# run this file via a shell glob. The full format of this file is given
# in regexp.t
# Prior to the implementation of autoloading of \N{}, tests that used \N{name}
-# could not go in this file, and were farmed out to other .t's.
+# could not go in this file, and were farmed out to other .t's, where they
+# remain
#
-# pat string y/n/etc expr expected-expr skip-reason
+# pat string y/n/etc expr expected-expr skip-reason comment
__END__
abc abc y $& abc
abc abc y $-[0] 0
a[b-d] aac y $& ac
a[-b] a- y $& a-
a[b-] a- y $& a-
-a[b-a] - c - Invalid [] range \"b-a\"
+a[b-a] - c - Invalid [] range
a[]b - c - Unmatched [
a[ - c - Unmatched [
a] a] y $& a]
\By\b xy y - -
\by\B yz y - -
\By\B xyz y - -
+\B y - -
\w a y - -
\w - n - -
\W a n - -
()ef def y $&-$1 ef-
()ef def y $-[0] 1
()ef def y $+[0] 3
+()\x{100}\x{1000} d\x{100}\x{1000} y $+[0] 3
()ef def y $-[1] 1
()ef def y $+[1] 1
*a - c - Quantifier follows nothing
((a)(b)c)(d) abcd y $-[4] 3
((a)(b)c)(d) abcd y $+[4] 4
[a-zA-Z_][a-zA-Z0-9_]* alpha y $& alpha
+[_A-Z] } n - - # This could match on EBCDIC if A-Z not excluding things in middle
^a(bc+|b[eh])g|.h$ abh y $&-$1 bh-
(bc+d$|ef*g.|h?i(j|k)) effgz y $&-$1-$2 effgz-effgz-
(bc+d$|ef*g.|h?i(j|k)) ij y $&-$1-$2 ij-ij-j
'a[b-d]'i AAC y $& AC
'a[-b]'i A- y $& A-
'a[b-]'i A- y $& A-
-'a[b-a]'i - c - Invalid [] range \"b-a\"
+'a[b-a]'i - c - Invalid [] range
'a[]b'i - c - Unmatched [
'a['i - c - Unmatched [
'a]'i A] y $& A]
(?<!c)b cb n - -
(?<!c)b b y - -
(?<!c)b b y $& b
-(?<%)b - c - Sequence (?<%...) not recognized
+(?<%)b - c - Group name must start with a non-digit word character
(?:..)*a aba y $& aba
(?:..)*?a aba y $& a
^(?:b|a(?=(.)))*\1 abc y $& ab
'(ab)\d\1'i ab4Ab y $1 ab
foo\w*\d{4}baz foobar1234baz y $& foobar1234baz
a(?{})b cabd y $& ab
-a(?{)b - c - Sequence (?{...}) not terminated or not {}-balanced
-a(?{{})b - c - Sequence (?{...}) not terminated or not {}-balanced
+a(?{f()+ - c - Missing right curly or square bracket
+a(?{{1}+ - c - Missing right curly or square bracket
+ (?{(^{}) - c - syntax error
a(?{}})b - c -
-a(?{"{"})b - c - Sequence (?{...}) not terminated or not {}-balanced
+a(?{"{"})b ab y - -
a(?{"\{"})b cabd y $& ab
-a(?{"{"}})b - c - Unmatched right curly bracket
+a(?{"{"}})b - c - Sequence (?{...}) not terminated with ')'
a(?{$::bl="\{"}).b caxbd y $::bl {
x(~~)*(?:(?:F)?)? x~~ y - -
^a(?#xxx){3}c aaac y $& aaac
^(\(+)?blah(?(1)(\)))$ blah y ($2) ()
^(\(+)?blah(?(1)(\)))$ blah) n - -
^(\(+)?blah(?(1)(\)))$ (blah n - -
-(?(1?)a|b) a c - Switch condition not recognized
-(?(1)a|b|c) a c - Switch (?(condition)... contains too many branches
+(?(1?)a|b) - c - Switch condition not recognized
+(?(1)a|b|c) - c - Switch (?(condition)... contains too many branches
(?(?{0})a|b) a n - -
(?(?{0})b|a) a y $& a
(?(?{1})b|a) a n - -
(?(?!a)b|a) a y $& a
(?(?=a)b|a) a n - -
(?(?=a)a|b) a y $& a
+(?(?!\x{100})\x{100}|b) \x{100} n - -
+(?(?!\x{100})b|\x{100}) \x{100} y $& \x{100}
+(?(?=\x{100})b|\x{100}) \x{100} n - -
+(?(?=\x{100})\x{100}|b) \x{100} y $& \x{100}
(?=(a+?))(\1ab) aaab y $2 aab
^(?=(a+?))\1ab aaab n - -
(\w+:)+ one: y $1 one:
^[^bcd]*(c+) aexycd y $1 c
(?{$a=2})a*aa(?{local$a=$a+1})k*c(?{$b=$a}) yaaxxaaaacd y $b 3
(?{$a=2})(a(?{local$a=$a+1}))*aak*c(?{$b=$a}) yaaxxaaaacd y $b 4
+# [perl #117917]
+^(a(?{ "x" =~ m{x}})b)(??{ $1 }) abab y $& abab
(>a+)ab aaab n - -
(?>a+)b aaab y - -
([[:]+) a:[b]: y $1 :[
[a[:xyz: - c - Unmatched [
[a[:xyz:] - c - POSIX class [:xyz:] unknown
[a[:]b[:c] abc y $& abc
-([a[:xyz:]b]+) pbaq c - POSIX class [:xyz:] unknown
-[a[:]b[:c] abc y $& abc
+([a[:xyz:]b]+) - c - POSIX class [:xyz:] unknown
([[:alpha:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ABcd
([[:alnum:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ABcd01Xy
([[:ascii:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ABcd01Xy__-- ${nulnul}
-([[:cntrl:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ${nulnul}
+([[:cntrl:]]+) ABcd01Xy__-- ${nulnul}${ffff} ya $1 ${nulnul} ff is a control on EBCDIC (non-POSIX-BC), so this fails under /u rules
