+++ /dev/null
-#!/usr/bin/env python\r
-# -*- mode: python -*-\r
-\r
-# Re test suite and benchmark suite v1.5\r
-\r
-# The 3 possible outcomes for each pattern\r
-[SUCCEED, FAIL, SYNTAX_ERROR] = range(3)\r
-\r
-# Benchmark suite (needs expansion)\r
-#\r
-# The benchmark suite does not test correctness, just speed. The\r
-# first element of each tuple is the regex pattern; the second is a\r
-# string to match it against. The benchmarking code will embed the\r
-# second string inside several sizes of padding, to test how regex\r
-# matching performs on large strings.\r
-\r
-benchmarks = [\r
-\r
- # test common prefix\r
- ('Python|Perl', 'Perl'), # Alternation\r
- ('(Python|Perl)', 'Perl'), # Grouped alternation\r
-\r
- ('Python|Perl|Tcl', 'Perl'), # Alternation\r
- ('(Python|Perl|Tcl)', 'Perl'), # Grouped alternation\r
-\r
- ('(Python)\\1', 'PythonPython'), # Backreference\r
- ('([0a-z][a-z0-9]*,)+', 'a5,b7,c9,'), # Disable the fastmap optimization\r
- ('([a-z][a-z0-9]*,)+', 'a5,b7,c9,'), # A few sets\r
-\r
- ('Python', 'Python'), # Simple text literal\r
- ('.*Python', 'Python'), # Bad text literal\r
- ('.*Python.*', 'Python'), # Worse text literal\r
- ('.*(Python)', 'Python'), # Bad text literal with grouping\r
-\r
-]\r
-\r
-# Test suite (for verifying correctness)\r
-#\r
-# The test suite is a list of 5- or 3-tuples. The 5 parts of a\r
-# complete tuple are:\r
-# element 0: a string containing the pattern\r
-# 1: the string to match against the pattern\r
-# 2: the expected result (SUCCEED, FAIL, SYNTAX_ERROR)\r
-# 3: a string that will be eval()'ed to produce a test string.\r
-# This is an arbitrary Python expression; the available\r
-# variables are "found" (the whole match), and "g1", "g2", ...\r
-# up to "g99" contain the contents of each group, or the\r
-# string 'None' if the group wasn't given a value, or the\r
-# string 'Error' if the group index was out of range;\r
-# also "groups", the return value of m.group() (a tuple).\r
-# 4: The expected result of evaluating the expression.\r
-# If the two don't match, an error is reported.\r
-#\r
-# If the regex isn't expected to work, the latter two elements can be omitted.\r
-\r
-tests = [\r
- # Test ?P< and ?P= extensions\r
- ('(?P<foo_123', '', SYNTAX_ERROR), # Unterminated group identifier\r
- ('(?P<1>a)', '', SYNTAX_ERROR), # Begins with a digit\r
- ('(?P<!>a)', '', SYNTAX_ERROR), # Begins with an illegal char\r
- ('(?P<foo!>a)', '', SYNTAX_ERROR), # Begins with an illegal char\r
-\r
- # Same tests, for the ?P= form\r
- ('(?P<foo_123>a)(?P=foo_123', 'aa', SYNTAX_ERROR),\r
- ('(?P<foo_123>a)(?P=1)', 'aa', SYNTAX_ERROR),\r
- ('(?P<foo_123>a)(?P=!)', 'aa', SYNTAX_ERROR),\r
- ('(?P<foo_123>a)(?P=foo_124', 'aa', SYNTAX_ERROR), # Backref to undefined group\r
-\r
- ('(?P<foo_123>a)', 'a', SUCCEED, 'g1', 'a'),\r
- ('(?P<foo_123>a)(?P=foo_123)', 'aa', SUCCEED, 'g1', 'a'),\r
-\r
- # Test octal escapes\r
- ('\\1', 'a', SYNTAX_ERROR), # Backreference\r
- ('[\\1]', '\1', SUCCEED, 'found', '\1'), # Character\r
- ('\\09', chr(0) + '9', SUCCEED, 'found', chr(0) + '9'),\r
- ('\\141', 'a', SUCCEED, 'found', 'a'),\r
- ('(a)(b)(c)(d)(e)(f)(g)(h)(i)(j)(k)(l)\\119', 'abcdefghijklk9', SUCCEED, 'found+"-"+g11', 'abcdefghijklk9-k'),\r
-\r
- # Test \0 is handled everywhere\r
- (r'\0', '\0', SUCCEED, 'found', '\0'),\r
- (r'[\0a]', '\0', SUCCEED, 'found', '\0'),\r
- (r'[a\0]', '\0', SUCCEED, 'found', '\0'),\r
- (r'[^a\0]', '\0', FAIL),\r
-\r
- # Test various letter escapes\r
- (r'\a[\b]\f\n\r\t\v', '\a\b\f\n\r\t\v', SUCCEED, 'found', '\a\b\f\n\r\t\v'),\r
- (r'[\a][\b][\f][\n][\r][\t][\v]', '\a\b\f\n\r\t\v', SUCCEED, 'found', '\a\b\f\n\r\t\v'),\r
- # NOTE: not an error under PCRE/PRE:\r
- # (r'\u', '', SYNTAX_ERROR), # A Perl escape\r
- (r'\c\e\g\h\i\j\k\m\o\p\q\y\z', 'ceghijkmopqyz', SUCCEED, 'found', 'ceghijkmopqyz'),\r
- (r'\xff', '\377', SUCCEED, 'found', chr(255)),\r
- # new \x semantics\r
- (r'\x00ffffffffffffff', '\377', FAIL, 'found', chr(255)),\r
- (r'\x00f', '\017', FAIL, 'found', chr(15)),\r
- (r'\x00fe', '\376', FAIL, 'found', chr(254)),\r
