#!/usr/bin/env python # -------------------------------------------------------------------------- # This file is a modified re module testsuite from Python 2.7.5. # # Modifications have been commented with "PCRE:". # Imports pcre instead of re. # Internal _sre tests have been commented out. # Not supported features have been commented out. # Regular expressions have been changed to PCRE format. # -------------------------------------------------------------------------- from test.test_support import verbose, run_unittest, import_module from test.test_support import precisionbigmemtest, _2G, cpython_only from test.test_support import captured_stdout # PCRE: import pcre instead of re #import re #from re import Scanner #import sre_constants import pcre as re re.enable_re_template_mode() sre_constants = re re._pattern_type = re.Pattern import sys import string import traceback from weakref import proxy # Misc tests from Tim Peters' re.doc # WARNING: Don't change details in these tests if you don't know # what you're doing. Some of these tests were carefully modeled to # cover most of the code. import unittest class ReTests(unittest.TestCase): def test_weakref(self): s = 'QabbbcR' x = re.compile('ab+c') y = proxy(x) self.assertEqual(x.findall('QabbbcR'), y.findall('QabbbcR')) def test_search_star_plus(self): self.assertEqual(re.search('x*', 'axx').span(0), (0, 0)) self.assertEqual(re.search('x*', 'axx').span(), (0, 0)) self.assertEqual(re.search('x+', 'axx').span(0), (1, 3)) self.assertEqual(re.search('x+', 'axx').span(), (1, 3)) self.assertEqual(re.search('x', 'aaa'), None) self.assertEqual(re.match('a*', 'xxx').span(0), (0, 0)) self.assertEqual(re.match('a*', 'xxx').span(), (0, 0)) self.assertEqual(re.match('x*', 'xxxa').span(0), (0, 3)) self.assertEqual(re.match('x*', 'xxxa').span(), (0, 3)) self.assertEqual(re.match('a+', 'xxx'), None) def bump_num(self, matchobj): int_value = int(matchobj.group(0)) return str(int_value + 1) def test_basic_re_sub(self): self.assertEqual(re.sub("(?i)b+", "x", "bbbb BBBB"), 'x x') self.assertEqual(re.sub(r'\d+', self.bump_num, '08.2 -2 23x99y'), '9.3 -3 24x100y') self.assertEqual(re.sub(r'\d+', self.bump_num, '08.2 -2 23x99y', 3), '9.3 -3 23x99y') self.assertEqual(re.sub('.', lambda m: r"\n", 'x'), '\\n') self.assertEqual(re.sub('.', r"\n", 'x'), '\n') s = r"\1\1" self.assertEqual(re.sub('(.)', s, 'x'), 'xx') self.assertEqual(re.sub('(.)', re.escape(s), 'x'), s) self.assertEqual(re.sub('(.)', lambda m: s, 'x'), s) self.assertEqual(re.sub('(?Px)', '\g\g', 'xx'), 'xxxx') self.assertEqual(re.sub('(?Px)', '\g\g<1>', 'xx'), 'xxxx') self.assertEqual(re.sub('(?Px)', '\g\g', 'xx'), 'xxxx') self.assertEqual(re.sub('(?Px)', '\g<1>\g<1>', 'xx'), 'xxxx') self.assertEqual(re.sub('a',r'\t\n\v\r\f\a\b\B\Z\a\A\w\W\s\S\d\D','a'), '\t\n\v\r\f\a\b\\B\\Z\a\\A\\w\\W\\s\\S\\d\\D') self.assertEqual(re.sub('a', '\t\n\v\r\f\a', 'a'), '\t\n\v\r\f\a') self.assertEqual(re.sub('a', '\t\n\v\r\f\a', 'a'), (chr(9)+chr(10)+chr(11)+chr(13)+chr(12)+chr(7))) self.assertEqual(re.sub('^\s*', 'X', 'test'), 'Xtest') def test_bug_449964(self): # fails for group followed by other escape self.assertEqual(re.sub(r'(?Px)', '\g<1>\g<1>\\b', 'xx'), 'xx\bxx\b') def test_bug_449000(self): # Test for sub() on escaped characters self.assertEqual(re.sub(r'\r\n', r'\n', 'abc\r\ndef\r\n'), 'abc\ndef\n') self.assertEqual(re.sub('\r\n', r'\n', 'abc\r\ndef\r\n'), 'abc\ndef\n') self.assertEqual(re.sub(r'\r\n', '\n', 'abc\r\ndef\r\n'), 'abc\ndef\n') self.assertEqual(re.sub('\r\n', '\n', 'abc\r\ndef\r\n'), 'abc\ndef\n') def test_bug_1140(self): # re.sub(x, y, u'') should return u'', not '', and # re.sub(x, y, '') should return '', not u''. # Also: # re.sub(x, y, unicode(x)) should return unicode(y), and # re.sub(x, y, str(x)) should return # str(y) if isinstance(y, str) else unicode(y). for x in 'x', u'x': for y in 'y', u'y': z = re.sub(x, y, u'') self.assertEqual(z, u'') self.assertEqual(type(z), unicode) # z = re.sub(x, y, '') self.assertEqual(z, '') self.assertEqual(type(z), str) # z = re.sub(x, y, unicode(x)) self.assertEqual(z, y) self.assertEqual(type(z), unicode) # z = re.sub(x, y, str(x)) self.assertEqual(z, y) self.assertEqual(type(z), type(y)) def test_bug_1661(self): # Verify that flags do not get silently ignored with compiled patterns pattern = re.compile('.') self.assertRaises(ValueError, re.match, pattern, 'A', re.I) self.assertRaises(ValueError, re.search, pattern, 'A', re.I) self.assertRaises(ValueError, re.findall, pattern, 'A', re.I) self.assertRaises(ValueError, re.compile, pattern, re.I) def test_bug_3629(self): # A regex that triggered a bug in the sre-code validator re.compile("(?P)(?