([[:digit:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 01
([[:graph:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ABcd01Xy__--
([[:lower:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 cd
([[:^alpha:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 01
((?a)[[:^alnum:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 __-- ${nulnul}${ffff}
([[:^ascii:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ${ffff}
+([:[:^ascii:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ${ffff} RT #120799
([[:^cntrl:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ABcd01Xy__--
([[:^digit:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 ABcd
([[:^lower:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 AB
([[:^xdigit:]]+) ABcd01Xy__-- ${nulnul}${ffff} y $1 Xy__-- ${nulnul}${ffff}
[[:foo:]] - c - POSIX class [:foo:] unknown
[[:^foo:]] - c - POSIX class [:^foo:] unknown
+'[[:^cntrl:]]+'u a\x80 y $& a \x80 was matching on EBCDIC platforms
((?>a+)b) aaab y $1 aaab
(?>(a+))b aaab y $1 aaa
((?>[^()]+)|\([^()]*\))+ ((abc(ade)ufh()()x y $& abc(ade)ufh()()x
(?<=x+)y - c - Variable length lookbehind not implemented
-a{37,17} - c - Can't do {n,m} with n > m
-a{37,0} - c - Can't do {n,m} with n > m
+((def){37,17})?ABC ABC y $& ABC
\Z a\nb\n y $-[0] 3
\z a\nb\n y $-[0] 4
$ a\nb\n y $-[0] 3
'abb\Z'm b\nca n - -
'abb\z'm b\nca n - -
'abb$'m b\nca n - -
+'\Aa$'m a\n\n y $& a
(^|x)(c) ca y $2 c
a*abc?xyz+pqr{3}ab{2,}xy{4,5}pq{0,6}AB{0,}zz x n - -
a(?{$a=2;$b=3;($b)=$a})b yabz y $b 2
.[X](.+)+[X][X] bbbbXXXaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa n - -
.[X][X](.+)+[X] bbbbXXXaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa n - -
tt+$ xxxtt y - -
-([a-\d]+) za-9z y $1 a-9
-([\d-z]+) a0-za y $1 0-z
-([\d-\s]+) a0- z y $1 0-
-([a-[:digit:]]+) za-9z y $1 a-9
-([[:digit:]-z]+) =0-z= y $1 0-z
-([[:digit:]-[:alpha:]]+) =0-z= y $1 0-z
+([a-\d]+) za-9z Sy $1 a-9
+([a-\d]+) - sc - False [] range
+([\d-z]+) a0-za Sy $1 0-z
+([\d-z]+) - sc $1 False [] range
+([\d-\s]+) a0- z Sy $1 0-
+([\d-\s]+) - sc $1 False [] range
+([a-[:digit:]]+) za-9z Sy $1 a-9
+([a-[:digit:]]+) - sc - False [] range
+([[:digit:]-z]+) =0-z= Sy $1 0-z
+([[:digit:]-z]+) - sc c False [] range
+([[:digit:]-[:alpha:]]+) =0-z= Sy $1 0-z
+([[:digit:]-[:alpha:]]+) - sc - False [] range
\GX.*X aaaXbX n - -
(\d+\.\d+) 3.1415926 y $1 3.1415926
(\ba.{0,10}br) have a web browser y $1 a web br
^.{2,3}?((?:b|a|r)+?)\1\z foobarbar y $1 bar
^(?:f|o|b){2,3}?((?:b|a|r)+?)\1\z foobarbar y $1 bar
.*a(?!(b|cd)*e).*f ......abef n - - # [perl #23030]
-x(?# x c - Sequence (?#... not terminated
-:x(?#: x c - Sequence (?#... not terminated
+x(?# - c - Sequence (?#... not terminated
+:x(?#: - c - Sequence (?#... not terminated
(WORDS|WORD)S WORDS y $1 WORD
(X.|WORDS|X.|WORD)S WORDS y $1 WORD
(WORDS|WORLD|WORD)S WORDS y $1 WORD
(?=foo) foo y pos 0
(?=foo) XfooY y pos 1
.*(?=foo) XfooY y pos 1
+(?=.*P)P aP y pos 2
(?<=foo) foo y pos 3
(?<=foo) XfooY y pos 4
.*(?<=foo) foo y pos 3
(?P<n>foo|bar|baz)(?P<m>[ew]+) snofooewa yM $+{m} ew miniperl cannot load Tie::Hash::NamedCapture
(?P<n>foo)|(?P<n>bar)|(?P<n>baz) snofooewa yM $+{n} foo miniperl cannot load Tie::Hash::NamedCapture
(?P<n>foo)(??{ $+{n} }) snofooefoofoowaa yM $+{n} foo miniperl cannot load Tie::Hash::NamedCapture
-(?P<=n>foo|bar|baz) snofooewa c - Sequence (?P<=...) not recognized
-(?P<!n>foo|bar|baz) snofooewa c - Sequence (?P<!...) not recognized
-(?PX<n>foo|bar|baz) snofooewa c - Sequence (?PX<...) not recognized
+(?P<=n>foo|bar|baz) - c - Group name must start with a non-digit word character
+(?P<!n>foo|bar|baz) - c - Group name must start with a non-digit word character
+(?PX<n>foo|bar|baz) - c - Sequence (?PX...) not recognized
/(?'n'foo|bar|baz)/ snofooewa y $1 foo
/(?'n'foo|bar|baz)/ snofooewa yM $+{n} foo miniperl cannot load Tie::Hash::NamedCapture
/(?'n'foo|bar|baz)(?'m'[ew]+)/ snofooewa yM $+{n} foo miniperl cannot load Tie::Hash::NamedCapture