- # (r'\x00ffffffffffffff', '\377', SUCCEED, 'found', chr(255)),\r
- # (r'\x00f', '\017', SUCCEED, 'found', chr(15)),\r
- # (r'\x00fe', '\376', SUCCEED, 'found', chr(254)),\r
-\r
- (r"^\w+=(\\[\000-\277]|[^\n\\])*", "SRC=eval.c g.c blah blah blah \\\\\n\tapes.c",\r
- SUCCEED, 'found', "SRC=eval.c g.c blah blah blah \\\\"),\r
-\r
- # Test that . only matches \n in DOTALL mode\r
- ('a.b', 'acb', SUCCEED, 'found', 'acb'),\r
- ('a.b', 'a\nb', FAIL),\r
- ('a.*b', 'acc\nccb', FAIL),\r
- ('a.{4,5}b', 'acc\nccb', FAIL),\r
- ('a.b', 'a\rb', SUCCEED, 'found', 'a\rb'),\r
- ('a.b(?s)', 'a\nb', SUCCEED, 'found', 'a\nb'),\r
- ('a.*(?s)b', 'acc\nccb', SUCCEED, 'found', 'acc\nccb'),\r
- ('(?s)a.{4,5}b', 'acc\nccb', SUCCEED, 'found', 'acc\nccb'),\r
- ('(?s)a.b', 'a\nb', SUCCEED, 'found', 'a\nb'),\r
-\r
- (')', '', SYNTAX_ERROR), # Unmatched right bracket\r
- ('', '', SUCCEED, 'found', ''), # Empty pattern\r
- ('abc', 'abc', SUCCEED, 'found', 'abc'),\r
- ('abc', 'xbc', FAIL),\r
- ('abc', 'axc', FAIL),\r
- ('abc', 'abx', FAIL),\r
- ('abc', 'xabcy', SUCCEED, 'found', 'abc'),\r
- ('abc', 'ababc', SUCCEED, 'found', 'abc'),\r
- ('ab*c', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab*bc', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab*bc', 'abbc', SUCCEED, 'found', 'abbc'),\r
- ('ab*bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab+bc', 'abbc', SUCCEED, 'found', 'abbc'),\r
- ('ab+bc', 'abc', FAIL),\r
- ('ab+bc', 'abq', FAIL),\r
- ('ab+bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab?bc', 'abbc', SUCCEED, 'found', 'abbc'),\r
- ('ab?bc', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab?bc', 'abbbbc', FAIL),\r
- ('ab?c', 'abc', SUCCEED, 'found', 'abc'),\r
- ('^abc$', 'abc', SUCCEED, 'found', 'abc'),\r
- ('^abc$', 'abcc', FAIL),\r
- ('^abc', 'abcc', SUCCEED, 'found', 'abc'),\r
- ('^abc$', 'aabc', FAIL),\r
- ('abc$', 'aabc', SUCCEED, 'found', 'abc'),\r
- ('^', 'abc', SUCCEED, 'found+"-"', '-'),\r
- ('$', 'abc', SUCCEED, 'found+"-"', '-'),\r
- ('a.c', 'abc', SUCCEED, 'found', 'abc'),\r
- ('a.c', 'axc', SUCCEED, 'found', 'axc'),\r
- ('a.*c', 'axyzc', SUCCEED, 'found', 'axyzc'),\r
- ('a.*c', 'axyzd', FAIL),\r
- ('a[bc]d', 'abc', FAIL),\r
- ('a[bc]d', 'abd', SUCCEED, 'found', 'abd'),\r
- ('a[b-d]e', 'abd', FAIL),\r
- ('a[b-d]e', 'ace', SUCCEED, 'found', 'ace'),\r
- ('a[b-d]', 'aac', SUCCEED, 'found', 'ac'),\r
- ('a[-b]', 'a-', SUCCEED, 'found', 'a-'),\r
- ('a[\\-b]', 'a-', SUCCEED, 'found', 'a-'),\r
- # NOTE: not an error under PCRE/PRE:\r
- # ('a[b-]', 'a-', SYNTAX_ERROR),\r
- ('a[]b', '-', SYNTAX_ERROR),\r
- ('a[', '-', SYNTAX_ERROR),\r
- ('a\\', '-', SYNTAX_ERROR),\r
- ('abc)', '-', SYNTAX_ERROR),\r
- ('(abc', '-', SYNTAX_ERROR),\r
- ('a]', 'a]', SUCCEED, 'found', 'a]'),\r
- ('a[]]b', 'a]b', SUCCEED, 'found', 'a]b'),\r
- ('a[\]]b', 'a]b', SUCCEED, 'found', 'a]b'),\r
- ('a[^bc]d', 'aed', SUCCEED, 'found', 'aed'),\r
- ('a[^bc]d', 'abd', FAIL),\r
- ('a[^-b]c', 'adc', SUCCEED, 'found', 'adc'),\r
- ('a[^-b]c', 'a-c', FAIL),\r
- ('a[^]b]c', 'a]c', FAIL),\r
- ('a[^]b]c', 'adc', SUCCEED, 'found', 'adc'),\r
- ('\\ba\\b', 'a-', SUCCEED, '"-"', '-'),\r
- ('\\ba\\b', '-a', SUCCEED, '"-"', '-'),\r
- ('\\ba\\b', '-a-', SUCCEED, '"-"', '-'),\r
- ('\\by\\b', 'xy', FAIL),\r
- ('\\by\\b', 'yz', FAIL),\r
- ('\\by\\b', 'xyz', FAIL),\r
- ('x\\b', 'xyz', FAIL),\r
- ('x\\B', 'xyz', SUCCEED, '"-"', '-'),\r
- ('\\Bz', 'xyz', SUCCEED, '"-"', '-'),\r
- ('z\\B', 'xyz', FAIL),\r
- ('\\Bx', 'xyz', FAIL),\r
- ('\\Ba\\B', 'a-', FAIL, '"-"', '-'),\r
- ('\\Ba\\B', '-a', FAIL, '"-"', '-'),\r
- ('\\Ba\\B', '-a-', FAIL, '"-"', '-'),\r
- ('\\By\\B', 'xy', FAIL),\r
- ('\\By\\B', 'yz', FAIL),\r
- ('\\By\\b', 'xy', SUCCEED, '"-"', '-'),\r
- ('\\by\\B', 'yz', SUCCEED, '"-"', '-'),\r
- ('\\By\\B', 'xyz', SUCCEED, '"-"', '-'),\r
- ('ab|cd', 'abc', SUCCEED, 'found', 'ab'),\r
- ('ab|cd', 'abcd', SUCCEED, 'found', 'ab'),\r
- ('()ef', 'def', SUCCEED, 'found+"-"+g1', 'ef-'),\r
- ('$b', 'b', FAIL),\r
- ('a\\(b', 'a(b', SUCCEED, 'found+"-"+g1', 'a(b-Error'),\r
- ('a\\(*b', 'ab', SUCCEED, 'found', 'ab'),\r
- ('a\\(*b', 'a((b', SUCCEED, 'found', 'a((b'),\r