(quote))") def test_sub_template_numeric_escape(self): # bug 776311 and friends self.assertEqual(re.sub('x', r'\0', 'x'), '\0') self.assertEqual(re.sub('x', r'\000', 'x'), '\000') self.assertEqual(re.sub('x', r'\001', 'x'), '\001') self.assertEqual(re.sub('x', r'\008', 'x'), '\0' + '8') self.assertEqual(re.sub('x', r'\009', 'x'), '\0' + '9') self.assertEqual(re.sub('x', r'\111', 'x'), '\111') self.assertEqual(re.sub('x', r'\117', 'x'), '\117') self.assertEqual(re.sub('x', r'\1111', 'x'), '\1111') self.assertEqual(re.sub('x', r'\1111', 'x'), '\111' + '1') self.assertEqual(re.sub('x', r'\00', 'x'), '\x00') self.assertEqual(re.sub('x', r'\07', 'x'), '\x07') self.assertEqual(re.sub('x', r'\08', 'x'), '\0' + '8') self.assertEqual(re.sub('x', r'\09', 'x'), '\0' + '9') self.assertEqual(re.sub('x', r'\0a', 'x'), '\0' + 'a') self.assertEqual(re.sub('x', r'\400', 'x'), '\0') self.assertEqual(re.sub('x', r'\777', 'x'), '\377') self.assertRaises(re.error, re.sub, 'x', r'\1', 'x') self.assertRaises(re.error, re.sub, 'x', r'\8', 'x') self.assertRaises(re.error, re.sub, 'x', r'\9', 'x') self.assertRaises(re.error, re.sub, 'x', r'\11', 'x') self.assertRaises(re.error, re.sub, 'x', r'\18', 'x') self.assertRaises(re.error, re.sub, 'x', r'\1a', 'x') self.assertRaises(re.error, re.sub, 'x', r'\90', 'x') self.assertRaises(re.error, re.sub, 'x', r'\99', 'x') self.assertRaises(re.error, re.sub, 'x', r'\118', 'x') # r'\11' + '8' self.assertRaises(re.error, re.sub, 'x', r'\11a', 'x') self.assertRaises(re.error, re.sub, 'x', r'\181', 'x') # r'\18' + '1' self.assertRaises(re.error, re.sub, 'x', r'\800', 'x') # r'\80' + '0' # in python2.3 (etc), these loop endlessly in sre_parser.py self.assertEqual(re.sub('(((((((((((x)))))))))))', r'\11', 'x'), 'x') self.assertEqual(re.sub('((((((((((y))))))))))(.)', r'\118', 'xyz'), 'xz8') self.assertEqual(re.sub('((((((((((y))))))))))(.)', r'\11a', 'xyz'), 'xza') def test_qualified_re_sub(self): self.assertEqual(re.sub('a', 'b', 'aaaaa'), 'bbbbb') self.assertEqual(re.sub('a', 'b', 'aaaaa', 1), 'baaaa') def test_bug_114660(self): self.assertEqual(re.sub(r'(\S)\s+(\S)', r'\1 \2', 'hello there'), 'hello there') def test_bug_462270(self): # Test for empty sub() behaviour, see SF bug #462270 self.assertEqual(re.sub('x*', '-', 'abxd'), '-a-b-d-') self.assertEqual(re.sub('x+', '-', 'abxd'), 'ab-d') def test_symbolic_groups(self): re.compile('(?Px)(?P=a)(?(a)y)') re.compile('(?Px)(?P=a1)(?(a1)y)') self.assertRaises(re.error, re.compile, '(?P)(?P)') self.assertRaises(re.error, re.compile, '(?Px)') self.assertRaises(re.error, re.compile, '(?P=)') self.assertRaises(re.error, re.compile, '(?P=1)') self.assertRaises(re.error, re.compile, '(?P=a)') self.assertRaises(re.error, re.compile, '(?P=a1)') self.assertRaises(re.error, re.compile, '(?P=a.)') self.assertRaises(re.error, re.compile, '(?P<)') self.assertRaises(re.error, re.compile, '(?P<>)') self.assertRaises(re.error, re.compile, '(?P<1>)') self.assertRaises(re.error, re.compile, '(?P)') self.assertRaises(re.error, re.compile, '(?())') self.assertRaises(re.error, re.compile, '(?(a))') self.assertRaises(re.error, re.compile, '(?(1a))') self.assertRaises(re.error, re.compile, '(?(a.))') def test_symbolic_refs(self): self.assertRaises(re.error, re.sub, '(?Px)', '\gx)', '\g<', 'xx') self.assertRaises(re.error, re.sub, '(?Px)', '\g', 'xx') self.assertRaises(re.error, re.sub, '(?Px)', '\g', 'xx') self.assertRaises(re.error, re.sub, '(?Px)', '\g<>', 'xx') self.assertRaises(re.error, re.sub, '(?Px)', '\g<1a1>', 'xx') self.assertRaises(IndexError, re.sub, '(?Px)', '\g', 'xx') self.assertRaises(re.error, re.sub, '(?Px)|(?Py)', '\g', 'xx') self.assertRaises(re.error, re.sub, '(?Px)|(?Py)', '\\2', 'xx') self.assertRaises(re.error, re.sub, '(?Px)', '\g<-1>', 'xx') def test_re_subn(self): self.assertEqual(re.subn("(?i)b+", "x", "bbbb BBBB"), ('x x', 2)) self.assertEqual(re.subn("b+", "x", "bbbb BBBB"), ('x BBBB', 1)) self.assertEqual(re.subn("b+", "x", "xyz"), ('xyz', 0)) self.assertEqual(re.subn("b*", "x", "xyz"), ('xxxyxzx', 4)) self.assertEqual(re.subn("b*", "x", "xyz", 2), ('xxxyz', 2)) def test_re_split(self): self.assertEqual(re.split(":", ":a:b::c"), ['', 'a', 'b', '', 'c']) self.assertEqual(re.split(":*", ":a:b::c"), ['', 'a', 'b', 'c']) self.assertEqual(re.split("(:*)", ":a:b::c"), ['', ':', 'a', ':', 'b', '::', 'c']) self.assertEqual(re.split("(?::*)", ":a:b::c"), ['', 'a', 'b', 'c']) self.assertEqual(re.split("(:)*", ":a:b::c"), ['', ':', 'a', ':', 'b', ':', 'c']) self.assertEqual(re.split("([b:]+)", ":a:b::c"), ['', ':', 'a', ':b::', 'c']) self.assertEqual(re.split("(b)|(:+)", ":a:b::c"), ['', None, ':', 'a', None, ':', '', 'b', None, '', None, '::', 'c']) self.assertEqual(re.split("(?:b)|(?::+)", ":a:b::c"), ['', 'a', '', '', 'c']) def test_qualified_re_split(self): self.assertEqual(re.split(":", ":a:b::c", 2), ['', 'a', 'b::c']) self.assertEqual(re.split(':', 'a:b:c:d', 2), ['a', 'b', 'c:d']) self.assertEqual(re.split("(:)", ":a:b::c", 2), ['', ':', 'a', ':', 'b::c']) self.assertEqual(re.split("(:*)", ":a:b::c", 2), ['', ':', 'a', ':', 'b::c']) def test_re_findall(self): self.assertEqual(re.findall(":+", "abc"), []) self.assertEqual(re.findall(":+", "a:b::c:::d"), [":", "::", ":::"]) self.assertEqual(re.findall("(:+)", "a:b::c:::d"), [":", "::", ":::"]) self.assertEqual(re.findall("(:)(:*)", "a:b::c:::d"), [(":", ""), (":", ":"), (":", "::")]) def test_bug_117612(self): self.assertEqual(re.findall(r"(a|(b))", "aba"), [("a", ""),("b", "b"),("a", "")]) def test_re_match(self): self.assertEqual(re.match('a', 'a').groups(), ()) self.assertEqual(re.match('(a)', 'a').groups(), ('a',)) self.assertEqual(re.match(r'(a)', 'a').group(0), 'a') self.assertEqual(re.match(r'(a)', 'a').group(1), 'a') self.assertEqual(re.match(r'(a)', 'a').group(1, 1), ('a', 'a')) pat = re.compile('((a)|(b))(c)?') self.assertEqual(pat.match('a').groups(), ('a', 'a', None, None)) self.assertEqual(pat.match('b').groups(), ('b', None, 'b', None)) self.assertEqual(pat.match('ac').groups(), ('a', 'a', None, 'c')) self.assertEqual(pat.match('bc').groups(), ('b', None, 'b', 'c')) self.assertEqual(pat.match('bc').groups(""), ('b', "", 'b', 'c')) # A single group m = re.match('(a)', 'a') self.assertEqual(m.group(0), 'a') self.assertEqual(m.group(0), 'a') self.assertEqual(m.group(1), 'a') self.assertEqual(m.group(1, 1), ('a', 'a')) pat = re.compile('(?:(?Pa)|(?Pb))(?Pc)?') self.assertEqual(pat.match('a').group(1, 2, 3), ('a', None, None)) self.assertEqual(pat.match('b').group('a1', 'b2', 'c3'), (None, 'b', None)) self.assertEqual(pat.match('ac').group(1, 'b2', 3), ('a', None, 'c')) def test_re_groupref_exists(self): self.assertEqual(re.match('^(\()?