/[pcre]/code/trunk/testdata/testinput2
ViewVC logotype

Diff of /code/trunk/testdata/testinput2

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 701 by ph10, Tue Sep 20 11:30:56 2011 UTC revision 963 by ph10, Sat Apr 21 18:06:31 2012 UTC
# Line 3  Line 3 
3      It also checks the non-Perl syntax the PCRE supports (Python, .NET,      It also checks the non-Perl syntax the PCRE supports (Python, .NET,
4      Oniguruma). Finally, there are some tests where PCRE and Perl differ,      Oniguruma). Finally, there are some tests where PCRE and Perl differ,
5      either because PCRE can't be compatible, or there is a possible Perl      either because PCRE can't be compatible, or there is a possible Perl
6      bug. --/      bug.
7    
8        NOTE: This is a non-UTF set of tests. When UTF support is needed, use
9        test 5, and if Unicode Property Support is needed, use test 7. --/
10    
 /-- Originally, the Perl >= 5.10 things were in here too, but now I have  
     separated many (most?) of them out into test 11. However, there may still  
     be some that were overlooked. --/  
   
11  /(a)b|/I  /(a)b|/I
12    
13  /abc/I  /abc/I
# Line 144  Line 143 
143      defabc      defabc
144      \Zdefabc      \Zdefabc
145    
 /abc/P  
     abc  
     *** Failers  
   
 /^abc|def/P  
     abcdef  
     abcdef\B  
   
 /.*((abc)$|(def))/P  
     defabc  
     \Zdefabc  
   
 /the quick brown fox/P  
     the quick brown fox  
     *** Failers  
     The Quick Brown Fox  
   
 /the quick brown fox/Pi  
     the quick brown fox  
     The Quick Brown Fox  
   
 /abc.def/P  
     *** Failers  
     abc\ndef  
   
 /abc$/P  
     abc  
     abc\n  
   
 /(abc)\2/P  
   
 /(abc\1)/P  
     abc  
   
146  /)/  /)/
147    
148  /a[]b/  /a[]b/
# Line 442  Line 407 
407    
408  /abc/\  /abc/\
409    
 /abc/\P  
   
410  /abc/\i  /abc/\i
411    
412  /(a)bc(d)/I  /(a)bc(d)/I
# Line 491  Line 454 
454  /\Biss\B/I+  /\Biss\B/I+
455      Mississippi      Mississippi
456    
 /\Biss\B/I+P  
     Mississippi  
   
457  /iss/IG+  /iss/IG+
458      Mississippi      Mississippi
459    
# Line 629  Line 589 
589      *** Failers      *** Failers
590      \Nabc      \Nabc
591    
 /a*(b+)(z)(z)/P  
     aaaabbbbzzzz  
     aaaabbbbzzzz\O0  
     aaaabbbbzzzz\O1  
     aaaabbbbzzzz\O2  
     aaaabbbbzzzz\O3  
     aaaabbbbzzzz\O4  
     aaaabbbbzzzz\O5  
   
592  /^.?abcd/IS  /^.?abcd/IS
593    
594  /\(             # ( at start  /\(             # ( at start
# Line 1491  Line 1442 
1442      ** Failers      ** Failers
1443      line one\nthis is a line\nbreak in the second line      line one\nthis is a line\nbreak in the second line
1444    
 /ab.cd/P  
     ab-cd  
     ab=cd  
     ** Failers  
     ab\ncd  
   
 /ab.cd/Ps  
     ab-cd  
     ab=cd  
     ab\ncd  
   
1445  /(?i)(?-i)AbCd/I  /(?i)(?-i)AbCd/I
1446      AbCd      AbCd
1447      ** Failers      ** Failers
# Line 1552  Line 1492 
1492      (this)      (this)
1493      ((this))      ((this))
1494    
 /a(b)c/PN  
     abc  
   
 /a(?P<name>b)c/PN  
     abc  
   
 /\x{100}/I  
   
1495  /\x{0000ff}/I  /\x{0000ff}/I
1496    
1497  /^((?P<A>a1)|(?P<A>a2)b)/I  /^((?P<A>a1)|(?P<A>a2)b)/I
# Line 1637  a random value. /Ix Line 1569  a random value. /Ix
1569  /()()()()()()()()()(?:(?(A)(?P=A)a|b)(?P<A>X|Y))+/I  /()()()()()()()()()(?:(?(A)(?P=A)a|b)(?P<A>X|Y))+/I
1570      bXXaYYaY      bXXaYYaY
1571    
 /\777/I  
   
