~ /\p{InKana1}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2880 print "\x{303F}" =~ /\P{InKana1}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2882 print "\x{3040}" =~ /\p{InKana2}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2883 print "\x{303F}" =~ /\P{InKana2}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2885 print "\x{3041}" =~ /\p{InKana3}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2886 print "\x{3040}" =~ /\P{InKana3}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2888 print "\x{3040}" =~ /\p{InNotKana}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2889 print "\x{3041}" =~ /\P{InNotKana}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2891 sub InConsonant { # Not EBCDIC-aware.
2902 print "d" =~ /\p{InConsonant}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2903 print "e" =~ /\P{InConsonant}/ ? "ok $test\n" : "not ok $test\n"; $test++;
2906 print "# [ID 20020630.002] utf8 regex only matches 32k\n";
2908 for ([ 'byte', "\x{ff}" ], [ 'utf8', "\x{1ff}" ]) {
2909 my($type, $char) = @$_;
2910 for my $len (32000, 32768, 33000) {
2911 my $s = $char . "f" x $len;
2912 my $r = $s =~ /$char([f]*)/gc;
2913 print $r ? "ok $test\n" : "not ok $test\t# <$type x $len> fail\n";
2915 print +(!$r or pos($s) == $len + 1) ? "ok $test\n"
2916 : "not ok $test\t# <$type x $len> pos @{[ pos($s) ]}\n";
2924 $a = bless qr/foo/, 'Foo';
2925 print(('goodfood' =~ $a ? '' : 'not '),
2926 "ok $test\t# reblessed qr// matches\n");
2929 print(($a eq '(?-xism:foo)' ? '' : 'not '),
2930 "ok $test\t# reblessed qr// stringizes\n");
2934 $z=$y = "\317\276"; # $y is byte representation of $x
2937 print(($x =~ $a ? '' : 'not '), "ok $test - utf8 interpolation in qr//\n");
2940 print(("a$a" =~ $x ? '' : 'not '),
2941 "ok $test - stringifed qr// preserves utf8\n");
2944 print(("a$x" =~ /^a$a\z/ ? '' : 'not '),
2945 "ok $test - interpolated qr// preserves utf8\n");
2948 print(("a$x" =~ /^a(??{$a})\z/ ? '' : 'not '),
2949 "ok $test - postponed interpolation of qr// preserves utf8\n");
2952 print((length(qr/##/x) == 12 ? '' : 'not '),
2953 "ok $test - ## in qr// doesn't corrupt memory [perl #17776]\n");
2958 print(("$x$x" =~ /^$x(??{$x})\z/ ? '' : 'not '),
2959 "ok $test - postponed utf8 string in utf8 re matches utf8\n");
2962 print(("$y$x" =~ /^$y(??{$x})\z/ ? '' : 'not '),
2963 "ok $test - postponed utf8 string in non-utf8 re matches utf8\n");
2966 print(("$y$x" !~ /^$y(??{$y})\z/ ? '' : 'not '),
2967 "ok $test - postponed non-utf8 string in non-utf8 re doesn't match utf8\n");
2970 print(("$x$x" !~ /^$x(??{$y})\z/ ? '' : 'not '),
2971 "ok $test - postponed non-utf8 string in utf8 re doesn't match utf8\n");
2974 print(("$y$y" =~ /^$y(??{$y})\z/ ? '' : 'not '),
2975 "ok $test - postponed non-utf8 string in non-utf8 re matches non-utf8\n");
2978 print(("$x$y" =~ /^$x(??{$y})\z/ ? '' : 'not '),
2979 "ok $test - postponed non-utf8 string in utf8 re matches non-utf8\n");
2981 $y = $z; # reset $y after upgrade
2983 print(("$x$y" !~ /^$x(??{$x})\z/ ? '' : 'not '),
2984 "ok $test - postponed utf8 string in utf8 re doesn't match non-utf8\n");
2986 $y = $z; # reset $y after upgrade
2988 print(("$y$y" !~ /^$y(??{$x})\z/ ? '' : 'not '),
2989 "ok $test - postponed utf8 string in non-utf8 re doesn't match non-utf8\n");
2994 print "# more user-defined character properties\n";
3003 ok("\x{0712}" =~ /\p{IsSyriac1}/, '\x{0712}, \p{IsSyriac1}');
3004 ok("\x{072F}" =~ /\P{IsSyriac1}/, '\x{072F}, \P{IsSyriac1}');
3013 ok("\x{0712}" =~ /\p{Syriac1}/, '\x{0712}, \p{Syriac1}');
3014 ok("\x{072F}" =~ /\P{Syriac1}/, '\x{072F}, \p{Syriac1}');
3016 print "# user-defined character properties may lack \\n at the end\n";
3017 sub InGreekSmall { return "03B1\t03C9" }
3018 sub InGreekCapital { return "0391\t03A9\n-03A2" }
3020 ok("\x{03C0}" =~ /\p{InGreekSmall}/, "Small pi");
3021 ok("\x{03C2}" =~ /\p{InGreekSmall}/, "Final sigma");
3022 ok("\x{03A0}" =~ /\p{InGreekCapital}/, "Capital PI");
3023 ok("\x{03A2}" =~ /\P{InGreekCapital}/, "Reserved");
3025 sub AsciiHexAndDash {
3027 +utf8::ASCII_Hex_Digit
3032 ok("-" =~ /\p{Dash}/, "'-' is Dash");
3033 ok("A" =~ /\p{ASCII_Hex_Digit}/, "'A' is ASCII_Hex_Digit");
3034 ok("-" =~ /\p{AsciiHexAndDash}/, "'-' is AsciiHexAndDash");
3035 ok("A" =~ /\p{AsciiHexAndDash}/, "'A' is AsciiHexAndDash");
3038 print "# Change #18179\n";
3039 # previously failed with "panic: end_shift
3040 my $s = "\x{100}" x 5;
3041 my $ok = $s =~ /(\x{100}{4})/;
3042 my($ord, $len) = (ord $1, length $1);
3043 print +($ok && $ord == 0x100 && $len == 4)
3044 ? "ok $test\n" : "not ok $test\t# [#18179] $ok/$ord/$len\n";
3049 print "# [perl #15763]\n";
3052 chop $a; # but leaves the UTF-8 flag
3053 $a .= "y"; # 1 byte before "y"
3055 ok($a =~ /^\C/, 'match one \C on 1-byte UTF-8');
3056 ok($a =~ /^\C{1}/, 'match \C{1}');
3058 ok($a =~ /^\Cy/, 'match \Cy');
3059 ok($a =~ /^\C{1}y/, 'match \C{1}y');
3061 $a = "\x{100}y"; # 2 bytes before "y"
3063 ok($a =~ /^\C/, 'match one \C on 2-byte UTF-8');
3064 ok($a =~ /^\C{1}/, 'match \C{1}');
3065 ok($a =~ /^\C\C/, 'match two \C');
3066 ok($a =~ /^\C{2}/, 'match \C{2}');
3068 ok($a =~ /^\C\C\C/, 'match three \C on 2-byte UTF-8 and a byte');
