1NOTE implicit vs. explicit repetitions : 2009-02-02 2 3# Glenn Fowler <gsf@research.att.com> 4# conforming matches (column 4) must match one of the following BREs 5# NOMATCH 6# (0,.)\((\(.\),\(.\))(?,?)(\2,\3)\)* 7# (0,.)\((\(.\),\(.\))(\2,\3)(?,?)\)* 8# i.e., each 3-tuple has two identical elements and one (?,?) 9 10E ((..)|(.)) NULL NOMATCH 11E ((..)|(.))((..)|(.)) NULL NOMATCH 12E ((..)|(.))((..)|(.))((..)|(.)) NULL NOMATCH 13 14E ((..)|(.)){1} NULL NOMATCH 15E ((..)|(.)){2} NULL NOMATCH 16E ((..)|(.)){3} NULL NOMATCH 17 18E ((..)|(.))* NULL (0,0) 19 20E ((..)|(.)) a (0,1)(0,1)(?,?)(0,1) 21E ((..)|(.))((..)|(.)) a NOMATCH 22E ((..)|(.))((..)|(.))((..)|(.)) a NOMATCH 23 24E ((..)|(.)){1} a (0,1)(0,1)(?,?)(0,1) 25E ((..)|(.)){2} a NOMATCH 26E ((..)|(.)){3} a NOMATCH 27 28E ((..)|(.))* a (0,1)(0,1)(?,?)(0,1) 29 30E ((..)|(.)) aa (0,2)(0,2)(0,2)(?,?) 31E ((..)|(.))((..)|(.)) aa (0,2)(0,1)(?,?)(0,1)(1,2)(?,?)(1,2) 32E ((..)|(.))((..)|(.))((..)|(.)) aa NOMATCH 33 34E ((..)|(.)){1} aa (0,2)(0,2)(0,2)(?,?) 35E ((..)|(.)){2} aa (0,2)(1,2)(?,?)(1,2) 36E ((..)|(.)){3} aa NOMATCH 37 38E ((..)|(.))* aa (0,2)(0,2)(0,2)(?,?) 39 40E ((..)|(.)) aaa (0,2)(0,2)(0,2)(?,?) 41E ((..)|(.))((..)|(.)) aaa (0,3)(0,2)(0,2)(?,?)(2,3)(?,?)(2,3) 42E ((..)|(.))((..)|(.))((..)|(.)) aaa (0,3)(0,1)(?,?)(0,1)(1,2)(?,?)(1,2)(2,3)(?,?)(2,3) 43 44E ((..)|(.)){1} aaa (0,2)(0,2)(0,2)(?,?) 45E ((..)|(.)){2} aaa (0,3)(2,3)(?,?)(2,3) 46E ((..)|(.)){3} aaa (0,3)(2,3)(?,?)(2,3) 47 48E ((..)|(.))* aaa (0,3)(2,3)(?,?)(2,3) 49 50E ((..)|(.)) aaaa (0,2)(0,2)(0,2)(?,?) 51E ((..)|(.))((..)|(.)) aaaa (0,4)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?) 52E ((..)|(.))((..)|(.))((..)|(.)) aaaa (0,4)(0,2)(0,2)(?,?)(2,3)(?,?)(2,3)(3,4)(?,?)(3,4) 53 54E ((..)|(.)){1} aaaa (0,2)(0,2)(0,2)(?,?) 55E ((..)|(.)){2} aaaa (0,4)(2,4)(2,4)(?,?) 56E ((..)|(.)){3} aaaa (0,4)(3,4)(?,?)(3,4) 57 58E ((..)|(.))* aaaa (0,4)(2,4)(2,4)(?,?) 59 60E ((..)|(.)) aaaaa (0,2)(0,2)(0,2)(?,?) 61E ((..)|(.))((..)|(.)) aaaaa (0,4)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?) 62E ((..)|(.))((..)|(.))((..)|(.)) aaaaa (0,5)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?)(4,5)(?,?)(4,5) 63 64E ((..)|(.)){1} aaaaa (0,2)(0,2)(0,2)(?,?) 65E ((..)|(.)){2} aaaaa (0,4)(2,4)(2,4)(?,?) 66E ((..)|(.)){3} aaaaa (0,5)(4,5)(?,?)(4,5) 67 68E ((..)|(.))* aaaaa (0,5)(4,5)(?,?)(4,5) 69 70E ((..)|(.)) aaaaaa (0,2)(0,2)(0,2)(?,?) 71E ((..)|(.))((..)|(.)) aaaaaa (0,4)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?) 72E ((..)|(.))((..)|(.))((..)|(.)) aaaaaa (0,6)(0,2)(0,2)(?,?)(2,4)(2,4)(?,?)(4,6)(4,6)(?,?) 73 74E ((..)|(.)){1} aaaaaa (0,2)(0,2)(0,2)(?,?) 75E ((..)|(.)){2} aaaaaa (0,4)(2,4)(2,4)(?,?) 76E ((..)|(.)){3} aaaaaa (0,6)(4,6)(4,6)(?,?) 77 78E ((..)|(.))* aaaaaa (0,6)(4,6)(4,6)(?,?) 79 80NOTE additional repetition tests graciously provided by Chris Kuklewicz www.haskell.org 2009-02-02 81 82# These test a bug in OS X / FreeBSD / NetBSD, and libtree. 