$| = 1;
-print "1..995\n";
+print "1..1187\n";
BEGIN {
chdir 't' if -d 't';
if ($x =~ /^abc/) {print "ok 1\n";} else {print "not ok 1\n";}
if ($x !~ /^def/) {print "ok 2\n";} else {print "not ok 2\n";}
-$* = 1;
-if ($x =~ /^def/) {print "ok 3\n";} else {print "not ok 3\n";}
-$* = 0;
+# used to be a test for $*
+if ($x =~ /^def/m) {print "ok 3\n";} else {print "not ok 3\n";}
$_ = '123';
if (/^([0-9][0-9]*)/) {print "ok 4\n";} else {print "not ok 4\n";}
if (/^$_$/) {print "ok 23\n";} else {print "not ok 23\n";}
-$* = 1; # test 3 only tested the optimized version--this one is for real
-if ("ab\ncd\n" =~ /^cd/) {print "ok 24\n";} else {print "not ok 24\n";}
-$* = 0;
+# used to be a test for $*
+if ("ab\ncd\n" =~ /^cd/m) {print "ok 24\n";} else {print "not ok 24\n";}
$XXX{123} = 123;
$XXX{234} = 234;
print "ok 263\n";
}
-{
+SKIP: {
my $test = 264; # till 575
- use charnames ':full';
+ use charnames ":full";
# This is far from complete testing, there are dozens of character
# classes in Unicode. The mixing of literals and \N{...} is
print "not " if $x =~ /[\x{100}]/;
print "ok 604\n";
- print "not " unless $x =~ /\p{InLatin1Supplement}/;
+ # the next two tests must be ignored on EBCDIC
+ print "not " unless $x =~ /\p{InLatin1Supplement}/ or ord("A") == 193;
print "ok 605\n";
- print "not " if $x =~ /\P{InLatin1Supplement}/;
+ print "not " if $x =~ /\P{InLatin1Supplement}/ and ord("A") != 193;
print "ok 606\n";
print "not " if $x =~ /\p{InLatinExtendedA}/;
print "not " unless chr(0xfb4f) =~ /\p{IsHebrew}/; # outside InHebrew
print "ok 664\n";
-print "not " unless chr(0xb5) =~ /\p{IsGreek}/; # singleton (not in a range)
-print "ok 665\n";
+# # singleton (not in a range, this test must be ignored on EBCDIC)
+# print "not " unless chr(0xb5) =~ /\p{IsGreek}/ or ord("A") == 193;
+# print "ok 665\n";
+print "ok 665 # 0xb5 moved from Greek to Common with Unicode 4.0.1\n";
print "not " unless chr(0x37a) =~ /\p{IsGreek}/; # singleton
print "ok 666\n";
}
{
- print "not " unless "a" =~ /\p{L&}/;
+ # L& and LC are the same
+ print "not " unless "a" =~ /\p{LC}/ and "a" =~ /\p{L&}/;
print "ok 743\n";
- print "not " if "1" =~ /\p{L&}/;
+ print "not " if "1" =~ /\p{LC}/ or "1" =~ /\p{L&}/;
print "ok 744\n";
}
{
my $i;
ok('-1-3-5-' eq join('', split /((??{$i++}))/, '-1-3-5-'),
- "[perl #21411] (??{ .. }) corrupts split's stack")
+ "[perl #21411] (??{ .. }) corrupts split's stack");
+ split /(?{'WOW'})/, 'abc';
+ ok('a|b|c' eq join ('|', @_),
+ "[perl #21411] (?{ .. }) version of the above");
+}
+
+{
+ split /(?{ split "" })/, "abc";
+ ok(1,'cache_re & "(?{": it dumps core in 5.6.1 & 5.8.0');
}
{
ok("\x{100}\n" =~ /\x{100}\n$/, "UTF8 length cache and fbm_compile");
}
-# last test 995
+{
+ package Str;
+ use overload q/""/ => sub { ${$_[0]}; };
+ sub new { my ($c, $v) = @_; bless \$v, $c; }
+
+ package main;
+ $_ = Str->new("a\x{100}/\x{100}b");
+ ok(join(":", /\b(.)\x{100}/g) eq "a:/", "re_intuit_start and PL_bostr");
+}
+
+{
+ $_ = "code: 'x' { '...' }\n"; study;
+ my @x; push @x, $& while m/'[^\']*'/gx;
+ ok(join(":", @x) eq "'x':'...'",
+ "[perl #17757] Parse::RecDescent triggers infinite loop");
+}
+
+{
+ my $re = qq/^([^X]*)X/;
+ utf8::upgrade($re);
+ ok("\x{100}X" =~ /$re/, "S_cl_and ANYOF_UNICODE & ANYOF_INVERTED");
+}
+
+# bug #22354
+sub func ($) {
+ ok( "a\nb" !~ /^b/, $_[0] );
+ ok( "a\nb" =~ /^b/m, "$_[0] - with /m" );
+}
+func "standalone";
+$_ = "x"; s/x/func "in subst"/e;
+$_ = "x"; s/x/func "in multiline subst"/em;
+#$_ = "x"; /x(?{func "in regexp"})/;
+#$_ = "x"; /x(?{func "in multiline regexp"})/m;
+
+# bug #19049
+$_="abcdef\n";
+@x = m/./g;
+ok("abcde" eq "$`", '# TODO #19049 - global match not setting $`');
+
+ok("123\x{100}" =~ /^.*1.*23\x{100}$/, 'uft8 + multiple floating substr');
+
+# LATIN SMALL/CAPITAL LETTER A WITH MACRON
+ok(" \x{101}" =~ qr/\x{100}/i,
+ "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+# LATIN SMALL/CAPITAL LETTER A WITH RING BELOW
+ok(" \x{1E01}" =~ qr/\x{1E00}/i,
+ "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+# DESERET SMALL/CAPITAL LETTER LONG I
+ok(" \x{10428}" =~ qr/\x{10400}/i,
+ "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+# LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'
+ok(" \x{1E01}x" =~ qr/\x{1E00}X/i,
+ "<20030808193656.5109.1@llama.ni-s.u-net.com>");
+
+{
+ # [perl #23769] Unicode regex broken on simple example
+ # regrepeat() didn't handle UTF-8 EXACT case right.