3069 ok($a =~ /^\C{3}/, 'match \C{3}');
3071 ok($a =~ /^\C\Cy/, 'match two \C');
3072 ok($a =~ /^\C{2}y/, 'match \C{2}');
3074 ok($a !~ /^\C\C\Cy/, q{don't match three \Cy});
3075 ok($a !~ /^\C{2}\Cy/, q{don't match \C{3}y});
3077 $a = "\x{1000}y"; # 3 bytes before "y"
3079 ok($a =~ /^\C/, 'match one \C on three-byte UTF-8');
3080 ok($a =~ /^\C{1}/, 'match \C{1}');
3081 ok($a =~ /^\C\C/, 'match two \C');
3082 ok($a =~ /^\C{2}/, 'match \C{2}');
3083 ok($a =~ /^\C\C\C/, 'match three \C');
3084 ok($a =~ /^\C{3}/, 'match \C{3}');
3086 ok($a =~ /^\C\C\C\C/, 'match four \C on three-byte UTF-8 and a byte');
3087 ok($a =~ /^\C{4}/, 'match \C{4}');
3089 ok($a =~ /^\C\C\Cy/, 'match three \Cy');
3090 ok($a =~ /^\C{3}y/, 'match \C{3}y');
3092 ok($a !~ /^\C\C\C\C\y/, q{don't match four \Cy});
3093 ok($a !~ /^\C{4}y/, q{don't match \C{4}y});
3099 utf8::upgrade($_); chop $_; $\="\n";
3100 ok(/[^\s]+/, "m/[^\s]/ utf8");
3101 ok(/[^\d]+/, "m/[^\d]/ utf8");
3102 ok(($a = $_, $_ =~ s/[^\s]+/./g), "s/[^\s]/ utf8");
3103 ok(($a = $_, $a =~ s/[^\d]+/./g), "s/[^\s]/ utf8");
3106 ok("\x{100}" =~ /\x{100}/, "[perl #15397]");
3107 ok("\x{100}" =~ /(\x{100})/, "[perl #15397]");
3108 ok("\x{100}" =~ /(\x{100}){1}/, "[perl #15397]");
3109 ok("\x{100}\x{100}" =~ /(\x{100}){2}/, "[perl #15397]");
3110 ok("\x{100}\x{100}" =~ /(\x{100})(\x{100})/, "[perl #15397]");
3114 ok(!defined $1, "[perl #7471]");
3118 ok(!defined $1, "[perl #7471]");
3120 $pattern = "^(b+?|a){1,2}c";
3121 ok("bac" =~ /$pattern/ && $1 eq 'a', "[perl #3547]");
3122 ok("bbac" =~ /$pattern/ && $1 eq 'a', "[perl #3547]");
3123 ok("bbbac" =~ /$pattern/ && $1 eq 'a', "[perl #3547]");
3124 ok("bbbbac" =~ /$pattern/ && $1 eq 'a', "[perl #3547]");
3129 ok( $1 eq "\x{100}", '$1 is utf-8 [perl #18232]' );
3131 ok( $1 eq "\x{100}", '$1 is still utf-8' );
3132 ok( $1 ne "\xC4\x80", '$1 is not non-utf-8' );
3137 my $attr = 'Name-1' ;
3139 my $NormalChar = qr/[\p{IsDigit}\p{IsLower}\p{IsUpper}]/;
3140 my $NormalWord = qr/${NormalChar}+?/;
3141 my $PredNameHyphen = qr/^${NormalWord}(\-${NormalWord})*?$/;
3144 ok( $attr =~ $PredNameHyphen, "[perl #19767] original test" );
3150 ok ( "0" =~ /\p{N}+\z/, "[perl #19767] variant test" );
3157 $p++ if /(??{ $p })/
3159 iseq ($p, 5, "[perl #20683] (??{ }) returns stale values");
3160 { package P; $a=1; sub TIESCALAR { bless[] } sub FETCH { $a++ } }
3165 iseq ( $p, 5, "(??{ }) returns stale values");
3169 # Subject: Odd regexp behavior
3170 # From: Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>
3171 # Date: Wed, 26 Feb 2003 16:53:12 +0000
3172 # Message-Id: <E18o4nw-0008Ly-00@wisbech.cl.cam.ac.uk>
3173 # To: perl-unicode@perl.org
3175 $x = "\x{2019}\nk"; $x =~ s/(\S)\n(\S)/$1 $2/sg;
3176 ok($x eq "\x{2019} k", "Markus Kuhn 2003-02-26");
3178 $x = "b\nk"; $x =~ s/(\S)\n(\S)/$1 $2/sg;
3179 ok($x eq "b k", "Markus Kuhn 2003-02-26");
3181 ok("\x{2019}" =~ /\S/, "Markus Kuhn 2003-02-26");
3186 ok('-1-3-5-' eq join('', split /((??{$i++}))/, '-1-3-5-'),
3187 "[perl #21411] (??{ .. }) corrupts split's stack");
3188 split /(?{'WOW'})/, 'abc';
3189 ok('a|b|c' eq join ('|', @_),
3190 "[perl #21411] (?{ .. }) version of the above");
3194 # XXX DAPM 13-Apr-06. Recursive split is still broken. It's only luck it
3195 # hasn't been crashing. Disable this test until it is fixed properly.
3196 # XXX also check what it returns rather than just doing ok(1,...)
3197 # split /(?{ split "" })/, "abc";
3198 ok(1,'cache_re & "(?{": it dumps core in 5.6.1 & 5.8.0');
3202 ok("\x{100}\n" =~ /\x{100}\n$/, "UTF8 length cache and fbm_compile");
3207 use overload q/""/ => sub { ${$_[0]}; };
3208 sub new { my ($c, $v) = @_; bless \$v, $c; }
3211 $_ = Str->new("a\x{100}/\x{100}b");
3212 ok(join(":", /\b(.)\x{100}/g) eq "a:/", "re_intuit_start and PL_bostr");
3216 $_ = "code: 'x' { '...' }\n"; study;
3217 my @x; push @x, $& while m/'[^\']*'/gx;
3218 ok(join(":", @x) eq "'x':'...'",
3219 "[perl #17757] Parse::RecDescent triggers infinite loop");
3223 my $re = qq/^([^X]*)X/;
3225 ok("\x{100}X" =~ /$re/, "S_cl_and ANYOF_UNICODE & ANYOF_INVERTED");
3230 ok( "a\nb" !~ /^b/, $_[0] );
3231 ok( "a\nb" =~ /^b/m, "$_[0] - with /m" );
3234 $_ = "x"; s/x/func "in subst"/e;
3235 $_ = "x"; s/x/func "in multiline subst"/em;
3236 #$_ = "x"; /x(?{func "in regexp"})/;
3237 #$_ = "x"; /x(?{func "in multiline regexp"})/m;
3242 ok("abcde" eq "$`", 'RT#19049 - global match not setting $`');
3244 ok("123\x{100}" =~ /^.*1.*23\x{100}$/, 'uft8 + multiple floating substr');
3246 # LATIN SMALL/CAPITAL LETTER A WITH MACRON
3247 ok(" \x{101}" =~ qr/\x{100}/i,
3248 "<20030808193656.5109.1@llama.ni-s.u-net.com>");
3250 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW
3251 ok(" \x{1E01}" =~ qr/\x{1E00}/i,
3252 "<20030808193656.5109.1@llama.ni-s.u-net.com>");
3254 # DESERET SMALL/CAPITAL LETTER LONG I
3255 ok(" \x{10428}" =~ qr/\x{10400}/i,
3256 "<20030808193656.5109.1@llama.ni-s.u-net.com>");
3258 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'
3259 ok(" \x{1E01}x" =~ qr/\x{1E00}X/i,
3260 "<20030808193656.5109.1@llama.ni-s.u-net.com>");
3263 # [perl #23769] Unicode regex broken on simple example
3264 # regrepeat() didn't handle UTF-8 EXACT case right.
