#!./perl
#
-# This is a home for regular expression tests that don't fit into
+# This is a home for regular expression tests that do not fit into
# the format supported by re/regexp.t. If you want to add a test
# that does fit that format, add it to re/re_tests, not here.
+BEGIN {
+ chdir 't' if -d 't';
+ @INC = qw(../lib .);
+ require './test.pl';
+ require './charset_tools.pl';
+ skip_all_if_miniperl("miniperl can't load Tie::Hash::NamedCapture, need for %+ and %-");
+}
+
use strict;
use warnings;
use 5.010;
-
+our ($REGMARK, $REGERROR);
sub run_tests;
$| = 1;
-
-BEGIN {
- chdir 't' if -d 't';
- @INC = ('../lib','.');
- require './test.pl';
- skip_all_if_miniperl("miniperl can't load Tie::Hash::NamedCapture, need for %+ and %-");
-}
-
run_tests() unless caller;
#
sub run_tests {
{
- my $message = '\C matches octet';
- $_ = "a\x{100}b";
- ok(/(.)(\C)(\C)(.)/, $message);
- is($1, "a", $message);
- if ($::IS_ASCII) { # ASCII (or equivalent), should be UTF-8
- is($2, "\xC4", $message);
- is($3, "\x80", $message);
- }
- elsif ($::IS_EBCDIC) { # EBCDIC (or equivalent), should be UTF-EBCDIC
- is($2, "\x8C", $message);
- is($3, "\x41", $message);
- }
- else {
- SKIP: {
- ok 0, "Unexpected platform", "ord ('A') =" . ord 'A';
- skip "Unexpected platform";
- }
- }
- is($4, "b", $message);
- }
-
- {
- my $message = '\C matches octet';
- $_ = "\x{100}";
- ok(/(\C)/g, $message);
- if ($::IS_ASCII) {
- is($1, "\xC4", $message);
- }
- elsif ($::IS_EBCDIC) {
- is($1, "\x8C", $message);
- }
- else {
- ok 0, "Unexpected platform", "ord ('A') = " . ord 'A';
- }
- ok(/(\C)/g, $message);
- if ($::IS_ASCII) {
- is($1, "\x80", $message);
- }
- elsif ($::IS_EBCDIC) {
- is($1, "\x41", $message);
- }
- else {
- ok 0, "Unexpected platform", "ord ('A') = " . ord 'A';
- }
- }
-
- {
# Japhy -- added 03/03/2001
() = (my $str = "abc") =~ /(...)/;
$str = "def";
# should not match.
ok "\x8e" =~ /[\x89-\x91]/, '"\x8e" =~ /[\x89-\x91]/';
ok "\xce" =~ /[\xc9-\xd1]/, '"\xce" =~ /[\xc9-\xd1]/';
+ ok "\xd0" =~ /[\xc9-\xd1]/, '"\xd0" =~ /[\xc9-\xd1]/';
skip "Not an EBCDIC platform", 2 unless ord ('i') == 0x89 &&
ord ('J') == 0xd1;
# the good of EBCDIC, so let's test these only there.
unlike("\x8e", qr/[i-j]/, '"\x8e" !~ /[i-j]/');
unlike("\xce", qr/[I-J]/, '"\xce" !~ /[I-J]/');
+ unlike("\xd0", qr/[I-J]/, '"\xd0" !~ /[I-J]/');
}
{
{
my $message = 'bug id 20001008.001';
- my @x = ("stra\337e 138", "stra\337e 138");
+ my $strasse = "stra" . uni_to_native("\337") . "e";
+ my @x = ("$strasse 138", "$strasse 138");
for (@x) {
ok(s/(\d+)\s*([\w\-]+)/$1 . uc $2/e, $message);
ok(my ($latin) = /^(.+)(?:\s+\d)/, $message);
- is($latin, "stra\337e", $message);
- ok($latin =~ s/stra\337e/straße/, $message);
+ is($latin, $strasse, $message);
+ ok($latin =~ s/$strasse/straße/, $message);
#
# Previous code follows, but outcommented - there were no tests.
#
## Should probably put in tests for all the POSIX stuff,
## but not sure how to guarantee a specific locale......
- skip "Not an ASCII platform", 2 unless $::IS_ASCII;
my $message = 'Test [[:cntrl:]]';
my $AllBytes = join "" => map {chr} 0 .. 255;
(my $x = $AllBytes) =~ s/[[:cntrl:]]//g;
+ $x = join "", sort { $a cmp $b }
+ map { chr utf8::native_to_unicode(ord $_) } split "", $x;
is($x, join("", map {chr} 0x20 .. 0x7E, 0x80 .. 0xFF), $message);
($x = $AllBytes) =~ s/[^[:cntrl:]]//g;
+ $x = join "", sort { $a cmp $b }
+ map { chr utf8::native_to_unicode(ord $_) } split "", $x;
is($x, (join "", map {chr} 0x00 .. 0x1F, 0x7F), $message);
}
}
{
- my $message = '. matches \n with /s';
- my $str1 = "foo\nbar";
- my $str2 = "foo\n\x{100}bar";
- my ($a, $b) = map {chr} $::IS_ASCII ? (0xc4, 0x80) : (0x8c, 0x41);
- my @a;
- @a = $str1 =~ /./g; is(@a, 6, $message); is("@a", "f o o b a r", $message);
- @a = $str1 =~ /./gs; is(@a, 7, $message); is("@a", "f o o \n b a r", $message);
- @a = $str1 =~ /\C/g; is(@a, 7, $message); is("@a", "f o o \n b a r", $message);
- @a = $str1 =~ /\C/gs; is(@a, 7, $message); is("@a", "f o o \n b a r", $message);
- @a = $str2 =~ /./g; is(@a, 7, $message); is("@a", "f o o \x{100} b a r", $message);
- @a = $str2 =~ /./gs; is(@a, 8, $message); is("@a", "f o o \n \x{100} b a r", $message);
- @a = $str2 =~ /\C/g; is(@a, 9, $message); is("@a", "f o o \n $a $b b a r", $message);
- @a = $str2 =~ /\C/gs; is(@a, 9, $message); is("@a", "f o o \n $a $b b a r", $message);
- }
-
- {
no warnings 'digit';