- ('a\\\\b', 'a\\b', SUCCEED, 'found', 'a\\b'),\r
- ('((a))', 'abc', SUCCEED, 'found+"-"+g1+"-"+g2', 'a-a-a'),\r
- ('(a)b(c)', 'abc', SUCCEED, 'found+"-"+g1+"-"+g2', 'abc-a-c'),\r
- ('a+b+c', 'aabbabc', SUCCEED, 'found', 'abc'),\r
- ('(a+|b)*', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'),\r
- ('(a+|b)+', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'),\r
- ('(a+|b)?', 'ab', SUCCEED, 'found+"-"+g1', 'a-a'),\r
- (')(', '-', SYNTAX_ERROR),\r
- ('[^ab]*', 'cde', SUCCEED, 'found', 'cde'),\r
- ('abc', '', FAIL),\r
- ('a*', '', SUCCEED, 'found', ''),\r
- ('a|b|c|d|e', 'e', SUCCEED, 'found', 'e'),\r
- ('(a|b|c|d|e)f', 'ef', SUCCEED, 'found+"-"+g1', 'ef-e'),\r
- ('abcd*efg', 'abcdefg', SUCCEED, 'found', 'abcdefg'),\r
- ('ab*', 'xabyabbbz', SUCCEED, 'found', 'ab'),\r
- ('ab*', 'xayabbbz', SUCCEED, 'found', 'a'),\r
- ('(ab|cd)e', 'abcde', SUCCEED, 'found+"-"+g1', 'cde-cd'),\r
- ('[abhgefdc]ij', 'hij', SUCCEED, 'found', 'hij'),\r
- ('^(ab|cd)e', 'abcde', FAIL, 'xg1y', 'xy'),\r
- ('(abc|)ef', 'abcdef', SUCCEED, 'found+"-"+g1', 'ef-'),\r
- ('(a|b)c*d', 'abcd', SUCCEED, 'found+"-"+g1', 'bcd-b'),\r
- ('(ab|ab*)bc', 'abc', SUCCEED, 'found+"-"+g1', 'abc-a'),\r
- ('a([bc]*)c*', 'abc', SUCCEED, 'found+"-"+g1', 'abc-bc'),\r
- ('a([bc]*)(c*d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-bc-d'),\r
- ('a([bc]+)(c*d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-bc-d'),\r
- ('a([bc]*)(c+d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-b-cd'),\r
- ('a[bcd]*dcdcde', 'adcdcde', SUCCEED, 'found', 'adcdcde'),\r
- ('a[bcd]+dcdcde', 'adcdcde', FAIL),\r
- ('(ab|a)b*c', 'abc', SUCCEED, 'found+"-"+g1', 'abc-ab'),\r
- ('((a)(b)c)(d)', 'abcd', SUCCEED, 'g1+"-"+g2+"-"+g3+"-"+g4', 'abc-a-b-d'),\r
- ('[a-zA-Z_][a-zA-Z0-9_]*', 'alpha', SUCCEED, 'found', 'alpha'),\r
- ('^a(bc+|b[eh])g|.h$', 'abh', SUCCEED, 'found+"-"+g1', 'bh-None'),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'effgz', SUCCEED, 'found+"-"+g1+"-"+g2', 'effgz-effgz-None'),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'ij', SUCCEED, 'found+"-"+g1+"-"+g2', 'ij-ij-j'),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'effg', FAIL),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'bcdd', FAIL),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'reffgz', SUCCEED, 'found+"-"+g1+"-"+g2', 'effgz-effgz-None'),\r
- ('(((((((((a)))))))))', 'a', SUCCEED, 'found', 'a'),\r
- ('multiple words of text', 'uh-uh', FAIL),\r
- ('multiple words', 'multiple words, yeah', SUCCEED, 'found', 'multiple words'),\r
- ('(.*)c(.*)', 'abcde', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcde-ab-de'),\r
- ('\\((.*), (.*)\\)', '(a, b)', SUCCEED, 'g2+"-"+g1', 'b-a'),\r
- ('[k]', 'ab', FAIL),\r
- ('a[-]?c', 'ac', SUCCEED, 'found', 'ac'),\r
- ('(abc)\\1', 'abcabc', SUCCEED, 'g1', 'abc'),\r
- ('([a-c]*)\\1', 'abcabc', SUCCEED, 'g1', 'abc'),\r
- ('^(.+)?B', 'AB', SUCCEED, 'g1', 'A'),\r
- ('(a+).\\1$', 'aaaaa', SUCCEED, 'found+"-"+g1', 'aaaaa-aa'),\r
- ('^(a+).\\1$', 'aaaa', FAIL),\r
- ('(abc)\\1', 'abcabc', SUCCEED, 'found+"-"+g1', 'abcabc-abc'),\r
- ('([a-c]+)\\1', 'abcabc', SUCCEED, 'found+"-"+g1', 'abcabc-abc'),\r
- ('(a)\\1', 'aa', SUCCEED, 'found+"-"+g1', 'aa-a'),\r
- ('(a+)\\1', 'aa', SUCCEED, 'found+"-"+g1', 'aa-a'),\r
- ('(a+)+\\1', 'aa', SUCCEED, 'found+"-"+g1', 'aa-a'),\r
- ('(a).+\\1', 'aba', SUCCEED, 'found+"-"+g1', 'aba-a'),\r
- ('(a)ba*\\1', 'aba', SUCCEED, 'found+"-"+g1', 'aba-a'),\r
- ('(aa|a)a\\1$', 'aaa', SUCCEED, 'found+"-"+g1', 'aaa-a'),\r
- ('(a|aa)a\\1$', 'aaa', SUCCEED, 'found+"-"+g1', 'aaa-a'),\r
- ('(a+)a\\1$', 'aaa', SUCCEED, 'found+"-"+g1', 'aaa-a'),\r
- ('([abc]*)\\1', 'abcabc', SUCCEED, 'found+"-"+g1', 'abcabc-abc'),\r
- ('(a)(b)c|ab', 'ab', SUCCEED, 'found+"-"+g1+"-"+g2', 'ab-None-None'),\r
- ('(a)+x', 'aaax', SUCCEED, 'found+"-"+g1', 'aaax-a'),\r
- ('([ac])+x', 'aacx', SUCCEED, 'found+"-"+g1', 'aacx-c'),\r
- ('([^/]*/)*sub1/', 'd:msgs/tdir/sub1/trial/away.cpp', SUCCEED, 'found+"-"+g1', 'd:msgs/tdir/sub1/-tdir/'),\r
- ('([^.]*)\\.([^:]*):[T ]+(.*)', 'track1.title:TBlah blah blah', SUCCEED, 'found+"-"+g1+"-"+g2+"-"+g3', 'track1.title:TBlah blah blah-track1-title-Blah blah blah'),\r