3266 my $s = "\x{a0}\x{a0}\x{a0}\x{100}"; chop $s;
3268 ok($s =~ /\x{a0}/, "[perl #23769]");
3269 ok($s =~ /\x{a0}+/, "[perl #23769]");
3270 ok($s =~ /\x{a0}\x{a0}/, "[perl #23769]");
3272 ok("aaa\x{100}" =~ /(a+)/, "[perl #23769] easy invariant");
3273 ok($1 eq "aaa", "[perl #23769]");
3275 ok("\xa0\xa0\xa0\x{100}" =~ /(\xa0+)/, "[perl #23769] regrepeat invariant");
3276 ok($1 eq "\xa0\xa0\xa0", "[perl #23769]");
3278 ok("ababab\x{100} " =~ /((?:ab)+)/, "[perl #23769] hard invariant");
3279 ok($1 eq "ababab", "[perl #23769]");
3281 ok("\xa0\xa1\xa0\xa1\xa0\xa1\x{100}" =~ /((?:\xa0\xa1)+)/, "[perl #23769] hard variant");
3282 ok($1 eq "\xa0\xa1\xa0\xa1\xa0\xa1", "[perl #23769]");
3284 ok("aaa\x{100} " =~ /(a+?)/, "[perl #23769] easy invariant");
3285 ok($1 eq "a", "[perl #23769]");
3287 ok("\xa0\xa0\xa0\x{100} " =~ /(\xa0+?)/, "[perl #23769] regrepeat variant");
3288 ok($1 eq "\xa0", "[perl #23769]");
3290 ok("ababab\x{100} " =~ /((?:ab)+?)/, "[perl #23769] hard invariant");
3291 ok($1 eq "ab", "[perl #23769]");
3293 ok("\xa0\xa1\xa0\xa1\xa0\xa1\x{100}" =~ /((?:\xa0\xa1)+?)/, "[perl #23769] hard variant");
3294 ok($1 eq "\xa0\xa1", "[perl #23769]");
3296 ok("\xc4\xc4\xc4" !~ /(\x{100}+)/, "[perl #23769] don't match first byte of utf8 representation");
3297 ok("\xc4\xc4\xc4" !~ /(\x{100}+?)/, "[perl #23769] don't match first byte of utf8 representation");
3301 my $head = 'x' x $_;
3302 for my $tail ('\x{0061}', '\x{1234}') {
3304 eval qq{ "$head$tail" =~ /$head$tail/ },
3305 '\x{...} misparsed in regexp near 127 char EXACT limit'
3310 # perl #25269: panic: pp_match start/end pointers
3312 my($x, $y) = "bca" =~ /^(?=.*(a)).*(bc)/;
3314 }, 'captures can move backwards in string');
3316 # perl #27940: \cA not recognized in character classes
3317 ok("a\cAb" =~ /\cA/, '\cA in pattern');
3318 ok("a\cAb" =~ /[\cA]/, '\cA in character class');
3319 ok("a\cAb" =~ /[\cA-\cB]/, '\cA in character class range');
3320 ok("abc" =~ /[^\cA-\cB]/, '\cA in negated character class range');
3321 ok("a\cBb" =~ /[\cA-\cC]/, '\cB in character class range');
3322 ok("a\cCbc" =~ /[^\cA-\cB]/, '\cC in negated character class range');
3323 ok("a\cAb" =~ /(??{"\cA"})/, '\cA in ??{} pattern');
3324 ok("ab" !~ /a\cIb/x, '\cI in pattern');
3326 # perl #28532: optional zero-width match at end of string is ignored
3327 ok(("abc" =~ /^abc(\z)?/) && defined($1),
3328 'optional zero-width match at end of string');
3329 ok(("abc" =~ /^abc(\z)??/) && !defined($1),
3330 'optional zero-width match at end of string');
3336 "words"=~/(word|word|word)(?{push @got,$1})s$/;
3337 ok(@got==1,"TRIE optimation is working") or warn "# @got";
3339 "words"=~/(word|word|word)(?{push @got,$1})s$/i;
3340 ok(@got==1,"TRIEF optimisation is working") or warn "# @got";
3342 my @nums=map {int rand 1000} 1..100;
3343 my $re="(".(join "|",@nums).")";
3347 ok($_=~/$re/,"Trie nums");
3351 push @got,$1 while /$re/g;
3354 $count{$_}++ for @got;
3357 $ok=0 if --$count{$_}<0;
3359 ok($ok,"Trie min count matches");
3364 # LATIN SMALL/CAPITAL LETTER A WITH MACRON
3365 ok(("foba \x{101}foo" =~ qr/(foo|\x{100}foo|bar)/i) && $1 eq "\x{101}foo",
3366 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH MACRON");
3368 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW
3369 ok(("foba \x{1E01}foo" =~ qr/(foo|\x{1E00}foo|bar)/i) && $1 eq "\x{1E01}foo",
3370 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW");
3372 # DESERET SMALL/CAPITAL LETTER LONG I
3373 ok(("foba \x{10428}foo" =~ qr/(foo|\x{10400}foo|bar)/i) && $1 eq "\x{10428}foo",
3374 "TRIEF + DESERET SMALL/CAPITAL LETTER LONG I");
3376 # LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'
3377 ok(("foba \x{1E01}xfoo" =~ qr/(foo|\x{1E00}Xfoo|bar)/i) && $1 eq "\x{1E01}xfoo",
3378 "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'");
3382 use charnames ':full';
3384 $s="\N{LATIN SMALL LETTER SHARP S}";
3385 ok(("foba ba$s" =~ qr/(foo|Ba$s|bar)/i)
3387 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
3388 ok(("foba ba$s" =~ qr/(Ba$s|foo|bar)/i)
3390 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
3391 ok(("foba ba$s" =~ qr/(foo|bar|Ba$s)/i)
3393 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
3395 ok(("foba ba$s" =~ qr/(foo|Bass|bar)/i)
3397 "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
3399 ok(("foba ba$s" =~ qr/(foo|BaSS|bar)/i)
3401 "TRIEF + LATIN SMALL LETTER SHARP S =~ SS");
3403 ok(("foba ba${s}pxySS$s$s" =~ qr/(b(?:a${s}t|a${s}f|a${s}p)[xy]+$s*)/i)
3404 && $1 eq "ba${s}pxySS$s$s",
3405 "COMMON PREFIX TRIEF + LATIN SMALL LETTER SHARP S");
3411 print "# set PERL_SKIP_PSYCHO_TEST to skip this test\n";
3412 if (!$ENV{PERL_SKIP_PSYCHO_TEST}){
3413 my @normal=qw(these are some normal words);
3414 my $psycho=join "|",@normal,map chr $_,255..20000;
3415 ok(('these'=~/($psycho)/) && $1 eq 'these','Pyscho');
3417 ok(1,'Skipped Psycho');
3420 # [perl #36207] mixed utf8 / latin-1 and case folding
3423 my $utf8 = "\xe9\x{100}"; chop $utf8;
3424 my $latin1 = "\xe9";
3426 ok($utf8 =~ /\xe9/i, "utf8/latin");
3427 ok($utf8 =~ /$latin1/i, "utf8/latin runtime");
3428 ok($utf8 =~ /(abc|\xe9)/i, "utf8/latin trie");
3429 ok($utf8 =~ /(abc|$latin1)/i, "utf8/latin trie runtime");
3431 ok("\xe9" =~ /$utf8/i, "# latin/utf8");
3432 ok("\xe9" =~ /(abc|$utf8)/i, "# latin/utf8 trie");
3433 ok($latin1 =~ /$utf8/i, "# latin/utf8 runtime");
3434 ok($latin1 =~ /(abc|$utf8)/i, "# latin/utf8 trie runtime");
3437 # [perl #37038] Global regular matches generate invalid pointers
3445 "# assigning to original string should not corrupt match vars");
3455 my $aeek = bless {}, 'wooosh';
3456 eval {$aeek->gloople() =~ /(.)/g;};
3457 ok($@ eq "", "//g match against return value of sub") or print "# $@\n";
3464 eval {gloople() =~ /(.)/g;};
3465 ok($@ eq "", "# 26410 didn't affect sub calls for some reason")
3470 local $TODO = "See changes 26925-26928, which reverted change 26410";
3473 sub variable : lvalue { $var }
3476 my $o = bless [], "lv";
3478 eval { for (1..2) { $f .= $1 if $o->variable =~ /(.)/g } };
3479 ok($f eq "ab", "pos retained between calls") or print "# $@\n";
3483 local $TODO = "See changes 26925-26928, which reverted change 26410";
3485 sub variable : lvalue { $var }
3488 eval { for (1..2) { $f .= $1 if variable() =~ /(.)/g } };
3489 ok($f eq "ab", "pos retained between calls") or print "# $@\n";
3492 # [perl #37836] Simple Regex causes SEGV when run on specific data
3494 print "ok $test # Skip: in EBCDIC\n"; $test++;
3497 $_ = pack('U0C2', 0xa2, 0xf8); # ill-formed UTF-8
3499 eval { $ret = s/[\0]+//g };
3500 ok($ret == 0, "ill-formed UTF-8 doesn't match NUL in class");
3503 { # [perl #38293] chr(65535) should be allowed in regexes
3504 no warnings 'utf8'; # to allow non-characters
3509 ok($c eq "", "U+FFFF, parsed as atom");
3514 ok($c eq "", "U+FFFF backslashed, parsed as atom");
3518 ok($c eq "", "U+FFFF, parsed in class");
3523 ok($c eq "", "U+FFFF backslashed, parsed in class");
3526 $s =~ s/\x{ffff}//i;
3527 ok($s eq "AB", "U+FFFF, EXACTF");
3531 ok($s eq "\x{ffff}", "U+FFFF, BOUND");
3535 ok($s eq "\x{ffff}", "U+FFFF, NBOUND");
3536 } # non-characters end
3539 # https://rt.perl.org/rt3/Ticket/Display.html?id=39583
3541 # The printing characters
3542 my @chars = ("A".."Z");
3544 my $size = 32771 - 4;
3547 # create some random junk. Inefficient, but it works.