# Check that \x## works. 5.6.1 and 5.005_03 fail some of these.
my $x;
=~ /^(\X)!/ &&
$1 eq "\N{LATIN CAPITAL LETTER E}\N{COMBINING GRAVE ACCENT}", $message);
- $message = '\C and \X';
- like("!abc!", qr/a\Cc/, $message);
- like("!abc!", qr/a\Xc/, $message);
}
{
$& eq "Francais", $message);
ok("Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~ /Fran.ais/ &&
$& eq "Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais", $message);
- ok("Fran\N{LATIN SMALL LETTER C}ais" =~ /Fran\Cais/ &&
- $& eq "Francais", $message);
- # COMBINING CEDILLA is two bytes when encoded
- like("Franc\N{COMBINING CEDILLA}ais", qr/Franc\C\Cais/, $message);
ok("Fran\N{LATIN SMALL LETTER C}ais" =~ /Fran\Xais/ &&
$& eq "Francais", $message);
ok("Fran\N{LATIN SMALL LETTER C WITH CEDILLA}ais" =~ /Fran\Xais/ &&
# More whitespace: U+0085, U+2028, U+2029\n";
# U+0085, U+00A0 need to be forced to be Unicode, the \x{100} does that.
- SKIP: {
- skip "EBCDIC platform", 4 if $::IS_EBCDIC;
- # Do \x{0015} and \x{0041} match \s in EBCDIC?
- ok "<\x{100}\x{0085}>" =~ /<\x{100}\s>/, '\x{0085} in \s';
- ok "<\x{0085}>" =~ /<\v>/, '\x{0085} in \v';
- ok "<\x{100}\x{00A0}>" =~ /<\x{100}\s>/, '\x{00A0} in \s';
- ok "<\x{00A0}>" =~ /<\h>/, '\x{00A0} in \h';
- }
- my @h = map {sprintf "%05x" => $_} 0x01680, 0x0180E, 0x02000 .. 0x0200A,
+ ok "<\x{100}" . uni_to_native("\x{0085}") . ">" =~ /<\x{100}\s>/, '\x{0085} in \s';
+ ok "<" . uni_to_native("\x{0085}") . ">" =~ /<\v>/, '\x{0085} in \v';
+ ok "<\x{100}" . uni_to_native("\x{00A0}") . ">" =~ /<\x{100}\s>/, '\x{00A0} in \s';
+ ok "<" . uni_to_native("\x{00A0}") . ">" =~ /<\h>/, '\x{00A0} in \h';
+ my @h = map {sprintf "%05x" => $_} 0x01680, 0x02000 .. 0x0200A,
0x0202F, 0x0205F, 0x03000;
my @v = map {sprintf "%05x" => $_} 0x02028, 0x02029;
my @H = map {sprintf "%05x" => $_} 0x01361, 0x0200B, 0x02408, 0x02420,
- 0x0303F, 0xE0020;
+ 0x0303F, 0xE0020, 0x180E;
my @V = map {sprintf "%05x" => $_} 0x0008A .. 0x0008D, 0x00348, 0x10100,
- 0xE005F, 0xE007C;
+ 0xE005F, 0xE007C, 0x180E;
for my $hex (@h) {
my $str = eval qq ["<\\x{$hex}>"];
}
{
+ # The second half of RT #114808
+ warning_is(sub {'aa' =~ /.+\x{100}/}, undef,
+ 'utf8-only floating substr, non-utf8 target, no warning');
+ }
+
+ {
my $message = "qr /.../x";
my $R = qr / A B C # D E/x;
ok("ABCDE" =~ $R && $& eq "ABC", $message);
}
{
- # XXX DAPM 13-Apr-06. Recursive split is still broken. It's only luck it
- # hasn't been crashing. Disable this test until it is fixed properly.
- # XXX also check what it returns rather than just doing ok(1,...)
- # split /(?{ split "" })/, "abc";
- local $::TODO = "Recursive split is still broken";
- ok 0, 'cache_re & "(?{": it dumps core in 5.6.1 & 5.8.0';
- }
-
- {
ok "\x{100}\n" =~ /\x{100}\n$/, "UTF-8 length cache and fbm_compile";
}
ok "\x{100}X" =~ /$re/, "S_cl_and ANYOF_UNICODE & ANYOF_INVERTED";
my $loc_re = qq /(?l:^([^X]*)X)/;
utf8::upgrade ($loc_re);
+ no warnings 'locale';
ok "\x{100}X" =~ /$loc_re/, "locale, S_cl_and ANYOF_UNICODE & ANYOF_INVERTED";
}
}
{
- for (120 .. 130) {
+ for (120 .. 130, 240 .. 260) {
my $head = 'x' x $_;
my $message = q [Don't misparse \x{...} in regexp ] .
- q [near 127 char EXACT limit];
+ q [near EXACT char count limit];
for my $tail ('\x{0061}', '\x{1234}', '\x61') {
eval qq{like("$head$tail", qr/$head$tail/, \$message)};
is($@, '', $message);
}
$message = q [Don't misparse \N{...} in regexp ] .
- q [near 127 char EXACT limit];
+ q [near EXACT char count limit];
for my $tail ('\N{SNOWFLAKE}') {
eval qq {use charnames ':full';
like("$head$tail", qr/$head$tail/, \$message)};
}
{
- BEGIN {
- unshift @INC, 'lib';
- }
+ BEGIN {
+ unshift @INC, 'lib';
+ }
use Cname;
ok 'fooB' =~ /\N{foo}[\N{B}\N{b}]/, "Passthrough charname";
+ my $name = "foo\xDF";
+ my $result = eval "'A${name}B' =~ /^A\\N{$name}B\$/";
+ ok !$@ && $result, "Passthrough charname of non-ASCII, Latin1";