/(ab)+((?1))(fox)/ ababfox y $1-$2-$3 ab-ab-fox
/(ab){1,100}((?1))(fox)/ ababfox y $1-$2-$3 ab-ab-fox
/(ab){0,100}((?1))(fox)/ ababfox y $1-$2-$3 ab-ab-fox
+foo(?0)?bar phoofoofoobarbarbarr y $& foofoobarbar
+foo(?R)?bar phoofoofoobarbarbarr y $& foofoobarbar
# possessive captures
a++a aaaaa n - -
a*+a aaaaa n - -
/(?<_>foo) \k<_>/ ..foo foo.. yM $+{_} foo miniperl cannot load Tie::Hash::NamedCapture
/(?'_0_'foo) \k'_0_'/ ..foo foo.. yM $+{_0_} foo miniperl cannot load Tie::Hash::NamedCapture
/(?<_0_>foo) \k<_0_>/ ..foo foo.. yM $+{_0_} foo miniperl cannot load Tie::Hash::NamedCapture
-/(?'0'foo) bar/ ..foo bar.. c - Sequence (?'
-/(?<0>foo) bar/ ..foo bar.. c - Sequence (?<
-/(?'12'foo) bar/ ..foo bar.. c - Sequence (?'
-/(?<12>foo) bar/ ..foo bar.. c - Sequence (?<
-/(?'1a'foo) bar/ ..foo bar.. c - Sequence (?'
-/(?<1a>foo) bar/ ..foo bar.. c - Sequence (?<
-/(?''foo) bar/ ..foo bar.. c - Sequence (?''
-/(?<>foo) bar/ ..foo bar.. c - Sequence (?<>
-/foo \k'n'/ foo foo c - Reference to nonexistent named group
-/foo \k<n>/ foo foo c - Reference to nonexistent named group
-/foo \k'a1'/ foo foo c - Reference to nonexistent named group
-/foo \k<a1>/ foo foo c - Reference to nonexistent named group
-/foo \k'_'/ foo foo c - Reference to nonexistent named group
-/foo \k<_>/ foo foo c - Reference to nonexistent named group
-/foo \k'_0_'/ foo foo c - Reference to nonexistent named group
-/foo \k<_0_>/ foo foo c - Reference to nonexistent named group
-/foo \k'0'/ foo foo c - Sequence \\k'
-/foo \k<0>/ foo foo c - Sequence \\k<
-/foo \k'12'/ foo foo c - Sequence \\k'
-/foo \k<12>/ foo foo c - Sequence \\k<
-/foo \k'1a'/ foo foo c - Sequence \\k'
-/foo \k<1a>/ foo foo c - Sequence \\k<
-/foo \k''/ foo foo c - Sequence \\k'
-/foo \k<>/ foo foo c - Sequence \\k<
+/(?'0'foo) bar/ - c - Group name must start with a non-digit word character
+/(?<0>foo) bar/ - c - Group name must start with a non-digit word character
+/(?'12'foo) bar/ - c - Group name must start with a non-digit word character
+/(?<12>foo) bar/ - c - Group name must start with a non-digit word character
+/(?'1a'foo) bar/ - c - Group name must start with a non-digit word character
+/(?<1a>foo) bar/ - c - Group name must start with a non-digit word character
+/(?''foo) bar/ - c - Group name must start with a non-digit word character
+/(?<>foo) bar/ - c - Group name must start with a non-digit word character
+/foo \k'n'/ - c - Reference to nonexistent named group
+/foo \k<n>/ - c - Reference to nonexistent named group
+/foo \k'a1'/ - c - Reference to nonexistent named group
+/foo \k<a1>/ - c - Reference to nonexistent named group
+/foo \k'_'/ - c - Reference to nonexistent named group
+/foo \k<_>/ - c - Reference to nonexistent named group
+/foo \k'_0_'/ - c - Reference to nonexistent named group
+/foo \k<_0_>/ - c - Reference to nonexistent named group
+/foo \k'0'/ - c - Group name must start with a non-digit word character
+/foo \k<0>/ - c - Group name must start with a non-digit word character
+/foo \k'12'/ - c - Group name must start with a non-digit word character
+/foo \k<12>/ - c - Group name must start with a non-digit word character
+/foo \k'1a'/ - c - Group name must start with a non-digit word character
+/foo \k<1a>/ - c - Group name must start with a non-digit word character
+/foo \k''/ - c - Group name must start with a non-digit word character
+/foo \k<>/ - c - Group name must start with a non-digit word character
/(?<as>as) (\w+) \k<as> (\w+)/ as easy as pie y $1-$2-$3 as-easy-pie
# \g{...} with a name as the argument
/[a\N{U+0100}]/ \x{100} y $& \x{100}
/[a\N{U+0100}]/ a y $& a
-# Verify that \N{U+...} forces Unicode semantics
+# Verify that \N{U+...} forces Unicode rules
/\N{U+41}\x{c1}/i a\x{e1} y $& a\x{e1}
/[\N{U+41}\x{c1}]/i \x{e1} y $& \x{e1}
+'\N{U+41}' A y $& A # Even for single quoted patterns
+/\N{}\xe4/i \xc4 y $& \xc4 # Empty \N{} should change /d to /u
[\s][\S] \x{a0}\x{a0} n - - # Unicode complements should not match same character
