Update bundled PCRE2-library to version 10.23
Some manual changes done to the library were lost with this update. They will be added in the next commit.
This commit is contained in:
Esa Korhonen
434
pcre2/testdata/testoutput5
vendored
434
pcre2/testdata/testoutput5
vendored
@ -3,6 +3,8 @@
|
||||
# results in 8-bit, 16-bit, and 32-bit modes are excluded (see tests 10 and
|
||||
# 12).
|
||||
|
||||
#newline_default lf any anycrlf
|
||||
|
||||
# PCRE2 and Perl disagree about the characteristics of certain Unicode
|
||||
# characters. For example, 061C is considered by Perl to be Arabic, though
|
||||
# is it not listed as such in the Unicode Scripts.txt file, and 2066-2069 are
|
||||
@ -11,14 +13,12 @@
|
||||
# test 4.
|
||||
|
||||
/^[\p{Arabic}]/utf
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x{061c}
|
||||
No match
|
||||
|
||||
/^[[:graph:]]+$/utf,ucp
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x{61c}
|
||||
No match
|
||||
\x{2066}
|
||||
@ -31,8 +31,7 @@ No match
|
||||
No match
|
||||
|
||||
/^[[:print:]]+$/utf,ucp
|
||||
** Failers
|
||||
0: ** Failers
|
||||
\= Expect no match
|
||||
\x{61c}
|
||||
No match
|
||||
\x{2066}
|
||||
@ -76,6 +75,7 @@ No match
|
||||
0: A\x{85}\x{2005}Z
|
||||
|
||||
/^[[:graph:]]+$/utf,ucp
|
||||
\= Expect no match
|
||||
\x{180e}
|
||||
No match
|
||||
|
||||
@ -88,6 +88,7 @@ No match
|
||||
0: \x{09}\x{0a}\x{1d} \x{85}\x{a0}\x{61c}\x{1680}\x{180e}
|
||||
|
||||
/^[[:^print:]]+$/utf,ucp
|
||||
\= Expect no match
|
||||
\x{180e}
|
||||
No match
|
||||
|
||||
@ -182,10 +183,6 @@ Subject length lower bound = 3
|
||||
\x{212ab}\x{212ab}\x{212ab}\x{861}
|
||||
0: \x{212ab}\x{212ab}\x{212ab}
|
||||
|
||||
/(?<=\C)X/utf
|
||||
Failed: error 136 at offset 6: \C is not allowed in a lookbehind assertion
|
||||
Should produce an error diagnostic
|
||||
|
||||
/^[ab]/IB,utf
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
@ -200,8 +197,7 @@ Overall options: anchored utf
|
||||
Subject length lower bound = 1
|
||||
bar
|
||||
0: b
|
||||
*** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
c
|
||||
No match
|
||||
\x{ff}
|
||||
@ -227,8 +223,7 @@ Subject length lower bound = 1
|
||||
0: \x{ff}
|
||||
\x{100}
|
||||
0: \x{100}
|
||||
*** Failers
|
||||
0: *
|
||||
\= Expect no match
|
||||
aaa
|
||||
No match
|
||||
|
||||
@ -251,8 +246,7 @@ No match
|
||||
\x{100}\x{100}"12"
|
||||
0: \x{100}\x{100}"12"
|
||||
1: "12"
|
||||
*** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x{100}\x{100}abcd
|
||||
No match
|
||||
|
||||
@ -303,8 +297,7 @@ Failed: error 108 at offset 15: range out of order in character class
|
||||
0: \x{100}
|
||||
\x{104}
|
||||
0: \x{104}
|
||||
*** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x{105}
|
||||
No match
|
||||
\x{ff}
|
||||
@ -581,8 +574,7 @@ Matched, but too many substrings
|
||||
0: a\x{2028}b
|
||||
a\x{2029}b