#
# Why doesn't must_warn work here?
#
my $w;
local $SIG {__WARN__} = sub {$w .= "@_"};
- eval 'q(xxWxx) =~ /[\N{WARN}]/';
- ok $w && $w =~ /Using just the first character returned by \\N{} in character class/,
- "single character in [\\N{}] warning";
+ $result = eval 'q(WARN) =~ /^[\N{WARN}]$/';
+ ok !$@ && $result && ! $w, '\N{} returning multi-char works';
undef $w;
eval q [ok "\0" !~ /[\N{EMPTY-STR}XY]/,
"Zerolength charname in charclass doesn't match \\\\0"];
ok $w && $w =~ /Ignoring zero length/,
'Ignoring zero length \N{} in character class warning';
+ undef $w;
+ eval q [ok 'xy' =~ /x[\N{EMPTY-STR} y]/x,
+ 'Empty string charname in [] is ignored; finds a following character'];
+ ok $w && $w =~ /Ignoring zero length/,
+ 'Ignoring zero length \N{} in character class warning';
+ undef $w;
+ eval q [ok 'x ' =~ /x[\N{EMPTY-STR} y]/,
+ 'Empty string charname in [] is ignored; finds a following blank under /x'];
+ ok $w && $w =~ /Ignoring zero length/,
+ 'Ignoring zero length \N{} in character class warning';
ok 'AB' =~ /(\N{EVIL})/ && $1 eq 'A', 'Charname caching $1';
ok 'ABC' =~ /(\N{EVIL})/, 'Charname caching $1';
ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/, 'Verify that long string works';
ok "\N{LONG-STR}" =~ /^\N{LONG-STR}$/i, 'Verify under folding that long string works';
- # If remove the limitation in regcomp code these should work
- # differently
+ eval '/(?[[\N{EMPTY-STR}]])/';
+ ok $@ && $@ =~ /Zero length \\N\{}/;
+
+ undef $w;
+ {
+ () = eval q ["\N{TOO MANY SPACES}"];
+ like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "Multiple spaces in a row in a charnames alias is fatal");
+ eval q [use utf8; () = "\N{TOO MANY SPACES}"];
+ like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/, "... same under utf8");
+ }
+
undef $w;
- eval q [ok "\N{LONG-STR}" =~ /^\N{TOO-LONG-STR}$/, 'Verify that too long a string fails gracefully'];
- ok $w && $w =~ /Using just the first characters returned/, 'Verify that got too-long string warning in \N{} that exceeds the limit';
+ {
+ () = eval q ["\N{TRAILING SPACE }"];
+ like ($@, qr/charnames alias definitions may not contain trailing white-space/, "Trailing white-space in a charnames alias is fatal");
+ eval q [use utf8; () = "\N{TRAILING SPACE }"];
+ like ($@, qr/charnames alias definitions may not contain trailing white-space/, "... same under utf8");
+ }
+
undef $w;
- eval q [ok "\N{LONG-STR}" =~ /^\N{TOO-LONG-STR}$/i, 'Verify under folding that too long a string fails gracefully'];
- ok $w && $w =~ /Using just the first characters returned/, 'Verify under folding that got too-long string warning in \N{} that exceeds the limit';
+ my $Cedilla_Latin1 = "GAR"
+ . uni_to_native("\xC7")
+ . "ON";
+ my $Cedilla_utf8 = $Cedilla_Latin1;
+ utf8::upgrade($Cedilla_utf8);
+ eval qq[is("\\N{$Cedilla_Latin1}", "$Cedilla_Latin1", "A cedilla in character name works")];
undef $w;
- eval q [ok "\N{TOO-LONG-STR}" !~ /^\N{TOO-LONG-STR}$/, 'Verify that too long a string doesnt work'];
- ok $w && $w =~ /Using just the first characters returned/, 'Verify that got too-long string warning in \N{} that exceeds the limit';
+ {
+ use feature 'unicode_eval';
+ eval qq[use utf8; is("\\N{$Cedilla_utf8}", "$Cedilla_utf8", "... same under 'use utf8': they work")];
+ }
+
undef $w;
- eval q [ok "\N{TOO-LONG-STR}" !~ /^\N{TOO-LONG-STR}$/i, 'Verify under folding that too long a string doesnt work'];
- ok $w && $w =~ /Using just the first characters returned/i, 'Verify under folding that got too-long string warning in \N{} that exceeds the limit';
+ my $NBSP_Latin1 = "NBSP"
+ . uni_to_native("\xA0")
+ . "SEPARATED"
+ . uni_to_native("\xA0")
+ . "SPACE";
+ my $NBSP_utf8 = $NBSP_Latin1;
+ utf8::upgrade($NBSP_utf8);
+ eval qq[is("\\N{$NBSP_Latin1}", "$NBSP_Latin1", "An NBSP in character name works")];
+ like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... but returns a deprecation warning");
undef $w;
- eval 'q(syntax error) =~ /\N{MALFORMED}/';
- ok $@ && $@ =~ /Malformed/, 'Verify that malformed utf8 gives an error';
+ {
+ use feature 'unicode_eval';
+ eval qq[use utf8; is("\\N{$NBSP_utf8}", "$NBSP_utf8", "Same under 'use utf8': they work")];
+ like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... but return a deprecation warning");
+ }
+ {
+ # disable lexical warnings
+ BEGIN { ${^WARNING_BITS} = undef; $^W = 0 }
+ undef $w;
+ () = eval qq["\\N{$NBSP_Latin1}"];
+ like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "And returns a deprecation warning outside of lexical warnings");
+ undef $w;
+ use feature 'unicode_eval';
+ eval qq[use utf8; () = "\\N{$NBSP_utf8}"];
+ like ($w, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... same under utf8");
+ }
+ {
+ no warnings 'deprecated';
+ undef $w;
+ eval qq["\\N{$NBSP_Latin1}"];
+ ok (! defined $w, "... and no warning if warnings are off");
+ use feature 'unicode_eval';
+ eval qq[use utf8; "\\N{$NBSP_utf8}"];