83# Linux/GLIBC gets the {8,} and {8,8} wrong. 84 85:HA#100:E X(.?){0,}Y X1234567Y (0,9)(7,8) 86:HA#101:E X(.?){1,}Y X1234567Y (0,9)(7,8) 87:HA#102:E X(.?){2,}Y X1234567Y (0,9)(7,8) 88:HA#103:E X(.?){3,}Y X1234567Y (0,9)(7,8) 89:HA#104:E X(.?){4,}Y X1234567Y (0,9)(7,8) 90:HA#105:E X(.?){5,}Y X1234567Y (0,9)(7,8) 91:HA#106:E X(.?){6,}Y X1234567Y (0,9)(7,8) 92:HA#107:E X(.?){7,}Y X1234567Y (0,9)(7,8) 93:HA#108:E X(.?){8,}Y X1234567Y (0,9)(8,8) 94:HA#110:E X(.?){0,8}Y X1234567Y (0,9)(7,8) 95:HA#111:E X(.?){1,8}Y X1234567Y (0,9)(7,8) 96:HA#112:E X(.?){2,8}Y X1234567Y (0,9)(7,8) 97:HA#113:E X(.?){3,8}Y X1234567Y (0,9)(7,8) 98:HA#114:E X(.?){4,8}Y X1234567Y (0,9)(7,8) 99:HA#115:E X(.?){5,8}Y X1234567Y (0,9)(7,8) 100:HA#116:E X(.?){6,8}Y X1234567Y (0,9)(7,8) 101:HA#117:E X(.?){7,8}Y X1234567Y (0,9)(7,8) 102:HA#118:E X(.?){8,8}Y X1234567Y (0,9)(8,8) 103 104# These test a fixed bug in my regex-tdfa that did not keep the expanded 105# form properly grouped, so right association did the wrong thing with 106# these ambiguous patterns (crafted just to test my code when I became 107# suspicious of my implementation). The first subexpression should use 108# "ab" then "a" then "bcd". 109 110# OS X / FreeBSD / NetBSD badly fail many of these, with impossible 111# results like (0,6)(4,5)(6,6). 112 113:HA#260:E (a|ab|c|bcd){0,}(d*) ababcd (0,6)(3,6)(6,6) 114:HA#261:E (a|ab|c|bcd){1,}(d*) ababcd (0,6)(3,6)(6,6) 115:HA#262:E (a|ab|c|bcd){2,}(d*) ababcd (0,6)(3,6)(6,6) 116:HA#263:E (a|ab|c|bcd){3,}(d*) ababcd (0,6)(3,6)(6,6) 117:HA#264:E (a|ab|c|bcd){4,}(d*) ababcd NOMATCH 118:HA#265:E (a|ab|c|bcd){0,10}(d*) ababcd (0,6)(3,6)(6,6) 119:HA#266:E (a|ab|c|bcd){1,10}(d*) ababcd (0,6)(3,6)(6,6) 120:HA#267:E (a|ab|c|bcd){2,10}(d*) ababcd (0,6)(3,6)(6,6) 121:HA#268:E (a|ab|c|bcd){3,10}(d*) ababcd (0,6)(3,6)(6,6) 122:HA#269:E (a|ab|c|bcd){4,10}(d*) ababcd NOMATCH 123:HA#270:E (a|ab|c|bcd)*(d*) ababcd (0,6)(3,6)(6,6) 124:HA#271:E (a|ab|c|bcd)+(d*) ababcd (0,6)(3,6)(6,6) 125 126# The above worked on Linux/GLIBC but the following often fail. 127# They also trip up OS X / FreeBSD / NetBSD: 128 129:HA#280:E (ab|a|c|bcd){0,}(d*) ababcd (0,6)(3,6)(6,6) 130:HA#281:E (ab|a|c|bcd){1,}(d*) ababcd (0,6)(3,6)(6,6) 131:HA#282:E (ab|a|c|bcd){2,}(d*) ababcd (0,6)(3,6)(6,6) 132:HA#283:E (ab|a|c|bcd){3,}(d*) ababcd (0,6)(3,6)(6,6) 133:HA#284:E (ab|a|c|bcd){4,}(d*) ababcd NOMATCH 134:HA#285:E (ab|a|c|bcd){0,10}(d*) ababcd (0,6)(3,6)(6,6) 135:HA#286:E (ab|a|c|bcd){1,10}(d*) ababcd (0,6)(3,6)(6,6) 136:HA#287:E (ab|a|c|bcd){2,10}(d*) ababcd (0,6)(3,6)(6,6) 137:HA#288:E (ab|a|c|bcd){3,10}(d*) ababcd (0,6)(3,6)(6,6) 138:HA#289:E (ab|a|c|bcd){4,10}(d*) ababcd NOMATCH 139:HA#290:E (ab|a|c|bcd)*(d*) ababcd (0,6)(3,6)(6,6) 140:HA#291:E (ab|a|c|bcd)+(d*) ababcd (0,6)(3,6)(6,6) 141