tests.txt

C++的一个好库。。。现在很流行

; 
; 
; this file contains a script of tests to run through regress.exe
;
; comments start with a semicolon and proceed to the end of the line
;
; changes to regular expression compile flags start with a "-" as the first
; non-whitespace character and consist of a list of the printable names
; of the flags, for example "match_default"
;
; Other lines contain a test to perform using the current flag status
; the first token contains the expression to compile, the second the string
; to match it against. If the second string is "!" then the expression should
; not compile, that is the first string is an invalid regular expression.
; This is then followed by a list of integers that specify what should match,
; each pair represents the starting and ending positions of a subexpression
; starting with the zeroth subexpression (the whole match).
; A value of -1 indicates that the subexpression should not take part in the
; match at all, if the first value is -1 then no part of the expression should
; match the string.
;

- match_default normal REG_EXTENDED

;
; try some really simple literals:
a a 0 1
Z Z 0 1
Z aaa -1 -1
Z xxxxZZxxx 4 5

; and some simple brackets:
(a) zzzaazz 3 4 3 4
() zzz 0 0 0 0
() "" 0 0 0 0
( !
) !
(aa !
aa) !
a b -1 -1
\(\) () 0 2
\(a\) (a) 0 3
\() !
(\) !
p(a)rameter ABCparameterXYZ 3 12 4 5
[pq](a)rameter ABCparameterXYZ 3 12 4 5

; now try escaped brackets:
- match_default bk_parens REG_BASIC
\(a\) zzzaazz 3 4 3 4
\(\) zzz 0 0 0 0
\(\) "" 0 0 0 0
\( !
\) !
\(aa !
aa\) !
() () 0 2
(a) (a) 0 3
(\) !
\() !

; now move on to "." wildcards
- match_default normal REG_EXTENDED REG_STARTEND
. a 0 1
. \n 0 1
. \r 0 1
. \0 0 1
- match_default normal match_not_dot_newline REG_EXTENDED REG_STARTEND REG_NEWLINE
. a 0 1
. \n -1 -1
. \r -1 -1
. \0 0 1
- match_default normal match_not_dot_null match_not_dot_newline REG_EXTENDED REG_STARTEND REG_NEWLINE
. \n -1 -1
. \r -1 -1
; this *WILL* produce an error from the POSIX API functions:
- match_default normal match_not_dot_null match_not_dot_newline REG_EXTENDED REG_STARTEND REG_NEWLINE REG_NO_POSIX_TEST
. \0 -1 -1


;
; now move on to the repetion ops,
; starting with operator *
- match_default normal REG_EXTENDED
a* b 0 0
ab* a 0 1
ab* ab 0 2
ab* sssabbbbbbsss 3 10
ab*c* a 0 1
ab*c* abbb 0 4
ab*c* accc 0 4
ab*c* abbcc 0 5
*a !
\<* !
\>* !
\n* \n\n 0 2
\** ** 0 2
\* * 0 1

; now try operator +
ab+ a -1 -1
ab+ ab 0 2
ab+ sssabbbbbbsss 3 10
ab+c+ a -1 -1
ab+c+ abbb -1 -1
ab+c+ accc -1 -1
ab+c+ abbcc 0 5
+a !
\<+ !
\>+ !
\n+ \n\n 0 2
\+ + 0 1
\+ ++ 0 1
\++ ++ 0 2
- match_default normal bk_plus_qm REG_EXTENDED REG_NO_POSIX_TEST
+ + 0 1
\+ !
a\+ aa 0 2

; now try operator ?
- match_default normal REG_EXTENDED
a? b 0 0
ab? a 0 1
ab? ab 0 2
ab? sssabbbbbbsss 3 5
ab?c? a 0 1
ab?c? abbb 0 2
ab?c? accc 0 2
ab?c? abcc 0 3
?a !
\<? !
\>? !
\n? \n\n 0 1
\? ? 0 1
\? ?? 0 1
\?? ?? 0 1
- match_default normal bk_plus_qm REG_EXTENDED REG_NO_POSIX_TEST
? ? 0 1
\? !
a\? aa 0 1
a\? b 0 0