# and bypasses the lexer.
/\N{U+}/ - c - Invalid hexadecimal number
# Below currently gives a misleading message
-/[\N{U+}]/ - c - Unmatched
+/[\N{U+}]/ - Sc - Unmatched
+/[\N{U+}]/ - sc - Syntax error in (?[...])
/abc\N{def/ - c - Missing right brace
-/\N{U+4AG3}/ - c - Illegal hexadecimal digit
-/[\N{U+4AG3}]/ - c - Illegal hexadecimal digit
+/\N{U+4AG3}/ - c - Invalid hexadecimal number
+/[\N{U+4AG3}]/ - Sc - Unmatched
+/[\N{U+4AG3}]/ - sc - Syntax error in (?[...])
# And verify that in single quotes which bypasses the lexer, the regex compiler
# figures it out.
\N{U+} - c - Invalid hexadecimal number
[\N{U+}] - c - Invalid hexadecimal number
-\N{U+4AG3} - c - Illegal hexadecimal digit
+\N{U+4AG3} - c - Invalid hexadecimal number
[\N{U+4AG3}] - c - Invalid hexadecimal number
abc\N{def - c - \\N{NAME} must be resolved by the lexer
/abc\N {U+41}/x - c - Missing braces
/abc\N {SPACE}/x - c - Missing braces
-# Verifies catches hex errors, and doesn't expose our . notation to the outside
-/\N{U+0xBEEF}/ - c - Illegal hexadecimal digit
-/\N{U+BEEF.BEAD}/ - c - Illegal hexadecimal digit
+# Verifies catches hex errors
+/\N{U+0xBEEF}/ - c - Invalid hexadecimal number
+# Used to be an error, but not any more:
+/\N{U+BEEF.BEAD}/ - c -
# Verify works in single quotish context; regex compiler delivers slightly different msg
# \N{U+BEEF.BEAD} succeeds here, because can't completely hide it from the outside.
-\N{U+0xBEEF} - c - Illegal hexadecimal digit
-\c` - c - \"\\c`\" is more clearly written simply as \"\\ \"
-\c1 - c - \"\\c1\" is more clearly written simply as \"q\"
-\cA \001 y $& \1
-
+\N{U+0xBEEF} - c - Invalid hexadecimal number
+\c` - ac - \"\\c`\" is more clearly written simply as \"\\ \"
+\c1 - ac - \"\\c1\" is more clearly written simply as \"q\"
+\c@ \0 y $& \0
+\cA \001 y $& \001
+\cB \002 y $& \002
+\cC \003 y $& \003
+\cI \011 y $& \011
+\cJ \012 y $& \012
+\cR \022 y $& \022
+\cS \023 y $& \023
+\cX \030 y $& \030
+\cY \031 y $& \031
+\cZ \032 y $& \032
+\c[ \033 y $& \033
+\c\X \034X y $& \034X
+\c] \035 y $& \035
+\c^ \036 y $& \036
+\c_ \037 y $& \037
+\c? \177 ay $& \177
+\c? \x9F ey $& \x9F \c? is APC on EBCDIC
\400 \x{100} y $& \x{100}
\600 \x{180} y $& \x{180}
\777 \x{1FF} y $& \x{1FF}
[a\o{400}] \x{100} y $& \x{100}
[a\o{1000}] \x{200} y $& \x{200}
+# The below were inserting a NULL
+\87 87 c - Reference to nonexistent group in regex
+a\87 a87 c - Reference to nonexistent group in regex
+a\97 a97 c - Reference to nonexistent group in regex
+
+# avoid problems with 32-bit signed integer overflow
+
+(.)\g2147483648} x c - Reference to nonexistent group in regex
+(.)\g2147483649} x c - Reference to nonexistent group in regex
+(.)\g2147483650} x c - Reference to nonexistent group in regex
+(.)\g4294967296} x c - Reference to nonexistent group in regex
+(.)\g4294967297} x c - Reference to nonexistent group in regex
+(.)\g4294967298} x c - Reference to nonexistent group in regex
+a(.)\g2147483648} x c - Reference to nonexistent group in regex
+a(.)\g2147483649} x c - Reference to nonexistent group in regex
+a(.)\g2147483650} x c - Reference to nonexistent group in regex
+a(.)\g4294967296} x c - Reference to nonexistent group in regex
+a(.)\g4294967297} x c - Reference to nonexistent group in regex
+a(.)\g4294967298} x c - Reference to nonexistent group in regex
+
+(.)\g{2147483648} x c - Reference to nonexistent group in regex
+(.)\g{2147483649} x c - Reference to nonexistent group in regex