1572  /\s*,\s*/IS  /\s*,\s*/IS
1573      \x0b,\x0b      \x0b,\x0b
1574      \x0c,\x0d      \x0c,\x0d
# Line 2241  a random value. /Ix Line 2171  a random value. /Ix
2171      xabcpqrx      xabcpqrx
2172      xxyzx      xxyzx
2173    
 /[\h]/BZ  
     >\x09<  
   
 /[\h]+/BZ  
     >\x09\x20\xa0<  
   
 /[\v]/BZ  
   
 /[\H]/BZ  
   
 /[^\h]/BZ  
   
 /[\V]/BZ  
   
 /[\x0a\V]/BZ  
   
2174  /\H++X/BZ  /\H++X/BZ
2175      ** Failers      ** Failers
2176      XXXX      XXXX
# Line 2616  a random value. /Ix Line 2530  a random value. /Ix
2530    
2531  /(?(?=.*b).*b|^d)/I  /(?(?=.*b).*b|^d)/I
2532    
 /a?|b?/P  
     abc  
     ** Failers  
     ddd\N  
   
2533  /xyz/C  /xyz/C
2534    xyz    xyz
2535    abcxyz    abcxyz
# Line 2812  a random value. /Ix Line 2721  a random value. /Ix
2721     abc\P     abc\P
2722     abc\P\P     abc\P\P
2723    
 /\w+A/P  
    CDAAAAB  
   
 /\w+A/PU  
    CDAAAAB  
   
2724  /abc\K123/  /abc\K123/
2725      xyzabc123pqr      xyzabc123pqr
2726      xyzabc12\P      xyzabc12\P
# Line 2961  a random value. /Ix Line 2864  a random value. /Ix
2864    
2865  /^From +([^ ]+) +[a-zA-Z][a-zA-Z][a-zA-Z] +[a-zA-Z][a-zA-Z][a-zA-Z] +[0-9]?[0-9] +[0-9][0-9]:[0-9][0-9]/SI  /^From +([^ ]+) +[a-zA-Z][a-zA-Z][a-zA-Z] +[a-zA-Z][a-zA-Z][a-zA-Z] +[0-9]?[0-9] +[0-9][0-9]:[0-9][0-9]/SI
2866    
 /  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*                          # optional leading comment  
 (?:    (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |  
 " (?:                      # opening quote...  
 [^\\\x80-\xff\n\015"]                #   Anything except backslash and quote  
 |                     #    or  
 \\ [^\x80-\xff]           #   Escaped something (something != CR)  
 )* "  # closing quote  
 )                    # initial word  
 (?:  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  \.  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*   (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |  
 " (?:                      # opening quote...  
 [^\\\x80-\xff\n\015"]                #   Anything except backslash and quote  
 |                     #    or  
 \\ [^\x80-\xff]           #   Escaped something (something != CR)  
 )* "  # closing quote  
 )  )* # further okay, if led by a period  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  @  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*    (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                           # initial subdomain  
 (?:                                  #  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  \.                        # if led by a period...  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*   (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                     #   ...further okay  
 )*  
 # address  
 |                     #  or  
 (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |  
 " (?:                      # opening quote...  
 [^\\\x80-\xff\n\015"]                #   Anything except backslash and quote  
 |                     #    or  
 \\ [^\x80-\xff]           #   Escaped something (something != CR)  
 )* "  # closing quote  
 )             # one word, optionally followed by....  
 (?:  
 [^()<>@,;:".\\\[\]\x80-\xff\000-\010\012-\037]  |  # atom and space parts, or...  
 \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)       |  # comments, or...  
   