- ('([^N]*N)+', 'abNNxyzN', SUCCEED, 'found+"-"+g1', 'abNNxyzN-xyzN'),\r
- ('([^N]*N)+', 'abNNxyz', SUCCEED, 'found+"-"+g1', 'abNN-N'),\r
- ('([abc]*)x', 'abcx', SUCCEED, 'found+"-"+g1', 'abcx-abc'),\r
- ('([abc]*)x', 'abc', FAIL),\r
- ('([xyz]*)x', 'abcx', SUCCEED, 'found+"-"+g1', 'x-'),\r
- ('(a)+b|aac', 'aac', SUCCEED, 'found+"-"+g1', 'aac-None'),\r
-\r
- # Test symbolic groups\r
-\r
- ('(?P<i d>aaa)a', 'aaaa', SYNTAX_ERROR),\r
- ('(?P<id>aaa)a', 'aaaa', SUCCEED, 'found+"-"+id', 'aaaa-aaa'),\r
- ('(?P<id>aa)(?P=id)', 'aaaa', SUCCEED, 'found+"-"+id', 'aaaa-aa'),\r
- ('(?P<id>aa)(?P=xd)', 'aaaa', SYNTAX_ERROR),\r
-\r
- # Test octal escapes/memory references\r
-\r
- ('\\1', 'a', SYNTAX_ERROR),\r
- ('\\09', chr(0) + '9', SUCCEED, 'found', chr(0) + '9'),\r
- ('\\141', 'a', SUCCEED, 'found', 'a'),\r
- ('(a)(b)(c)(d)(e)(f)(g)(h)(i)(j)(k)(l)\\119', 'abcdefghijklk9', SUCCEED, 'found+"-"+g11', 'abcdefghijklk9-k'),\r
-\r
- # All tests from Perl\r
-\r
- ('abc', 'abc', SUCCEED, 'found', 'abc'),\r
- ('abc', 'xbc', FAIL),\r
- ('abc', 'axc', FAIL),\r
- ('abc', 'abx', FAIL),\r
- ('abc', 'xabcy', SUCCEED, 'found', 'abc'),\r
- ('abc', 'ababc', SUCCEED, 'found', 'abc'),\r
- ('ab*c', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab*bc', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab*bc', 'abbc', SUCCEED, 'found', 'abbc'),\r
- ('ab*bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab{0,}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab+bc', 'abbc', SUCCEED, 'found', 'abbc'),\r
- ('ab+bc', 'abc', FAIL),\r
- ('ab+bc', 'abq', FAIL),\r
- ('ab{1,}bc', 'abq', FAIL),\r
- ('ab+bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab{1,}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab{1,3}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab{3,4}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'),\r
- ('ab{4,5}bc', 'abbbbc', FAIL),\r
- ('ab?bc', 'abbc', SUCCEED, 'found', 'abbc'),\r
- ('ab?bc', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab{0,1}bc', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab?bc', 'abbbbc', FAIL),\r
- ('ab?c', 'abc', SUCCEED, 'found', 'abc'),\r
- ('ab{0,1}c', 'abc', SUCCEED, 'found', 'abc'),\r
- ('^abc$', 'abc', SUCCEED, 'found', 'abc'),\r
- ('^abc$', 'abcc', FAIL),\r
- ('^abc', 'abcc', SUCCEED, 'found', 'abc'),\r
- ('^abc$', 'aabc', FAIL),\r
- ('abc$', 'aabc', SUCCEED, 'found', 'abc'),\r
- ('^', 'abc', SUCCEED, 'found', ''),\r
- ('$', 'abc', SUCCEED, 'found', ''),\r
- ('a.c', 'abc', SUCCEED, 'found', 'abc'),\r
- ('a.c', 'axc', SUCCEED, 'found', 'axc'),\r
- ('a.*c', 'axyzc', SUCCEED, 'found', 'axyzc'),\r
- ('a.*c', 'axyzd', FAIL),\r
- ('a[bc]d', 'abc', FAIL),\r
- ('a[bc]d', 'abd', SUCCEED, 'found', 'abd'),\r
- ('a[b-d]e', 'abd', FAIL),\r
- ('a[b-d]e', 'ace', SUCCEED, 'found', 'ace'),\r
- ('a[b-d]', 'aac', SUCCEED, 'found', 'ac'),\r
- ('a[-b]', 'a-', SUCCEED, 'found', 'a-'),\r
- ('a[b-]', 'a-', SUCCEED, 'found', 'a-'),\r
- ('a[b-a]', '-', SYNTAX_ERROR),\r
- ('a[]b', '-', SYNTAX_ERROR),\r
- ('a[', '-', SYNTAX_ERROR),\r
- ('a]', 'a]', SUCCEED, 'found', 'a]'),\r
- ('a[]]b', 'a]b', SUCCEED, 'found', 'a]b'),\r
- ('a[^bc]d', 'aed', SUCCEED, 'found', 'aed'),\r
- ('a[^bc]d', 'abd', FAIL),\r
- ('a[^-b]c', 'adc', SUCCEED, 'found', 'adc'),\r
- ('a[^-b]c', 'a-c', FAIL),\r
- ('a[^]b]c', 'a]c', FAIL),\r
- ('a[^]b]c', 'adc', SUCCEED, 'found', 'adc'),\r
- ('ab|cd', 'abc', SUCCEED, 'found', 'ab'),\r
- ('ab|cd', 'abcd', SUCCEED, 'found', 'ab'),\r
- ('()ef', 'def', SUCCEED, 'found+"-"+g1', 'ef-'),\r
- ('*a', '-', SYNTAX_ERROR),\r
- ('(*)b', '-', SYNTAX_ERROR),\r
- ('$b', 'b', FAIL),\r
- ('a\\', '-', SYNTAX_ERROR),\r
- ('a\\(b', 'a(b', SUCCEED, 'found+"-"+g1', 'a(b-Error'),\r
- ('a\\(*b', 'ab', SUCCEED, 'found', 'ab'),\r
- ('a\\(*b', 'a((b', SUCCEED, 'found', 'a((b'),\r
- ('a\\\\b', 'a\\b', SUCCEED, 'found', 'a\\b'),\r
- ('abc)', '-', SYNTAX_ERROR),\r
- ('(abc', '-', SYNTAX_ERROR),\r
- ('((a))', 'abc', SUCCEED, 'found+"-"+g1+"-"+g2', 'a-a-a'),\r
- ('(a)b(c)', 'abc', SUCCEED, 'found+"-"+g1+"-"+g2', 'abc-a-c'),\r