3548 for ($i = 0 ; $i < $size ; $i++) {
3549 $str .= $chars[int(rand(@chars))];
3552 $str .= ($delim x 4);
3555 if ($str =~ s/^(.*?)${delim}{4}//s) {
3559 ok($matched,'pattern matches');
3560 ok(length($str)==0,"Empty string");
3561 ok(defined($res) && length($res)==$size,"\$1 is correct size");
3564 { # related to [perl #27940]
3565 ok("\0-A" =~ /\c@-A/, '@- should not be interpolated in a pattern');
3566 ok("\0\0A" =~ /\c@+A/, '@+ should not be interpolated in a pattern');
3567 ok("X\@-A" =~ /X@-A/, '@- should not be interpolated in a pattern');
3568 ok("X\@\@A" =~ /X@+A/, '@+ should not be interpolated in a pattern');
3570 ok("X\0A" =~ /X\c@?A/, '\c@?');
3571 ok("X\0A" =~ /X\c@*A/, '\c@*');
3572 ok("X\0A" =~ /X\c@(A)/, '\c@(');
3573 ok("X\0A" =~ /X(\c@)A/, '\c@)');
3574 ok("X\0A" =~ /X\c@|ZA/, '\c@|');
3576 ok("X\@A" =~ /X@?A/, '@?');
3577 ok("X\@A" =~ /X@*A/, '@*');
3578 ok("X\@A" =~ /X@(A)/, '@(');
3579 ok("X\@A" =~ /X(@)A/, '@)');
3580 ok("X\@A" =~ /X@|ZA/, '@|');
3582 local $" = ','; # non-whitespace and non-RE-specific
3583 ok('abc' =~ /(.)(.)(.)/, 'the last successful match is bogus');
3584 ok("A@+B" =~ /A@{+}B/, 'interpolation of @+ in /@{+}/');
3585 ok("A@-B" =~ /A@{-}B/, 'interpolation of @- in /@{-}/');
3586 ok("A@+B" =~ /A@{+}B/x, 'interpolation of @+ in /@{+}/x');
3587 ok("A@-B" =~ /A@{-}B/x, 'interpolation of @- in /@{-}/x');
3594 ok('fooB'=~/\N{foo}[\N{B}\N{b}]/,"Passthrough charname");
3595 $test=1233; my $handle=make_must_warn('Ignoring excess chars from');
3596 $handle->('q(xxWxx) =~ /[\N{WARN}]/');
3600 local $SIG{__WARN__} = sub { $w.=shift };
3601 eval($code=<<'EOFTEST') or die "$@\n$code\n";
3606 ok("\0" !~ /[\N{EMPTY-STR}XY]/,
3607 "Zerolength charname in charclass doesnt match \0");
3611 ok($w=~/Ignoring zero length/,
3612 "Got expected zero length warning");
3616 $handle= make_must_warn('Ignoring zero length');
3617 $handle->('qq(\\0) =~ /[\N{EMPTY-STR}XY]/');
3618 ok('AB'=~/(\N{EVIL})/ && $1 eq 'A',"Charname caching $1");
3619 ok('ABC'=~/(\N{EVIL})/,"Charname caching $1");
3620 ok('xy'=~/x\N{EMPTY-STR}y/, 'Empty string charname produces NOTHING node');
3621 ok(''=~/\N{EMPTY-STR}/, 'Empty string charname produces NOTHING node 2');
3625 print "# MORE LATIN SMALL LETTER SHARP S\n";
3627 use charnames ':full';
3630 ok("ss" =~ /[\N{LATIN SMALL LETTER SHARP S}x]/i,
3631 "unoptimized named sequence in class 1");
3632 ok("SS" =~ /[\N{LATIN SMALL LETTER SHARP S}x]/i,
3633 "unoptimized named sequence in class 2");
3634 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /[\N{LATIN SMALL LETTER SHARP S}x]/,
3635 "unoptimized named sequence in class 3");
3636 ok("\N{LATIN SMALL LETTER SHARP S}" =~ /[\N{LATIN SMALL LETTER SHARP S}x]/i,
3637 "unoptimized named sequence in class 4");
3639 ok('aabc' !~ /a\N{PLUS SIGN}b/,'/a\N{PLUS SIGN}b/ against aabc');
3640 ok('a+bc' =~ /a\N{PLUS SIGN}b/,'/a\N{PLUS SIGN}b/ against a+bc');
3641 ok('a+bc' =~ /a\N{PLUS SIGN}b/,'/a\N{PLUS SIGN}b/ against a+bc');
3643 ok(' A B'=~/\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/,
3644 'Intermixed named and unicode escapes 1');
3645 ok("\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}"=~
3646 /\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042}/,
3647 'Intermixed named and unicode escapes 2');
3648 ok("\N{SPACE}\N{U+0041}\N{SPACE}\N{U+0042} 3"=~
3649 /[\N{SPACE}\N{U+0041}][\N{SPACE}\N{U+0042}]/,
3650 'Intermixed named and unicode escapes');
3653 { (?> [^{}]+ | (??{ $brackets }) )* }
3655 ok("{b{c}d" !~ m/^((??{ $brackets }))/, "bracket mismatch");
3667 <<stuff1>and<stuff2>><<<<right>>>>
3670 local $_='<<<stuff1>and<stuff2>><<<<right>>>>>';
3671 ok(/^(<((?:(?>[^<>]+)|(?1))*)>(?{push @stack, $2 }))$/,
3672 "Recursion should match");
3674 or skip("Won't test individual results as count isn't equal",
3676 foreach my $idx (@expect) {
3677 ok($expect[$idx] eq $stack[$idx],
3678 "Expecting '$expect' at stack pos #$idx");
3684 $s=~s/(?<digits>\d+)\k<digits>/$+{digits}/;
3685 ok($s eq '123456','Named capture (angle brackets) s///');
3687 $s=~s/(?'digits'\d+)\k'digits'/$+{digits}/;
3688 ok($s eq '123456','Named capture (single quotes) s///');
3693 pack('U', 0x00F1), # n-tilde
3694 '_'.pack('U', 0x00F1), # _ + n-tilde
3695 'c'.pack('U', 0x0327), # c + cedilla
3696 pack('U*', 0x00F1, 0x0327), # n-tilde + cedilla
3697 'a'.pack('U', 0x00B2), # a + superscript two
3698 pack('U', 0x0391), # ALPHA
3699 pack('U', 0x0391).'2', # ALPHA + 2
3700 pack('U', 0x0391).'_', # ALPHA + _
3702 for my $uni (@ary) {
3703 my ($r1, $c1, $r2, $c2) = eval qq{
3705 scalar("..foo foo.." =~ /(?'${uni}'foo) \\k'${uni}'/),
3707 scalar("..bar bar.." =~ /(?<${uni}>bar) \\k<${uni}>/),
3710 ok($r1, "Named capture UTF (?'')");
3711 ok(defined $c1 && $c1 eq 'foo', "Named capture UTF \%+");
3712 ok($r2, "Named capture UTF (?<>)");
3713 ok(defined $c2 && $c2 eq 'bar', "Named capture UTF \%+");
3718 my ( $got, $expect, $name)=@_;
3719 my $todo = $TODO ? " # TODO $TODO" : '';
3721 $_=defined($_) ? "'$_'" : "undef"
3724 my $ok= $got eq $expect;
3726 printf "%sok %d - %s$todo\n", ($ok ? "" : "not "), $test,
3727 ($name||$Message)."\tLine ".((caller)[2]);
3729 printf "# Failed test at line %d\n".