|
||||
0: a\x{2029}b
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
a\n\rb
|
||||
No match
|
||||
|
||||
@ -623,8 +615,7 @@ No match
|
||||
0: a\x{0a}\x{0d}b
|
||||
a\n\r\x{85}\x0cb
|
||||
0: a\x{0a}\x{0d}\x{85}\x{0c}b
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
ab
|
||||
No match
|
||||
|
||||
@ -643,8 +634,7 @@ No match
|
||||
0: a\x{0a}\x{0d}\x{0a}\x{0d}b
|
||||
a\n\n\r\nb
|
||||
0: a\x{0a}\x{0a}\x{0d}\x{0a}b
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
a\n\n\n\rb
|
||||
No match
|
||||
a\r
|
||||
@ -655,8 +645,7 @@ No match
|
||||
0: X X\x{0a}
|
||||
X\x09X\x0b
|
||||
0: X\x{09}X\x{0b}
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x{a0} X\x0a
|
||||
No match
|
||||
|
||||
@ -667,8 +656,7 @@ No match
|
||||
0: \x{09} \x{a0}\x{0a}\x{0b}\x{0c}\x{0d}
|
||||
\x09\x20\x{a0}\x0a\x0b\x0c
|
||||
0: \x{09} \x{a0}\x{0a}\x{0b}\x{0c}
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x09\x20\x{a0}\x0a\x0b
|
||||
No match
|
||||
|
||||
@ -677,8 +665,7 @@ No match
|
||||
0: \x{3001}\x{3000}\x{2030}\x{2028}
|
||||
X\x{180e}X\x{85}
|
||||
0: X\x{180e}X\x{85}
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x{2009} X\x0a
|
||||
No match
|
||||
|
||||
@ -689,8 +676,7 @@ No match
|
||||
0: \x{09}\x{205f}\x{a0}\x{0a}\x{2029}\x{0c}\x{2028}
|
||||
\x09\x20\x{202f}\x0a\x0b\x0c
|
||||
0: \x{09} \x{202f}\x{0a}\x{0b}\x{0c}
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
\x09\x{200a}\x{a0}\x{2028}\x0b
|
||||
No match
|
||||
|
||||
@ -755,8 +741,7 @@ Subject length lower bound = 3
|
||||
0: a\x{0a}b
|
||||
a\r\nb
|
||||
0: a\x{0d}\x{0a}b
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
a\x{85}b
|
||||
No match
|
||||
a\x0bb
|
||||
@ -793,8 +778,7 @@ Subject length lower bound = 2
|
||||
0: a\x{0a}b
|
||||
a\r\nb
|
||||
0: a\x{0d}\x{0a}b
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
a\x{85}b
|
||||
No match
|
||||
a\x0bb
|
||||
@ -817,14 +801,11 @@ Subject length lower bound = 2
|
||||
0: a\x{85}b
|
||||
a\x0bb
|
||||
0: a\x{0b}b
|
||||
** Failers
|
||||
No match
|
||||
|
||||
/.*a.*=.b.*/utf,newline=any
|
||||
QQQ\x{2029}ABCaXYZ=!bPQR
|
||||
0: ABCaXYZ=!bPQR
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
a\x{2029}b
|
||||
No match
|
||||
\x61\xe2\x80\xa9\x62
|
||||
@ -838,8 +819,7 @@ Failed: error 130 at offset 3: unknown POSIX class name
|
||||
0: a\x{1234}b
|
||||
a\nb
|
||||
0: a\x{0a}b
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
ab
|
||||
No match
|
||||
|
||||
@ -848,8 +828,7 @@ No match
|
||||
0: aXb
|
||||
a\nX\nX\x{1234}b
|
||||
0: a\x{0a}X\x{0a}X\x{1234}b
|
||||
** Failers
|
||||
No match
|
||||
\= Expect no match
|
||||
ab
|
||||
No match
|
||||
|
||||
@ -935,6 +914,7 @@ Partial match: X\x{123}\x{123}\x{123}
|
||||
Partial match: X\x{123}\x{123}\x{123}\x{123}
|
||||
|
||||
/X\x{123}{2,4}b/utf
|
||||
\= Expect no match
|
||||
Xx\=ps
|
||||
No match