 " (?:                      # opening quote...  
 [^\\\x80-\xff\n\015"]                #   Anything except backslash and quote  
 |                     #    or  
 \\ [^\x80-\xff]           #   Escaped something (something != CR)  
 )* "  # closing quote  
 # quoted strings  
 )*  
 <  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*                     # leading <  
 (?:  @  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*    (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                           # initial subdomain  
 (?:                                  #  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  \.                        # if led by a period...  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*   (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                     #   ...further okay  
 )*  
   
 (?:  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  ,  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  @  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*    (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                           # initial subdomain  
 (?:                                  #  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  \.                        # if led by a period...  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*   (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                     #   ...further okay  
 )*  
 )* # further okay, if led by comma  
 :                                # closing colon  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  )? #       optional route  
 (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |  
 " (?:                      # opening quote...  
 [^\\\x80-\xff\n\015"]                #   Anything except backslash and quote  
 |                     #    or  
 \\ [^\x80-\xff]           #   Escaped something (something != CR)  
 )* "  # closing quote  
 )                    # initial word  
 (?:  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  \.  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*   (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |  
 " (?:                      # opening quote...  
 [^\\\x80-\xff\n\015"]                #   Anything except backslash and quote  
 |                     #    or  
 \\ [^\x80-\xff]           #   Escaped something (something != CR)  
 )* "  # closing quote  
 )  )* # further okay, if led by a period  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  @  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*    (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                           # initial subdomain  
 (?:                                  #  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  \.                        # if led by a period...  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*   (?:  
 [^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]+    # some number of atom characters...  
 (?![^(\040)<>@,;:".\\\[\]\000-\037\x80-\xff]) # ..not followed by something that could be part of an atom  
 |   \[                         # [  
 (?: [^\\\x80-\xff\n\015\[\]] |  \\ [^\x80-\xff]  )*    #    stuff  
 \]                        #           ]  
 )                     #   ...further okay  
 )*  
 #       address spec  
 (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*  > #                  trailing >  
 # name and address  
 )  (?: [\040\t] |  \(  
 (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  |  \( (?:  [^\\\x80-\xff\n\015()]  |  \\ [^\x80-\xff]  )* \)  )*  
 \)  )*                       # optional trailing comment  
 /xSI  
   
2867  /<tr([\w\W\s\d][^<>]{0,})><TD([\w\W\s\d][^<>]{0,})>([\d]{0,}\.)(.*)((<BR>([\w\W\s\d][^<>]{0,})|[\s]{0,}))<\/a><\/TD><TD([\w\W\s\d][^<>]{0,})>([\w\W\s\d][^<>]{0,})<\/TD><TD([\w\W\s\d][^<>]{0,})>([\w\W\s\d][^<>]{0,})<\/TD><\/TR>/isIS  /<tr([\w\W\s\d][^<>]{0,})><TD([\w\W\s\d][^<>]{0,})>([\d]{0,}\.)(.*)((<BR>([\w\W\s\d][^<>]{0,})|[\s]{0,}))<\/a><\/TD><TD([\w\W\s\d][^<>]{0,})>([\w\W\s\d][^<>]{0,})<\/TD><TD([\w\W\s\d][^<>]{0,})>([\w\W\s\d][^<>]{0,})<\/TD><\/TR>/isIS
2868    
2869  "(?>.*/)foo"SI  "(?>.*/)foo"SI
# Line 3322  a random value. /Ix Line 3030  a random value. /Ix
3030  /A(*PRUNE)B|A(*PRUNE)C/K  /A(*PRUNE)B|A(*PRUNE)C/K
3031      AC      AC
3032    
 /--- A whole lot of tests of verbs with arguments are here rather than in test  
      11 because Perl doesn't seem to follow its specification entirely  
      correctly. ---/  
   
 /--- Perl 5.11 sets $REGERROR on the AC failure case here; PCRE does not. It is  
      not clear how Perl defines "involved in the failure of the match". ---/  
   
 /^(A(*THEN:A)B|C(*THEN:B)D)/K  
     AB  
     CD  
     ** Failers  
     AC  
     CB  
   
 /--- Check the use of names for success and failure. PCRE doesn't show these  
 names for success, though Perl does, contrary to its spec. ---/  
   
 /^(A(*PRUNE:A)B|C(*PRUNE:B)D)/K  
     AB  
     CD  
     ** Failers  
     AC  
     CB  
   
 /--- An empty name does not pass back an empty string. It is the same as if no  
 name were given. ---/  
   