([^()]+)(?(1)\))$', '(a)').groups(), ('(', 'a')) self.assertEqual(re.match('^(\()?([^()]+)(?(1)\))$', 'a').groups(), (None, 'a')) self.assertEqual(re.match('^(\()?([^()]+)(?(1)\))$', 'a)'), None) self.assertEqual(re.match('^(\()?([^()]+)(?(1)\))$', '(a'), None) self.assertEqual(re.match('^(?:(a)|c)((?(1)b|d))$', 'ab').groups(), ('a', 'b')) self.assertEqual(re.match('^(?:(a)|c)((?(1)b|d))$', 'cd').groups(), (None, 'd')) self.assertEqual(re.match('^(?:(a)|c)((?(1)|d))$', 'cd').groups(), (None, 'd')) self.assertEqual(re.match('^(?:(a)|c)((?(1)|d))$', 'a').groups(), ('a', '')) # Tests for bug #1177831: exercise groups other than the first group p = re.compile('(?Pa)(?Pb)?((?(g2)c|d))') self.assertEqual(p.match('abc').groups(), ('a', 'b', 'c')) self.assertEqual(p.match('ad').groups(), ('a', None, 'd')) self.assertEqual(p.match('abd'), None) self.assertEqual(p.match('ac'), None) def test_re_groupref(self): self.assertEqual(re.match(r'^(\|)?([^()]+)\1$', '|a|').groups(), ('|', 'a')) self.assertEqual(re.match(r'^(\|)?([^()]+)\1?$', 'a').groups(), (None, 'a')) self.assertEqual(re.match(r'^(\|)?([^()]+)\1$', 'a|'), None) self.assertEqual(re.match(r'^(\|)?([^()]+)\1$', '|a'), None) self.assertEqual(re.match(r'^(?:(a)|c)(\1)$', 'aa').groups(), ('a', 'a')) self.assertEqual(re.match(r'^(?:(a)|c)(\1)?$', 'c').groups(), (None, None)) def test_groupdict(self): self.assertEqual(re.match('(?Pfirst) (?Psecond)', 'first second').groupdict(), {'first':'first', 'second':'second'}) def test_expand(self): self.assertEqual(re.match("(?Pfirst) (?Psecond)", "first second") .expand(r"\2 \1 \g \g"), "second first second first") def test_repeat_minmax(self): self.assertEqual(re.match("^(\w){1}$", "abc"), None) self.assertEqual(re.match("^(\w){1}?$", "abc"), None) self.assertEqual(re.match("^(\w){1,2}$", "abc"), None) self.assertEqual(re.match("^(\w){1,2}?$", "abc"), None) self.assertEqual(re.match("^(\w){3}$", "abc").group(1), "c") self.assertEqual(re.match("^(\w){1,3}$", "abc").group(1), "c") self.assertEqual(re.match("^(\w){1,4}$", "abc").group(1), "c") self.assertEqual(re.match("^(\w){3,4}?$", "abc").group(1), "c") self.assertEqual(re.match("^(\w){3}?$", "abc").group(1), "c") self.assertEqual(re.match("^(\w){1,3}?$", "abc").group(1), "c") self.assertEqual(re.match("^(\w){1,4}?$", "abc").group(1), "c") self.assertEqual(re.match("^(\w){3,4}?$", "abc").group(1), "c") self.assertEqual(re.match("^x{1}$", "xxx"), None) self.assertEqual(re.match("^x{1}?$", "xxx"), None) self.assertEqual(re.match("^x{1,2}$", "xxx"), None) self.assertEqual(re.match("^x{1,2}?$", "xxx"), None) self.assertNotEqual(re.match("^x{3}$", "xxx"), None) self.assertNotEqual(re.match("^x{1,3}$", "xxx"), None) self.assertNotEqual(re.match("^x{1,4}$", "xxx"), None) self.assertNotEqual(re.match("^x{3,4}?$", "xxx"), None) self.assertNotEqual(re.match("^x{3}?$", "xxx"), None) self.assertNotEqual(re.match("^x{1,3}?$", "xxx"), None) # PCRE: need to make ungreedy for this case to work in PCRE #self.assertNotEqual(re.match("^x{1,4}?$", "xxx"), None) self.assertNotEqual(re.match("(?U)^x{1,4}?$", "xxx"), None) self.assertNotEqual(re.match("^x{3,4}?$", "xxx"), None) self.assertEqual(re.match("^x{}$", "xxx"), None) self.assertNotEqual(re.match("^x{}$", "x{}"), None) def test_getattr(self): self.assertEqual(re.match("(a)", "a").pos, 0) self.assertEqual(re.match("(a)", "a").endpos, 1) self.assertEqual(re.match("(a)", "a").string, "a") self.assertEqual(re.match("(a)", "a").regs, ((0, 1), (0, 1))) self.assertNotEqual(re.match("(a)", "a").re, None) def test_special_escapes(self): self.assertEqual(re.search(r"\b(b.)\b", "abcd abc bcd bx").group(1), "bx") self.assertEqual(re.search(r"\B(b.)\B", "abc bcd bc abxd").group(1), "bx") # PCRE: python-pcre doesn't support LOCALE flag #self.assertEqual(re.search(r"\b(b.)\b", # "abcd abc bcd bx", re.LOCALE).group(1), "bx") #self.assertEqual(re.search(r"\B(b.)\B", # "abc bcd bc abxd", re.LOCALE).group(1), "bx") self.assertEqual(re.search(r"\b(b.)\b", "abcd abc bcd bx", re.UNICODE).group(1), "bx") self.assertEqual(re.search(r"\B(b.)\B", "abc bcd bc abxd", re.UNICODE).group(1), "bx") self.assertEqual(re.search(r"^abc$", "\nabc\n", re.M).group(0), "abc") self.assertEqual(re.search(r"^\Aabc\Z$", "abc", re.M).group(0), "abc") self.assertEqual(re.search(r"^\Aabc\Z$", "\nabc\n", re.M), None) self.assertEqual(re.search(r"\b(b.)\b", u"abcd abc bcd bx").group(1), "bx") self.assertEqual(re.search(r"\B(b.)