- ('a+b+c', 'aabbabc', SUCCEED, 'found', 'abc'),\r
- ('a{1,}b{1,}c', 'aabbabc', SUCCEED, 'found', 'abc'),\r
- ('a**', '-', SYNTAX_ERROR),\r
- ('a.+?c', 'abcabc', SUCCEED, 'found', 'abc'),\r
- ('(a+|b)*', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'),\r
- ('(a+|b){0,}', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'),\r
- ('(a+|b)+', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'),\r
- ('(a+|b){1,}', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'),\r
- ('(a+|b)?', 'ab', SUCCEED, 'found+"-"+g1', 'a-a'),\r
- ('(a+|b){0,1}', 'ab', SUCCEED, 'found+"-"+g1', 'a-a'),\r
- (')(', '-', SYNTAX_ERROR),\r
- ('[^ab]*', 'cde', SUCCEED, 'found', 'cde'),\r
- ('abc', '', FAIL),\r
- ('a*', '', SUCCEED, 'found', ''),\r
- ('([abc])*d', 'abbbcd', SUCCEED, 'found+"-"+g1', 'abbbcd-c'),\r
- ('([abc])*bcd', 'abcd', SUCCEED, 'found+"-"+g1', 'abcd-a'),\r
- ('a|b|c|d|e', 'e', SUCCEED, 'found', 'e'),\r
- ('(a|b|c|d|e)f', 'ef', SUCCEED, 'found+"-"+g1', 'ef-e'),\r
- ('abcd*efg', 'abcdefg', SUCCEED, 'found', 'abcdefg'),\r
- ('ab*', 'xabyabbbz', SUCCEED, 'found', 'ab'),\r
- ('ab*', 'xayabbbz', SUCCEED, 'found', 'a'),\r
- ('(ab|cd)e', 'abcde', SUCCEED, 'found+"-"+g1', 'cde-cd'),\r
- ('[abhgefdc]ij', 'hij', SUCCEED, 'found', 'hij'),\r
- ('^(ab|cd)e', 'abcde', FAIL),\r
- ('(abc|)ef', 'abcdef', SUCCEED, 'found+"-"+g1', 'ef-'),\r
- ('(a|b)c*d', 'abcd', SUCCEED, 'found+"-"+g1', 'bcd-b'),\r
- ('(ab|ab*)bc', 'abc', SUCCEED, 'found+"-"+g1', 'abc-a'),\r
- ('a([bc]*)c*', 'abc', SUCCEED, 'found+"-"+g1', 'abc-bc'),\r
- ('a([bc]*)(c*d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-bc-d'),\r
- ('a([bc]+)(c*d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-bc-d'),\r
- ('a([bc]*)(c+d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-b-cd'),\r
- ('a[bcd]*dcdcde', 'adcdcde', SUCCEED, 'found', 'adcdcde'),\r
- ('a[bcd]+dcdcde', 'adcdcde', FAIL),\r
- ('(ab|a)b*c', 'abc', SUCCEED, 'found+"-"+g1', 'abc-ab'),\r
- ('((a)(b)c)(d)', 'abcd', SUCCEED, 'g1+"-"+g2+"-"+g3+"-"+g4', 'abc-a-b-d'),\r
- ('[a-zA-Z_][a-zA-Z0-9_]*', 'alpha', SUCCEED, 'found', 'alpha'),\r
- ('^a(bc+|b[eh])g|.h$', 'abh', SUCCEED, 'found+"-"+g1', 'bh-None'),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'effgz', SUCCEED, 'found+"-"+g1+"-"+g2', 'effgz-effgz-None'),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'ij', SUCCEED, 'found+"-"+g1+"-"+g2', 'ij-ij-j'),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'effg', FAIL),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'bcdd', FAIL),\r
- ('(bc+d$|ef*g.|h?i(j|k))', 'reffgz', SUCCEED, 'found+"-"+g1+"-"+g2', 'effgz-effgz-None'),\r
- ('((((((((((a))))))))))', 'a', SUCCEED, 'g10', 'a'),\r
- ('((((((((((a))))))))))\\10', 'aa', SUCCEED, 'found', 'aa'),\r
-# Python does not have the same rules for \\41 so this is a syntax error\r
-# ('((((((((((a))))))))))\\41', 'aa', FAIL),\r
-# ('((((((((((a))))))))))\\41', 'a!', SUCCEED, 'found', 'a!'),\r
- ('((((((((((a))))))))))\\41', '', SYNTAX_ERROR),\r
- ('(?i)((((((((((a))))))))))\\41', '', SYNTAX_ERROR),\r
- ('(((((((((a)))))))))', 'a', SUCCEED, 'found', 'a'),\r
- ('multiple words of text', 'uh-uh', FAIL),\r
- ('multiple words', 'multiple words, yeah', SUCCEED, 'found', 'multiple words'),\r
- ('(.*)c(.*)', 'abcde', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcde-ab-de'),\r
- ('\\((.*), (.*)\\)', '(a, b)', SUCCEED, 'g2+"-"+g1', 'b-a'),\r
- ('[k]', 'ab', FAIL),\r
- ('a[-]?c', 'ac', SUCCEED, 'found', 'ac'),\r
- ('(abc)\\1', 'abcabc', SUCCEED, 'g1', 'abc'),\r
- ('([a-c]*)\\1', 'abcabc', SUCCEED, 'g1', 'abc'),\r
- ('(?i)abc', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)abc', 'XBC', FAIL),\r
- ('(?i)abc', 'AXC', FAIL),\r
- ('(?i)abc', 'ABX', FAIL),\r
- ('(?i)abc', 'XABCY', SUCCEED, 'found', 'ABC'),\r
- ('(?i)abc', 'ABABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)ab*c', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)ab*bc', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)ab*bc', 'ABBC', SUCCEED, 'found', 'ABBC'),\r
- ('(?i)ab*?bc', 'ABBBBC', SUCCEED, 'found', 'ABBBBC'),\r
- ('(?i)ab{0,}?bc', 'ABBBBC', SUCCEED, 'found', 'ABBBBC'),\r