3732 (caller)[2], $expect, $got
3739 my $s='foo bar baz';
3740 my (@k,@v,@fetch,$res);
3742 my @names=qw($+{A} $+{B} $+{C});
3743 if ($s=~/(?<A>foo)\s+(?<B>bar)?\s+(?<C>baz)/) {
3744 while (my ($k,$v)=each(%+)) {
3758 iseq($fetch[$_][0],$fetch[$_][1],$names[$_]);
3763 iseq($res,1,"$s~=/(?<A>foo)\s+(?<B>bar)?\s+(?<C>baz)/");
3764 iseq($count,3,"Got 3 keys in %+ via each");
3765 iseq(0+@k, 3, 'Got 3 keys in %+ via keys');
3766 iseq("@k","A B C", "Got expected keys");
3767 iseq("@v","bar baz foo", "Got expected values");
3769 print for $+{this_key_doesnt_exist};
3771 ok(!$@,'lvalue $+{...} should not throw an exception');
3774 my $s='foo bar baz';
3776 if ('1234'=~/(?<A>1)(?<B>2)(?<A>3)(?<B>4)/) {
3777 foreach my $name (sort keys(%-)) {
3778 my $ary = $-{$name};
3779 foreach my $idx (0..$#$ary) {
3780 push @res,"$name:$idx:$ary->[$idx]";
3784 my @expect=qw(A:0:1 A:1:3 B:0:2 B:1:4);
3785 iseq("@res","@expect","Check %-");
3787 print for $-{this_key_doesnt_exist};
3789 ok(!$@,'lvalue $-{...} should not throw an exception');
3791 # stress test CURLYX/WHILEM.
3793 # This test includes varying levels of nesting, and according to
3794 # profiling done against build 28905, exercises every code line in the
3795 # CURLYX and WHILEM blocks, except those related to LONGJMP, the
3796 # super-linear cache and warnings. It executes about 0.5M regexes
3798 if ($ENV{PERL_SKIP_PSYCHO_TEST}){
3799 printf "ok %d Skip: No psycho tests\n", $test++;
3801 print "# set PERL_SKIP_PSYCHO_TEST to skip this test\n";
3837 my $msg = "CURLYX stress test";
3839 for my $a ("x","a","aa") {
3840 for my $b ("x","bbb","bbbb") {
3842 for my $c ("x","c","cc") {
3844 for my $d ("x","d","dd") {
3846 for my $e ("x","e","ee") {
3848 for my $f ("x","f","ff") {
3850 for my $g ("x","g","gg") {
3852 for my $h ("x","h","hh") {
3854 for my $i ("x","i","ii") {
3856 for my $j ("x","j","jj") {
3858 for my $k ("x","k","kk") {
3860 for my $l ("x","l","ll") {
3864 $msg .= ": unexpected match for [$ls]";
3868 my $cap = "$1$2$3$4$5$6$7$8$9$10$11$12";
3869 unless ($ls eq $cap) {
3870 $msg .= ": capture: [$ls], got [$cap]";
3876 unless ($ls =~ /x/) {
3877 $msg = ": failed for [$ls]";
3898 ok("xaaay" !~ /xa{3\,4}y/, "\, in a pattern");
3899 ok("xa{3,4}y" =~ /xa{3\,4}y/, "\, in a pattern");
3902 ok("x\c_y" !~ /x\c\_y/, "\_ in a pattern");
3903 ok("x\c\_y" =~ /x\c\_y/, "\_ in a pattern");
3905 # \c\ followed by other characters
3906 for my $c ("z", "\0", "!", chr(254), chr(256)) {
3907 my $targ = "a\034$c";
3908 my $reg = "a\\c\\$c";
3909 ok(eval("qq/$targ/ =~ /$reg/"), "\\c\\ in pattern");
3917 while ($str=~/b/g) { $mval=$#-; $pval=$#+; $count++ }
3918 iseq($mval,0,"\@- should be empty [RT#36046]");
3919 iseq($pval,0,"\@+ should be empty [RT#36046]");
3920 iseq($count,1,"should have matched once only [RT#36046]");
3923 { # Test the (*PRUNE) pattern
3925 'aaab'=~/a+b?(?{$count++})(*FAIL)/;
3926 iseq($count,9,"expect 9 for no (*PRUNE)");
3928 'aaab'=~/a+b?(*PRUNE)(?{$count++})(*FAIL)/;
3929 iseq($count,3,"expect 3 with (*PRUNE)");
3932 1 while /.(*PRUNE)(?{$count++})(*FAIL)/g;
3933 iseq($count,4,"/.(*PRUNE)/");
3935 'aaab'=~/a+b?(??{'(*PRUNE)'})(?{$count++})(*FAIL)/;
3936 iseq($count,3,"expect 3 with (*PRUNE)");
3939 1 while /.(??{'(*PRUNE)'})(?{$count++})(*FAIL)/g;
3940 iseq($count,4,"/.(*PRUNE)/");
3942 { # Test the (*SKIP) pattern
3944 'aaab'=~/a+b?(*SKIP)(?{$count++})(*FAIL)/;
3945 iseq($count,1,"expect 1 with (*SKIP)");
3948 1 while /.(*SKIP)(?{$count++})(*FAIL)/g;
3949 iseq($count,4,"/.(*SKIP)/");
3953 1 while /(a+b?)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g;
3954 iseq($count,2,"Expect 2 with (*SKIP)" );
3955 iseq("@res","aaab aaab","adjacent (*SKIP) works as expected" );
3957 { # Test the (*SKIP) pattern
3959 'aaab'=~/a+b?(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/;
3960 iseq($count,1,"expect 1 with (*SKIP)");
3963 1 while /.(*MARK:foo)(*SKIP)(?{$count++})(*FAIL)/g;
3964 iseq($count,4,"/.(*SKIP)/");