|
||||
X\x{123}x\=ps
|
||||
@ -947,6 +927,7 @@ No match
|
||||
No match
|
||||
|
||||
/X\x{123}{2,4}?b/utf
|
||||
\= Expect no match
|
||||
Xx\=ps
|
||||
No match
|
||||
X\x{123}x\=ps
|
||||
@ -959,6 +940,7 @@ No match
|
||||
No match
|
||||
|
||||
/X\x{123}{2,4}+b/utf
|
||||
\= Expect no match
|
||||
Xx\=ps
|
||||
No match
|
||||
X\x{123}x\=ps
|
||||
@ -1745,6 +1727,7 @@ Last code unit = 'y'
|
||||
First code unit = 'x'
|
||||
Last code unit = 'y'
|
||||
Subject length lower bound = 2
|
||||
|
||||
/(?<!^)ETA/utf
|
||||
\= Expect no match
|
||||
ETA
|
||||
@ -1765,7 +1748,7 @@ No match
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
|
||||
/\ud800/utf,alt_bsux,allow_empty_class,match_unset_backref
|
||||
Failed: error 173 at offset 6: disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
|
||||
|
||||
@ -1874,8 +1857,7 @@ Subject length lower bound = 1
|
||||
0: 1234
|
||||
12-34
|
||||
0: 12-34
|
||||
12+\x{661}-34
|
||||
0: 12+\x{661}-34
|
||||
12+\x{661}-34
|
||||
0: 12+\x{661}-34
|
||||
\= Expect no match
|
||||
abcd
|
||||
@ -1995,8 +1977,7 @@ No match
|
||||
0: \x{2069}
|
||||
|
||||
/^\p{Cs}/utf
|
||||
\x{dfff}\=no_utf_check
|
||||
0: \x{dfff}
|
||||
\x{dfff}\=no_utf_check
|
||||
0: \x{dfff}
|
||||
\= Expect no match
|
||||
\x{09f}
|
||||
@ -2021,8 +2002,7 @@ No match
|
||||
/^\p{Sc}+/utf
|
||||
$\x{a2}\x{a3}\x{a4}\x{a5}\x{a6}
|
||||
0: $\x{a2}\x{a3}\x{a4}\x{a5}
|
||||
\x{9f2}
|
||||
0: \x{9f2}
|
||||
\x{9f2}
|
||||
0: \x{9f2}
|
||||
\= Expect no match
|
||||
X
|
||||
@ -2039,8 +2019,7 @@ No match
|
||||
0: \x{1680}
|
||||
\x{2000}
|
||||
0: \x{2000}
|
||||
\x{2001}
|
||||
0: \x{2001}
|
||||
\x{2001}
|
||||
0: \x{2001}
|
||||
\= Expect no match
|
||||
\x{2028}
|
||||
@ -2052,16 +2031,14 @@ No match
|
||||
# properties and has changed how it behaves for caseless matching.
|
||||
|
||||
/\p{^Lu}/i,utf
|
||||
1234
|
||||
0: 1
|
||||
1234
|
||||
0: 1
|
||||
\= Expect no match
|
||||
ABC
|
||||
No match
|
||||
|
||||
/\P{Lu}/i,utf
|
||||
1234
|
||||
0: 1
|
||||
1234
|
||||
0: 1
|
||||
\= Expect no match
|
||||
ABC
|
||||
@ -2070,8 +2047,7 @@ No match
|
||||
/\p{Ll}/i,utf
|
||||
a
|
||||
0: a
|
||||
Az
|
||||
0: z
|
||||
Az
|
||||
0: z
|
||||
\= Expect no match
|
||||
ABC
|
||||
@ -2080,8 +2056,7 @@ No match
|
||||
/\p{Lu}/i,utf
|
||||
A
|
||||
0: A
|
||||
a\x{10a0}B
|
||||
0: \x{10a0}
|
||||
a\x{10a0}B
|
||||
0: \x{10a0}
|
||||
\= Expect no match
|
||||
a
|
||||
@ -2092,8 +2067,7 @@ No match
|
||||
/\p{Lu}/i,utf
|
||||
A
|
||||
0: A
|
||||
aZ
|
||||
0: Z
|
||||
aZ
|
||||
0: Z
|
||||
\= Expect no match
|
||||
abc
|
||||
@ -2182,16 +2156,14 @@ No match
|
||||
0: \x{6ca}
|
||||
\x{a6c}
|
||||
0: \x{a6c}
|
||||
\x{10a7}
|
||||
0: \x{10a7}
|
||||
\x{10a7}
|
||||
0: \x{10a7}
|
||||
\= Expect no match
|
||||
_ABC
|
||||
No match
|
||||
|
||||
/^\p{Xan}+/utf
|
||||
ABCD1234\x{6ca}\x{a6c}\x{10a7}_
|
||||