+ ok (! defined $w, "... same under 'use utf8'");
+ }
+ {
+ use warnings FATAL=>'deprecated';
+ () = eval qq["\\N{$NBSP_Latin1}"];
+ like ($@, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... the warning can be fatal");
+ use feature 'unicode_eval';
+ eval qq[use utf8; () = "\\N{$NBSP_utf8}"];
+ like ($@, qr/NO-BREAK SPACE in a charnames alias definition is deprecated/, "... same under utf8");
+ }
+
+ {
+ BEGIN { no strict; *CnameTest:: = *{"_charnames\0A::" } }
+ package CnameTest { sub translator { pop } }
+ BEGIN { $^H{charnames} = \&CnameTest::translator }
+ undef $w;
+ () = eval q ["\N{TOO MANY SPACES}"];
+ like ($@, qr/charnames alias definitions may not contain a sequence of multiple spaces/,
+ 'translators in _charnames\0* packages get validated');
+ }
+
+ # If remove the limitation in regcomp code these should work
+ # differently
undef $w;
+ eval q [ok "\N{TOO-LONG-STR}" =~ /^\N{TOO-LONG-STR}$/, 'Verify that what once was too long a string works'];
eval 'q() =~ /\N{4F}/';
- ok $w && $w =~ /Deprecated/, 'Verify that leading digit in name gives warning';
- undef $w;
+ ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in name gives error';
eval 'q() =~ /\N{COM,MA}/';
- ok $w && $w =~ /Deprecated/, 'Verify that comma in name gives warning';
- undef $w;
- my $name = "A\x{D7}O";
+ ok $@ && $@ =~ /Invalid character/, 'Verify that comma in name gives error';
+ $name = "A" . uni_to_native("\x{D7}") . "O";
eval "q(W) =~ /\\N{$name}/";
- ok $w && $w =~ /Deprecated/, 'Verify that latin1 symbol in name gives warning';
+ ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in name gives error';
+ my $utf8_name = "7 CITIES OF GOLD";
+ utf8::upgrade($utf8_name);
+ eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
+ ok $@ && $@ =~ /Invalid character/, 'Verify that leading digit in utf8 name gives error';
+ $utf8_name = "SHARP #";
+ utf8::upgrade($utf8_name);
+ eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
+ ok $@ && $@ =~ /Invalid character/, 'Verify that ASCII symbol in utf8 name gives error';
+ $utf8_name = "A HOUSE " . uni_to_native("\xF7") . " AGAINST ITSELF";
+ utf8::upgrade($utf8_name);
+ eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
+ ok $@ && $@ =~ /Invalid character/, 'Verify that latin1 symbol in utf8 name gives error';
+ $utf8_name = "\x{664} HORSEMEN}";
+ eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
+ ok $@ && $@ =~ /Invalid character/, 'Verify that leading above Latin1 digit in utf8 name gives error';
+ $utf8_name = "A \x{1F4A9} WOULD SMELL AS SWEET}";
+ eval "use utf8; q(W) =~ /\\N{$utf8_name}/";
+ ok $@ && $@ =~ /Invalid character/, 'Verify that above Latin1 symbol in utf8 name gives error';
+
undef $w;
- $name = "A\x{D1}O";
+ $name = "A" . uni_to_native("\x{D1}") . "O";
eval "q(W) =~ /\\N{$name}/";
ok ! $w, 'Verify that latin1 letter in name doesnt give warning';
+ # This tests the code path that restarts the parse when the recursive
+ # call to S_reg() from within S_grok_bslash_N() discovers that the
+ # pattern needs to be recalculated as UTF-8. use eval to avoid
+ # needing literal Unicode in this source file:
+ my $r = eval "qr/\\N{\x{100}\x{100}}/";
+ isnt $r, undef, "Generated regex for multi-char UTF-8 charname"
+ or diag($@);
+ ok "\x{100}\x{100}" =~ $r, "which matches";
}
{
{
my @ary = (
- pack('U', 0x00F1), # n-tilde
- '_'.pack('U', 0x00F1), # _ + n-tilde
+ pack('U', 0x00F1), # n-tilde
+ '_'.pack('U', 0x00F1), # _ + n-tilde
'c'.pack('U', 0x0327), # c + cedilla
- pack('U*', 0x00F1, 0x0327), # n-tilde + cedilla
+ pack('U*', 0x00F1, 0x0327),# n-tilde + cedilla
pack('U', 0x0391), # ALPHA
pack('U', 0x0391).'2', # ALPHA + 2
pack('U', 0x0391).'_', # ALPHA + _
{
# \, breaks {3,4}
+ no warnings qw{deprecated regexp};
ok "xaaay" !~ /xa{3\,4}y/, '\, in a pattern';
ok "xa{3,4}y" =~ /xa{3\,4}y/, '\, in a pattern';
# \c\ followed by other characters
for my $c ("z", "\0", "!", chr(254), chr(256)) {
- my $targ = "a\034$c";
+ my $targ = "a" . uni_to_native("\034") . "$c";
my $reg = "a\\c\\$c";
ok eval ("qq/$targ/ =~ /$reg/"), "\\c\\ in pattern";
}
1 while /(a+b?)(*COMMIT)(?{$count++; push @res,$1})(*FAIL)/g;
is($count, 1, "Expect 1 with (*COMMIT)");
is("@res", "aaab", "Adjacent (*COMMIT) works as expected");
+
+ ok("1\n2a\n" !~ /^\d+(*COMMIT)\w+/m, "COMMIT and anchors");
}
{
# Test named commits and the $REGERROR var
- our $REGERROR;
+ local $REGERROR;
for my $name ('', ':foo') {
for my $pat ("(*PRUNE$name)",
($name ? "(*MARK$name)" : "") . "(*SKIP$name)",
# Test named commits and the $REGERROR var
package Fnorble;
our $REGERROR;
+ local $REGERROR;
for my $name ('', ':foo') {
for my $pat ("(*PRUNE$name)",
($name ? "(*MARK$name)" : "") . "(*SKIP$name)",
{
# Test named commits and the $REGERROR var
my $message = '$REGERROR';
- our $REGERROR;
+ local $REGERROR;
for my $word (qw (bar baz bop)) {
$REGERROR = "";
"aaaaa$word" =~
{
my $message = '$REGMARK';