3968 1 while /(a+b?)(*MARK:foo)(*SKIP)(?{$count++; push @res,$1})(*FAIL)/g;
3969 iseq($count,2,"Expect 2 with (*SKIP)" );
3970 iseq("@res","aaab aaab","adjacent (*SKIP) works as expected" );
3972 { # Test the (*SKIP) pattern
3974 'aaab'=~/a*(*MARK:a)b?(*MARK:b)(*SKIP:a)(?{$count++})(*FAIL)/;
3975 iseq($count,3,"expect 3 with *MARK:a)b?(*MARK:b)(*SKIP:a)");
3976 local $_='aaabaaab';
3979 1 while /(a*(*MARK:a)b?)(*MARK:x)(*SKIP:a)(?{$count++; push @res,$1})(*FAIL)/g;
3980 iseq($count,5,"Expect 5 with (*MARK:a)b?)(*MARK:x)(*SKIP:a)" );
3981 iseq("@res","aaab b aaab b ","adjacent (*MARK:a)b?)(*MARK:x)(*SKIP:a) works as expected" );
3983 { # Test the (*COMMIT) pattern
3985 'aaabaaab'=~/a+b?(*COMMIT)(?{$count++})(*FAIL)/;
3986 iseq($count,1,"expect 1 with (*COMMIT)");
3989 1 while /.(*COMMIT)(?{$count++})(*FAIL)/g;
3990 iseq($count,1,"/.(*COMMIT)/");
3994 1 while /(a+b?)(*COMMIT)(?{$count++; push @res,$1})(*FAIL)/g;
3995 iseq($count,1,"Expect 1 with (*COMMIT)" );
3996 iseq("@res","aaab","adjacent (*COMMIT) works as expected" );
3999 # Test named commits and the $REGERROR var
4001 for my $name ('',':foo')
4003 for my $pat ("(*PRUNE$name)",
4004 ($name? "(*MARK$name)" : "")
4008 for my $suffix ('(*FAIL)','')
4010 'aaaab'=~/a+b$pat$suffix/;
4013 ($suffix ? ($name ? 'foo' : "1") : ""),
4014 "Test $pat and \$REGERROR $suffix"
4021 # Test named commits and the $REGERROR var
4024 for my $name ('',':foo')
4026 for my $pat ("(*PRUNE$name)",
4027 ($name? "(*MARK$name)" : "")
4031 for my $suffix ('(*FAIL)','')
4033 'aaaab'=~/a+b$pat$suffix/;
4036 ($suffix ? ($name ? 'foo' : "1") : ""),
4037 "Test $pat and \$REGERROR $suffix"
4044 # Test named commits and the $REGERROR var
4045 local $Message = "\$REGERROR";
4047 for $word (qw(bar baz bop)) {
4049 "aaaaa$word"=~/a+(?:bar(*COMMIT:bar)|baz(*COMMIT:baz)|bop(*COMMIT:bop))(*FAIL)/;
4050 iseq($REGERROR,$word);
4053 { #Regression test for perlbug 40684
4054 local $Message = "RT#40684 tests:";
4056 my $rex = qr'^abc$'m;
4058 ok($s =~ m/^abc$/m);
4061 #Mindnumbingly simple test of (*THEN)
4063 ok(/A (*THEN) X | B (*THEN) C/x,"Simple (*THEN) test");
4068 local $Message = "Relative Recursion";
4069 my $parens=qr/(\((?:[^()]++|(?-1))*+\))/;
4070 local $_='foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))';
4071 my ($all,$one,$two)=('','','');
4072 if (/foo $parens \s* \+ \s* bar $parens/x) {
4077 iseq($one, '((2*3)+4-3)');
4078 iseq($two, '(2*(3+4)-1*(2-3))');
4079 iseq($all, 'foo((2*3)+4-3) + bar(2*(3+4)-1*(2-3))');
4084 local $_=join 'bar',$spaces,$spaces;
4086 s/(?>\s+bar)(?{$count++})//g;
4087 iseq($_,$spaces,"SUSPEND final string");
4088 iseq($count,1,"Optimiser should have prevented more than one match");
4091 local $Message = "RT#36909 test";
4095 ok('x foofoo y' =~ m{
4096 (foo) # $^R correctly set
4097 (?{ "last regexp code result" })
4099 iseq($^R,'last regexp code result');
4101 iseq($^R,'Nothing');
4105 ok('x foofoo y' =~ m{
4106 (?:foo|bar)+ # $^R correctly set
4107 (?{"last regexp code result"})
4109 iseq($^R,'last regexp code result');
4111 iseq($^R,'Nothing');
4115 ok('x foofoo y' =~ m{
4116 (foo|bar)\1+ # $^R undefined
4117 (?{"last regexp code result"})
4119 iseq($^R,'last regexp code result');
4121 iseq($^R,'Nothing');
4125 ok('x foofoo y' =~ m{
4126 (foo|bar)\1 # this time without the +
4127 (?{"last regexp code result"})
4129 iseq($^R,'last regexp code result');
4131 iseq($^R,'Nothing');
4134 local $Message="RT 22395";
4135 local $TODO = "Should be L+1 not L*(L+3)/2 (L=$l)";
4137 for my $l (10,100,1000) {
4139 ('a' x $l) =~ /(.*)(?{$count++})[bc]/;
4140 iseq( $count, $l + 1);
4144 local $Message = "RT#22614";
4147 /(.){1,}(?{push @len,0+@-})(.){1,}(?{})^/;
4148 iseq("@len","2 2 2");
4151 local $Message = "RT#18209";
4152 my $text = ' word1 word2 word3 word4 word5 word6 ';
4154 my @words = ('word1', 'word3', 'word5');
4156 foreach my $word (@words){
4157 $text =~ s/$word\s//gi; # Leave a space to seperate words in the resultant str.
4158 # The following block is not working.