+
+ my $s = "\x{a0}\x{a0}\x{a0}\x{100}"; chop $s;
+
+ ok($s =~ /\x{a0}/, "[perl #23769]");
+ ok($s =~ /\x{a0}+/, "[perl #23769]");
+ ok($s =~ /\x{a0}\x{a0}/, "[perl #23769]");
+
+ ok("aaa\x{100}" =~ /(a+)/, "[perl #23769] easy invariant");
+ ok($1 eq "aaa", "[perl #23769]");
+
+ ok("\xa0\xa0\xa0\x{100}" =~ /(\xa0+)/, "[perl #23769] regrepeat invariant");
+ ok($1 eq "\xa0\xa0\xa0", "[perl #23769]");
+
+ ok("ababab\x{100} " =~ /((?:ab)+)/, "[perl #23769] hard invariant");
+ ok($1 eq "ababab", "[perl #23769]");
+
+ ok("\xa0\xa1\xa0\xa1\xa0\xa1\x{100}" =~ /((?:\xa0\xa1)+)/, "[perl #23769] hard variant");
+ ok($1 eq "\xa0\xa1\xa0\xa1\xa0\xa1", "[perl #23769]");
+
+ ok("aaa\x{100} " =~ /(a+?)/, "[perl #23769] easy invariant");
+ ok($1 eq "a", "[perl #23769]");
+
+ ok("\xa0\xa0\xa0\x{100} " =~ /(\xa0+?)/, "[perl #23769] regrepeat variant");
+ ok($1 eq "\xa0", "[perl #23769]");
+
+ ok("ababab\x{100} " =~ /((?:ab)+?)/, "[perl #23769] hard invariant");
+ ok($1 eq "ab", "[perl #23769]");
+
+ ok("\xa0\xa1\xa0\xa1\xa0\xa1\x{100}" =~ /((?:\xa0\xa1)+?)/, "[perl #23769] hard variant");
+ ok($1 eq "\xa0\xa1", "[perl #23769]");
+
+ ok("\xc4\xc4\xc4" !~ /(\x{100}+)/, "[perl #23769] don't match first byte of utf8 representation");
+ ok("\xc4\xc4\xc4" !~ /(\x{100}+?)/, "[perl #23769] don't match first byte of utf8 representation");
+}
+
+for (120 .. 130) {
+ my $head = 'x' x $_;
+ for my $tail ('\x{0061}', '\x{1234}') {
+ ok(
+ eval qq{ "$head$tail" =~ /$head$tail/ },
+ '\x{...} misparsed in regexp near 127 char EXACT limit'
+ );
+ }
+}
+
+# perl #25269: panic: pp_match start/end pointers
+ok("a-bc" eq eval {
+ my($x, $y) = "bca" =~ /^(?=.*(a)).*(bc)/;
+ "$x-$y";
+}, 'captures can move backwards in string');
+
+# perl #27940: \cA not recognized in character classes
+ok("a\cAb" =~ /\cA/, '\cA in pattern');
+ok("a\cAb" =~ /[\cA]/, '\cA in character class');
+ok("a\cAb" =~ /[\cA-\cB]/, '\cA in character class range');
+ok("abc" =~ /[^\cA-\cB]/, '\cA in negated character class range');
+ok("a\cBb" =~ /[\cA-\cC]/, '\cB in character class range');
+ok("a\cCbc" =~ /[^\cA-\cB]/, '\cC in negated character class range');
+ok("a\cAb" =~ /(??{"\cA"})/, '\cA in ??{} pattern');
+
+# perl #28532: optional zero-width match at end of string is ignored
+ok(("abc" =~ /^abc(\z)?/) && defined($1),
+ 'optional zero-width match at end of string');
+ok(("abc" =~ /^abc(\z)??/) && !defined($1),
+ 'optional zero-width match at end of string');
+
+
+
+{ # TRIE related
+ my @got=();
+ "words"=~/(word|word|word)(?{push @got,$1})s$/;
+ ok(@got==1,"TRIE optimation is working") or warn "# @got";
+ @got=();
+ "words"=~/(word|word|word)(?{push @got,$1})s$/i;
+ ok(@got==1,"TRIEF optimisation is working") or warn "# @got";
+
+ my @nums=map {int rand 1000} 1..100;