- ('(?i)ab+?bc', 'ABBC', SUCCEED, 'found', 'ABBC'),\r
- ('(?i)ab+bc', 'ABC', FAIL),\r
- ('(?i)ab+bc', 'ABQ', FAIL),\r
- ('(?i)ab{1,}bc', 'ABQ', FAIL),\r
- ('(?i)ab+bc', 'ABBBBC', SUCCEED, 'found', 'ABBBBC'),\r
- ('(?i)ab{1,}?bc', 'ABBBBC', SUCCEED, 'found', 'ABBBBC'),\r
- ('(?i)ab{1,3}?bc', 'ABBBBC', SUCCEED, 'found', 'ABBBBC'),\r
- ('(?i)ab{3,4}?bc', 'ABBBBC', SUCCEED, 'found', 'ABBBBC'),\r
- ('(?i)ab{4,5}?bc', 'ABBBBC', FAIL),\r
- ('(?i)ab??bc', 'ABBC', SUCCEED, 'found', 'ABBC'),\r
- ('(?i)ab??bc', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)ab{0,1}?bc', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)ab??bc', 'ABBBBC', FAIL),\r
- ('(?i)ab??c', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)ab{0,1}?c', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)^abc$', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)^abc$', 'ABCC', FAIL),\r
- ('(?i)^abc', 'ABCC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)^abc$', 'AABC', FAIL),\r
- ('(?i)abc$', 'AABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)^', 'ABC', SUCCEED, 'found', ''),\r
- ('(?i)$', 'ABC', SUCCEED, 'found', ''),\r
- ('(?i)a.c', 'ABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)a.c', 'AXC', SUCCEED, 'found', 'AXC'),\r
- ('(?i)a.*?c', 'AXYZC', SUCCEED, 'found', 'AXYZC'),\r
- ('(?i)a.*c', 'AXYZD', FAIL),\r
- ('(?i)a[bc]d', 'ABC', FAIL),\r
- ('(?i)a[bc]d', 'ABD', SUCCEED, 'found', 'ABD'),\r
- ('(?i)a[b-d]e', 'ABD', FAIL),\r
- ('(?i)a[b-d]e', 'ACE', SUCCEED, 'found', 'ACE'),\r
- ('(?i)a[b-d]', 'AAC', SUCCEED, 'found', 'AC'),\r
- ('(?i)a[-b]', 'A-', SUCCEED, 'found', 'A-'),\r
- ('(?i)a[b-]', 'A-', SUCCEED, 'found', 'A-'),\r
- ('(?i)a[b-a]', '-', SYNTAX_ERROR),\r
- ('(?i)a[]b', '-', SYNTAX_ERROR),\r
- ('(?i)a[', '-', SYNTAX_ERROR),\r
- ('(?i)a]', 'A]', SUCCEED, 'found', 'A]'),\r
- ('(?i)a[]]b', 'A]B', SUCCEED, 'found', 'A]B'),\r
- ('(?i)a[^bc]d', 'AED', SUCCEED, 'found', 'AED'),\r
- ('(?i)a[^bc]d', 'ABD', FAIL),\r
- ('(?i)a[^-b]c', 'ADC', SUCCEED, 'found', 'ADC'),\r
- ('(?i)a[^-b]c', 'A-C', FAIL),\r
- ('(?i)a[^]b]c', 'A]C', FAIL),\r
- ('(?i)a[^]b]c', 'ADC', SUCCEED, 'found', 'ADC'),\r
- ('(?i)ab|cd', 'ABC', SUCCEED, 'found', 'AB'),\r
- ('(?i)ab|cd', 'ABCD', SUCCEED, 'found', 'AB'),\r
- ('(?i)()ef', 'DEF', SUCCEED, 'found+"-"+g1', 'EF-'),\r
- ('(?i)*a', '-', SYNTAX_ERROR),\r
- ('(?i)(*)b', '-', SYNTAX_ERROR),\r
- ('(?i)$b', 'B', FAIL),\r
- ('(?i)a\\', '-', SYNTAX_ERROR),\r
- ('(?i)a\\(b', 'A(B', SUCCEED, 'found+"-"+g1', 'A(B-Error'),\r
- ('(?i)a\\(*b', 'AB', SUCCEED, 'found', 'AB'),\r
- ('(?i)a\\(*b', 'A((B', SUCCEED, 'found', 'A((B'),\r
- ('(?i)a\\\\b', 'A\\B', SUCCEED, 'found', 'A\\B'),\r
- ('(?i)abc)', '-', SYNTAX_ERROR),\r
- ('(?i)(abc', '-', SYNTAX_ERROR),\r
- ('(?i)((a))', 'ABC', SUCCEED, 'found+"-"+g1+"-"+g2', 'A-A-A'),\r
- ('(?i)(a)b(c)', 'ABC', SUCCEED, 'found+"-"+g1+"-"+g2', 'ABC-A-C'),\r
- ('(?i)a+b+c', 'AABBABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)a{1,}b{1,}c', 'AABBABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)a**', '-', SYNTAX_ERROR),\r
- ('(?i)a.+?c', 'ABCABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)a.*?c', 'ABCABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)a.{0,5}?c', 'ABCABC', SUCCEED, 'found', 'ABC'),\r
- ('(?i)(a+|b)*', 'AB', SUCCEED, 'found+"-"+g1', 'AB-B'),\r
- ('(?i)(a+|b){0,}', 'AB', SUCCEED, 'found+"-"+g1', 'AB-B'),\r
- ('(?i)(a+|b)+', 'AB', SUCCEED, 'found+"-"+g1', 'AB-B'),\r
- ('(?i)(a+|b){1,}', 'AB', SUCCEED, 'found+"-"+g1', 'AB-B'),\r
- ('(?i)(a+|b)?', 'AB', SUCCEED, 'found+"-"+g1', 'A-A'),\r
- ('(?i)(a+|b){0,1}', 'AB', SUCCEED, 'found+"-"+g1', 'A-A'),\r
- ('(?i)(a+|b){0,1}?', 'AB', SUCCEED, 'found+"-"+g1', '-None'),\r
- ('(?i))(', '-', SYNTAX_ERROR),\r
- ('(?i)[^ab]*', 'CDE', SUCCEED, 'found', 'CDE'),\r
- ('(?i)abc', '', FAIL),\r
- ('(?i)a*', '', SUCCEED, 'found', ''),\r
- ('(?i)([abc])*d', 'ABBBCD', SUCCEED, 'found+"-"+g1', 'ABBBCD-C'),\r
- ('(?i)([abc])*bcd', 'ABCD', SUCCEED, 'found+"-"+g1', 'ABCD-A'),\r
- ('(?i)a|b|c|d|e', 'E', SUCCEED, 'found', 'E'),\r
- ('(?i)(a|b|c|d|e)f', 'EF', SUCCEED, 'found+"-"+g1', 'EF-E'),\r