\B", u"abc bcd bc abxd").group(1), "bx") self.assertEqual(re.search(r"^abc$", u"\nabc\n", re.M).group(0), "abc") self.assertEqual(re.search(r"^\Aabc\Z$", u"abc", re.M).group(0), "abc") self.assertEqual(re.search(r"^\Aabc\Z$", u"\nabc\n", re.M), None) self.assertEqual(re.search(r"\d\D\w\W\s\S", "1aa! a").group(0), "1aa! a") # PCRE: python-pcre doesn't support LOCALE flag #self.assertEqual(re.search(r"\d\D\w\W\s\S", # "1aa! a", re.LOCALE).group(0), "1aa! a") self.assertEqual(re.search(r"\d\D\w\W\s\S", "1aa! a", re.UNICODE).group(0), "1aa! a") def test_string_boundaries(self): # See http://bugs.python.org/issue10713 self.assertEqual(re.search(r"\b(abc)\b", "abc").group(1), "abc") # There's a word boundary at the start of a string. self.assertTrue(re.match(r"\b", "abc")) # A non-empty string includes a non-boundary zero-length match. self.assertTrue(re.search(r"\B", "abc")) # There is no non-boundary match at the start of a string. self.assertFalse(re.match(r"\B", "abc")) # However, an empty string contains no word boundaries, and also no # non-boundaries. # PCRE: PCRE needs the NOTEMPTY_ATSTART flag to pass #self.assertEqual(re.search(r"\B", ""), None) self.assertEqual(re.compile(r"\B").search("", flags=re.NOTEMPTY_ATSTART), None) # This one is questionable and different from the perlre behaviour, # but describes current behavior. self.assertEqual(re.search(r"\b", ""), None) # A single word-character string has two boundaries, but no # non-boundary gaps. self.assertEqual(len(re.findall(r"\b", "a")), 2) self.assertEqual(len(re.findall(r"\B", "a")), 0) # If there are no words, there are no boundaries self.assertEqual(len(re.findall(r"\b", " ")), 0) self.assertEqual(len(re.findall(r"\b", " ")), 0) # Can match around the whitespace. self.assertEqual(len(re.findall(r"\B", " ")), 2) def test_bigcharset(self): self.assertEqual(re.match(u"([\u2222\u2223])", u"\u2222").group(1), u"\u2222") self.assertEqual(re.match(u"([\u2222\u2223])", u"\u2222", re.UNICODE).group(1), u"\u2222") r = u'[%s]' % u''.join(map(unichr, range(256, 2**16, 255))) self.assertEqual(re.match(r, u"\uff01", re.UNICODE).group(), u"\uff01") def test_big_codesize(self): # Issue #1160 # PCRE: Size needs to be reduced to pass. #r = re.compile('|'.join(('%d'%x for x in range(10000)))) r = re.compile('|'.join(('%d'%x for x in range(6000)))) # PCRE self.assertIsNotNone(r.match('1000')) self.assertIsNotNone(r.match('9999')) def test_anyall(self): self.assertEqual(re.match("a.b", "a\nb", re.DOTALL).group(0), "a\nb") self.assertEqual(re.match("a.*b", "a\n\nb", re.DOTALL).group(0), "a\n\nb") def test_non_consuming(self): self.assertEqual(re.match("(a(?=\s[^a]))", "a b").group(1), "a") self.assertEqual(re.match("(a(?=\s[^a]*))", "a b").group(1), "a") self.assertEqual(re.match("(a(?=\s[abc]))", "a b").group(1), "a") self.assertEqual(re.match("(a(?=\s[abc]*))", "a bc").group(1), "a") self.assertEqual(re.match(r"(a)(?=\s\1)", "a a").group(1), "a") self.assertEqual(re.match(r"(a)(?=\s\1*)", "a aa").group(1), "a") self.assertEqual(re.match(r"(a)(?=\s(abc|a))", "a a").group(1), "a") self.assertEqual(re.match(r"(a(?!\s[^a]))", "a a").group(1), "a") self.assertEqual(re.match(r"(a(?!\s[abc]))", "a d").group(1), "a") self.assertEqual(re.match(r"(a)(?!\s\1)", "a b").group(1), "a") self.assertEqual(re.match(r"(a)(?!\s(abc|a))", "a b").group(1), "a") def test_ignore_case(self): self.assertEqual(re.match("abc", "ABC", re.I).group(0), "ABC") self.assertEqual(re.match("abc", u"ABC", re.I).group(0), "ABC") self.assertEqual(re.match(r"(a\s[^a])", "a b", re.I).group(1), "a b") self.assertEqual(re.match(r"(a\s[^a]*)", "a bb", re.I).group(1), "a bb") self.assertEqual(re.match(r"(a\s[abc])", "a b", re.I).group(1), "a b") self.assertEqual(re.match(r"(a\s[abc]*)", "a bb", re.I).group(1), "a bb") self.assertEqual(re.match(r"((a)\s\2)", "a a", re.I).group(1), "a a") self.assertEqual(re.match(r"((a)\s\2*)", "a aa", re.I).group(1), "a aa") self.assertEqual(re.match(r"((a)\s(abc|a))", "a a", re.I).group(1), "a a") self.assertEqual(re.match(r"((a)\s(abc|a)*)", "a aa", re.I).group(1), "a aa") def test_category(self): self.assertEqual(re.match(r"(\s)", " ").group(1), " ") def test_getlower(self): # PCRE: we're not testing _sre here #import _sre #self.assertEqual(_sre.getlower(ord('A'), 0), ord('a')) #self.assertEqual(_sre.getlower(ord('A'), re.LOCALE), ord('a')) #self.assertEqual(_sre.getlower(ord('A'), re.UNICODE), ord('a')) self.assertEqual(re.match("abc", "ABC", re.I).group(0), "ABC") self.assertEqual(re.match("abc", u"ABC", re.I).group(0), "ABC") def test_not_literal(self): self.assertEqual(re.search("\s([^a])", " b").group(1), "b") self.assertEqual(re.search("\s([^a]*)", " bb").group(1), "bb") def test_search_coverage(self): self.assertEqual(re.search("\s(b)", " b").group(1), "b") self.assertEqual(re.search("a\s", "a ").group(0), "a ") def assertMatch(self, pattern, text, match=None, span=None, matcher=re.match): if match is None and span is None: # the pattern matches the whole text match = text span = (0, len(text)) elif match is None or span is None: raise ValueError('If match is not None, span should be specified ' '(and vice versa).') m = matcher(pattern, text) self.assertTrue(m) self.assertEqual(m.group(), match) self.assertEqual(m.span(), span) def test_re_escape(self): alnum_chars = string.ascii_letters + string.digits p = u''.join(unichr(i) for i in range(256)) for c in p: if c in alnum_chars: self.assertEqual(re.escape(c), c) elif c == u'\x00': self.assertEqual(re.escape(c), u'\\000') else: self.assertEqual(re.escape(c), u'\\' + c) self.assertMatch(re.escape(c), c) self.assertMatch(re.escape(p), p) def test_re_escape_byte(self): alnum_chars = (string.ascii_letters + string.digits).encode('ascii') p = ''.join(chr(i) for i in range(256)) for b in p: if b in alnum_chars: self.assertEqual(re.escape(b), b) elif b == b'\x00': self.assertEqual(re.escape(b), b'\\000') else: self.assertEqual(re.escape(b), b'\\' + b) self.assertMatch(re.escape(b), b) self.assertMatch(re.escape(p), p) def test_re_escape_non_ascii(self): s = u'xxx\u2620\u2620\u2620xxx' s_escaped = re.escape(s) self.assertEqual(s_escaped, u'xxx\\\u2620\\\u2620\\\u2620xxx') self.assertMatch(s_escaped, s) self.assertMatch(u'.