4165 iseq($text,' word2 word4 word6 ');
4169 local $_= qq(A\nB\nC\n);
4171 while (m#(\G|\n)([^\n]*)\n#gsx)
4176 iseq("@res","A B C","RT#6893: /g pattern shouldn't infinite loop");
4180 # From Message-ID: <877ixs6oa6.fsf@k75.linux.bogus>
4181 my $dow_name= "nada";
4182 my $parser = "(\$dow_name) = \$time_string =~ /(D\x{e9}\\ C\x{e9}adaoin|D\x{e9}\\ Sathairn|\\w+|\x{100})/";
4183 my $time_string = "D\x{e9} C\x{e9}adaoin";
4185 ok(!$@,"Test Eval worked");
4186 iseq($dow_name,$time_string,"UTF8 trie common prefix extraction");
4191 ($v='bar')=~/(\w+)/g;
4193 iseq("$1",'bar','$1 is safe after /g - may fail due to specialized config in pp_hot.c')
4196 local $Message = "http://nntp.perl.org/group/perl.perl5.porters/118663";
4197 my $qr_barR1 = qr/(bar)\g-1/;
4198 ok("foobarbarxyz" =~ $qr_barR1);
4199 ok("foobarbarxyz" =~ qr/foo${qr_barR1}xyz/);
4200 ok("foobarbarxyz" =~ qr/(foo)${qr_barR1}xyz/);
4201 ok("foobarbarxyz" =~ qr/(foo)(bar)\g{-1}xyz/);
4202 ok("foobarbarxyz" =~ qr/(foo${qr_barR1})xyz/);
4203 ok("foobarbarxyz" =~ qr/(foo(bar)\g{-1})xyz/);
4206 local $Message = "RT#41010";
4207 my @tails=('','(?(1))','(|)','()?');
4208 my @quants=('*','+');
4212 for my $pat (@$pats) {
4213 for my $quant (@quants) {
4214 for my $tail (@tails) {
4215 my $re = "($pat$quant\$)$tail";
4216 ok(/$re/ && $1 eq $_,"'$_'=~/$re/");
4217 ok(/$re/m && $1 eq $_,"'$_'=~/$re/m");
4227 '(1|[23]|4|[56]|[78]|[90])',
4228 '(?:1|[23]|4|[56]|[78]|[90])',
4229 '(1|2|3|4|5|6|7|8|9|0)',
4230 '(?:1|2|3|4|5|6|7|8|9|0)',
4232 my @spats=('[ ]',' ','( |\t)','(?: |\t)','[ \t]','\s');
4234 my @dstrs=('12345');
4235 $doit->(\@spats,@sstrs);
4236 $doit->(\@dpats,@dstrs);
4239 local $Message = "\$REGMARK";
4241 ok('foofoo' =~ /foo (*MARK:foo) (?{push @r,$REGMARK}) /x);
4243 iseq($REGMARK,"foo");
4244 ok('foofoo' !~ /foo (*MARK:foo) (*FAIL) /x);
4246 iseq($REGERROR,'foo');
4250 $x = "abc.def.ghi.jkl";
4252 ok($x eq "abc.def.ghi");
4254 $x = "one two three four";
4255 $x =~ s/o+ \Kthree//g;
4256 ok($x eq "one two four");
4260 ok($x eq "aabbccddee");
4264 return '4' if $_[0] eq '09028623';
4267 { # Nested EVAL using PL_curpm (via $1 or friends)
4269 our $grabit = qr/ ([0-6][0-9]{7}) (??{ kt $1 }) [890] /x;
4270 $re = qr/^ ( (??{ $grabit }) ) $ /x;
4271 my @res = '0902862349' =~ $re;
4272 iseq(join("-",@res),"0902862349",
4273 'PL_curpm is set properly on nested eval');
4275 our $qr = qr/ (o) (??{ $1 }) /x;
4276 ok( 'boob'=~/( b (??{ $qr }) b )/x && 1,
4277 "PL_curpm, nested eval");
4281 use charnames ":full";
4282 ok("\N{ROMAN NUMERAL ONE}" =~ /\p{Alphabetic}/, "I =~ Alphabetic");
4283 ok("\N{ROMAN NUMERAL ONE}" =~ /\p{Uppercase}/, "I =~ Uppercase");
4284 ok("\N{ROMAN NUMERAL ONE}" !~ /\p{Lowercase}/, "I !~ Lowercase");
4285 ok("\N{ROMAN NUMERAL ONE}" =~ /\p{IDStart}/, "I =~ ID_Start");
4286 ok("\N{ROMAN NUMERAL ONE}" =~ /\p{IDContinue}/, "I =~ ID_Continue");
4287 ok("\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{Alphabetic}/, "i =~ Alphabetic");
4288 ok("\N{SMALL ROMAN NUMERAL ONE}" !~ /\p{Uppercase}/, "i !~ Uppercase");
4289 ok("\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{Lowercase}/, "i =~ Lowercase");
4290 ok("\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{IDStart}/, "i =~ ID_Start");
4291 ok("\N{SMALL ROMAN NUMERAL ONE}" =~ /\p{IDContinue}/, "i =~ ID_Continue");
4295 # requirement of Unicode Technical Standard #18, 1.7 Code Points
4296 # cf. http://www.unicode.org/reports/tr18/#Supplementary_Characters
4297 for my $u (0x7FF, 0x800, 0xFFFF, 0x10000) {
4298 no warnings 'utf8'; # oops
4300 my $x = sprintf '%04X', $u;
4301 ok( "A${c}B" =~ /A[\0-\x{10000}]B/, "unicode range - $x");
4308 if ('1' =~ /(?|(?<digit>1)|(?<digit>2))/) {
4309 $res = "@{$- {digit}}";
4312 "Check that (?|...) doesnt cause dupe entries in the names array");
4315 if ('11' =~ /(?|(?<digit>1)|(?<digit>2))(?&digit)/) {
4316 $res = "@{$- {digit}}";
4319 "Check that (?&..) to a buffer inside a (?|...) goes to the leftmost");
4324 local $Message = "ASCII pattern that really is utf8";
4326 local $SIG{__WARN__}=sub{push @w,"@_"};
4328 ok($c=~/${c}|\x{100}/);
4332 local $Message = "corruption of match results of qr// across scopes";
4333 my $qr=qr/(fo+)(ba+r)/;
4335 iseq("$1$2","foobar");
4337 'foooooobaaaaar'=~/$qr/;
4338 iseq("$1$2",'foooooobaaaaar');
4340 iseq("$1$2","foobar");
4343 local $Message = "HORIZWS";
4344 local $_="\t \r\n \n \t".chr(11)."\n";
4347 iseq($_,"hhHHhHhhHH");
4348 $_="\t \r\n \n \t".chr(11)."\n";
4352 iseq($_,"hhHHhHhhHH");
4355 local $Message = "Various whitespace special patterns";
4356 my @h=map { chr( $_ ) } (
4357 0x09, 0x20, 0xa0, 0x1680, 0x180e, 0x2000, 0x2001, 0x2002,
4358 0x2003, 0x2004, 0x2005, 0x2006, 0x2007, 0x2008, 0x2009, 0x200a,
4359 0x202f, 0x205f, 0x3000
4361 my @v=map { chr( $_ ) } ( 0x0a, 0x0b, 0x0c, 0x0d, 0x85, 0x2028, 0x2029 );
4362 my @lb=( "\x0D\x0A",
4363 map { chr( $_ ) } ( 0x0A..0x0D,0x85,0x2028,0x2029 ));
4364 foreach my $t ([\@h,qr/\h/,qr/\h+/],[\@v,qr/\v/,qr/\v+/],[\@lb,qr/\R/,qr/\R+/],){
4366 foreach my $pat (@$t) {
4367 foreach my $str (@$ary) {
4368 ok($str=~/($pat)/,$pat);
4370 utf8::upgrade($str);
4371 ok($str=~/($pat)/,"Upgraded string - $pat");
4372 iseq($1,$str,"Upgraded string - $pat");
4378 local $Message = "Check that \\xDF match properly in its various forms";
4379 # test that \xDF matches properly. this is pretty hacky stuff,
4380 # but its actually needed. the malarky with '-' is to prevent
4381 # compilation caching from playing any role in the test.