our @r = ();
- our ($REGMARK, $REGERROR);
+ local $REGMARK;
+ local $REGERROR;
like('foofoo', qr/foo (*MARK:foo) (?{push @r,$REGMARK}) /x, $message);
is("@r","foo", $message);
is($REGMARK, "foo", $message);
ok "a" =~ /\p{PosixLower}/, "a =~ PosixLower";
ok "a" =~ /\p{PosixLower}/i, "a =~ PosixLower under /i";
- ok "\xC0" =~ /\p{XPosixUpper}/, "\\xC0 =~ XPosixUpper";
- ok "\xC0" =~ /\p{XPosixUpper}/i, "\\xC0 =~ XPosixUpper under /i";
- ok "\xC0" !~ /\p{XPosixLower}/, "\\xC0 !~ XPosixLower";
- ok "\xC0" =~ /\p{XPosixLower}/i, "\\xC0 =~ XPosixLower under /i";
- ok "\xE0" !~ /\p{XPosixUpper}/, "\\xE0 !~ XPosixUpper";
- ok "\xE0" =~ /\p{XPosixUpper}/i, "\\xE0 =~ XPosixUpper under /i";
- ok "\xE0" =~ /\p{XPosixLower}/, "\\xE0 =~ XPosixLower";
- ok "\xE0" =~ /\p{XPosixLower}/i, "\\xE0 =~ XPosixLower under /i";
-
- ok "\xC0" =~ /\p{UppercaseLetter}/, "\\xC0 =~ UppercaseLetter";
- ok "\xC0" =~ /\p{UppercaseLetter}/i, "\\xC0 =~ UppercaseLetter under /i";
- ok "\xC0" !~ /\p{LowercaseLetter}/, "\\xC0 !~ LowercaseLetter";
- ok "\xC0" =~ /\p{LowercaseLetter}/i, "\\xC0 =~ LowercaseLetter under /i";
- ok "\xC0" !~ /\p{TitlecaseLetter}/, "\\xC0 !~ TitlecaseLetter";
- ok "\xC0" =~ /\p{TitlecaseLetter}/i, "\\xC0 =~ TitlecaseLetter under /i";
- ok "\xE0" !~ /\p{UppercaseLetter}/, "\\xE0 !~ UppercaseLetter";
- ok "\xE0" =~ /\p{UppercaseLetter}/i, "\\xE0 =~ UppercaseLetter under /i";
- ok "\xE0" =~ /\p{LowercaseLetter}/, "\\xE0 =~ LowercaseLetter";
- ok "\xE0" =~ /\p{LowercaseLetter}/i, "\\xE0 =~ LowercaseLetter under /i";
- ok "\xE0" !~ /\p{TitlecaseLetter}/, "\\xE0 !~ TitlecaseLetter";
- ok "\xE0" =~ /\p{TitlecaseLetter}/i, "\\xE0 =~ TitlecaseLetter under /i";
+ ok uni_to_native("\xC0") =~ /\p{XPosixUpper}/, "\\xC0 =~ XPosixUpper";
+ ok uni_to_native("\xC0") =~ /\p{XPosixUpper}/i, "\\xC0 =~ XPosixUpper under /i";
+ ok uni_to_native("\xC0") !~ /\p{XPosixLower}/, "\\xC0 !~ XPosixLower";
+ ok uni_to_native("\xC0") =~ /\p{XPosixLower}/i, "\\xC0 =~ XPosixLower under /i";
+ ok uni_to_native("\xE0") !~ /\p{XPosixUpper}/, "\\xE0 !~ XPosixUpper";
+ ok uni_to_native("\xE0") =~ /\p{XPosixUpper}/i, "\\xE0 =~ XPosixUpper under /i";
+ ok uni_to_native("\xE0") =~ /\p{XPosixLower}/, "\\xE0 =~ XPosixLower";
+ ok uni_to_native("\xE0") =~ /\p{XPosixLower}/i, "\\xE0 =~ XPosixLower under /i";
+
+ ok uni_to_native("\xC0") =~ /\p{UppercaseLetter}/, "\\xC0 =~ UppercaseLetter";
+ ok uni_to_native("\xC0") =~ /\p{UppercaseLetter}/i, "\\xC0 =~ UppercaseLetter under /i";
+ ok uni_to_native("\xC0") !~ /\p{LowercaseLetter}/, "\\xC0 !~ LowercaseLetter";
+ ok uni_to_native("\xC0") =~ /\p{LowercaseLetter}/i, "\\xC0 =~ LowercaseLetter under /i";
+ ok uni_to_native("\xC0") !~ /\p{TitlecaseLetter}/, "\\xC0 !~ TitlecaseLetter";
+ ok uni_to_native("\xC0") =~ /\p{TitlecaseLetter}/i, "\\xC0 =~ TitlecaseLetter under /i";
+ ok uni_to_native("\xE0") !~ /\p{UppercaseLetter}/, "\\xE0 !~ UppercaseLetter";
+ ok uni_to_native("\xE0") =~ /\p{UppercaseLetter}/i, "\\xE0 =~ UppercaseLetter under /i";
+ ok uni_to_native("\xE0") =~ /\p{LowercaseLetter}/, "\\xE0 =~ LowercaseLetter";
+ ok uni_to_native("\xE0") =~ /\p{LowercaseLetter}/i, "\\xE0 =~ LowercaseLetter under /i";
+ ok uni_to_native("\xE0") !~ /\p{TitlecaseLetter}/, "\\xE0 !~ TitlecaseLetter";
+ ok uni_to_native("\xE0") =~ /\p{TitlecaseLetter}/i, "\\xE0 =~ TitlecaseLetter under /i";
ok "\x{1C5}" !~ /\p{UppercaseLetter}/, "\\x{1C5} !~ UppercaseLetter";
ok "\x{1C5}" =~ /\p{UppercaseLetter}/i, "\\x{1C5} =~ UppercaseLetter under /i";
ok "\x{1C5}" !~ /\p{LowercaseLetter}/, "\\x{1C5} !~ LowercaseLetter";
{
# Various whitespace special patterns
- my @h = map {chr $_} 0x09, 0x20, 0xa0, 0x1680, 0x180e, 0x2000,
+ my @h = map {chr utf8::unicode_to_native($_) }
+ 0x09, 0x20, 0xa0, 0x1680, 0x2000,
0x2001, 0x2002, 0x2003, 0x2004, 0x2005, 0x2006,
0x2007, 0x2008, 0x2009, 0x200a, 0x202f, 0x205f,
0x3000;
- my @v = map {chr $_} 0x0a, 0x0b, 0x0c, 0x0d, 0x85, 0x2028,
+ my @v = map {chr utf8::unicode_to_native($_) }
+ 0x0a, 0x0b, 0x0c, 0x0d, 0x85, 0x2028,
0x2029;
- my @lb = ("\x0D\x0A", map {chr $_} 0x0A .. 0x0D, 0x85, 0x2028, 0x2029);
+ my @lb = (uni_to_native("\x0D\x0A"),
+ map {chr utf8::unicode_to_native($_) }
+ 0x0A .. 0x0D, 0x85, 0x2028, 0x2029);
foreach my $t ([\@h, qr/\h/, qr/\h+/],
[\@v, qr/\v/, qr/\v+/],
[\@lb, qr/\R/, qr/\R+/],) {
my $ary = shift @$t;
foreach my $pat (@$t) {
foreach my $str (@$ary) {
- ok $str =~ /($pat)/, $pat;
- is($1, $str, $pat);
+ my $temp_str = $str;
+ $temp_str = display($temp_str);
+ ok $str =~ /($pat)/, $temp_str . " =~ /($pat)";
+ my $temp_1 = $1;
+ is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "' after ($pat)");
utf8::upgrade ($str);
- ok $str =~ /($pat)/, "Upgraded string - $pat";
- is($1, $str, "Upgraded string - $pat");
+ ok $str =~ /($pat)/, "Upgraded " . $temp_str . " =~ /($pat)/";
+ is($1, $str, "\$1='" . display($temp_1) . "' eq '" . $temp_str . "'(upgraded) after ($pat)");
