NOTE implicit vs. explicit repetitions : 2009-02-02 # Glenn Fowler # conforming matches (column 4) must match one of the following BREs # NOMATCH # (0,.)\((\(.\),\(.\))(?,?)(\2,\3)\)* # (0,.)\((\(.\),\(.\))(\2,\3)(?,?)\)* # i.e., each 3-tuple has two identical elements and one (?,?) E ((..)|(.)) NULL NOMATCH E ((..)|(.))((..)|(.)) NULL NOMATCH E ((..)|(.))((..)|(.))((..)|(.)) NULL NOMATCH E ((..)|(.)){1} NULL NOMATCH E ((..)|(.)){2} NULL NOMATCH E ((..)|(.)){3} NULL NOMATCH E ((..)|(.))* NULL (0,0) E ((..)|(.)) a (0,1)(0,1)(?,?)(0,1) E ((..)|(.))((..)|(.)) a NOMATCH E ((..)|(.))((..)|(.))((..)|(.)) a NOMATCH E ((..)|(.)){1} a (0,1)(0,1)(?,?)(0,1) E ((..)|(.)){2} a NOMATCH E ((..)|(.)){3} a NOMATCH E ((..)|(.))* a (0,1)(0,1)(?,?)(0,1) E ((..)|(.)) aa (0,2)(0,2)(0,2)(?,?) E ((..)|(.))((..)|(.)) aa (0,2)(0,1)(?,?)(0,1)(1,2)(?,?)(1,2) E ((..)|(.))((..)|(.))((..)|(.)) aa NOMATCH E ((..)|(.)){1} aa (0,2)(0,2)(0,2)(?,?) E ((..)|(.)){2} aa (0,2)(1,2)(?,?)(1,2) E ((..)|(.)){3} aa NOMATCH E ((..)|(.))* aa (0,2)(0,2)(0,2)(?,?) E ((..)|(.)) aaa (0,2)(0,2)(0,2)(?,?) E ((..)|(.))((..)|(.)) aaa (0,3)(0,2)(0,2)(?,?)(2,3)(?,?)(2,3) E ((..)|(.))((..)|(.))((..)|(.)) aaa (0,3)(0,1)(?,?)(0,1)(1,2)(?,?)(1,2)(2,3)(?,?)(2,3) E ((..)|(.)){1} aaa (0,2)(0,2)(0,2)(?,?) #E ((..)|(.)){2} aaa (0,3)(2,3)(?,?)(2,3) E ((..)|(.)){2} aaa (0,3)(2,3)(0,2)(2,3) RE2/Go E ((..)|(.)){3} aaa (0,3)(2,3)(?,?)(2,3) #E ((..)|(.))* aaa (0,3)(2,3)(?,?)(2,3) E ((..)|(.))* aaa (0,3)(2,3)(0,2)(2,3) RE2/Go E ((..)|(.)) aaaa (0,2)(0,2)(0,2)(?,?) E ((..)|(.))((..)|(.)) aaaa (0,4)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?) E ((..)|(.))((..)|(.))((..)|(.)) aaaa (0,4)(0,2)(0,2)(?,?)(2,3)(?,?)(2,3)(3,4)(?,?)(3,4) E ((..)|(.)){1} aaaa (0,2)(0,2)(0,2)(?,?) E ((..)|(.)){2} aaaa (0,4)(2,4)(2,4)(?,?) #E ((..)|(.)){3} aaaa (0,4)(3,4)(?,?)(3,4) E ((..)|(.)){3} aaaa (0,4)(3,4)(0,2)(3,4) RE2/Go E ((..)|(.))* aaaa (0,4)(2,4)(2,4)(?,?) E ((..)|(.)) aaaaa (0,2)(0,2)(0,2)(?,?) E ((..)|(.))((..)|(.)) aaaaa (0,4)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?) E ((..)|(.))((..)|(.))((..)|(.)) aaaaa (0,5)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?)(4,5)(?,?)(4,5) E ((..)|(.)){1} aaaaa (0,2)(0,2)(0,2)(?,?) E ((..)|(.)){2} aaaaa (0,4)(2,4)(2,4)(?,?) #E ((..)|(.)){3} aaaaa (0,5)(4,5)(?,?)(4,5) E ((..)|(.)){3} aaaaa (0,5)(4,5)(2,4)(4,5) RE2/Go #E ((..)|(.))* aaaaa (0,5)(4,5)(?,?)(4,5) E ((..)|(.))* aaaaa (0,5)(4,5)(2,4)(4,5) RE2/Go E ((..)