diff options
Diffstat (limited to 'Lib/test/re_tests.py')
| -rwxr-xr-x | Lib/test/re_tests.py | 140 |
1 files changed, 125 insertions, 15 deletions
diff --git a/Lib/test/re_tests.py b/Lib/test/re_tests.py index 85b0267..0cf9241 100755 --- a/Lib/test/re_tests.py +++ b/Lib/test/re_tests.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python3 +#!/usr/bin/env python # -*- mode: python -*- # Re test suite and benchmark suite v1.5 @@ -86,8 +86,9 @@ tests = [ (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'[\a][\b][\f][\n][\r][\t][\v]', '\a\b\f\n\r\t\v', SUCCEED, 'found', '\a\b\f\n\r\t\v'), # NOTE: not an error under PCRE/PRE: - (r'\u', '', SYNTAX_ERROR), # A Perl escape - # (r'\c\e\g\h\i\j\k\m\o\p\q\y\z', 'ceghijkmopqyz', SUCCEED, 'found', 'ceghijkmopqyz'), + # (r'\u', '', SYNTAX_ERROR), # A Perl escape + (r'\c\e\g\h\i\j\k\m\o\p\q\y\z', 'ceghijkmopqyz', SUCCEED, 'found', 'ceghijkmopqyz'), + (r'\xff', '\377', SUCCEED, 'found', chr(255)), # new \x semantics (r'\x00ffffffffffffff', '\377', FAIL, 'found', chr(255)), (r'\x00f', '\017', FAIL, 'found', chr(15)), @@ -105,10 +106,10 @@ tests = [ ('a.*b', 'acc\nccb', FAIL), ('a.{4,5}b', 'acc\nccb', FAIL), ('a.b', 'a\rb', SUCCEED, 'found', 'a\rb'), - ('(?s)a.b', 'a\nb', SUCCEED, 'found', 'a\nb'), - ('(?s)a.*b', 'acc\nccb', SUCCEED, 'found', 'acc\nccb'), + ('a.b(?s)', 'a\nb', SUCCEED, 'found', 'a\nb'), + ('a.*(?s)b', 'acc\nccb', SUCCEED, 'found', 'acc\nccb'), ('(?s)a.{4,5}b', 'acc\nccb', SUCCEED, 'found', 'acc\nccb'), - ('(?s)a.b', 'a\rb', SUCCEED, 'found', 'a\rb'), + ('(?s)a.b', 'a\nb', SUCCEED, 'found', 'a\nb'), (')', '', SYNTAX_ERROR), # Unmatched right bracket ('', '', SUCCEED, 'found', ''), # Empty pattern @@ -157,7 +158,7 @@ tests = [ ('(abc', '-', SYNTAX_ERROR), ('a]', 'a]', SUCCEED, 'found', 'a]'), ('a[]]b', 'a]b', SUCCEED, 'found', 'a]b'), - ('a[\\]]b', 'a]b', SUCCEED, 'found', 'a]b'), + ('a[\]]b', 'a]b', SUCCEED, 'found', 'a]b'), ('a[^bc]d', 'aed', SUCCEED, 'found', 'aed'), ('a[^bc]d', 'abd', FAIL), ('a[^-b]c', 'adc', SUCCEED, 'found', 'adc'), @@ -272,32 +273,124 @@ tests = [ # Test octal escapes/memory references ('\\1', 'a', SYNTAX_ERROR), + ('\\09', chr(0) + '9', SUCCEED, 'found', chr(0) + '9'), + ('\\141', 'a', SUCCEED, 'found', 'a'), + ('(a)(b)(c)(d)(e)(f)(g)(h)(i)(j)(k)(l)\\119', 'abcdefghijklk9', SUCCEED, 'found+"-"+g11', 'abcdefghijklk9-k'), # All tests from Perl + ('abc', 'abc', SUCCEED, 'found', 'abc'), + ('abc', 'xbc', FAIL), + ('abc', 'axc', FAIL), + ('abc', 'abx', FAIL), + ('abc', 'xabcy', SUCCEED, 'found', 'abc'), + ('abc', 'ababc', SUCCEED, 'found', 'abc'), + ('ab*c', 'abc', SUCCEED, 'found', 'abc'), + ('ab*bc', 'abc', SUCCEED, 'found', 'abc'), + ('ab*bc', 'abbc', SUCCEED, 'found', 'abbc'), + ('ab*bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'), ('ab{0,}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'), + ('ab+bc', 