- match_default normal limited_ops
a? a? 0 2
a+ a+ 0 2
a\? a? 0 2
a\+ a+ 0 2

; now try operator {}
- match_default normal REG_EXTENDED
a{2} a -1 -1
a{2} aa 0 2
a{2} aaa 0 2
a{2,} a -1 -1
a{2,} aa 0 2
a{2,} aaaaa 0 5
a{2,4} a -1 -1
a{2,4} aa 0 2
a{2,4} aaa 0 3
a{2,4} aaaa 0 4
a{2,4} aaaaa 0 4
; spaces are now allowed inside {}
"a{ 2 , 4 }" aaaaa 0 4
a{} !
"a{ }" !
a{2 !
a} !
\{\} {} 0 2

- match_default normal bk_braces
a\{2\} a -1 -1
a\{2\} aa 0 2
a\{2\} aaa 0 2
a\{2,\} a -1 -1
a\{2,\} aa 0 2
a\{2,\} aaaaa 0 5
a\{2,4\} a -1 -1
a\{2,4\} aa 0 2
a\{2,4\} aaa 0 3
a\{2,4\} aaaa 0 4
a\{2,4\} aaaaa 0 4
"a\{ 2 , 4 \}" aaaaa 0 4
{} {} 0 2

; now test the alternation operator |
- match_default normal REG_EXTENDED
a|b a 0 1
a|b b 0 1
a(b|c) ab 0 2 1 2
a(b|c) ac 0 2 1 2
a(b|c) ad -1 -1 -1 -1
|c !
c| !
(|) !
(a|) !
(|a) !
a\| a| 0 2
- match_default normal limited_ops
a| a| 0 2
a\| a| 0 2
| | 0 1
- match_default normal bk_vbar REG_NO_POSIX_TEST
a| a| 0 2
a\|b a 0 1
a\|b b 0 1

; now test the set operator []
- match_default normal REG_EXTENDED
; try some literals first
[abc] a 0 1
[abc] b 0 1
[abc] c 0 1
[abc] d -1 -1
[^bcd] a 0 1
[^bcd] b -1 -1
[^bcd] d -1 -1
[^bcd] e 0 1
a[b]c abc 0 3
a[ab]c abc 0 3
a[^ab]c adc 0 3
a[]b]c a]c 0 3
a[[b]c a[c 0 3
a[-b]c a-c 0 3
a[^]b]c adc 0 3
a[^-b]c adc 0 3
a[b-]c a-c 0 3
a[b !
a[] !

; then some ranges
[b-e] a -1 -1
[b-e] b 0 1
[b-e] e 0 1
[b-e] f -1 -1
[^b-e] a 0 1
[^b-e] b -1 -1
[^b-e] e -1 -1
[^b-e] f 0 1
a[1-3]c a2c 0 3
a[3-1]c !
a[1-3-5]c !
a[1- !

; and some classes
a[[:alpha:]]c abc 0 3
a[[:unknown:]]c !
a[[: !
a[[:alpha !
a[[:alpha:] !
a[[:alpha,:] !
a[[:]:]]b !
a[[:-:]]b !
a[[:alph:]] !
a[[:alphabet:]] !
[[:alnum:]]+ -%@a0X_- 3 6
[[:alpha:]]+ -%@aX_0- 3 5
[[:blank:]]+ "a  \tb" 1 4
[[:cntrl:]]+ a\n\tb 1 3
[[:digit:]]+ a019b 1 4
[[:graph:]]+ " a%b " 1 4
[[:lower:]]+ AabC 1 3
; This test fails with STLPort, disable for now as this is a corner case anyway...
;[[:print:]]+ "\na b\n" 1 4
[[:punct:]]+ " %-&\t" 1 4
[[:space:]]+ "a \n\t\rb" 1 5
[[:upper:]]+ aBCd 1 3
[[:xdigit:]]+ p0f3Cx 1 5

; now test flag settings:
- escape_in_lists REG_NO_POSIX_TEST
[\n] \n 0 1
- REG_NO_POSIX_TEST
[\n] \n -1 -1
[\n] \\ 0 1
[[:class:] : 0 1
[[:class:] [ 0 1
[[:class:] c 0 1

; line anchors
- match_default normal REG_EXTENDED
^ab ab 0 2
^ab xxabxx -1 -1
^ab xx\nabzz 3 5
ab$ ab 0 2
ab$ abxx -1 -1
ab$ ab\nzz 0 2
- match_default match_not_bol match_not_eol normal REG_EXTENDED REG_NOTBOL REG_NOTEOL
^ab ab -1 -1
^ab xxabxx -1 -1
^ab xx\nabzz 3 5
ab$ ab -1 -1
ab$ abxx -1 -1
ab$ ab\nzz 0 2