0: ABCD1234\x{6ca}\x{a6c}\x{10a7}
|
||||
ABCD1234\x{6ca}\x{a6c}\x{10a7}_
|
||||
0: ABCD1234\x{6ca}\x{a6c}\x{10a7}
|
||||
\= Expect no match
|
||||
_ABC
|
||||
@ -2222,16 +2194,14 @@ No match
|
||||
0: \x{6ca}
|
||||
\x{a6c}
|
||||
0: \x{a6c}
|
||||
\x{10a7}
|
||||
0: \x{10a7}
|
||||
\x{10a7}
|
||||
0: \x{10a7}
|
||||
\= Expect no match
|
||||
_ABC
|
||||
No match
|
||||
|
||||
/^[\p{Xan}]+/utf
|
||||
ABCD1234\x{6ca}\x{a6c}\x{10a7}_
|
||||
0: ABCD1234\x{6ca}\x{a6c}\x{10a7}
|
||||
ABCD1234\x{6ca}\x{a6c}\x{10a7}_
|
||||
0: ABCD1234\x{6ca}\x{a6c}\x{10a7}
|
||||
\= Expect no match
|
||||
_ABC
|
||||
@ -2240,8 +2210,7 @@ No match
|
||||
/^>\p{Xsp}/utf
|
||||
>\x{1680}\x{2028}\x{0b}
|
||||
0: >\x{1680}
|
||||
>\x{a0}
|
||||
0: >\x{a0}
|
||||
>\x{a0}
|
||||
0: >\x{a0}
|
||||
\= Expect no match
|
||||
\x{0b}
|
||||
@ -2278,8 +2247,7 @@ No match
|
||||
/^>\p{Xps}/utf
|
||||
>\x{1680}\x{2028}\x{0b}
|
||||
0: >\x{1680}
|
||||
>\x{a0}
|
||||
0: >\x{a0}
|
||||
>\x{a0}
|
||||
0: >\x{a0}
|
||||
\= Expect no match
|
||||
\x{0b}
|
||||
@ -2324,8 +2292,7 @@ No match
|
||||
0: \x{a6c}
|
||||
\x{10a7}
|
||||
0: \x{10a7}
|
||||
_ABC
|
||||
0: _
|
||||
_ABC
|
||||
0: _
|
||||
\= Expect no match
|
||||
[]
|
||||
@ -2362,8 +2329,7 @@ No match
|
||||
0: \x{a6c}
|
||||
\x{10a7}
|
||||
0: \x{10a7}
|
||||
_ABC
|
||||
0: _
|
||||
_ABC
|
||||
0: _
|
||||
\= Expect no match
|
||||
[]
|
||||
@ -2630,8 +2596,7 @@ No match
|
||||
# Without PCRE_UCP, non-ASCII always fail, even if < 256
|
||||
|
||||
/\b...\B/utf
|
||||
abc_
|
||||
0: abc
|
||||
abc_
|
||||
0: abc
|
||||
\= Expect no match
|
||||
\x{37e}abc\x{376}
|
||||
@ -2825,10 +2790,12 @@ No match
|
||||
------------------------------------------------------------------
|
||||
|
||||
# These behaved oddly in Perl, so they are kept in this test
|
||||
|
||||
/(\x{23a}\x{23a}\x{23a})?\1/i,utf
|
||||
\= Expect no match
|
||||
\x{23a}\x{23a}\x{23a}\x{2c65}\x{2c65}
|
||||
No match
|
||||
|
||||
/(ȺȺȺ)?\1/i,utf
|
||||
\= Expect no match
|
||||
ȺȺȺⱥⱥ
|
||||
@ -2843,10 +2810,12 @@ No match
|
||||
ȺȺȺⱥⱥⱥ
|
||||
0: \x{23a}\x{23a}\x{23a}\x{2c65}\x{2c65}\x{2c65}
|
||||
1: \x{23a}\x{23a}\x{23a}
|
||||
|
||||
/(\x{23a}\x{23a}\x{23a})\1/i,utf
|
||||
\= Expect no match
|
||||
\x{23a}\x{23a}\x{23a}\x{2c65}\x{2c65}
|
||||
No match
|
||||
|
||||
/(ȺȺȺ)\1/i,utf
|
||||
\= Expect no match
|
||||
ȺȺȺⱥⱥ
|
||||
@ -2887,8 +2856,7 @@ No match
|
||||
/^[\p{Batak}]/utf
|
||||
\x{1bc0}
|
||||
0: \x{1bc0}
|
||||
\x{1bff}
|
||||
0: \x{1bff}
|
||||
\x{1bff}
|
||||
0: \x{1bff}
|
||||
\= Expect no match
|
||||
\x{1bf4}
|
||||
@ -2897,8 +2865,7 @@ No match
|
||||
/^[\p{Brahmi}]/utf
|
||||
\x{11000}
|
||||
0: \x{11000}
|
||||
\x{1106f}
|
||||
0: \x{1106f}
|
||||
\x{1106f}
|
||||
0: \x{1106f}
|
||||
\= Expect no match
|
||||
\x{1104e}
|
||||
@ -2907,8 +2874,7 @@ No match
|
||||
/^[\p{Mandaic}]/utf
|
||||
\x{840}
|
||||
0: \x{840}