}
}
}
# Test that \xDF matches properly. this is pretty hacky stuff,
# but its actually needed. The malarky with '-' is to prevent
# compilation caching from playing any role in the test.
- my @df = (chr (0xDF), '-', chr (0xDF));
+ my @df = (chr utf8::unicode_to_native(0xDF), '-', chr utf8::unicode_to_native(0xDF));
utf8::upgrade ($df [2]);
- my @strs = ('ss', 'sS', 'Ss', 'SS', chr (0xDF));
+ my @strs = ('ss', 'sS', 'Ss', 'SS', chr utf8::unicode_to_native(0xDF));
my @ss = map {("$_", "$_")} @strs;
utf8::upgrade ($ss [$_ * 2 + 1]) for 0 .. $#strs;
my $str = $ss [$ssi];
my $utf_df = ($dfi > 1) ? 'utf8' : '';
my $utf_ss = ($ssi % 2) ? 'utf8' : '';
- (my $sstr = $str) =~ s/\xDF/\\xDF/;
+ my $sstr; # We hard-code the ebcdic value below to avoid
+ # perturbing the test
+ ($sstr = $str) =~ s/\xDF/\\xDF/ if $::IS_ASCII;
+ ($sstr = $str) =~ s/\x59/\\x59/ if $::IS_EBCDIC;
if ($utf_df || $utf_ss || length ($ss [$ssi]) == 1) {
my $ret = $str =~ /$pat/i;
next if $pat eq '-';
- ok $ret, "\"$sstr\" =~ /\\xDF/i " .
+ if ($::IS_ASCII) {
+ ok $ret, "\"$sstr\" =~ /\\xDF/i " .
+ "(str is @{[$utf_ss||'latin']}, pat is " .
+ "@{[$utf_df||'latin']})";
+ }
+ else {
+ ok $ret, "\"$sstr\" =~ /\\x59/i " .
"(str is @{[$utf_ss||'latin']}, pat is " .
"@{[$utf_df||'latin']})";
+ }
}
else {
my $ret = $str !~ /$pat/i;
next if $pat eq '-';
- ok $ret, "\"$sstr\" !~ /\\xDF/i " .
+ if ($::IS_EBCDIC) {
+ ok $ret, "\"$sstr\" !~ /\\x59/i " .
+ "(str is @{[$utf_ss||'latin']}, pat is " .
+ "@{[$utf_df||'latin']})";
+ }
+ else {
+ ok $ret, "\"$sstr\" !~ /\\xDF/i " .
"(str is @{[$utf_ss||'latin']}, pat is " .
"@{[$utf_df||'latin']})";
+ }
}
}
}
# Test for keys in %+ and %-
my $message = 'Test keys in %+ and %-';
no warnings 'uninitialized';
- my $_ = "abcdef";
+ local $_ = "abcdef";
/(?<foo>a)|(?<foo>b)/;
is((join ",", sort keys %+), "foo", $message);
is((join ",", sort keys %-), "foo", $message);
{
# length() on captures, the numbered ones end up in Perl_magic_len
- my $_ = "aoeu \xe6var ook";
+ local $_ = "aoeu " . uni_to_native("\xe6") . "var ook";
/^ \w+ \s (?<eek>\S+)/x;
is(length $`, 0, q[length $`]);
print ">$1<\n";
EOP
- local $::TODO = 'RT #86042';
fresh_perl_is(<<'EOP', ">abc<\n", {}, 'no mention of $&');
my $x;
($x='abc')=~/(abc)/g;
{
# more TRIE/AHOCORASICK problems with mixed utf8 / latin-1 and case folding
- for my $chr (160 .. 255) {
+ for my $ord (160 .. 255) {
+ my $chr = utf8::unicode_to_native($ord);
my $chr_byte = chr($chr);
my $chr_utf8 = chr($chr); utf8::upgrade($chr_utf8);
my $rx = qr{$chr_byte|X}i;
my @isPunct = grep {/[[:punct:]]/ != /\p{IsPunct}/}
map {chr} 0x80 .. 0xff;
- is(join ('', @isPunct), "\xa1\xab\xb7\xbb\xbf", # ¡ « · » ¿
+ is(join ('', @isPunct), "\xa1\xa7\xab\xb6\xb7\xbb\xbf", # ¡ « · » ¿
'IsPunct disagrees with [:punct:] outside ASCII');
my @isPunctLatin1 = eval q {
- use encoding 'latin1';
- grep {/[[:punct:]]/ != /\p{IsPunct}/} map {chr} 0x80 .. 0xff;
+ grep {/[[:punct:]]/u != /\p{IsPunct}/} map {chr} 0x80 .. 0xff;
};
skip "Eval failed ($@)", 1 if $@;
skip "PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS set to 0", 1
(?<=[=&]) (?=.)