+ my $re="(".(join "|",@nums).")";
+ $re=qr/\b$re\b/;
+
+ foreach (@nums) {
+ ok($_=~/$re/,"Trie nums");
+ }
+ $_=join " ", @nums;
+ @got=();
+ push @got,$1 while /$re/g;
+
+ my %count;
+ $count{$_}++ for @got;
+ my $ok=1;
+ for (@nums) {
+ $ok=0 if --$count{$_}<0;
+ }
+ ok($ok,"Trie min count matches");
+}
+
+
+# TRIE related
+# LATIN SMALL/CAPITAL LETTER A WITH MACRON
+ok(("foba \x{101}foo" =~ qr/(foo|\x{100}foo|bar)/i) && $1 eq "\x{101}foo",
+ "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH MACRON");
+
+# LATIN SMALL/CAPITAL LETTER A WITH RING BELOW
+ok(("foba \x{1E01}foo" =~ qr/(foo|\x{1E00}foo|bar)/i) && $1 eq "\x{1E01}foo",
+ "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW");
+
+# DESERET SMALL/CAPITAL LETTER LONG I
+ok(("foba \x{10428}foo" =~ qr/(foo|\x{10400}foo|bar)/i) && $1 eq "\x{10428}foo",
+ "TRIEF + DESERET SMALL/CAPITAL LETTER LONG I");
+
+# LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'
+ok(("foba \x{1E01}xfoo" =~ qr/(foo|\x{1E00}Xfoo|bar)/i) && $1 eq "\x{1E01}xfoo",
+ "TRIEF + LATIN SMALL/CAPITAL LETTER A WITH RING BELOW + 'X'");
+
+{# TRIE related
+
+use charnames ':full';
+
+$s="\N{LATIN SMALL LETTER SHARP S}";
+ok(("foba ba$s" =~ qr/(foo|Ba$s|bar)/i)
+ && $1 eq "ba$s",
+ "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
+ok(("foba ba$s" =~ qr/(Ba$s|foo|bar)/i)
+ && $1 eq "ba$s",
+ "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
+ok(("foba ba$s" =~ qr/(foo|bar|Ba$s)/i)
+ && $1 eq "ba$s",
+ "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
+
+ok(("foba ba$s" =~ qr/(foo|Bass|bar)/i)
+ && $1 eq "ba$s",
+ "TRIEF + LATIN SMALL LETTER SHARP S =~ ss");
+
+ok(("foba ba$s" =~ qr/(foo|BaSS|bar)/i)
+ && $1 eq "ba$s",
+ "TRIEF + LATIN SMALL LETTER SHARP S =~ SS");
+}
+
+
+
+{
+ my @normal=qw(these are some normal words);
+ my $psycho=join "|",@normal,map chr $_,255..20000;
+ ok(('these'=~/($psycho)/) && $1 eq 'these','Pyscho');
+}
+
+# [perl #36207] mixed utf8 / latin-1 and case folding
+
+{
+ my $utf8 = "\xe9\x{100}"; chop $utf8;
+ my $latin1 = "\xe9";
+
+ ok($utf8 =~ /\xe9/i, "utf8/latin");
+ ok($utf8 =~ /$latin1/i, "utf8/latin runtime");
+ ok($utf8 =~ /(abc|\xe9)/i, "utf8/latin trie");
+ ok($utf8 =~ /(abc|$latin1)/i, "utf8/latin trie runtime");
+
+ ok("\xe9" =~ /$utf8/i, "# TODO latin/utf8");
+ ok("\xe9" =~ /(abc|$utf8)/i, "# latin/utf8 trie");
+ ok($latin1 =~ /$utf8/i, "# TODO latin/utf8 runtime");
+ ok($latin1 =~ /(abc|$utf8)/i, "# latin/utf8 trie runtime");
+}
+
+# [perl #37038] Global regular matches generate invalid pointers
+
+{
+ my $s = "abcd";
+ $s =~ /(..)(..)/g;
+ $s = $1;
+ $s = $2;
+ ok($s eq 'cd',
+ "# TODO assigning to original string should not corrupt match vars");
+}
+
+# last test 1187
+
https://perl5.git.perl.org / perl5.git / blobdiff