'abbc', SUCCEED, 'found', 'abbc'), + ('ab+bc', 'abc', FAIL), + ('ab+bc', 'abq', FAIL), ('ab{1,}bc', 'abq', FAIL), + ('ab+bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'), ('ab{1,}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'), ('ab{1,3}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'), ('ab{3,4}bc', 'abbbbc', SUCCEED, 'found', 'abbbbc'), ('ab{4,5}bc', 'abbbbc', FAIL), + ('ab?bc', 'abbc', SUCCEED, 'found', 'abbc'), + ('ab?bc', 'abc', SUCCEED, 'found', 'abc'), ('ab{0,1}bc', 'abc', SUCCEED, 'found', 'abc'), + ('ab?bc', 'abbbbc', FAIL), + ('ab?c', 'abc', SUCCEED, 'found', 'abc'), ('ab{0,1}c', 'abc', SUCCEED, 'found', 'abc'), + ('^abc$', 'abc', SUCCEED, 'found', 'abc'), + ('^abc$', 'abcc', FAIL), + ('^abc', 'abcc', SUCCEED, 'found', 'abc'), + ('^abc$', 'aabc', FAIL), + ('abc$', 'aabc', SUCCEED, 'found', 'abc'), ('^', 'abc', SUCCEED, 'found', ''), ('$', 'abc', SUCCEED, 'found', ''), + ('a.c', 'abc', SUCCEED, 'found', 'abc'), + ('a.c', 'axc', SUCCEED, 'found', 'axc'), + ('a.*c', 'axyzc', SUCCEED, 'found', 'axyzc'), + ('a.*c', 'axyzd', FAIL), + ('a[bc]d', 'abc', FAIL), + ('a[bc]d', 'abd', SUCCEED, 'found', 'abd'), + ('a[b-d]e', 'abd', FAIL), + ('a[b-d]e', 'ace', SUCCEED, 'found', 'ace'), + ('a[b-d]', 'aac', SUCCEED, 'found', 'ac'), + ('a[-b]', 'a-', SUCCEED, 'found', 'a-'), ('a[b-]', 'a-', SUCCEED, 'found', 'a-'), ('a[b-a]', '-', SYNTAX_ERROR), + ('a[]b', '-', SYNTAX_ERROR), + ('a[', '-', SYNTAX_ERROR), + ('a]', 'a]', SUCCEED, 'found', 'a]'), + ('a[]]b', 'a]b', SUCCEED, 'found', 'a]b'), + ('a[^bc]d', 'aed', SUCCEED, 'found', 'aed'), + ('a[^bc]d', 'abd', FAIL), + ('a[^-b]c', 'adc', SUCCEED, 'found', 'adc'), + ('a[^-b]c', 'a-c', FAIL), + ('a[^]b]c', 'a]c', FAIL), + ('a[^]b]c', 'adc', SUCCEED, 'found', 'adc'), + ('ab|cd', 'abc', SUCCEED, 'found', 'ab'), + ('ab|cd', 'abcd', SUCCEED, 'found', 'ab'), + ('()ef', 'def', SUCCEED, 'found+"-"+g1', 'ef-'), ('*a', '-', SYNTAX_ERROR), ('(*)b', '-', SYNTAX_ERROR), + ('$b', 'b', FAIL), + ('a\\', '-', SYNTAX_ERROR), + ('a\\(b', 'a(b', SUCCEED, 'found+"-"+g1', 'a(b-Error'), + ('a\\(*b', 'ab', SUCCEED, 'found', 'ab'), + ('a\\(*b', 'a((b', SUCCEED, 'found', 'a((b'), + ('a\\\\b', 'a\\b', SUCCEED, 'found', 'a\\b'), + ('abc)', '-', SYNTAX_ERROR), + ('(abc', '-', SYNTAX_ERROR), + ('((a))', 'abc', SUCCEED, 'found+"-"+g1+"-"+g2', 'a-a-a'), + ('(a)b(c)', 'abc', SUCCEED, 'found+"-"+g1+"-"+g2', 'abc-a-c'), + ('a+b+c', 'aabbabc', SUCCEED, 'found', 'abc'), ('a{1,}b{1,}c', 'aabbabc', SUCCEED, 'found', 'abc'), ('a**', '-', SYNTAX_ERROR), ('a.