+(.)\g{2147483650} x c - Reference to nonexistent group in regex
+(.)\g{4294967296} x c - Reference to nonexistent group in regex
+(.)\g{4294967297} x c - Reference to nonexistent group in regex
+(.)\g{4294967298} x c - Reference to nonexistent group in regex
+a(.)\g{2147483648} x c - Reference to nonexistent group in regex
+a(.)\g{2147483649} x c - Reference to nonexistent group in regex
+a(.)\g{2147483650} x c - Reference to nonexistent group in regex
+a(.)\g{4294967296} x c - Reference to nonexistent group in regex
+a(.)\g{4294967297} x c - Reference to nonexistent group in regex
+a(.)\g{4294967298} x c - Reference to nonexistent group in regex
+
+(.)\2147483648 b\o{214}7483648 ya $1 b \d not converted to native; \o{} is
+(.)\2147483649 b\o{214}7483649 ya $1 b \d not converted to native; \o{} is
+(.)\2147483650 b\o{214}7483650 ya $1 b \d not converted to native; \o{} is
+(.)\4294967296 b\o{42}94967296 ya $1 b \d not converted to native; \o{} is
+(.)\4294967297 b\o{42}94967297 ya $1 b \d not converted to native; \o{} is
+(.)\4294967298 b\o{42}94967298 ya $1 b \d not converted to native; \o{} is
+a(.)\2147483648 ab\o{214}7483648 ya $1 b \d not converted to native; \o{} is
+a(.)\2147483649 ab\o{214}7483649 ya $1 b \d not converted to native; \o{} is
+a(.)\2147483650 ab\o{214}7483650 ya $1 b \d not converted to native; \o{} is
+a(.)\4294967296 ab\o{42}94967296 ya $1 b \d not converted to native; \o{} is
+a(.)\4294967297 ab\o{42}94967297 ya $1 b \d not converted to native; \o{} is
+a(.)\4294967298 ab\o{42}94967298 ya $1 b \d not converted to native; \o{} is
+
# The below was inserting a NULL into the character class.
-[\8\9] \000 n - -
-[\8\9] 8 y $& 8
-[\8\9] 9 y $& 9
+[\8\9] \000 Sn - -
+[\8\9] - sc $& Unrecognized escape \\8 in character class
+[\8\9] 8 Sy $& 8
+[\8\9] 9 Sy $& 9
# Verify that reads 1-3 octal digits, and that \_ works in char class
-[\0] \000 y $& \000
-[\07] \007 y $& \007
-[\07] 7\000 n - -
+[\0] \000 Sy $& \000
+[\0] - sc - Need exactly 3 octal digits
+[\07] \007 Sy $& \007
+[\07] - sc - Need exactly 3 octal digits
+[\07] 7\000 Sn - -
+[\07] - sc - Need exactly 3 octal digits
[\006] \006 y $& \006
[\006] 6\000 n - -
[\0005] \0005 y $& \000
# Normally 1E9E generates a multi-char fold, but not in inverted class;
# See [perl #89750]. This makes sure that the simple fold gets generated
# in that case, to DF.
-/[^\x{1E9E}]/i \x{DF} n - -
+/[^\x{1E9E}]/i \x{DF} Sn - -
# RT #96354
/^.*\d\H/ X1 n - -
/^\p{L}/ \x{3400} y $& \x{3400}
# RT #89774
-/[s\xDF]/ui \xDFs ybT $& \xDFs
+/[s\xDF]a/ui ssa Sy $& ssa
+/[s\xDF]a/ui sa y $& sa
# RT #99928
/^\R\x0A$/ \x0D\x0A n - -
/ff/i \x{FB01}\x{FB00} y $& \x{FB00}
/fi/i \x{FB01}\x{FB00} y $& \x{FB01}
/fi/i \x{FB00}\x{FB01} y $& \x{FB01}
+#
+# Make sure we don't see code blocks where there aren't, and vice-versa
+(?#( (?{1+)a a y - -
+'a# (?{1+'x a y - -
+ab[(?{1] ab1 y - -
+ab[(?{1\](?{2] ab2 y - -
+ab(?{"["})cd abcd y - -
+ab(??{"[x]"})cd abxcd y - -
+ab\[(??{1})c ab[1c y - -
+ab\\[(??{1;})]c ab\\;c y - -
+ab\\\[(??{1})c ab\\[1c y - -
+ab[c\](??{"]d abcd y - -
+ab[c\\](??{"[x]"})d ab\\xd y - -
+ab[c\\\](??{"x"})]{3}d ab\\](d y - -
# These test that doesn't cut-off matching too soon in the string for
# multi-char folds
/st/i \x{DF}\x{FB05} y $& \x{FB05}
/ssst/i \x{DF}\x{FB05} y $& \x{DF}\x{FB05}
+/[s]s/i \x{DF} n - -
+/s[s]/i \x{DF} n - -
# [perl #101970]
/[[:lower:]]/i \x{100} y $& \x{100}
# [perl #110648]
[^\p{Alphabetic}] \x{100} n - -
+# [perl #111400]. Tests the first Y/N boundary above 255 for each of these.