 /^(A(*PRUNE:)B|C(*PRUNE:B)D)/K  
     AB  
     CD  
   
 /--- PRUNE goes to next bumpalong; COMMIT does not. ---/  
   
 /A(*PRUNE:A)B/K  
     ACAB  
   
 /(*MARK:A)(*PRUNE:B)(C|X)/KS  
     C  
     D  
   
 /(*MARK:A)(*PRUNE:B)(C|X)/KSS  
     C  
     D  
   
 /(*MARK:A)(*THEN:B)(C|X)/KS  
     C  
     D  
   
 /(*MARK:A)(*THEN:B)(C|X)/KSS  
     C  
     D  
   
 /--- This should fail, as the skip causes a bump to offset 3 (the skip) ---/  
   
 /A(*MARK:A)A+(*SKIP)(B|Z) | AC/xK  
     AAAC  
   
 /--- Same --/  
   
 /A(*MARK:A)A+(*MARK:B)(*SKIP:B)(B|Z) | AC/xK  
     AAAC  
   
3033  /--- This should fail; the SKIP advances by one, but when we get to AC, the  /--- This should fail; the SKIP advances by one, but when we get to AC, the
3034       PRUNE kills it. ---/       PRUNE kills it. Perl behaves differently. ---/
3035    
3036  /A(*PRUNE:A)A+(*SKIP:A)(B|Z) | AC/xK  /A(*PRUNE:A)A+(*SKIP:A)(B|Z) | AC/xK
3037      AAAC      AAAC
3038    
3039  /A(*:A)A+(*SKIP)(B|Z) | AC/xK  /--- Mark names can be duplicated. Perl doesn't give a mark for this one,
3040      AAAC  though PCRE does. ---/
   
 /--- This should fail, as a null name is the same as no name ---/  
   
 /A(*MARK:A)A+(*SKIP:)(B|Z) | AC/xK  
     AAAC  
   
 /--- This fails in PCRE, and I think that is in accordance with Perl's  
      documentation, though in Perl it succeeds. ---/  
   
 /A(*MARK:A)A+(*SKIP:B)(B|Z) | AAC/xK  
     AAAC  
   
 /--- Mark names can be duplicated ---/  
3041    
 /A(*:A)B|X(*:A)Y/K  
     AABC  
     XXYZ  
   
3042  /^A(*:A)B|^X(*:A)Y/K  /^A(*:A)B|^X(*:A)Y/K
3043      ** Failers      ** Failers
3044      XAQQ      XAQQ
3045    
 /--- A check on what happens after hitting a mark and them bumping along to  
 something that does not even start. Perl reports tags after the failures here,  
 though it does not when the individual letters are made into something  
 more complicated. ---/  
   
 /A(*:A)B|XX(*:B)Y/K  
     AABC  
     XXYZ  
     ** Failers  
     XAQQ  
     XAQQXZZ  
     AXQQQ  
     AXXQQQ  
   
3046  /--- COMMIT at the start of a pattern should be the same as an anchor. Perl  /--- COMMIT at the start of a pattern should be the same as an anchor. Perl
3047  optimizations defeat this. So does the PCRE optimization unless we disable it  optimizations defeat this. So does the PCRE optimization unless we disable it
3048  with \Y. ---/  with \Y. ---/
# Line 3437  with \Y. ---/ Line 3052  with \Y. ---/
3052      ** Failers      ** Failers
3053      DEFGABC\Y      DEFGABC\Y
3054    
 /--- Repeat some tests with added studying. ---/  
   