)}iox';
is($@, '', $message);
- isa_ok($r, 'Regexp', $message);
+ object_ok($r, 'Regexp', $message);
}
# RT #82610
ok 'foo/file.fob' =~ m,^(?=[^\.])[^/]*/(?=[^\.])[^/]*\.fo[^/]$,;
- #
- # Keep the following tests last -- they may crash perl
- #
+ { # This was failing unless an explicit /d was added
+ my $E0 = uni_to_native("\xE0");
+ my $p = qr/[_$E0]/i;
+ utf8::upgrade($p);
+ like(uni_to_native("\xC0"), qr/$p/, "Verify \"\\xC0\" =~ /[\\xE0_]/i; pattern in utf8");
+ }
+
+ ok "x" =~ /\A(?>(?:(?:)A|B|C?x))\z/,
+ "Check TRIE does not overwrite EXACT following NOTHING at start - RT #111842";
+
+ {
+ my $single = ":";
+ my $upper = "\x{390}"; # Fold is 3 chars.
+ my $multi = CORE::fc($upper);
+
+ my $failed = 0;
+
+ # Try forcing a node to be split, with a multi-char fold at the
+ # boundary
+ for my $repeat (1 .. 300) {
+ my $string = $single x $repeat;
+ my $lhs = $string . $upper;
+ if ($lhs !~ m/$string$multi/i) {
+ $failed = $repeat;
+ last;
+ }
+ }
+ ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed");
+
+ $failed = 0;
+ for my $repeat (1 .. 300) {
+ my $string = $single x $repeat;
+ my $lhs = $string . "\N{LATIN SMALL LIGATURE FFI}";
+ if ($lhs !~ m/${string}ff\N{LATIN SMALL LETTER I}/i) {
+ $failed = $repeat;
+ last;
+ }
+ }
+ ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed");
+
+ $failed = 0;
+ for my $repeat (1 .. 300) {
+ my $string = $single x $repeat;
+ my $lhs = $string . "\N{LATIN SMALL LIGATURE FFL}";
+ if ($lhs !~ m/${string}ff\N{U+6c}/i) {
+ $failed = $repeat;
+ last;
+ }
+ }
+ ok(! $failed, "Matched multi-char fold across EXACTFish node boundaries; if failed, was at count $failed");
+
+ # This tests that under /d matching that an 'ss' split across two
+ # parts of a node doesn't end up turning into something that matches
+ # \xDF unless it is in utf8.
+ $failed = 0;
+ $single = 'a'; # Is non-terminal multi-char fold char
+ for my $repeat (1 .. 300) {
+ my $string = $single x $repeat;
+ my $lhs = "$string\N{LATIN SMALL LETTER SHARP S}";
+ utf8::downgrade($lhs);
+ $string .= "s";
+ if ($lhs =~ m/${string}s/di) {
+ $failed = $repeat;
+ last;
+ }
+ }
+ ok(! $failed, "Matched multi-char fold 'ss' across EXACTF node boundaries; if failed, was at count $failed");
+ }
+
+ {
+ fresh_perl_is('print eval "\"\x{101}\" =~ /[[:lower:]]/", "\n"; print eval "\"\x{100}\" =~ /[[:lower:]]/i", "\n";',
+ "1\n1", # Both re's should match
+ {},
+ "get [:lower:] swash in first eval; test under /i in second");
+ }
+
+ {
+ fresh_perl_is(<<'EOF',
+ my $s = "\x{41c}";
+ $s =~ /(.*)/ or die;
+ $ls = lc $1;
+ print $ls eq lc $s ? "good\n" : "bad: [$ls]\n";
+EOF
+ "good\n",
+ {},
+ "swash triggered by lc() doesn't corrupt \$1"
+ );
+ }
+
+ {
+ #' RT #119075
+ no warnings 'regexp'; # Silence "has useless greediness modifier"
+ local $@;
+ eval { /a{0}?/; };
+ ok(! $@,
+ "PCRE regression test: No 'Quantifier follows nothing in regex' warning");
+
+ }
+
+ {
+ unlike("\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
+ like("\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
+ unlike("\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
+ like("\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
+ unlike("\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
+ like("\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
+
+ unlike("_\xB5", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
+ like("_\xB6", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
+ unlike("_\xB7", qr/^_?\p{IsMyRuntimeProperty}\z/, "yadayada");
+ like("_\xB5", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
+ unlike("_\xB6", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
+ like("_\xB7", qr/^_?\P{IsMyRuntimeProperty}\z/, "yadayada");
+ }
+
+ # These are defined later, so won't be known at regex compile time above
+ sub IsMyRuntimeProperty {
+ return "B6\n";
+ }
+
+ sub IsntMyRuntimeProperty {
+ return "!B6\n";
+ }
+
+ { # [perl 121777]
+ my $regex;
+ { package Some;
+ # define a Unicode propertyIs_q
+ sub Is_q
+ {
+ sprintf '%x', ord 'q'
+ }
+ $regex = qr/\p{Is_q}/;
+
+ # If we uncomment the following line, prior to the patch that
+ # fixed this, everything would work because we would have expanded
+ # the property by the time the regex in the 'like' below got
+ # compiled.
+ #'q' =~ $regex;
+ }
+
+ like('q', $regex, 'User-defined property matches outside package');
+
+ package Some {
+ main::like('abcq', qr/abc$regex/, 'Run-time compiled in-package user-defined property matches');
+ }
+ }
+
+ { # From Lingua::Stem::UniNE; no ticket filed but related to #121778
+ use utf8;
+ my $word = 'рабта';
+ $word =~ s{ (?:
+ ия # definite articles for nouns:
+ | ът # ∙ masculine
+ | та # ∙ feminine
+ | то # ∙ neutral
+ | те # ∙ plural
+ ) $ }{}x;
+ is($word, 'раб', "Handles UTF8 trie correctly");
+ }
+
+ { # [perl #122460]
+ my $a = "rdvark";
+ $a =~ /(?{})(?=[A-Za-z0-9_])a*?/g;
+ is (pos $a, 0, "optimizer correctly thinks (?=...) is 0-length");
+ }
+
+ { # [perl #123417] multi-char \N{...} tripping roundly
+ use Cname;
+ my $qr = qr$(\N{foo})$;
+ "afoot" =~ eval "qr/$qr/";
+ is "$1" || $@, "foo", 'multichar \N{...} stringified and retoked';
+ }
+ { # empty \N{...} tripping roundly
+ BEGIN { $^H{charnames} = sub { "" } }
+ my $qr = qr$(a\N{foo}t)$;
+ "at" =~ eval "qr/$qr/";
+ is "$1" || $@, "at", 'empty \N{...} stringified and retoked';
+ }
+
+ is (scalar split(/\b{sb}/, "Don't think twice. It's all right."),
+ 2, '\b{wb} splits sentences correctly');
+
+
+ # !!! NOTE! Keep the following tests last -- they may crash perl
+
print "# Tests that follow may crash perl\n";
{
eval '/\k/';
ok(1, $message); # If it didn't crash, it worked.