%s+.' % re.escape(u'\u2620'), s, u'x\u2620\u2620\u2620x', (2, 7), re.search) def test_re_escape_non_ascii_bytes(self): b = u'y\u2620y\u2620y'.encode('utf-8') b_escaped = re.escape(b) self.assertEqual(b_escaped, b'y\\\xe2\\\x98\\\xa0y\\\xe2\\\x98\\\xa0y') self.assertMatch(b_escaped, b) res = re.findall(re.escape(u'\u2620'.encode('utf-8')), b) self.assertEqual(len(res), 2) def test_pickling(self): import pickle self.pickle_test(pickle) import cPickle self.pickle_test(cPickle) # PCRE: we're not testing _sre here # old pickles expect the _compile() reconstructor in sre module #import_module("sre", deprecated=True) #from sre import _compile def pickle_test(self, pickle): oldpat = re.compile('a(?:b|(c|e){1,2}?|d)+?(.)') s = pickle.dumps(oldpat) newpat = pickle.loads(s) self.assertEqual(oldpat, newpat) def test_constants(self): self.assertEqual(re.I, re.IGNORECASE) # PCRE: python-pcre doesn't support LOCALE flag #self.assertEqual(re.L, re.LOCALE) self.assertEqual(re.M, re.MULTILINE) self.assertEqual(re.S, re.DOTALL) self.assertEqual(re.X, re.VERBOSE) def test_flags(self): # PCRE: python-pcre doesn't support LOCALE flag #for flag in [re.I, re.M, re.X, re.S, re.L]: for flag in [re.I, re.M, re.X, re.S]: self.assertNotEqual(re.compile('^pattern$', flag), None) def test_sre_character_literals(self): for i in [0, 8, 16, 32, 64, 127, 128, 255]: self.assertNotEqual(re.match(r"\%03o" % i, chr(i)), None) self.assertNotEqual(re.match(r"\%03o0" % i, chr(i)+"0"), None) self.assertNotEqual(re.match(r"\%03o8" % i, chr(i)+"8"), None) self.assertNotEqual(re.match(r"\x%02x" % i, chr(i)), None) self.assertNotEqual(re.match(r"\x%02x0" % i, chr(i)+"0"), None) self.assertNotEqual(re.match(r"\x%02xz" % i, chr(i)+"z"), None) # PCRE: PCRE doesn't see this as an error #self.assertRaises(re.error, re.match, "\911", "") def test_sre_character_class_literals(self): for i in [0, 8, 16, 32, 64, 127, 128, 255]: self.assertNotEqual(re.match(r"[\%03o]" % i, chr(i)), None) self.assertNotEqual(re.match(r"[\%03o0]" % i, chr(i)), None) self.assertNotEqual(re.match(r"[\%03o8]" % i, chr(i)), None) self.assertNotEqual(re.match(r"[\x%02x]" % i, chr(i)), None) self.assertNotEqual(re.match(r"[\x%02x0]" % i, chr(i)), None) self.assertNotEqual(re.match(r"[\x%02xz]" % i, chr(i)), None) # PCRE: PCRE doesn't see this as an error #self.assertRaises(re.error, re.match, "[\911]", "") def test_bug_113254(self): self.assertEqual(re.match(r'(a)|(b)', 'b').start(1), -1) self.assertEqual(re.match(r'(a)|(b)', 'b').end(1), -1) self.assertEqual(re.match(r'(a)|(b)', 'b').span(1), (-1, -1)) def test_bug_527371(self): # bug described in patches 527371/672491 self.assertEqual(re.match(r'(a)?a','a').lastindex, None) self.assertEqual(re.match(r'(a)(b)?b','ab').lastindex, 1) self.assertEqual(re.match(r'(?Pa)(?Pb)?b','ab').lastgroup, 'a') # PCRE: PCRE's lastgroup/lastindex is based on opening parentheses #self.assertEqual(re.match("(?Pa(b))", "ab").lastgroup, 'a') #self.assertEqual(re.match("((a))", "a").lastindex, 1) self.assertEqual(re.match("(?Pa(b))", "ab").lastgroup, None) self.assertEqual(re.match("((a))", "a").lastindex, 2) def test_bug_545855(self): # bug 545855 -- This pattern failed to cause a compile error as it # should, instead provoking a TypeError. self.assertRaises(re.error, re.compile, 'foo[a-') def test_bug_418626(self): # bugs 418626 at al. -- Testing Greg Chapman's addition of op code # SRE_OP_MIN_REPEAT_ONE for eliminating recursion on simple uses of # pattern '*?' on a long string. self.assertEqual(re.match('.*?c', 10000*'ab'+'cd').end(0), 20001) self.assertEqual(re.match('.*?cd', 5000*'ab'+'c'+5000*'ab'+'cde').end(0), 20003) self.assertEqual(re.match('.*?cd', 20000*'abc'+'de').end(0), 60001) # non-simple '*?' still used to hit the recursion limit, before the # non-recursive scheme was implemented. self.assertEqual(re.search('(a|b)*?c', 10000*'ab'+'cd').end(0), 20001) def test_bug_612074(self): pat=u"["+re.escape(u"\u2039")+u"]" self.assertEqual(re.compile(pat) and 1, 1) def test_stack_overflow(self): # nasty cases that used to overflow the straightforward recursive # implementation of repeated groups. self.assertEqual(re.match('(x)*', 50000*'x').group(1), 'x') self.assertEqual(re.match('(x)*y', 50000*'x'+'y').group(1), 'x') self.assertEqual(re.match('(x)*?y', 50000*'x'+'y').group(1), 'x') def test_unlimited_zero_width_repeat(self): # Issue #9669 self.assertIsNone(re.match(r'(?:a?)*y', 'z')) self.assertIsNone(re.match(r'(?:a?)+y', 'z')) self.assertIsNone(re.match(r'(?:a?){2,}y', 'z')) self.assertIsNone(re.match(r'(?:a?)*?y', 'z')) self.assertIsNone(re.match(r'(?:a?)