; line anchors, single line mode
- match_default normal match_single_line REG_NO_POSIX_TEST
^ab ab 0 2
^ab xxabxx -1 -1
^ab xx\nabzz -1 -1
ab$ ab 0 2
ab$ abxx -1 -1
ab$ ab\nzz -1 -1
- match_default match_not_bol match_not_eol normal REG_NO_POSIX_TEST match_single_line
^ab ab -1 -1
^ab xxabxx -1 -1
^ab xx\nabzz -1 -1
ab$ ab -1 -1
ab$ abxx -1 -1
ab$ ab\nzz -1 -1

; back references
- match_default normal REG_PERL
a(b)\2c	!
a(b\1)c	!
a(b*)c\1d abbcbbd 0 7 1 3
a(b*)c\1d abbcbd -1 -1
a(b*)c\1d abbcbbbd -1 -1
^(.)\1 abc -1 -1
a([bc])\1d abcdabbd	4 8 5 6
; strictly speaking this is at best ambiguous, at worst wrong, this is what most
; re implimentations will match though.
a(([bc])\2)*d abbccd 0 6 3 5 3 4

a(([bc])\2)*d abbcbd -1 -1
a((b)*\2)*d abbbd 0 5 1 4 2 3
; perl only:
(ab*)[ab]*\1 ababaaa 0 4 0 2
(a)\1bcd aabcd 0 5 0 1
(a)\1bc*d aabcd 0 5 0 1
(a)\1bc*d aabd 0 4 0 1
(a)\1bc*d aabcccd 0 7 0 1
(a)\1bc*[ce]d aabcccd 0 7 0 1
^(a)\1b(c)*cd$ aabcccd 0 7 0 1 4 5

; posix only: 
- match_default extended REG_EXTENDED
(ab*)[ab]*\1 ababaaa 0 7 0 1

;
; characters by code:
- match_default normal REG_PERL REG_STARTEND
\0101 A 0 1
\00 \0 0 1
\0 \0 0 1
\0172 z 0 1

;
; word operators:
\w a 0 1
\w z 0 1
\w A 0 1
\w Z 0 1
\w _ 0 1
\w } -1 -1
\w ` -1 -1
\w [ -1 -1
\w @ -1 -1
; non-word:
\W a -1 -1
\W z -1 -1
\W A -1 -1
\W Z -1 -1
\W _ -1 -1
\W } 0 1
\W ` 0 1
\W [ 0 1
\W @ 0 1
; word start:
\<abcd "  abcd" 2 6
\<ab cab -1 -1
\<ab "\nab" 1 3
\<tag ::tag 2 5
;word end:
abc\> abc 0 3
abc\> abcd -1 -1
abc\> abc\n 0 3
abc\> abc:: 0 3
; word boundary:
\babcd "  abcd" 2 6
\bab cab -1 -1
\bab "\nab" 1 3
\btag ::tag 2 5
abc\b abc 0 3
abc\b abcd -1 -1
abc\b abc\n 0 3
abc\b abc:: 0 3
; within word:
\B ab 1 1
a\Bb ab 0 2
a\B ab 0 1
a\B a -1 -1
a\B "a " -1 -1

;
; buffer operators:
\`abc abc 0 3
\`abc \nabc -1 -1
\`abc " abc" -1 -1
abc\' abc 0 3
abc\' abc\n -1 -1
abc\' "abc " -1 -1

;
; extra escape sequences:
\a \a 0 1
\f \f 0 1
\n \n 0 1
\r \r 0 1
\t \t 0 1
\v \v 0 1


;
; now follows various complex expressions designed to try and bust the matcher:
a(((b)))c abc 0 3 1 2 1 2 1 2
a(b|(c))d abd 0 3 1 2 -1 -1
a(b|(c))d acd 0 3 1 2 1 2
a(b*|c)d abbd 0 4 1 3
; just gotta have one DFA-buster, of course
a[ab]{20} aaaaabaaaabaaaabaaaab 0 21
; and an inline expansion in case somebody gets tricky
a[ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab] aaaaabaaaabaaaabaaaab 0 21
; and in case somebody just slips in an NFA...
a[ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab](wee|week)(knights|night) aaaaabaaaabaaaabaaaabweeknights 0 31 21 24 24 31
; one really big one