- ('(?i)abcd*efg', 'ABCDEFG', SUCCEED, 'found', 'ABCDEFG'),\r
- ('(?i)ab*', 'XABYABBBZ', SUCCEED, 'found', 'AB'),\r
- ('(?i)ab*', 'XAYABBBZ', SUCCEED, 'found', 'A'),\r
- ('(?i)(ab|cd)e', 'ABCDE', SUCCEED, 'found+"-"+g1', 'CDE-CD'),\r
- ('(?i)[abhgefdc]ij', 'HIJ', SUCCEED, 'found', 'HIJ'),\r
- ('(?i)^(ab|cd)e', 'ABCDE', FAIL),\r
- ('(?i)(abc|)ef', 'ABCDEF', SUCCEED, 'found+"-"+g1', 'EF-'),\r
- ('(?i)(a|b)c*d', 'ABCD', SUCCEED, 'found+"-"+g1', 'BCD-B'),\r
- ('(?i)(ab|ab*)bc', 'ABC', SUCCEED, 'found+"-"+g1', 'ABC-A'),\r
- ('(?i)a([bc]*)c*', 'ABC', SUCCEED, 'found+"-"+g1', 'ABC-BC'),\r
- ('(?i)a([bc]*)(c*d)', 'ABCD', SUCCEED, 'found+"-"+g1+"-"+g2', 'ABCD-BC-D'),\r
- ('(?i)a([bc]+)(c*d)', 'ABCD', SUCCEED, 'found+"-"+g1+"-"+g2', 'ABCD-BC-D'),\r
- ('(?i)a([bc]*)(c+d)', 'ABCD', SUCCEED, 'found+"-"+g1+"-"+g2', 'ABCD-B-CD'),\r
- ('(?i)a[bcd]*dcdcde', 'ADCDCDE', SUCCEED, 'found', 'ADCDCDE'),\r
- ('(?i)a[bcd]+dcdcde', 'ADCDCDE', FAIL),\r
- ('(?i)(ab|a)b*c', 'ABC', SUCCEED, 'found+"-"+g1', 'ABC-AB'),\r
- ('(?i)((a)(b)c)(d)', 'ABCD', SUCCEED, 'g1+"-"+g2+"-"+g3+"-"+g4', 'ABC-A-B-D'),\r
- ('(?i)[a-zA-Z_][a-zA-Z0-9_]*', 'ALPHA', SUCCEED, 'found', 'ALPHA'),\r
- ('(?i)^a(bc+|b[eh])g|.h$', 'ABH', SUCCEED, 'found+"-"+g1', 'BH-None'),\r
- ('(?i)(bc+d$|ef*g.|h?i(j|k))', 'EFFGZ', SUCCEED, 'found+"-"+g1+"-"+g2', 'EFFGZ-EFFGZ-None'),\r
- ('(?i)(bc+d$|ef*g.|h?i(j|k))', 'IJ', SUCCEED, 'found+"-"+g1+"-"+g2', 'IJ-IJ-J'),\r
- ('(?i)(bc+d$|ef*g.|h?i(j|k))', 'EFFG', FAIL),\r
- ('(?i)(bc+d$|ef*g.|h?i(j|k))', 'BCDD', FAIL),\r
- ('(?i)(bc+d$|ef*g.|h?i(j|k))', 'REFFGZ', SUCCEED, 'found+"-"+g1+"-"+g2', 'EFFGZ-EFFGZ-None'),\r
- ('(?i)((((((((((a))))))))))', 'A', SUCCEED, 'g10', 'A'),\r
- ('(?i)((((((((((a))))))))))\\10', 'AA', SUCCEED, 'found', 'AA'),\r
- #('(?i)((((((((((a))))))))))\\41', 'AA', FAIL),\r
- #('(?i)((((((((((a))))))))))\\41', 'A!', SUCCEED, 'found', 'A!'),\r
- ('(?i)(((((((((a)))))))))', 'A', SUCCEED, 'found', 'A'),\r
- ('(?i)(?:(?:(?:(?:(?:(?:(?:(?:(?:(a))))))))))', 'A', SUCCEED, 'g1', 'A'),\r
- ('(?i)(?:(?:(?:(?:(?:(?:(?:(?:(?:(a|b|c))))))))))', 'C', SUCCEED, 'g1', 'C'),\r
- ('(?i)multiple words of text', 'UH-UH', FAIL),\r
- ('(?i)multiple words', 'MULTIPLE WORDS, YEAH', SUCCEED, 'found', 'MULTIPLE WORDS'),\r
- ('(?i)(.*)c(.*)', 'ABCDE', SUCCEED, 'found+"-"+g1+"-"+g2', 'ABCDE-AB-DE'),\r
- ('(?i)\\((.*), (.*)\\)', '(A, B)', SUCCEED, 'g2+"-"+g1', 'B-A'),\r
- ('(?i)[k]', 'AB', FAIL),\r
-# ('(?i)abcd', 'ABCD', SUCCEED, 'found+"-"+\\found+"-"+\\\\found', 'ABCD-$&-\\ABCD'),\r
-# ('(?i)a(bc)d', 'ABCD', SUCCEED, 'g1+"-"+\\g1+"-"+\\\\g1', 'BC-$1-\\BC'),\r
- ('(?i)a[-]?c', 'AC', SUCCEED, 'found', 'AC'),\r
- ('(?i)(abc)\\1', 'ABCABC', SUCCEED, 'g1', 'ABC'),\r
- ('(?i)([a-c]*)\\1', 'ABCABC', SUCCEED, 'g1', 'ABC'),\r
- ('a(?!b).', 'abad', SUCCEED, 'found', 'ad'),\r
- ('a(?=d).', 'abad', SUCCEED, 'found', 'ad'),\r
- ('a(?=c|d).', 'abad', SUCCEED, 'found', 'ad'),\r
- ('a(?:b|c|d)(.)', 'ace', SUCCEED, 'g1', 'e'),\r
- ('a(?:b|c|d)*(.)', 'ace', SUCCEED, 'g1', 'e'),\r
- ('a(?:b|c|d)+?(.)', 'ace', SUCCEED, 'g1', 'e'),\r
- ('a(?:b|(c|e){1,2}?|d)+?(.)', 'ace', SUCCEED, 'g1 + g2', 'ce'),\r
- ('^(.+)?B', 'AB', SUCCEED, 'g1', 'A'),\r
-\r
- # lookbehind: split by : but not if it is escaped by -.\r
- ('(?<!-):(.*?)(?<!-):', 'a:bc-:de:f', SUCCEED, 'g1', 'bc-:de' ),\r
- # escaping with \ as we know it\r
- ('(?<!\\\):(.*?)(?<!\\\):', 'a:bc\\:de:f', SUCCEED, 'g1', 'bc\\:de' ),\r
- # terminating with ' and escaping with ? as in edifact\r
- ("(?<!\\?)'(.*?)(?<!\\?)'", "a'bc?'de'f", SUCCEED, 'g1', "bc?'de" ),\r
-\r
- # Comments using the (?#...) syntax\r
-\r
- ('w(?# comment', 'w', SYNTAX_ERROR),\r
- ('w(?# comment 1)xy(?# comment 2)z', 'wxyz', SUCCEED, 'found', 'wxyz'),\r
-\r
- # Check odd placement of embedded pattern modifiers\r
-\r
- # not an error under PCRE/PRE:\r
- ('w(?i)', 'W', SUCCEED, 'found', 'W'),\r
- # ('w(?i)', 'W', SYNTAX_ERROR),\r
-\r
- # Comments using the x embedded pattern modifier\r
-\r
- ("""(?x)w# comment 1\r
- x y\r
- # comment 2\r
- z""", 'wxyz', SUCCEED, 'found', 'wxyz'),\r
-\r
- # using the m embedded pattern modifier\r
-\r