}
+ TODO: { # Was looping
+ todo_skip('Triggers thread clone SEGV. See #86550')
+ if $::running_as_thread && $::running_as_thread;
+ watchdog(10); # Use a bigger value for busy systems
+ like("\x{00DF}", qr/[\x{1E9E}_]*/i, "\"\\x{00DF}\" =~ /[\\x{1E9E}_]*/i was looping");
+ }
+
+ { # Bug #90536, caused failed assertion
+ unlike("s\N{U+DF}", qr/^\x{00DF}/i, "\"s\\N{U+DF}\", qr/^\\x{00DF}/i");
+ }
+
+ # User-defined Unicode properties to match above-Unicode code points
+ sub Is_32_Bit_Super { return "110000\tFFFFFFFF\n" }
+ sub Is_Portable_Super { return '!utf8::Any' } # Matches beyond 32 bits
+
+ { # Assertion was failing on on 64-bit platforms; just didn't work on 32.
+ no warnings qw(non_unicode portable);
+ no warnings 'deprecated'; # These are above IV_MAX
+ use Config;
+
+ # We use 'ok' instead of 'like' because the warnings are lexically
+ # scoped, and want to turn them off, so have to do the match in this
+ # scope.
+ if ($Config{uvsize} < 8) {
+ ok(chr(0xFFFF_FFFE) =~ /\p{Is_32_Bit_Super}/,
+ "chr(0xFFFF_FFFE) can match a Unicode property");
+ ok(chr(0xFFFF_FFFF) =~ /\p{Is_32_Bit_Super}/,
+ "chr(0xFFFF_FFFF) can match a Unicode property");
+ my $p = qr/^[\x{FFFF_FFFF}]$/;
+ ok(chr(0xFFFF_FFFF) =~ $p,
+ "chr(0xFFFF_FFFF) can match itself in a [class]");
+ ok(chr(0xFFFF_FFFF) =~ $p, # Tests any caching
+ "chr(0xFFFF_FFFF) can match itself in a [class] subsequently");
+ }
+ else {
+ no warnings 'overflow';
+ ok(chr(0xFFFF_FFFF_FFFF_FFFE) =~ qr/\p{Is_Portable_Super}/,
+ "chr(0xFFFF_FFFF_FFFF_FFFE) can match a Unicode property");
+ ok(chr(0xFFFF_FFFF_FFFF_FFFF) =~ qr/^\p{Is_Portable_Super}$/,
+ "chr(0xFFFF_FFFF_FFFF_FFFF) can match a Unicode property");
+
+ my $p = qr/^[\x{FFFF_FFFF_FFFF_FFFF}]$/;
+ ok(chr(0xFFFF_FFFF_FFFF_FFFF) =~ $p,
+ "chr(0xFFFF_FFFF_FFFF_FFFF) can match itself in a [class]");
+ ok(chr(0xFFFF_FFFF_FFFF_FFFF) =~ $p, # Tests any caching
+ "chr(0xFFFF_FFFF_FFFF_FFFF) can match itself in a [class] subsequently");
+
+ # This test is because something was declared as 32 bits, but
+ # should have been cast to 64; only a problem where
+ # sizeof(STRLEN) != sizeof(UV)
+ ok(chr(0xFFFF_FFFF_FFFF_FFFE) !~ qr/\p{Is_32_Bit_Super}/, "chr(0xFFFF_FFFF_FFFF_FFFE) shouldn't match a range ending in 0xFFFF_FFFF");
+ }
+ }
+
+ { # [perl #112530], the code below caused a panic
+ sub InFoo { "a\tb\n9\ta\n" }
+ like(chr(0xA), qr/\p{InFoo}/,
+ "Overlapping ranges in user-defined properties");
+ }
+
+ { # [perl #125990], the final 2 tests below each caused a panic.
+ # The \0's are not necessary; it could be a printable character
+ # instead, but were in the ticket, so using them.
+ my $sharp_s = chr utf8::unicode_to_native(0xdf);
+ my $string = ("\0" x 8)
+ . ($sharp_s x 3)
+ . ("\0" x 42)
+ . "ý";
+ my $folded_string = ("\0" x 8)
+ . ("ss" x 3)
+ . ("\0" x 42)
+ . "ý";
+ utf8::downgrade($string);
+ utf8::downgrade($folded_string);
+
+ like($string, qr/$string/i, "LATIN SMALL SHARP S matches itself under /id");
+ unlike($folded_string, qr/$string/i, "LATIN SMALL SHARP S doesn't match 'ss' under /di");
+ like($folded_string, qr/\N{}$string/i, "\\N{} earlier than LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'");
+ like($folded_string, qr/$string\N{}/i, "\\N{} after LATIN SMALL SHARP S transforms /di into /ui, matches 'ss'");
+ }
+
+ { # Regexp:Grammars was broken:
+ # http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2013-06/msg01290.html
+ fresh_perl_like('use warnings; "abc" =~ qr{(?&foo){0}abc(?<foo>)}',
+ qr/Quantifier unexpected on zero-length expression/,
+ {},
+ 'No segfault on qr{(?&foo){0}abc(?<foo>)}');
+ }
+
+ SKIP:
+ { # [perl #125826] buffer overflow in TRIE_STORE_REVCHAR
+ # (during compilation, so use a fresh perl)
+ $Config{uvsize} == 8
+ or skip("need large code-points for this test", 1);
+
+ # This is above IV_MAX on 32 bit machines, so turn off those warnings
+ fresh_perl_is('no warnings "deprecated"; /\x{E000000000}|/ and print qq(ok\n)', "ok\n", {},
+ "buffer overflow in TRIE_STORE_REVCHAR");
+ }
+
# !!! NOTE that tests that aren't at all likely to crash perl should go
- # a ways above, above these last ones.
+ # a ways above, above these last ones. There's a comment there that, like
+ # this comment, contains the word 'NOTE'
done_testing();
} # End of sub run_tests