|
||||
\x{85e}
|
||||
0: \x{85e}
|
||||
\x{85e}
|
||||
0: \x{85e}
|
||||
\= Expect no match
|
||||
\x{85c}
|
||||
@ -2933,14 +2899,10 @@ No match
|
||||
0: \x{301}
|
||||
|
||||
/^a\X41z/alt_bsux,allow_empty_class,match_unset_backref,dupnames
|
||||
aX41z
|
||||
0: aX41z
|
||||
aX41z
|
||||
0: aX41z
|
||||
\= Expect no match
|
||||
aAz
|
||||
No match
|
||||
|
||||
/(?<=ab\Cde)X/utf
|
||||
No match
|
||||
|
||||
/\X/
|
||||
@ -3138,8 +3100,7 @@ Subject length lower bound = 3
|
||||
\x{3a3}\x{3c3}\x{3c2}\x{3a3}\x{3c3}\x{3c2}
|
||||
0: \x{3a3}\x{3c3}\x{3c2}\x{3a3}\x{3c3}\x{3c2}
|
||||
0+
|
||||
|
||||
/\x{3a3}++./i,utf,aftertext
|
||||
|
||||
/\x{3a3}++./i,utf,aftertext
|
||||
\= Expect no match
|
||||
\x{3a3}\x{3c3}\x{3c2}\x{3a3}\x{3c3}\x{3c2}
|
||||
@ -3179,24 +3140,29 @@ No match
|
||||
clist 0053 0073 017f
|
||||
/i t
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
\= Expect no match
|
||||
ikt
|
||||
No match
|
||||
|
||||
/is+t/i,utf
|
||||
iSs\x{17f}t
|
||||
0: iSs\x{17f}t
|
||||
\= Expect no match
|
||||
ikt
|
||||
No match
|
||||
|
||||
/is+?t/i,utf
|
||||
\= Expect no match
|
||||
ikt
|
||||
No match
|
||||
|
||||
/is?t/i,utf
|
||||
\= Expect no match
|
||||
ikt
|
||||
No match
|
||||
|
||||
/is{2}t/i,utf
|
||||
\= Expect no match
|
||||
iskt
|
||||
@ -3211,80 +3177,70 @@ No match
|
||||
0: @
|
||||
`abc
|
||||
0: `
|
||||
\x{1234}abc
|
||||
0: \x{1234}
|
||||
\x{1234}abc
|
||||
0: \x{1234}
|
||||
\= Expect no match
|
||||
abc
|
||||
No match
|
||||
|
||||
/^\p{Xuc}+/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^\p{Xuc}+?/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^\p{Xuc}+?\*/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}*
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}*
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^\p{Xuc}++/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^\p{Xuc}{3,5}/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^\p{Xuc}{3,5}?/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^[\p{Xuc}]/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^[\p{Xuc}]+/utf
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}
|
||||
$@`\x{a0}\x{1234}\x{e000}**
|
||||
0: $@`\x{a0}\x{1234}\x{e000}
|
||||
\= Expect no match
|
||||
\x{9f}
|
||||
No match
|
||||
|
||||
/^\P{Xuc}/utf
|
||||
abc
|
||||
0: a
|
||||
abc
|
||||
0: a
|
||||
\= Expect no match
|
||||
$abc
|
||||
@ -3297,8 +3253,7 @@ No match
|
||||
No match
|
||||
|
||||
/^[\P{Xuc}]/utf
|
||||
abc
|
||||
0: a
|
||||
abc
|
||||
0: a
|
||||
\= Expect no match
|
||||
$abc
|
||||
@ -3843,7 +3798,7 @@ No match
|
||||
[ab\p{L}]{2,3}+
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
------------------------------------------------------------------
|
||||
|
||||
/\D+\X \d+\X \S+\X \s+\X \W+\X \w+\X \R+\X \H+\X \h+\X \V+\X \v+\X a+\X \n+\X .+\X/Bx
|
||||