+/[[:alnum:]]/ \x{2c1} y - -
+/[[:alnum:]]/ \x{2c2} n - -
+/[[:alpha:]]/ \x{2c1} y - -
+/[[:alpha:]]/ \x{2c2} n - -
+/[[:graph:]]/ \x{377} y - -
+/[[:graph:]]/ \x{378} n - -
+/[[:lower:]]/ \x{100} n - -
+/[[:lower:]]/ \x{101} y - -
+/[[:lower:]]/ \x{102} n - -
+/[[:print:]]/ \x{377} y - -
+/[[:print:]]/ \x{378} n - -
+/[[:punct:]]/ \x{37D} n - -
+/[[:punct:]]/ \x{37E} y - -
+/[[:punct:]]/ \x{388} n - -
+/[[:upper:]]/ \x{100} y - -
+/[[:upper:]]/ \x{101} n - -
+/[[:word:]]/ \x{2c1} y - -
+/[[:word:]]/ \x{2c2} n - -
+
+# [perl #113400]
+/syntax OK\s+\z/si t/bin/good.pl syntax OK\n y - -
+
+/^(.*?)\s*\|\s*(?:\/\s*|)'(.+)'$/ text|'sec' y <$1><$2> <text><sec>
+/^(foo|)bar$/ bar y <$&> <bar>
+/^(foo||baz)bar$/ bar y <$&> <bar>
+/^(foo||baz)bar$/ bazbar y <$1> <baz>
+/^(foo||baz)bar$/ foobar y <$1> <foo>
+
+/^(?:foo|)bar$/ bar y <$&> <bar>
+/^(?:foo||baz)bar$/ bar y <$&> <bar>
+/^(?:foo||baz)bar$/ bazbar y <$&> <bazbar>
+/^(?:foo||baz)bar$/ foobar y <$&> <foobar>
+
+/^(?i:foo|)bar$/ bar y <$&> <bar>
+/^(?i:foo||baz)bar$/ bar y <$&> <bar>
+/^(?i:foo||baz)bar$/ bazbar y <$&> <bazbar>
+/^(?i:foo||baz)bar$/ foobar y <$&> <foobar>
+
+# $^N, $+ on backtrackracking
+# BRANCH
+^(.)(?:(..)|B)[CX] ABCDE y $^N-$+ A-A -
+# TRIE
+^(.)(?:BC(.)|B)[CX] ABCDE y $^N-$+ A-A -
+# CURLYX
+^(.)(?:(.)+)*[BX] ABCDE y $^N-$+ A-A -
+# CURLYM
+^(.)(BC)* ABCDE y $^N-$+ BC-BC -
+^(.)(BC)*[BX] ABCDE y $^N-$+ A-A -
+# CURLYN
+^(.)(B)*.[DX] ABCDE y $^N-$+ B-B -
+^(.)(B)*.[CX] ABCDE y $^N-$+ A-A -
+
+# using 'return' in code blocks
+^(A)(?{"xyz"})B$ AB y $1-$^R A-xyz -
+^(A)(?{return "xyz"})B$ AB y $1-$^R A-xyz -
+^(A)((??{"xyz"}))$ Axyz y $1-$2 A-xyz -
+^(A)((??{return "xyz"}))$ Axyz y $1-$2 A-xyz -
+^(A)((?(?{1})abc|xyz))$ Aabc y $1-$2 A-abc -
+^(A)((?(?{0})abc|xyz))$ Axyz y $1-$2 A-xyz -
+^(A)((?(?{return 1})abc|xyz))$ Aabc y $1-$2 A-abc -
+^(A)((?(?{return 0})abc|xyz))$ Axyz y $1-$2 A-xyz -
+
+# pattern modifier flags should propagate into returned (??{}) pattern
+# p,d,l not tested
+
+/^(a)((??{"b"}))$/i AB y $1-$2 A-B -
+/^(A)((??{'B$'}))(\nC)$/m AB\nC y $1-$2-$3 A-B-\nC -
+/^(A)((??{'.'}))(B)$/s A\nB y $1-$2-$3 A-\n-B -
+/^(A) ((??{' .'}))(B)$/x A B y $1-$2-$3 A- -B -
+/^((??{'\d'}))$/a \x{660} n - -
+/^(??{"s"})$/i \x{17F} y - -
+/^(??{"s"})$/ia \x{17F} y - -
+/^(??{"s"})$/iaa \x{17F} n - -
+/^(??{'\w'})$/u \x{AA} y - -
+
+# #113670 ensure any captures to the right are invalidated when CURLY
+# and CURLYM backtrack
+
+^(?:(X)?(\d)|(X)?(\d\d))$ X12 y $1-$2-$3-$4 --X-12
+^(?:(XX)?(\d)|(XX)?(\d\d))$ XX12 y $1-$2-$3-$4 --XX-12
+
+# rt 113770
+\A(?>\[(?:(?:)(?:R){1}|T|V?|A)\])\z [A] y $& [A]
+# rt 114068
+/( [^z] $ [^z]+)/xm aa\nbb\ncc\n y $1 a\nbb\ncc\n
+
+# [perl #114220]
+/[\h]/ \x{A0} y $& \xA0
+/[\H]/ \x{BF} y $& \xBF
+/[\H]/ \x{A0} n - -
+/[\H]/ \x{A1} y $& \xA1