+?c', 'abcabc', SUCCEED, 'found', 'abc'), + ('(a+|b)*', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'), ('(a+|b){0,}', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'), + ('(a+|b)+', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'), ('(a+|b){1,}', 'ab', SUCCEED, 'found+"-"+g1', 'ab-b'), + ('(a+|b)?', 'ab', SUCCEED, 'found+"-"+g1', 'a-a'), ('(a+|b){0,1}', 'ab', SUCCEED, 'found+"-"+g1', 'a-a'), + (')(', '-', SYNTAX_ERROR), + ('[^ab]*', 'cde', SUCCEED, 'found', 'cde'), + ('abc', '', FAIL), + ('a*', '', SUCCEED, 'found', ''), ('([abc])*d', 'abbbcd', SUCCEED, 'found+"-"+g1', 'abbbcd-c'), ('([abc])*bcd', 'abcd', SUCCEED, 'found+"-"+g1', 'abcd-a'), + ('a|b|c|d|e', 'e', SUCCEED, 'found', 'e'), + ('(a|b|c|d|e)f', 'ef', SUCCEED, 'found+"-"+g1', 'ef-e'), + ('abcd*efg', 'abcdefg', SUCCEED, 'found', 'abcdefg'), + ('ab*', 'xabyabbbz', SUCCEED, 'found', 'ab'), + ('ab*', 'xayabbbz', SUCCEED, 'found', 'a'), + ('(ab|cd)e', 'abcde', SUCCEED, 'found+"-"+g1', 'cde-cd'), + ('[abhgefdc]ij', 'hij', SUCCEED, 'found', 'hij'), ('^(ab|cd)e', 'abcde', FAIL), + ('(abc|)ef', 'abcdef', SUCCEED, 'found+"-"+g1', 'ef-'), + ('(a|b)c*d', 'abcd', SUCCEED, 'found+"-"+g1', 'bcd-b'), + ('(ab|ab*)bc', 'abc', SUCCEED, 'found+"-"+g1', 'abc-a'), + ('a([bc]*)c*', 'abc', SUCCEED, 'found+"-"+g1', 'abc-bc'), + ('a([bc]*)(c*d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-bc-d'), + ('a([bc]+)(c*d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-bc-d'), + ('a([bc]*)(c+d)', 'abcd', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcd-b-cd'), + ('a[bcd]*dcdcde', 'adcdcde', SUCCEED, 'found', 'adcdcde'), + ('a[bcd]+dcdcde', 'adcdcde', FAIL), + ('(ab|a)b*c', 'abc', SUCCEED, 'found+"-"+g1', 'abc-ab'), + ('((a)(b)c)(d)', 'abcd', SUCCEED, 'g1+"-"+g2+"-"+g3+"-"+g4', 'abc-a-b-d'), + ('[a-zA-Z_][a-zA-Z0-9_]*', 'alpha', SUCCEED, 'found', 'alpha'), + ('^a(bc+|b[eh])g|.h$', 'abh', SUCCEED, 'found+"-"+g1', 'bh-None'), + ('(bc+d$|ef*g.|h?i(j|k))', 'effgz', SUCCEED, 'found+"-"+g1+"-"+g2', 'effgz-effgz-None'), + ('(bc+d$|ef*g.|h?i(j|k))', 'ij', SUCCEED, 'found+"-"+g1+"-"+g2', 'ij-ij-j'), + ('(bc+d$|ef*g.|h?i(j|k))', 'effg', FAIL), + ('(bc+d$|ef*g.|h?i(j|k))', 'bcdd', FAIL), + ('(bc+d$|ef*g.|h?i(j|k))', 'reffgz', SUCCEED, 'found+"-"+g1+"-"+g2', 'effgz-effgz-None'), ('((((((((((a))))))))))', 'a', SUCCEED, 'g10', 'a'), ('((((((((((a))))))))))\\10', 'aa', SUCCEED, 'found', 'aa'), # Python does not have the same rules for \\41 so this is a syntax error @@ -305,6 +398,15 @@ tests = [ # ('((((((((((a))))))))))\\41', 'a!', SUCCEED, 'found', 'a!'), ('((((((((((a))))))))))\\41', '', SYNTAX_ERROR), ('(?i)((((((((((a))))))))))\\41', '', SYNTAX_ERROR), + ('(((((((((a)))))))))', 'a', SUCCEED, 'found', 'a'), + ('multiple words of text', 'uh-uh', FAIL), + ('multiple words', 'multiple words, yeah', SUCCEED, 'found', 'multiple words'), + ('(.*)c(.*)', 'abcde', SUCCEED, 'found+"-"+g1+"-"+g2', 'abcde-ab-de'), + ('\\((.*), (.