+?y', 'z')) self.assertIsNone(re.match(r'(?:a?){2,}?y', 'z')) # PCRE: python-pcre doesn't support scanner APIs def __test_scanner(self): def s_ident(scanner, token): return token def s_operator(scanner, token): return "op%s" % token def s_float(scanner, token): return float(token) def s_int(scanner, token): return int(token) scanner = Scanner([ (r"[a-zA-Z_]\w*", s_ident), (r"\d+\.\d*", s_float), (r"\d+", s_int), (r"=|\+|-|\*|/", s_operator), (r"\s+", None), ]) self.assertNotEqual(scanner.scanner.scanner("").pattern, None) self.assertEqual(scanner.scan("sum = 3*foo + 312.50 + bar"), (['sum', 'op=', 3, 'op*', 'foo', 'op+', 312.5, 'op+', 'bar'], '')) def test_bug_448951(self): # bug 448951 (similar to 429357, but with single char match) # (Also test greedy matches.) for op in '','?','*': self.assertEqual(re.match(r'((.%s):)?z'%op, 'z').groups(), (None, None)) self.assertEqual(re.match(r'((.%s):)?z'%op, 'a:z').groups(), ('a:', 'a')) def test_bug_725106(self): # capturing groups in alternatives in repeats self.assertEqual(re.match('^((a)|b)*', 'abc').groups(), ('b', 'a')) self.assertEqual(re.match('^(([ab])|c)*', 'abc').groups(), ('c', 'b')) self.assertEqual(re.match('^((d)|[ab])*', 'abc').groups(), ('b', None)) self.assertEqual(re.match('^((a)c|[ab])*', 'abc').groups(), ('b', None)) self.assertEqual(re.match('^((a)|b)*?c', 'abc').groups(), ('b', 'a')) self.assertEqual(re.match('^(([ab])|c)*?d', 'abcd').groups(), ('c', 'b')) self.assertEqual(re.match('^((d)|[ab])*?c', 'abc').groups(), ('b', None)) self.assertEqual(re.match('^((a)c|[ab])*?c', 'abc').groups(), ('b', None)) def test_bug_725149(self): # mark_stack_base restoring before restoring marks self.assertEqual(re.match('(a)(?:(?=(b)*)c)*', 'abb').groups(), ('a', None)) self.assertEqual(re.match('(a)((?!(b)*))*', 'abb').groups(), ('a', None, None)) def test_bug_764548(self): # bug 764548, re.compile() barfs on str/unicode subclasses try: unicode except NameError: self.skipTest('no problem if we have no unicode') class my_unicode(unicode): pass pat = re.compile(my_unicode("abc")) self.assertEqual(pat.match("xyz"), None) def test_finditer(self): iter = re.finditer(r":+", "a:b::c:::d") self.assertEqual([item.group(0) for item in iter], [":", "::", ":::"]) def test_bug_926075(self): try: unicode except NameError: self.skipTest('no problem if we have no unicode') self.assertTrue(re.compile('bug_926075') is not re.compile(eval("u'bug_926075'"))) def test_bug_931848(self): try: unicode except NameError: self.skipTest('no problem if we have no unicode') pattern = eval('u"[\u002E\u3002\uFF0E\uFF61]"') self.assertEqual(re.compile(pattern).split("a.b.c"), ['a','b','c']) def test_bug_581080(self): iter = re.finditer(r"\s", "a b") self.assertEqual(iter.next().span(), (1,2)) self.assertRaises(StopIteration, iter.next) # PCRE: python-pcre doesn't support scanner APIs return scanner = re.compile(r"\s").scanner("a b") self.assertEqual(scanner.search().span(), (1, 2)) self.assertEqual(scanner.search(), None) def test_bug_817234(self): iter = re.finditer(r".*", "asdf") self.assertEqual(iter.next().span(), (0, 4)) self.assertEqual(iter.next().span(), (4, 4)) self.assertRaises(StopIteration, iter.next) def test_bug_6561(self): # '\d' should match characters in Unicode category 'Nd' # (Number, Decimal Digit), but not those in 'Nl' (Number, # Letter) or 'No' (Number, Other). decimal_digits = [ u'\u0037', # '\N{DIGIT SEVEN}', category 'Nd' u'\u0e58', # '\N{THAI DIGIT SIX}', category 'Nd' u'\uff10', # '\N{FULLWIDTH DIGIT ZERO}', category 'Nd' ] for x in decimal_digits: self.assertEqual(re.match('^\d$', x, re.UNICODE).group(0), x) not_decimal_digits = [ u'\u2165', # '\N{ROMAN NUMERAL SIX}', category 'Nl' u'\u3039', # '\N{HANGZHOU NUMERAL TWENTY}', category 'Nl' u'\u2082', # '\N{SUBSCRIPT TWO}', category 'No' u'\u32b4', # '\N{CIRCLED NUMBER THIRTY NINE}', category 'No' ] for x in not_decimal_digits: self.assertIsNone(re.match('^\d$', x, re.UNICODE)) def test_empty_array(self): # SF buf 1647541 import array for typecode in 'cbBuhHiIlLfd': a = array.array(typecode) self.assertEqual(re.compile("bla").match(a), None) self.assertEqual(re.compile("").match(a).groups(), ()) def test_inline_flags(self): # Bug #1700 upper_char = unichr(0x1ea0) # Latin Capital Letter A with Dot Bellow lower_char = unichr(0x1ea1) # Latin Small Letter A with Dot Bellow p = re.compile(upper_char, re.I | re.U) q = p.match(lower_char) self.assertNotEqual(q, None) p = re.compile(lower_char, re.I | re.U) q = p.match(upper_char) self.assertNotEqual(q, None) p = re.compile('(?i)' + upper_char, re.U) q = p.match(lower_char) self.assertNotEqual(q, None) p = re.compile('(?i)' + lower_char, re.U) q = p.match(upper_char) self.assertNotEqual(q, None) # PCRE: UNICODE mode can be enabled using (*UCP) instead of (?u) #p = re.compile('(?iu)' + upper_char) p = re.compile('(*UCP)(?i)' + upper_char) q = p.match(lower_char) self.assertNotEqual(q, None) # PCRE: UNICODE mode can be enabled using (*UCP) instead of (?u) #p = re.compile('(?iu)' + lower_char) p = re.compile('(*UCP)(?i)' + lower_char) q = p.match(upper_char) self.assertNotEqual(q, None) def test_dollar_matches_twice(self): "$ matches the end of string, and just before the terminating \n" pattern = re.compile('$') self.assertEqual(pattern.sub('#', 'a\nb\n'), 'a\nb#\n#') self.assertEqual(pattern.sub('#', 