 /A(*COMMIT)B/+KS  
     ACABX  
   
 /A(*THEN)B|A(*THEN)C/KS  
     AC  
   
 /A(*PRUNE)B|A(*PRUNE)C/KS  
     AC  
   
 /^(A(*THEN:A)B|C(*THEN:B)D)/KS  
     AB  
     CD  
     ** Failers  
     AC  
     CB  
   
 /^(A(*PRUNE:A)B|C(*PRUNE:B)D)/KS  
     AB  
     CD  
     ** Failers  
     AC  
     CB  
   
 /^(A(*PRUNE:)B|C(*PRUNE:B)D)/KS  
     AB  
     CD  
   
 /A(*PRUNE:A)B/KS  
     ACAB  
   
 /(*MARK:A)(*PRUNE:B)(C|X)/KS  
     C  
     D  
   
 /(*MARK:A)(*THEN:B)(C|X)/KS  
     C  
     D  
   
 /A(*MARK:A)A+(*SKIP)(B|Z) | AC/xKS  
     AAAC  
   
 /A(*MARK:A)A+(*MARK:B)(*SKIP:B)(B|Z) | AC/xKS  
     AAAC  
   
 /A(*PRUNE:A)A+(*SKIP:A)(B|Z) | AC/xKS  
     AAAC  
   
 /A(*:A)A+(*SKIP)(B|Z) | AC/xKS  
     AAAC  
   
 /A(*MARK:A)A+(*SKIP:)(B|Z) | AC/xKS  
     AAAC  
   
 /A(*MARK:A)A+(*SKIP:B)(B|Z) | AAC/xKS  
     AAAC  
   
 /A(*:A)B|XX(*:B)Y/KS  
     AABC  
     XXYZ  
     ** Failers  
     XAQQ  
     XAQQXZZ  
     AXQQQ  
     AXXQQQ  
   
 /(*COMMIT)ABC/  
     ABCDEFG  
     ** Failers  
     DEFGABC\Y  
   
3055  /^(ab (c+(*THEN)cd) | xyz)/x  /^(ab (c+(*THEN)cd) | xyz)/x
3056      abcccd      abcccd
3057    
# Line 3559  with \Y. ---/ Line 3102  with \Y. ---/
3102  /\d*\R/BZ  /\d*\R/BZ
3103    
3104  /\s*\R/BZ  /\s*\R/BZ
3105        \x20\x0a
3106        \x20\x0d
3107        \x20\x0d\x0a
3108    
3109    /\S*\R/BZ
3110        a\x0a
3111    
3112    /X\h*\R/BZ
3113        X\x20\x0a
3114    
3115    /X\H*\R/BZ
3116        X\x0d\x0a
3117    
3118    /X\H+\R/BZ
3119        X\x0d\x0a
3120    
3121    /X\H++\R/BZ
3122        X\x0d\x0a
3123    
3124  /-- Perl treats this one differently, not failing the second string. I believe  /-- Perl treats this one differently, not failing the second string. I believe
3125      that is a bug in Perl. --/      that is a bug in Perl. --/
# Line 3741  with \Y. ---/ Line 3302  with \Y. ---/
3302    
3303  /^.*?(a(*THEN)b)c/  /^.*?(a(*THEN)b)c/
3304      aabc      aabc
3305    
3306    /^.*? (?1) c (?(DEFINE)(a(*THEN)b))/x
3307        aabc
3308    
3309  /^.*?(a(*THEN)b|z)c/  /^.*?(a(*THEN)b|z)c/
3310      aabc      aabc
# Line 3843  with \Y. ---/ Line 3407  with \Y. ---/
3407  /((?2))((?1))/SS  /((?2))((?1))/SS
3408      abc      abc
3409    
3410  /((?(R2)a+|(?1)b))/  /((?(R2)a+|(?1)b))/SS
3411      aaaabcde      aaaabcde
3412    
3413  /(?(R)a*(?1)|((?R))b)/  /(?(R)a*(?1)|((?R))b)/SS
3414      aaaabcde      aaaabcde
3415    
3416  /(a+|(?R)b)/  /(a+|(?R)b)/
# Line 3880  with \Y. ---/ Line 3444  with \Y. ---/
3444  /z(*ACCEPT)a/+I  /z(*ACCEPT)a/+I
3445      baxzbx      baxzbx
3446    
3447    /a(?:.)*?a/ims
3448        \Mabbbbbbbbbbbbbbbbbbbbba
3449    
3450    /a(?:.(*THEN))*?a/ims
3451        \Mabbbbbbbbbbbbbbbbbbbbba
3452    
3453    /a(?:.(*THEN:ABC))*?a/ims
3454        \Mabbbbbbbbbbbbbbbbbbbbba
3455    
3456    /-- These tests are in agreement with development Perl 5.015, which has fixed
3457        some things, but they don't all work with 5.012, so they aren't in the
3458        Perl-compatible tests. Those after the first come from Perl's own test
3459        files. --/
3460    
3461    /^((yes|no)(*THEN)(*F))?/
3462      yes
3463    
3464    /(A (.*)   C? (*THEN)  | A D) (*FAIL)/x
3465    AbcdCBefgBhiBqz
3466    
3467    /(A (.*)   C? (*THEN)  | A D) z/x