*)\\)', '(a, b)', SUCCEED, 'g2+"-"+g1', 'b-a'), + ('[k]', 'ab', FAIL), + ('a[-]?c', 'ac', SUCCEED, 'found', 'ac'), + ('(abc)\\1', 'abcabc', SUCCEED, 'g1', 'abc'), + ('([a-c]*)\\1', 'abcabc', SUCCEED, 'g1', 'abc'), ('(?i)abc', 'ABC', SUCCEED, 'found', 'ABC'), ('(?i)abc', 'XBC', FAIL), ('(?i)abc', 'AXC', FAIL), @@ -444,11 +546,12 @@ tests = [ ('a(?:b|c|d)*(.)', 'ace', SUCCEED, 'g1', 'e'), ('a(?:b|c|d)+?(.)', 'ace', SUCCEED, 'g1', 'e'), ('a(?:b|(c|e){1,2}?|d)+?(.)', 'ace', SUCCEED, 'g1 + g2', 'ce'), + ('^(.+)?B', 'AB', SUCCEED, 'g1', 'A'), # lookbehind: split by : but not if it is escaped by -. ('(?<!-):(.*?)(?<!-):', 'a:bc-:de:f', SUCCEED, 'g1', 'bc-:de' ), # escaping with \ as we know it - ('(?<!\\\\):(.*?)(?<!\\\\):', 'a:bc\\:de:f', SUCCEED, 'g1', 'bc\\:de' ), + ('(?<!\\\):(.*?)(?<!\\\):', 'a:bc\\:de:f', SUCCEED, 'g1', 'bc\\:de' ), # terminating with ' and escaping with ? as in edifact ("(?<!\\?)'(.*?)(?<!\\?)'", "a'bc?'de'f", SUCCEED, 'g1', "bc?'de" ), @@ -460,7 +563,7 @@ tests = [ # Check odd placement of embedded pattern modifiers # not an error under PCRE/PRE: - ('(?i)w', 'W', SUCCEED, 'found', 'W'), + ('w(?i)', 'W', SUCCEED, 'found', 'W'), # ('w(?i)', 'W', SYNTAX_ERROR), # Comments using the x embedded pattern modifier @@ -483,7 +586,10 @@ xyz""", SUCCEED, 'found', 'abc'), xyzabc 123""", SUCCEED, 'found', 'abc'), + # using the s embedded pattern modifier + ('a.b', 'a\nb', FAIL), + ('(?s)a.b', 'a\nb', SUCCEED, 'found', 'a\nb'), # test \w, etc. both inside and outside character classes @@ -501,8 +607,8 @@ xyzabc # new \x semantics (r'\x00ff', '\377', FAIL), # (r'\x00ff', '\377', SUCCEED, 'found', chr(255)), - (r'\t\n\v\r\f\a', '\t\n\v\r\f\a', SUCCEED, 'found', '\t\n\v\r\f\a'), - ('\t\n\v\r\f\a', '\t\n\v\r\f\a', SUCCEED, 'found', '\t\n\v\r\f\a'), + (r'\t\n\v\r\f\a\g', '\t\n\v\r\f\ag', SUCCEED, 'found', '\t\n\v\r\f\ag'), + ('\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', '\t\n\v\r\f\a', SUCCEED, 'found', chr(9)+chr(10)+chr(11)+chr(13)+chr(12)+chr(7)), (r'[\t][\n][\v][\r][\f][\b]', '\t\n\v\r\f\b', SUCCEED, 'found', '\t\n\v\r\f\b'), @@ -521,7 +627,7 @@ xyzabc # bug 114033: nothing to repeat (r'(x?)?', 'x', SUCCEED, 'found', 'x'), # bug 115040: rescan if flags are modified inside pattern - (r'(?x) foo ', 'foo', SUCCEED, 'found', 'foo'), + (r' (?x)foo ', 'foo', SUCCEED, 'found', 'foo'), # bug 115618: negative lookahead (r'(?<!abc)(d.f)', 'abcdefdof', SUCCEED, 'found', 'dof'), # bug 116251: character class bug @@ -555,10 +661,14 @@ xyzabc ('^([ab]*?)(?<!(a))c', 'abc', SUCCEED, 'g1+"-"+g2', 'ab-None'), ] -u = '\N{LATIN CAPITAL LETTER A WITH DIAERESIS}' -tests.extend([ +try: + u = eval("u'\N{LATIN CAPITAL LETTER A WITH DIAERESIS}'") +except (SyntaxError, ValueError): + pass +else: + tests.extend([ # bug 410271: \b broken under locales (r'\b.\b', 'a', SUCCEED, 'found', 'a'), (r'(?u)\b.\b', u, SUCCEED, 'found', u), (r'(?u)\w', u, SUCCEED, 'found', u), -]) + ]) |