- ('^abc', """jkl\r
-abc\r
-xyz""", FAIL),\r
- ('(?m)^abc', """jkl\r
-abc\r
-xyz""", SUCCEED, 'found', 'abc'),\r
-\r
- ('(?m)abc$', """jkl\r
-xyzabc\r
-123""", SUCCEED, 'found', 'abc'),\r
-\r
- # using the s embedded pattern modifier\r
-\r
- ('a.b', 'a\nb', FAIL),\r
- ('(?s)a.b', 'a\nb', SUCCEED, 'found', 'a\nb'),\r
-\r
- # test \w, etc. both inside and outside character classes\r
-\r
- ('\\w+', '--ab_cd0123--', SUCCEED, 'found', 'ab_cd0123'),\r
- ('[\\w]+', '--ab_cd0123--', SUCCEED, 'found', 'ab_cd0123'),\r
- ('\\D+', '1234abc5678', SUCCEED, 'found', 'abc'),\r
- ('[\\D]+', '1234abc5678', SUCCEED, 'found', 'abc'),\r
- ('[\\da-fA-F]+', '123abc', SUCCEED, 'found', '123abc'),\r
- # not an error under PCRE/PRE:\r
- # ('[\\d-x]', '-', SYNTAX_ERROR),\r
- (r'([\s]*)([\S]*)([\s]*)', ' testing!1972', SUCCEED, 'g3+g2+g1', 'testing!1972 '),\r
- (r'(\s*)(\S*)(\s*)', ' testing!1972', SUCCEED, 'g3+g2+g1', 'testing!1972 '),\r
-\r
- (r'\xff', '\377', SUCCEED, 'found', chr(255)),\r
- # new \x semantics\r
- (r'\x00ff', '\377', FAIL),\r
- # (r'\x00ff', '\377', SUCCEED, 'found', chr(255)),\r
- (r'\t\n\v\r\f\a\g', '\t\n\v\r\f\ag', SUCCEED, 'found', '\t\n\v\r\f\ag'),\r
- ('\t\n\v\r\f\a\g', '\t\n\v\r\f\ag', SUCCEED, 'found', '\t\n\v\r\f\ag'),\r
- (r'\t\n\v\r\f\a', '\t\n\v\r\f\a', SUCCEED, 'found', chr(9)+chr(10)+chr(11)+chr(13)+chr(12)+chr(7)),\r
- (r'[\t][\n][\v][\r][\f][\b]', '\t\n\v\r\f\b', SUCCEED, 'found', '\t\n\v\r\f\b'),\r
-\r
- #\r
- # post-1.5.2 additions\r
-\r
- # xmllib problem\r
- (r'(([a-z]+):)?([a-z]+)$', 'smil', SUCCEED, 'g1+"-"+g2+"-"+g3', 'None-None-smil'),\r
- # bug 110866: reference to undefined group\r
- (r'((.)\1+)', '', SYNTAX_ERROR),\r
- # bug 111869: search (PRE/PCRE fails on this one, SRE doesn't)\r
- (r'.*d', 'abc\nabd', SUCCEED, 'found', 'abd'),\r
- # bug 112468: various expected syntax errors\r
- (r'(', '', SYNTAX_ERROR),\r
- (r'[\41]', '!', SUCCEED, 'found', '!'),\r
- # bug 114033: nothing to repeat\r
- (r'(x?)?', 'x', SUCCEED, 'found', 'x'),\r
- # bug 115040: rescan if flags are modified inside pattern\r
- (r' (?x)foo ', 'foo', SUCCEED, 'found', 'foo'),\r
- # bug 115618: negative lookahead\r
- (r'(?<!abc)(d.f)', 'abcdefdof', SUCCEED, 'found', 'dof'),\r
- # bug 116251: character class bug\r
- (r'[\w-]+', 'laser_beam', SUCCEED, 'found', 'laser_beam'),\r
- # bug 123769+127259: non-greedy backtracking bug\r
- (r'.*?\S *:', 'xx:', SUCCEED, 'found', 'xx:'),\r
- (r'a[ ]*?\ (\d+).*', 'a 10', SUCCEED, 'found', 'a 10'),\r
- (r'a[ ]*?\ (\d+).*', 'a 10', SUCCEED, 'found', 'a 10'),\r
- # bug 127259: \Z shouldn't depend on multiline mode\r
- (r'(?ms).*?x\s*\Z(.*)','xx\nx\n', SUCCEED, 'g1', ''),\r
- # bug 128899: uppercase literals under the ignorecase flag\r
- (r'(?i)M+', 'MMM', SUCCEED, 'found', 'MMM'),\r
- (r'(?i)m+', 'MMM', SUCCEED, 'found', 'MMM'),\r
- (r'(?i)[M]+', 'MMM', SUCCEED, 'found', 'MMM'),\r
- (r'(?i)[m]+', 'MMM', SUCCEED, 'found', 'MMM'),\r
- # bug 130748: ^* should be an error (nothing to repeat)\r
- (r'^*', '', SYNTAX_ERROR),\r
- # bug 133283: minimizing repeat problem\r
- (r'"(?:\\"|[^"])*?"', r'"\""', SUCCEED, 'found', r'"\""'),\r
- # bug 477728: minimizing repeat problem\r
- (r'^.*?$', 'one\ntwo\nthree\n', FAIL),\r
- # bug 483789: minimizing repeat problem\r
- (r'a[^>]*?b', 'a>b', FAIL),\r
- # bug 490573: minimizing repeat problem\r
- (r'^a*?$', 'foo', FAIL),\r
- # bug 470582: nested groups problem\r
- (r'^((a)c)?(ab)$', 'ab', SUCCEED, 'g1+"-"+g2+"-"+g3', 'None-None-ab'),\r
- # another minimizing repeat problem (capturing groups in assertions)\r
- ('^([ab]*?)(?=(b)?)c', 'abc', SUCCEED, 'g1+"-"+g2', 'ab-None'),\r
- ('^([ab]*?)(?!(b))c', 'abc', SUCCEED, 'g1+"-"+g2', 'ab-None'),\r
- ('^([ab]*?)(?<!(a))c', 'abc', SUCCEED, 'g1+"-"+g2', 'ab-None'),\r
-]\r
-\r
-try:\r
- u = eval("u'\N{LATIN CAPITAL LETTER A WITH DIAERESIS}'")\r
-except SyntaxError:\r
- pass\r
-else:\r
- tests.extend([\r
- # bug 410271: \b broken under locales\r
- (r'\b.\b', 'a', SUCCEED, 'found', 'a'),\r
- (r'(?u)\b.\b', u, SUCCEED, 'found', u),\r
- (r'(?u)\w', u, SUCCEED, 'found', u),\r
- ])\r
projects / mirror_edk2.git / blobdiff