+
+[^\n]+ \nb y $& b
+[^\n]+ a\n y $& a
+
+# /a has no effect on properties
+(?a:\p{Any}) \x{100} y $& \x{100}
+(?aa:\p{Any}) \x{100} y $& \x{100}
+
+\w \x{200C} y $& \x{200C}
+\W \x{200C} n - -
+\w \x{200D} y $& \x{200D}
+\W \x{200D} n - -
+
+/^(?d:\xdf|_)*_/i \x{17f}\x{17f}_ y $& \x{17f}\x{17f}_
+#
+# check that @-, @+ count chars, not bytes; especially if beginning of
+# string is not copied
+
+(\x{100}) \x{2000}\x{2000}\x{2000}\x{100} y $-[0]:$-[1]:$+[0]:$+[1] 3:3:4:4
+
+^\R{2}$ \r\n\r\n y $& \r\n\r\n
+
+/^\D{11}/a \x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF} n - -
+/^\S{11}/a \x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF} n - -
+/^\W{11}/a \x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF}\x{10FFFF} n - -
+
+# [ perl #114272]
+\Vn \xFFn/ y $& \xFFn
+
+# These mostly exercize various paths in the optimizer
+/(?l:a?\w)/ b y $& b
+/(?l:a?\D)/ b y $& b
+^_?[^\S_0-9]\w*\z \t y $& \t
+m?^xy\?$? xy? y $& xy?
+a?\X a\x{100} y $& a\x{100}
+a?\R a\n y $& a\n
+^a?\N$ a\n y $& a
+\N?a \na y $& a
+/\s?\s/l \t y $& \t
+/\s?\d/l 3 y $& 3
+/\s?\S/l A y $& A
+
+# check we have the right utf8ness as we switch back and forth between
+# patterns
+^(\x{100}|a)(??{ qr/.?\xF7/d}) a_\xF7 y - -
+
+# [perl #118375]
+/(a|(bc)){0,0}?xyz/ xyz y $& xyz
+/( a | ( bc ) ) {0,0} ? xyz/x xyz y $& xyz
+
+/(a|(bc)){0,0}+xyz/ xyz y $& xyz
+/( a | ( bc ) ) {0,0} + xyz/x xyz y $& xyz
+
+(A(*PRUNE)B|A(*PRUNE)C) AC n - -
+(A(*PRUNE)B|A(*PRUNE)D|A(*PRUNE)C) AC n - -
+(A(*PRUNE)B|A(*PRUNE)C|A(*PRUNE)D) AC n - -
+((A(*PRUNE)B|A(*PRUNE)C)) AC n - -
+((A(*PRUNE)B|A(*PRUNE)D|A(*PRUNE)C)) AC n - -
+((A(*PRUNE)B|A(*PRUNE)C|A(*PRUNE)D)) AC n - -
+
+A+?(*THEN)BC AAABC y $& ABC
+A+?(*PRUNE)BC AAABC y $& ABC
+A+(*THEN)BC AAABC y $& AAABC
+A+(*PRUNE)BC AAABC y $& AAABC
+A+?(*THEN)BC(?{}) AAABC y $& ABC
+A+?(*PRUNE)BC(?{}) AAABC y $& ABC
+A+(*THEN)BC(?{}) AAABC y $& AAABC
+A+(*PRUNE)BC(?{}) AAABC y $& AAABC
+/[#]/ a#b y $& #
+/[#]b/ a#b y $& #b
+/[#]/x a#b y $& #
+/[#]b/x a#b y $& #b
+/[#](?{})/x a#b y $& #
+/[#](??{'b'})/x a#b y $& #b
+/(?#)(?{})b/ a#b y $& b
+/(?#)(??{'b'})/ a#b y $& b
+/[(?#](?{})b/ a#b y $& #b
+/[(?#](??{'b'})/ a#b y $& #b
+/(?#)(?{})b/x a#b y $& b
+/(?#)(??{'b'})/x a#b y $& b
+
+# RT #119069: PCRE regression test: no segfault
+^(?=(a)){0}b(?1) back y $& ba
+# RT #119071: PCRE regression test: No warning of infinite regression
+(?:(a(*SKIP)b)){0}(?:(?1)|ac) x n - -
+# RT #119073: PCRE regression test: {0} => NOTHING optimization
+(?1)(?:(b)){0} b y $& b
+# RT #120041
+^A*\z y $&
+/\N* /x ab y $& ab # Under /x was ignoring the '*'
+/\N (?#comment) * /x ab y $& ab # likewise
+
+# RT #120600: Variable length lookbehind is not variable
+(?<W>a)(?<BB>(?=(?&W))(?<=(?&W)))(?&BB) aa y $& a # test repeated recursive patterns