3468    AbcdCBefgBhiBqz
3469    
3470    /(A (.*)   C? (*THEN)  | A D) \s* (*FAIL)/x
3471    AbcdCBefgBhiBqz
3472    
3473    /(A (.*)   C? (*THEN)  | A D) \s* z/x
3474    AbcdCBefgBhiBqz
3475    
3476    /(A (.*)   (?:C|) (*THEN)  | A D) (*FAIL)/x
3477    AbcdCBefgBhiBqz
3478    
3479    /(A (.*)   (?:C|) (*THEN)  | A D) z/x
3480    AbcdCBefgBhiBqz
3481    
3482    /(A (.*)   C{0,6} (*THEN)  | A D) (*FAIL)/x
3483    AbcdCBefgBhiBqz
3484    
3485    /(A (.*)   C{0,6} (*THEN)  | A D) z/x
3486    AbcdCBefgBhiBqz
3487    
3488    /(A (.*)   (CE){0,6} (*THEN)  | A D) (*FAIL)/x
3489    AbcdCEBefgBhiBqz
3490    
3491    /(A (.*)   (CE){0,6} (*THEN)  | A D) z/x
3492    AbcdCEBefgBhiBqz
3493    
3494    /(A (.*)   (CE*){0,6} (*THEN)  | A D) (*FAIL)/x
3495    AbcdCBefgBhiBqz
3496    
3497    /(A (.*)   (CE*){0,6} (*THEN)  | A D) z/x
3498    AbcdCBefgBhiBqz
3499    
3500    /-----------------------------------------------/
3501    
3502    /^(?>a+)(?>(z+))\w/BZ
3503        aaaazzzzb
3504        ** Failers
3505        aazz
3506    
3507    /(.)(\1|a(?2))/
3508        bab
3509    
3510    /\1|(.)(?R)\1/
3511        cbbbc
3512    
3513    /(.)((?(1)c|a)|a(?2))/
3514        baa
3515    
3516    /(?P<abn>(?P=abn)xxx)/BZ
3517    
3518    /(a\1z)/BZ
3519    
3520    /^(?>a+)(?>b+)(?>c+)(?>d+)(?>e+)/
3521         \Maabbccddee
3522    
3523    /^(?>(a+))(?>(b+))(?>(c+))(?>(d+))(?>(e+))/
3524         \Maabbccddee
3525    
3526    /^(?>(a+))(?>b+)(?>(c+))(?>d+)(?>(e+))/
3527         \Maabbccddee
3528    
3529    /^a\x41z/<JS>
3530        aAz
3531        *** Failers
3532        ax41z
3533    
3534    /^a[m\x41]z/<JS>
3535        aAz
3536    
3537    /^a\x1z/<JS>
3538        ax1z
3539    
3540    /^a\u0041z/<JS>
3541        aAz
3542        *** Failers
3543        au0041z
3544    
3545    /^a[m\u0041]z/<JS>
3546        aAz
3547    
3548    /^a\u041z/<JS>
3549        au041z
3550        *** Failers
3551        aAz
3552    
3553    /^a\U0041z/<JS>
3554        aU0041z
3555        *** Failers
3556        aAz
3557    
3558    /(?(?=c)c|d)++Y/BZ
3559    
3560    /(?(?=c)c|d)*+Y/BZ
3561    
3562    /a[\NB]c/
3563        aNc
3564    
3565    /a[B-\Nc]/
3566    
3567    /(a)(?2){0,1999}?(b)/
3568    
3569    /(a)(?(DEFINE)(b))(?2){0,1999}?(?2)/
3570    
3571    /--- This test, with something more complicated than individual letters, causes
3572    different behaviour in Perl. Perhaps it disables some optimization; no tag is
3573    passed back for the failures, whereas in PCRE there is a tag. ---/
3574    
3575    /(A|P)(*:A)(B|P) | (X|P)(X|P)(*:B)(Y|P)/xK
3576        AABC
3577        XXYZ
3578        ** Failers
3579        XAQQ
3580        XAQQXZZ
3581        AXQQQ
3582        AXXQQQ
3583    
3584    /-- Perl doesn't give marks for these, though it does if the alternatives are
3585    replaced by single letters. --/
3586    
3587    /(b|q)(*:m)f|a(*:n)w/K
3588        aw
3589        ** Failers
3590        abc
3591    
3592    /(q|b)(*:m)f|a(*:n)w/K
3593        aw
3594        ** Failers
3595        abc
3596    
3597    /-- After a partial match, the behaviour is as for a failure. --/
3598    
3599    /^a(*:X)bcde/K
3600       abc\P
3601    
3602    /-- These are here because Perl doesn't return a mark, except for the first --/