'a\nb\nc'), 'a\nb\nc#') self.assertEqual(pattern.sub('#', '\n'), '#\n#') pattern = re.compile('$', re.MULTILINE) self.assertEqual(pattern.sub('#', 'a\nb\n' ), 'a#\nb#\n#' ) self.assertEqual(pattern.sub('#', 'a\nb\nc'), 'a#\nb#\nc#') self.assertEqual(pattern.sub('#', '\n'), '#\n#') def test_dealloc(self): # PCRE: we're not testing _sre here # issue 3299: check for segfault in debug build #import _sre # the overflow limit is different on wide and narrow builds and it # depends on the definition of SRE_CODE (see sre.h). # 2**128 should be big enough to overflow on both. For smaller values # a RuntimeError is raised instead of OverflowError. #long_overflow = 2**128 # PCRE: finditer is implemented as generator function -- next() has to be called #self.assertRaises(TypeError, re.finditer, "a", {}) self.assertRaises(TypeError, re.finditer("a", {}).next) #self.assertRaises(OverflowError, _sre.compile, "abc", 0, [long_overflow]) def test_compile(self): # Test return value when given string and pattern as parameter pattern = re.compile('random pattern') self.assertIsInstance(pattern, re._pattern_type) same_pattern = re.compile(pattern) self.assertIsInstance(same_pattern, re._pattern_type) self.assertIs(same_pattern, pattern) # Test behaviour when not given a string or pattern as parameter self.assertRaises(TypeError, re.compile, 0) def test_bug_13899(self): # Issue #13899: re pattern r"[\A]" should work like "A" but matches # nothing. Ditto B and Z. self.assertEqual(re.findall(r'[\A\B\b\C\Z]', 'AB\bCZ'), ['A', 'B', '\b', 'C', 'Z']) @precisionbigmemtest(size=_2G, memuse=1) def test_large_search(self, size): # Issue #10182: indices were 32-bit-truncated. s = 'a' * size m = re.search('$', s) self.assertIsNotNone(m) self.assertEqual(m.start(), size) self.assertEqual(m.end(), size) # The huge memuse is because of re.sub() using a list and a join() # to create the replacement result. @precisionbigmemtest(size=_2G, memuse=16 + 2) def test_large_subn(self, size): # Issue #10182: indices were 32-bit-truncated. s = 'a' * size r, n = re.subn('', '', s) self.assertEqual(r, s) self.assertEqual(n, size + 1) def test_repeat_minmax_overflow(self): # Issue #13169 # PCRE: PCRE requires "{0,x}" instead of "{,x}" # PCRE: PCRE doesn't support repetition over 64K # PCRE: 2**128 repetition raises NO error -- PCRE bug? string = "x" * 100000 self.assertEqual(re.match(r".{65535}", string).span(), (0, 65535)) self.assertEqual(re.match(r".{0,65535}", string).span(), (0, 65535)) self.assertEqual(re.match(r".{65534,}?", string).span(), (0, 65534)) self.assertEqual(re.match(r".{65535}", string).span(), (0, 65535)) self.assertEqual(re.match(r".{0,65535}", string).span(), (0, 65535)) self.assertEqual(re.match(r".{65535,}?", string).span(), (0, 65535)) self.assertRaises(OverflowError, re.compile, r".{%d}" % 65536) self.assertRaises(OverflowError, re.compile, r".{0,%d}" % 65536) self.assertRaises(OverflowError, re.compile, r".{%d,}?" % 65536) self.assertRaises(OverflowError, re.compile, r".{%d,%d}" % (65537, 65536)) return string = "x" * 100000 self.assertEqual(re.match(r".{65535}", string).span(), (0, 65535)) self.assertEqual(re.match(r".{,65535}", string).span(), (0, 65535)) self.assertEqual(re.match(r".{65535,}?", string).span(), (0, 65535)) self.assertEqual(re.match(r".{65536}", string).span(), (0, 65536)) self.assertEqual(re.match(r".{,65536}", string).span(), (0, 65536)) self.assertEqual(re.match(r".{65536,}?", string).span(), (0, 65536)) # 2**128 should be big enough to overflow both SRE_CODE and Py_ssize_t. self.assertRaises(OverflowError, re.compile, r".{%d}" % 2**128) self.assertRaises(OverflowError, re.compile, r".{,%d}" % 2**128) self.assertRaises(OverflowError, re.compile, r".{%d,}?" % 2**128) self.assertRaises(OverflowError, re.compile, r".{%d,%d}" % (2**129, 2**128)) @cpython_only def test_repeat_minmax_overflow_maxrepeat(self): try: # PCRE: this is defined in pcre module (and is always 64K) #from _sre import MAXREPEAT from pcre import MAXREPEAT except ImportError: self.skipTest('requires _sre.MAXREPEAT constant') # PCRE: PCRE requires "{0,x}" instead of "{,x}" string = "x" * 10000 self.assertIsNone(re.match(r".{%d}" % (MAXREPEAT - 1), string)) self.assertEqual(re.match(r".{0,%d}" % (MAXREPEAT - 1), string).span(), (0, 10000)) self.assertIsNone(re.match(r".{%d,}?" % (MAXREPEAT - 1), string)) self.assertRaises(OverflowError, re.compile, r".{%d}" % MAXREPEAT) self.assertRaises(OverflowError, re.compile, r".{0,%d}" % MAXREPEAT) self.assertRaises(OverflowError, re.compile, r".{%d,}?" % MAXREPEAT) return string = "x" * 100000 self.assertIsNone(re.match(r".{%d}" % (MAXREPEAT - 1), string)) self.assertEqual(re.match(r".{,%d}" % (MAXREPEAT - 1), string).span(), (0, 100000)) self.assertIsNone(re.match(r".{%d,}?" % (MAXREPEAT - 1), string)) self.assertRaises(OverflowError, re.compile, r".{%d}" % MAXREPEAT) self.assertRaises(OverflowError, re.compile, r".{,%d}" % MAXREPEAT) self.assertRaises(OverflowError, re.compile, r".