------------------------------------------------------------------
|
||||
@ -3858,8 +3813,6 @@ No match
|
||||
extuni
|
||||
\W+
|
||||
extuni
|
||||
\w+
|
||||
extuni
|
||||
\w+
|
||||
extuni
|
||||
\R+
|
||||
@ -3898,7 +3851,7 @@ No match
|
||||
/m $
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
------------------------------------------------------------------
|
||||
|
||||
/\X+\D \X+\d \X+\S \X+\s \X+\W \X+\w \X+. \X+\R \X+\H \X+\h \X+\V \X+\v \X+\X \X+\Z \X+\z \X+$/Bx
|
||||
------------------------------------------------------------------
|
||||
@ -3916,8 +3869,6 @@ No match
|
||||
extuni+
|
||||
\w
|
||||
extuni+
|
||||
Any
|
||||
extuni+
|
||||
Any
|
||||
extuni+
|
||||
\R
|
||||
@ -4003,12 +3954,9 @@ Subject length lower bound = 1
|
||||
/ábc/utf,replace=XሴZ
|
||||
123ábc123
|
||||
1: 123X\x{1234}Z123
|
||||
|
||||
|
||||
/(?<=abc)(|def)/g,utf,replace=<$0>
|
||||
123abcáyzabcdef789abcሴqr
|
||||
4: 123abc<>\x{e1}yzabc<><def>789abc<>\x{1234}qr
|
||||
|
||||
/[^\xff]((?1))/utf,debug
|
||||
4: 123abc<>\x{e1}yzabc<><def>789abc<>\x{1234}qr
|
||||
|
||||
/[A-`]/iB,utf
|
||||
@ -4050,4 +3998,238 @@ Failed: error 122 at offset 1227: unmatched closing parenthesis
|
||||
"\xa\xf<(.\pZ*\P{Xwd}+^\xa8\3'3yq.::?(?J:()\xd1+!~:3'(8?:)':(?'d'(?'d'^u]!.+.+\\A\Ah(n+?9){7}+\K;(?'X'u'(?'c'(?'z'(?<y>\xb::\xf0'|\xd3(\xae?'w(z\x8?P>l)\x8?P>a)'\H\R\xd1+!!~:3'(?:h$N{26875}\W+?\\=D{2}\x89(?i:Uy0\N({2\xa(\v\x85*){y*\A(()\p{L}+?\P{^Xan}'+?\xff\+pS\?|).{;y*\A(()\p{L}+?\8}\d?1(|)(/1){7}.+[Lp{Me}].\s\xdcC*?(?(<y>))(?<!^)$C((;*?(R))+(\xbf(R))\x8a\X*?\x8a\xb\xd1^9\3*+(\xc1,\k'R'\xb4)\xcc(z\z(?J)(?'X'\x1b(\xb\xd1^9\?'3*+P{^Xan}+?\xff\+(\xc1.]k+\xb'Pm'\xb4)\xcc4f\xa7'\xd1V(?i:U,{2,2})'(?'X'))?-%--\x95$9*\4'|\xd1(\x9c''%\x94$9)#(?'R')3\x7?('P\xed7'\xa8\xb1^u\xeaw\1\0\0\(|(?1){7}.+[\p{Me}].\s\xdcC*^\x14?(?(<y>))(?<!^)$C((;*?(R*?))+(?(R)\x8a\X*?\x8a\xb\xd1^9\3*+|(\xc1,\k'R'\xb4)\xcc! z)\z(?JJ)(?'X';(\xb\xd1^9\?'3*+(\xc1.]k+\xb'Pm'\xb4))':(?'d')(?'RD'(d')|)|$)'|(?<x>\g{d});\g{x}\x11\g{d}\x81\|$((?'X'\'X'(?'W''\x92()'9'\x83*))\xba*\!?^ <){)':;\xcc4'\xd1'(?'X'28))?-%--\x95$9*\4'|\xd1((''e\x94*$9:)*#(?'R')3)\x7?('P\xed')\\x16:;()\x1e\x10*:(?<y>)\xd1+0!~:(?)'d'E:yD!\s(?'R'\x1e;\x10:U))|'\x9g!\xb0*){)\\x16:;()\x1e\x10\x87*:(?<y>)\xd1+!~:(?)'}'\d'E:yD!\s(?'R'\x1e;\x10:U))|'))|)g!\xb0*R+9{29+)#(?'P'})*?pS\{3,}\x85,{0,}l{*UTF)(\xe{7}){3722,{9,}d{2,?|))|{)\(A?&d}}{\xa,}2}){3,}7,l{)22}(,}l:7{2,4}}29\x19+)#?'P'})*v?))\x5"
|
||||
Failed: error 122 at offset 1227: unmatched closing parenthesis
|
||||
|
||||
/$(&.+[\p{Me}].\s\xdcC*?(?(<y>))(?<!^)$C((;*?(R))+(?(R)){0,6}?|){12\x8a\X*?\x8a\x0b\xd1^9\3*+(\xc1,\k'P'\xb4)\xcc(z\z(?JJ)(?'X'8};(\x0b\xd1^9\?'3*+(\xc1.]k+\x0b'Pm'\xb4\xcc4'\xd1'(?'X'))?-%--\x95$9*\4'|\xd1(''%\x95*$9)#(?'R')3\x07?('P\xed')\\x16:;()\x1e\x10*:(?<y>)\xd1+!~:(?)''(d'E:yD!\s(?'R'\x1e;\x10:U))|')g!\xb0*){29+))#(?'P'})*?/
|
||||
|
||||