+
+# This group is from RT #121144
+/^\S+=/d \x{3a3}=\x{3a0} y $& \x{3a3}=
+/^\S+=/u \x{3a3}=\x{3a0} y $& \x{3a3}=
+
+# utf8 cache length panics
+\x{100}[xy]\x{100}{2} \x{100}y\x{100}\x{100} y $& \x{100}y\x{100}\x{100}
+\x{100}a{2,3} \x{100}aaa y $& \x{100}aaa
+^x?abc?de abcde y $& abcde
+
+'(?-m:^abc)'m abcde y $& abc
+'(?-m:^abc)'m x\nabcde n - -
+
+# RT #121248 - segfault
+/^[+]([^(]+)$/mi li n - -
+
+# RT #121299 - Inconsistent behavior with backreferences nested inside subpattern references
+/(.)((o)\1)(?2)/ fofof y $2 of
+/(.)(?2)((\1)(?4)(\1))/ fffffff y $1 f
+/(.)(?2)((.)(?4)(\1))/ foffoff y $2 off
+/^(.\2?)(.)(?1)$/ abcb y $2 b
+/^(.\1?)(?1)$/ aba y $1 a
+/^ (\3(?2)\3)? ((.)) (?1) $/x aaba y $2 a
+/^ (a|\3(?1)\2|(?2)) ((b|c)(?4)?) (?1) (d(?1)) $/x abbcdcabbda y $2 b
+
+# RT #121321 - perl 5.19.10 infinite loops on backrefs > number of capture buffers (npar) where npar>9
+/(a)\2/ a\2 c - Reference to nonexistent group in regex
+/(a)(b)\3/ ab\3 c - Reference to nonexistent group in regex
+/(a)(b)(c)\4/ abc\4 c - Reference to nonexistent group in regex
+/(a)(b)(c)(d)\5/ abcd\5 c - Reference to nonexistent group in regex
+/(a)(b)(c)(d)(e)\6/ abcde\6 c - Reference to nonexistent group in regex
+/(a)(b)(c)(d)(e)(f)\7/ abcdef\7 c - Reference to nonexistent group in regex
+/(a)(b)(c)(d)(e)(f)(g)(h)(i)\10/ abcdefghi\10 y $& abcdefghi\10
+/(a)(b)(c)(d)(e)(f)(g)(h)(i)(j)\11/ abcdefghij\11 y $& abcdefghij\11
+/(a)(b)(c)(d)(e)(f)(g)(h)(i)(j)(k)\12/ abcdefghijk\12 y $& abcdefghijk\12
+/(a)(b)(c)(d)(e)(f)(g)(h)(i)(j)(k)\12\13\14/ abcdefghijk\12\13\14 y $& abcdefghijk\12\13\14
+
+\d<(.*?)> a<> n - -
+[bcd].{2,3}aaaa XbXaaaaa y - -
+[bcd].{2,3}aaaa Xb\x{100}aaaaa y - -
+'\Awibble\z'm wibble y - -
+/(alias|status)es$/i Statuses y $1 Status # [perl #121778]
+/( (?&solution) | % ) \Z (?(DEFINE) (?<solution>7\%\ solution) )/x 7% solution y $1 7% solution # [perl #122890]
+(.)(?{$~=$^N}) \x{100} y $~ \x{100} # [perl #123135]
+
+/[a-z]/i \N{KELVIN SIGN} y $& \N{KELVIN SIGN}
+/[A-Z]/ia \N{KELVIN SIGN} y $& \N{KELVIN SIGN}
+/[A-Z]/iaa \N{KELVIN SIGN} n - -
+/[A-Z]/i \N{LATIN SMALL LETTER LONG S} y $& \N{LATIN SMALL LETTER LONG S}
+
+# RT #123840: these used to hang in re_intuit_start
+/w\zxy?\z/i \x{100}a\x{80}a n - -
+/w\z\R\z/i \x{100}a\x{80}a n - -
+
+/(a+){1}+a/ aaa n - - # [perl #125825]
+
+^((?(?=x)xb|ya)z) xbz y $1 xbz
+^((?(?=x)xb|ya)z) yaz y $1 yaz
+^((?(?!y)xb|ya)z) xbz y $1 xbz
+^((?(?!y)xb|ya)z) yaz y $1 yaz
+^((?(?!)xb|ya)z) xbz n - -
+^((?(?!)xb|ya)z) yaz y $1 yaz # [perl-126222]
+
+foo(*ACCEPT:foo) foo y $::REGMARK foo
+(foo(*ACCEPT:foo)) foo y $::REGMARK foo
+A(*FAIL:foo)[BC] A n $::REGERROR foo
+# Keep these lines at the end of the file
# vim: softtabstop=0 noexpandtab
https://perl5.git.perl.org / perl5.git / blobdiff