4382 my @df= (chr(0xDF),'-',chr(0xDF));
4383 utf8::upgrade($df[2]);
4384 my @strs= ('ss','sS','Ss','SS',chr(0xDF));
4385 my @ss= map { ("$_", "$_") } @strs;
4386 utf8::upgrade($ss[$_*2+1]) for 0..$#strs;
4388 for my $ssi (0..$#ss) {
4389 for my $dfi (0..$#df) {
4392 my $utf_df= ($dfi > 1) ? 'utf8' : '';
4393 my $utf_ss= ($ssi % 2) ? 'utf8' : '';
4394 (my $sstr=$str)=~s/\xDF/\\xDF/;
4396 if ($utf_df || $utf_ss || length($ss[$ssi])==1) {
4397 my $ret= $str=~/$pat/i;
4398 next if $pat eq '-';
4400 "\"$sstr\"=~/\\xDF/i (str is @{[$utf_ss||'latin']}, pat is @{[$utf_df||'latin']})");
4402 my $ret= $str !~ /$pat/i;
4403 next if $pat eq '-';
4405 "\"$sstr\"!~/\\xDF/i (str is @{[$utf_ss||'latin']}, pat is @{[$utf_df||'latin']})");
4411 local $Message = "BBC(Bleadperl Breaks CPAN) Today: String::Multibyte";
4412 my $re = qr/(?:[\x00-\xFF]{4})/;
4413 my $hyp = "\0\0\0-";
4414 my $esc = "\0\0\0\\";
4416 my $str = "$esc$hyp$hyp$esc$esc";
4417 my @a = ($str =~ /\G(?:\Q$esc$esc\E|\Q$esc$hyp\E|$re)/g);
4420 iseq(join('=', @a),"$esc$hyp=$hyp=$esc$esc");
4422 # test for keys in %+ and %-
4425 /(?<foo>a)|(?<foo>b)/;
4426 iseq( (join ",", sort keys %+), "foo" );
4427 iseq( (join ",", sort keys %-), "foo" );
4428 iseq( (join ",", sort values %+), "a" );
4429 iseq( (join ",", sort map "@$_", values %-), "a " );
4430 /(?<bar>a)(?<bar>b)(?<quux>.)/;
4431 iseq( (join ",", sort keys %+), "bar,quux" );
4432 iseq( (join ",", sort keys %-), "bar,quux" );
4433 iseq( (join ",", sort values %+), "a,c" ); # leftmost
4434 iseq( (join ",", sort map "@$_", values %-), "a b,c" );
4435 /(?<un>a)(?<deux>c)?/; # second buffer won't capture
4436 iseq( (join ",", sort keys %+), "un" );
4437 iseq( (join ",", sort keys %-), "deux,un" );
4438 iseq( (join ",", sort values %+), "a" );
4439 iseq( (join ",", sort map "@$_", values %-), ",a" );
4442 # length() on captures, the numbered ones end up in Perl_magic_len
4444 my $_ = "aoeu \xe6var ook";
4445 /^ \w+ \s (?<eek>\S+)/x;
4447 iseq( length($`), 0, 'length $`' );
4448 iseq( length($'), 4, q[length $'] );
4449 iseq( length($&), 9, 'length $&' );
4450 iseq( length($1), 4, 'length $1' );
4451 iseq( length($+{eek}), 4, 'length $+{eek} == length $1' );
4457 $ok=exists($-{x}) ? 1 : 0
4458 if 'bar'=~/(?<x>foo)|bar/;
4459 iseq($ok,1,'$-{x} exists after "bar"=~/(?<x>foo)|bar/');
4460 iseq(scalar(%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/');
4461 iseq(scalar(%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/');
4464 $ok=exists($+{x}) ? 1 : 0
4465 if 'bar'=~/(?<x>foo)|bar/;
4466 iseq($ok,0,'$+{x} not exists after "bar"=~/(?<x>foo)|bar/');
4467 iseq(scalar(%+), 0, 'scalar %+ == 0 after "bar"=~/(?<x>foo)|bar/');
4468 iseq(scalar(%-), 1, 'scalar %- == 1 after "bar"=~/(?<x>foo)|bar/');
4471 $ok=exists($-{x}) ? 1 : 0
4472 if 'foo'=~/(?<x>foo)|bar/;
4473 iseq($ok,1,'$-{x} exists after "foo"=~/(?<x>foo)|bar/');
4474 iseq(scalar(%+), 1, 'scalar %+ == 1 after "foo"=~/(?<x>foo)|bar/');
4475 iseq(scalar(%-), 1, 'scalar %- == 1 after "foo"=~/(?<x>foo)|bar/');
4478 $ok=exists($+{x}) ? 1 : 0
4479 if 'foo'=~/(?<x>foo)|bar/;
4480 iseq($ok,1,'$+{x} exists after "foo"=~/(?<x>foo)|bar/');
4484 ($_ = 'abc')=~/(abc)/g;
4486 iseq("$1",'abc',"/g leads to unsafe match vars: $1");
4489 local $Message="Message-ID: <20070818091501.7eff4831@r2d2>";
4493 $str .= "@x"; # this should ALWAYS be the empty string
4497 iseq(length($str),"0","Trie scope error, string should be empty");
4503 s/a|/push @bar, 1/e;
4505 iseq(length($str),"0","Trie scope error, string should be empty");
4508 # [perl #45605] Regexp failure with utf8-flagged and byte-flagged string
4510 my $utf_8 = "\xd6schel";
4511 utf8::upgrade($utf_8);
4512 $utf_8 =~ m{(\xd6|Ö)schel};
4513 iseq($1,"\xd6","#45605");
4517 # Regardless of utf8ness any character matches itself when
4518 # doing a case insensitive match. See also [perl #36207]
4519 for my $o (0..255) {
4520 my @ch=(chr($o),chr($o));
4521 utf8::upgrade($ch[1]);
4522 for my $u_str (0,1) {
4523 for my $u_pat (0,1) {
4524 ok( $ch[$u_str]=~/\Q$ch[$u_pat]\E/i,
4525 "\$c=~/\$c/i : chr($o) : u_str=$u_str u_pat=$u_pat");
4526 ok( $ch[$u_str]=~/\Q$ch[$u_pat]\E|xyz/i,
4527 "# \$c=~/\$c|xyz/i : chr($o) : u_str=$u_str u_pat=$u_pat");
4533 my $a = 3; "" =~ /(??{ $a })/;
4535 iseq($b, $a, "copy of scalar used for postponed subexpression");
4538 local $Message = "\$REGMARK in replacement -- Bug #49190";
4543 s/(*:X)A+|(*:Y)B+|(*:Z)C+/$REGMARK/g;
4546 if ($::running_as_thread) {
4548 print "not ok $test # TODO & SKIP: croaks when threaded\n";
4567 my $match= '<bla><blubb></blubb></bla>' =~ m/^$nested_tags$/;
4568 ok($match,'nested construct matches');
4569 iseq("@ctl_n","bla blubb",'$^N inside of (?{}) works as expected');
4570 iseq("@plus","bla blubb",'$+ inside of (?{}) works as expected');
4576 # Test counter is at bottom of file. Put new tests above here.
4577 #-------------------------------------------------------------------
4578 # Keep the following tests last -- they may crash perl
4580 # RT#19049 / RT#38869
4582 'ab cdef', # matches regex
4583 ( 'e' x 40000 ) .'ab c' # matches not, but 'ab c' matches part of it
4588 m/ab(.+)cd/i; # the ignore-case seems to be important
4589 $y = $1; # use $1, which might not be from the last match!
4590 $x = substr($list[0],$-[0],$+[0]-$-[0]);
4593 'pattern in a loop, failure should not affect previous success');
4595 'pattern in a loop, failure should not affect previous success');
4598 ok(("a" x (2**15 - 10)) =~ /^()(a|bb)*$/, "Recursive stack cracker: #24274")
4599 or print "# Unexpected outcome: should pass or crash perl\n";
4601 ok((q(a)x 100) =~ /^(??{'(.)'x 100})/,
4602 "Regexp /^(??{'(.)'x 100})/ crashes older perls")
4603 or print "# Unexpected outcome: should pass or crash perl\n";
4606 ok($@=~/\QSequence \k... not terminated in regex;\E/);
4609 local $Message = "substitution with lookahead (possible segv)";
4612 iseq($_,"ns_1ns_1ns_1");
4617 s/(?=\d+)|(?<=\d)/!Bang!/g;
4618 iseq($_,"!Bang!1!Bang!2!Bang!3!Bang!");
4621 # [perl #45337] utf8 + "[a]a{2}" + /$.../ = panic: sv_len_utf8 cache
4624 local ${^UTF8CACHE} = -1;
4627 ok("aaa" =~ /$s/, "#45337");
4630 # Put new tests above the dotted line about a page above this comment
4631 iseq(0+$::test,$::TestCount,"Got the right number of tests!");
4633 } # end of sub pat_tests
4635 # Don't forget to update this!
4637 $::TestCount = 4019;
4638 print "1..$::TestCount\n";