"(*UTF)(*UCP)(.UTF).+X(\V+;\^(\D|)!999}(?(?C{7(?C')\H*\S*/^\x5\xa\\xd3\x85n?(;\D*(?m).[^mH+((*UCP)(*U:F)})(?!^)(?'"
|
||||
Failed: error 162 at offset 113: subpattern name expected
|
||||
|
||||
/[\pS#moq]/
|
||||
=
|
||||
0: =
|
||||
|
||||
/(*:a\x{12345}b\t(d\)c)xxx/utf,alt_verbnames,mark
|
||||
cxxxz
|
||||
0: xxx
|
||||
MK: a\x{12345}b\x{09}(d)c
|
||||
|
||||
/abcd/utf,replace=x\x{824}y\o{3333}z(\Q12\$34$$\x34\E5$$),substitute_extended
|
||||
abcd
|
||||
1: x\x{824}y\x{6db}z(12\$34$$\x345$)
|
||||
|
||||
/a(\x{e0}\x{101})(\x{c0}\x{102})/utf,replace=a\u$1\U$1\E$1\l$2\L$2\Eab\U\x{e0}\x{101}\L\x{d0}\x{160}\EDone,substitute_extended
|
||||
a\x{e0}\x{101}\x{c0}\x{102}
|
||||
1: a\x{c0}\x{101}\x{c0}\x{100}\x{e0}\x{101}\x{e0}\x{102}\x{e0}\x{103}ab\x{c0}\x{100}\x{f0}\x{161}Done
|
||||
|
||||
/((?<digit>\d)|(?<letter>\p{L}))/g,substitute_extended,replace=<${digit:+digit; :not digit; }${letter:+letter:not a letter}>
|
||||
ab12cde
|
||||
7: <not digit; letter><not digit; letter><digit; not a letter><digit; not a letter><not digit; letter><not digit; letter><not digit; letter>
|
||||
|
||||
/(*UCP)(*UTF)[[:>:]]X/B
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
\b
|
||||
AssertB
|
||||
Reverse
|
||||
prop Xwd
|
||||
Ket
|
||||
X
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/abc/utf,replace=xyz
|
||||
abc\=zero_terminate
|
||||
1: xyz
|
||||
|
||||
/a[[:punct:]b]/ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
a
|
||||
[b[:punct:]]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/a[[:punct:]b]/utf,ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
a
|
||||
[b[:punct:]]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/a[b[:punct:]]/utf,ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
a
|
||||
[b[:punct:]]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/[[:^ascii:]]/utf,ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
[\x80-\xff] (neg)
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/[[:^ascii:]\w]/utf,ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
[\x80-\xff\p{Xwd}\x{100}-\x{10ffff}]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/[\w[:^ascii:]]/utf,ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
[\x80-\xff\p{Xwd}\x{100}-\x{10ffff}]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/[^[:ascii:]\W]/utf,ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
[^\x00-\x7f\P{Xwd}]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
\x{de}
|
||||
0: \x{de}
|
||||
\x{200}
|
||||
0: \x{200}
|
||||
\= Expect no match
|
||||
\x{300}
|
||||
No match
|
||||
\x{37e}
|
||||
No match
|
||||
|
||||
/[[:^ascii:]a]/utf,ucp,bincode
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
[a\x80-\xff] (neg)
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/L(?#(|++<!(2)?/B,utf,no_auto_possess,auto_callout
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
Callout 255 0 14
|
||||
L?