{%d,}?" % MAXREPEAT) # PCRE: PCRE's errors don't contain group names def __test_backref_group_name_in_exception(self): # Issue 17341: Poor error message when compiling invalid regex with self.assertRaisesRegexp(sre_constants.error, ''): re.compile('(?P=)') # PCRE: PCRE's errors don't contain group names def __test_group_name_in_exception(self): # Issue 17341: Poor error message when compiling invalid regex with self.assertRaisesRegexp(sre_constants.error, '\?foo'): re.compile('(?P)') def test_issue17998(self): for reps in '*', '+', '?', '{1}': for mod in '', '?': pattern = '.' + reps + mod + 'yz' self.assertEqual(re.compile(pattern, re.S).findall('xyz'), ['xyz'], msg=pattern) pattern = pattern.encode() self.assertEqual(re.compile(pattern, re.S).findall(b'xyz'), [b'xyz'], msg=pattern) def test_bug_2537(self): # issue 2537: empty submatches for outer_op in ('{0,}', '*', '+', '{1,187}'): for inner_op in ('{0,}', '*', '?'): r = re.compile("^((x|y)%s)%s" % (inner_op, outer_op)) m = r.match("xyyzy") self.assertEqual(m.group(0), "xyy") self.assertEqual(m.group(1), "") self.assertEqual(m.group(2), "y") # PCRE: python-pcre doesn't support DEBUG flag def __test_debug_flag(self): with captured_stdout() as out: re.compile('foo', re.DEBUG) self.assertEqual(out.getvalue().splitlines(), ['literal 102', 'literal 111', 'literal 111']) # Debug output is output again even a second time (bypassing # the cache -- issue #20426). with captured_stdout() as out: re.compile('foo', re.DEBUG) self.assertEqual(out.getvalue().splitlines(), ['literal 102', 'literal 111', 'literal 111']) def test_keyword_parameters(self): # Issue #20283: Accepting the string keyword parameter. pat = re.compile(r'(ab)') self.assertEqual( pat.match(string='abracadabra', pos=7, endpos=10).span(), (7, 9)) self.assertEqual( pat.search(string='abracadabra', pos=3, endpos=10).span(), (7, 9)) self.assertEqual( pat.findall(string='abracadabra', pos=3, endpos=10), ['ab']) self.assertEqual( pat.split(string='abracadabra', maxsplit=1), ['', 'ab', 'racadabra']) def run_re_tests(): # PCRE: use pcre_tests instead of re_tests #from test.re_tests import tests, SUCCEED, FAIL, SYNTAX_ERROR from test.pcre_tests import tests, SUCCEED, FAIL, SYNTAX_ERROR if verbose: #print 'Running re_tests test suite' print 'Running pcre_tests test suite' else: # To save time, only run the first and last 10 tests #tests = tests[:10] + tests[-10:] pass for t in tests: sys.stdout.flush() pattern = s = outcome = repl = expected = None if len(t) == 5: pattern, s, outcome, repl, expected = t elif len(t) == 3: pattern, s, outcome = t else: raise ValueError, ('Test tuples should have 3 or 5 fields', t) try: obj = re.compile(pattern) except re.error: if outcome == SYNTAX_ERROR: pass # Expected a syntax error else: print '=== Syntax error:', t except KeyboardInterrupt: raise KeyboardInterrupt except: print '*** Unexpected error ***', t if verbose: traceback.print_exc(file=sys.stdout) else: try: result = obj.search(s) except re.error, msg: print '=== Unexpected exception', t, repr(msg) if outcome == SYNTAX_ERROR: # This should have been a syntax error; forget it. pass elif outcome == FAIL: if result is None: pass # No match, as expected else: print '=== Succeeded incorrectly', t elif outcome == SUCCEED: if result is not None: # Matched, as expected, so now we compute the # result string and compare it to our expected result. start, end = result.span(0) vardict={'found': result.group(0), 'groups': result.group(), 'flags': result.re.flags} for i in range(1, 100): try: gi = result.group(i) # Special hack because else the string concat fails: if gi is None: gi = "None" except IndexError: gi = "Error" vardict['g%d' % i] = gi for i in result.re.groupindex.keys(): try: gi = result.group(i) if gi is None: gi = "None" except IndexError: gi = "Error" vardict[i] = gi repl = eval(repl, vardict) if repl != expected: print '=== grouping error', t, print repr(repl) + ' should be ' + repr(expected) else: print '=== Failed incorrectly', t # Try the match on a unicode string, and check that it # still succeeds. try: result = obj.search(unicode(s, "latin-1")) if result is None: print '=== Fails on unicode match', t except NameError: continue # 1.5.2 except TypeError: continue # unicode test case # Try the match on a unicode pattern, and check that it # still succeeds. obj=re.compile(unicode(pattern, "latin-1")) result = obj.search(s) if result is None: print '=== Fails on unicode pattern match', t # Try the match with the search area limited to the extent # of the match and see if it still succeeds. \B will # break (because it won't match at the end or start of a # string), so we'll ignore patterns that feature it. if pattern[:2] != '\\B' and pattern[-2:] != '\\B' \ and result is not None: obj = re.compile(pattern) result = obj.search(s, result.start(0), result.end(0) + 1) if result is None: print '=== Failed on range-limited match', t # Try the match with IGNORECASE enabled, and check that it # still succeeds. obj = re.compile(pattern, re.IGNORECASE) result = obj.search(s) if result is None: print '=== Fails on case-insensitive match', t # Try the match with LOCALE enabled, and check that it # still succeeds. # PCRE: python-pcre doesn't support LOCALE flag #obj = re.compile(pattern, re.LOCALE) #result = obj.search(s) #if result is None: # print '=== Fails on locale-sensitive match', t # Try the match with UNICODE locale enabled, and check # that it still succeeds. obj = re.compile(pattern, re.UNICODE) result = obj.search(s) if result is None: print '=== Fails on unicode-sensitive match', t def test_main(): run_unittest(ReTests) run_re_tests() if __name__ == "__main__": test_main()