3603    
3604    /(?=(*:x))(q|)/K+
3605        abc
3606    
3607    /(?=(*:x))((*:y)q|)/K+
3608        abc
3609    
3610    /(?=(*:x))(?:(*:y)q|)/K+
3611        abc
3612    
3613    /(?=(*:x))(?>(*:y)q|)/K+
3614        abc
3615    
3616    /(?=a(*:x))(?!a(*:y)c)/K+
3617        ab
3618    
3619    /(?=a(*:x))(?=a(*:y)c|)/K+
3620        ab
3621    
3622    /(..)\1/
3623        ab\P
3624        aba\P
3625        abab\P
3626    
3627    /(..)\1/i
3628        ab\P
3629        abA\P
3630        aBAb\P
3631    
3632    /(..)\1{2,}/
3633        ab\P
3634        aba\P
3635        abab\P
3636        ababa\P
3637        ababab\P
3638        ababab\P\P
3639        abababa\P
3640        abababa\P\P
3641    
3642    /(..)\1{2,}/i
3643        ab\P
3644        aBa\P
3645        aBAb\P
3646        AbaBA\P
3647        abABAb\P
3648        aBAbaB\P\P
3649        abABabA\P
3650        abaBABa\P\P
3651    
3652    /(..)\1{2,}?x/i
3653        ab\P
3654        abA\P
3655        aBAb\P
3656        abaBA\P
3657        abAbaB\P
3658        abaBabA\P
3659        abAbABaBx\P
3660    
3661    /^(..)\1/
3662        aba\P
3663    
3664    /^(..)\1{2,3}x/
3665        aba\P
3666        ababa\P
3667        ababa\P\P
3668        abababx
3669        ababababx
3670    
3671    /^(..)\1{2,3}?x/
3672        aba\P
3673        ababa\P
3674        ababa\P\P
3675        abababx
3676        ababababx
3677    
3678    /^(..)(\1{2,3})ab/
3679        abababab
3680    
3681    /^\R/
3682        \r\P
3683        \r\P\P
3684    
3685    /^\R{2,3}x/
3686        \r\P
3687        \r\P\P
3688        \r\r\P
3689        \r\r\P\P
3690        \r\r\r\P
3691        \r\r\r\P\P
3692        \r\rx
3693        \r\r\rx
3694    
3695    /^\R{2,3}?x/
3696        \r\P
3697        \r\P\P
3698        \r\r\P
3699        \r\r\P\P
3700        \r\r\r\P
3701        \r\r\r\P\P
3702        \r\rx
3703        \r\r\rx
3704    
3705    /^\R?x/
3706        \r\P
3707        \r\P\P
3708        x
3709        \rx
3710    
3711    /^\R+x/
3712        \r\P
3713        \r\P\P
3714        \r\n\P
3715        \r\n\P\P
3716        \rx
3717    
3718    /^a$/<CRLF>
3719        a\r\P
3720        a\r\P\P
3721    
3722    /^a$/m<CRLF>
3723        a\r\P
3724        a\r\P\P
3725    
3726    /^(a$|a\r)/<CRLF>
3727        a\r\P
3728        a\r\P\P
3729    
3730    /^(a$|a\r)/m<CRLF>
3731        a\r\P
3732        a\r\P\P
3733    
3734    /./<CRLF>
3735        \r\P
3736        \r\P\P
3737    
3738    /.{2,3}/<CRLF>
3739        \r\P
3740        \r\P\P
3741        \r\r\P
3742        \r\r\P\P
3743        \r\r\r\P
3744        \r\r\r\P\P
3745    
3746    /.{2,3}?/<CRLF>
3747        \r\P
3748        \r\P\P
3749        \r\r\P
3750        \r\r\P\P
3751        \r\r\r\P
3752        \r\r\r\P\P
3753    
3754    /-- These two are here because Perl does not match: it seems to allow the
3755    COMMIT to escape from the assertion. --/
3756    
3757    /(?=a(*COMMIT)b|ac)ac|ac/
3758        ac
3759    
3760    /(?=a(*COMMIT)b|(ac)) ac | (a)c/x
3761        ac
3762    
3763    "AB(C(D))(E(F))?(?(?=\2)(?=\4))"
3764        ABCDGHI\O03
3765    
3766  /-- End of testinput2 --/  /-- End of testinput2 --/
Legend:
Removed from v.701  
changed lines
  Added in v.963
webmaster@exim.org
 ViewVC Help
Powered by ViewVC 1.1.12