|
||||
Callout 255 14 0
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/L(?#(|++<!(2)?/B,utf,ucp,auto_callout
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
Callout 255 0 14
|
||||
L?+
|
||||
Callout 255 14 0
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/(*UTF)C\x09((?<!'(?x)!*H? #\xcc\x9a[^$]/
|
||||
Failed: error 114 at offset 39: missing closing parenthesis
|
||||
|
||||
/[\D]/utf
|
||||
\x{1d7cf}
|
||||
0: \x{1d7cf}
|
||||
|
||||
/[\D\P{Nd}]/utf
|
||||
\x{1d7cf}
|
||||
0: \x{1d7cf}
|
||||
|
||||
/[^\D]/utf
|
||||
a9b
|
||||
0: 9
|
||||
\= Expect no match
|
||||
\x{1d7cf}
|
||||
No match
|
||||
|
||||
/[^\D\P{Nd}]/utf
|
||||
a9b
|
||||
0: 9
|
||||
\x{1d7cf}
|
||||
0: \x{1d7cf}
|
||||
\= Expect no match
|
||||
\x{10000}
|
||||
No match
|
||||
|
||||
# Hex uses pattern length, not zero-terminated. This tests for overrunning
|
||||
# the given length of a pattern.
|
||||
|
||||
/'(*UTF)'/hex
|
||||
|
||||
/'#('/hex,extended,utf
|
||||
|
||||
/a(?<=A\XB)/utf
|
||||
Failed: error 125 at offset 1: lookbehind assertion is not fixed length
|
||||
|
||||
/ab(?<=A\RB)/utf
|
||||
Failed: error 125 at offset 2: lookbehind assertion is not fixed length
|
||||
|
||||
/../utf,auto_callout
|
||||
\n\x{123}\x{123}\x{123}\x{123}
|
||||
--->\x{0a}\x{123}\x{123}\x{123}\x{123}
|
||||
+0 ^ .
|
||||
+0 ^ .
|
||||
+1 ^ ^ .
|
||||
+2 ^ ^
|
||||
0: \x{123}\x{123}
|
||||
|
||||
# This tests processing wide characters in extended mode.
|
||||
|
||||
/XȀ/x,utf
|
||||
|
||||
# These three test a bug fix that was not clearing up after a locale setting
|
||||
# when the test or a subsequent one matched a wide character.
|
||||
|
||||
//locale=C
|
||||
|
||||
/[\P{Yi}]/utf
|
||||
\x{2f000}
|
||||
0: \x{2f000}
|
||||
|
||||
/[\P{Yi}]/utf,locale=C
|
||||
\x{2f000}
|
||||
0: \x{2f000}
|
||||
|
||||
/^(?<!(?=))/B,utf
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
^
|
||||
AssertB not
|
||||
Assert
|
||||
\x{10385c}
|
||||
Ket
|
||||
Ket
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
# Horizontal and vertical space lists ignore caseless
|
||||
|
||||
/[\HH]/Bi,utf
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
[\x00-\x08\x0a-\x1f!-\x9f\xa1-\xff\x{100}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{10ffff}]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
/[^\HH]/Bi,utf
|
||||
------------------------------------------------------------------
|
||||
Bra
|
||||
[^\x00-\x08\x0a-\x1f!-\x9f\xa1-\xff\x{100}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{10ffff}]
|
||||
Ket
|
||||
End
|
||||
------------------------------------------------------------------
|
||||
|
||||
Reference in New Issue
Block a user