|(.)) aaaaaa (0,2)(0,2)(0,2)(?,?) E ((..)|(.))((..)|(.)) aaaaaa (0,4)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?) E ((..)|(.))((..)|(.))((..)|(.)) aaaaaa (0,6)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?)(4,6)(4,6)(?,?) E ((..)|(.)){1} aaaaaa (0,2)(0,2)(0,2)(?,?) E ((..)|(.)){2} aaaaaa (0,4)(2,4)(2,4)(?,?) E ((..)|(.)){3} aaaaaa (0,6)(4,6)(4,6)(?,?) E ((..)|(.))* aaaaaa (0,6)(4,6)(4,6)(?,?) NOTE additional repetition tests graciously provided by Chris Kuklewicz www.haskell.org 2009-02-02 # These test a bug in OS X / FreeBSD / NetBSD, and libtree. # Linux/GLIBC gets the {8,} and {8,8} wrong. :HA#100:E X(.?){0,}Y X1234567Y (0,9)(7,8) :HA#101:E X(.?){1,}Y X1234567Y (0,9)(7,8) :HA#102:E X(.?){2,}Y X1234567Y (0,9)(7,8) :HA#103:E X(.?){3,}Y X1234567Y (0,9)(7,8) :HA#104:E X(.?){4,}Y X1234567Y (0,9)(7,8) :HA#105:E X(.?){5,}Y X1234567Y (0,9)(7,8) :HA#106:E X(.?){6,}Y X1234567Y (0,9)(7,8) :HA#107:E X(.?){7,}Y X1234567Y (0,9)(7,8) :HA#108:E X(.?){8,}Y X1234567Y (0,9)(8,8) #:HA#110:E X(.?){0,8}Y X1234567Y (0,9)(7,8) :HA#110:E X(.?){0,8}Y X1234567Y (0,9)(8,8) RE2/Go #:HA#111:E X(.?){1,8}Y X1234567Y (0,9)(7,8) :HA#111:E X(.?){1,8}Y X1234567Y (0,9)(8,8) RE2/Go #:HA#112:E X(.?){2,8}Y X1234567Y (0,9)(7,8) :HA#112:E X(.?){2,8}Y X1234567Y (0,9)(8,8) RE2/Go #:HA#113:E X(.?){3,8}Y X1234567Y (0,9)(7,8) :HA#113:E X(.?){3,8}Y X1234567Y (0,9)(8,8) RE2/Go #:HA#114:E X(.?){4,8}Y X1234567Y (0,9)(7,8) :HA#114:E X(.?){4,8}Y X1234567Y (0,9)(8,8) RE2/Go #:HA#115:E X(.?){5,8}Y X1234567Y (0,9)(7,8) :HA#115:E X(.?){5,8}Y X1234567Y (0,9)(8,8) RE2/Go #:HA#116:E X(.?){6,8}Y X1234567Y (0,9)(7,8) :HA#116:E X(.?){6,8}Y X1234567Y (0,9)(8,8) RE2/Go #:HA#117:E X(.?){7,8}Y X1234567Y (0,9)(7,8) :HA#117:E X(.?){7,8}Y X1234567Y (0,9)(8,8) RE2/Go :HA#118:E X(.?){8,8}Y X1234567Y (0,9)(8,8) # These test a fixed bug in my regex-tdfa that did not keep the expanded # form properly grouped, so right association did the wrong thing with # these ambiguous patterns (crafted just to test my code when I became # suspicious of my implementation). The first subexpression should use # "ab" then "a" then "bcd". # OS X / FreeBSD / NetBSD badly fail many of these, with impossible # results like (0,6)(4,5)(6,6). #:HA#260:E (a|ab|c|bcd){0,}(d*) ababcd (0,6)(3,6)(6,6) :HA#260:E (a|ab|c|bcd){0,}(d*) ababcd (0,1)(0,1)(1,1) Rust #:HA#261:E (a|ab|c|bcd){1,}(d*) ababcd (0,6)(3,6)(6,6) :HA#261:E (a|ab|c|bcd){1,}(d*) ababcd (0,1)(0,1)(1,1) Rust :HA#262:E (a|ab|c|bcd){2,}(d*) ababcd (0,6)(3,6)(6,6) :HA#263:E (a|ab|c|bcd){3,}(d*) ababcd (0,6)(3,6)(6,6) :HA#264:E (a|ab|c|bcd){4,}(d*) ababcd NOMATCH #:HA#265:E (a|ab|c|bcd){0,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#265:E (a|ab|c|bcd){0,10}(d*) ababcd (0,1)(0,1)(1,1) Rust #:HA#266:E (a|ab|c|bcd){1,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#266:E (a|ab|c|bcd){1,10}(d*) ababcd (0,1)(0,1)(1,1) Rust :HA#267:E (a|ab|c|bcd){2,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#268:E (a|ab|c|bcd){3,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#269:E (a|ab|c|bcd){4,10}(d*) ababcd NOMATCH #:HA#270:E (a|ab|c|bcd)*(d*) ababcd (0,6)(3,6)(6,6) :HA#270:E (a|ab|c|bcd)*(d*) ababcd (0,1)(0,1)(1,1) Rust #:HA#271:E (a|ab|c|bcd)+(d*) ababcd (0,6)(3,6)(6,6) :HA#271:E (a|ab|c|bcd)+(d*) ababcd (0,1)(0,1)(1,1) Rust # The above worked on Linux/GLIBC but the following often fail. # They also trip up OS X / FreeBSD / NetBSD: #:HA#280:E (ab|a|c|bcd){0,}(d*) ababcd (0,6)(3,6)(6,6) :HA#280:E (ab|a|c|bcd){0,}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go #:HA#281:E (ab|a|c|bcd){1,}(d*) ababcd (0,6)(3,6)(6,6) :HA#281:E (ab|a|c|bcd){1,}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go #:HA#282:E (ab|a|c|bcd){2,}(d*) ababcd (0,6)(3,6)(6,6) :HA#282:E (ab|a|c|bcd){2,}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go #:HA#283:E (ab|a|c|bcd){3,}(d*) ababcd (0,6)(3,6)(6,6) :HA#283:E (ab|a|c|bcd){3,}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go :HA#284:E (ab|a|c|bcd){4,}(d*) ababcd NOMATCH #:HA#285:E (ab|a|c|bcd){0,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#285:E (ab|a|c|bcd){0,10}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go #:HA#286:E (ab|a|c|bcd){1,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#286:E (ab|a|c|bcd){1,10}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go #:HA#287:E (ab|a|c|bcd){2,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#287:E (ab|a|c|bcd){2,10}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go #:HA#288:E (ab|a|c|bcd){3,10}(d*) ababcd (0,6)(3,6)(6,6) :HA#288:E (ab|a|c|bcd){3,10}(d*) ababcd (0,6)(4,5)(5,6) RE2/Go :HA#289:E (ab|a|c|bcd){4,10}(d*) ababcd NOMATCH #:HA#290:E (ab|a|c|bcd)*(d*) ababcd (0,6)(3,6)(6,6) :HA#290:E (ab|a|c|bcd)*(d*) ababcd (0,6)(4,5)(5,6) RE2/Go #:HA#291:E (ab|a|c|bcd)+(d*) ababcd (0,6)(3,6)(6,6) :HA#291:E (ab|a|c|bcd)+(d*) ababcd (0,6)(4,5)(5,6) RE2/Go