Python _sre.compile() Examples
The following are 30
code examples of _sre.compile().
You can vote up the ones you like or vote down the ones you don't like,
and go to the original project or source file by following the links above each example.
You may also want to check out all available functions/classes of the module
_sre
, or try the search function
.
.
Example #1
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 6 votes |
|
def test_re_groupref_exists(self):
self.assertEqual(re.match('^(\()?([^()]+)(?(1)\))$', '(a)').groups(),
('(', 'a'))
self.assertEqual(re.match('^(\()?([^()]+)(?(1)\))$', 'a').groups(),
(None, 'a'))
self.assertIsNone(re.match('^(\()?([^()]+)(?(1)\))$', 'a)'))
self.assertIsNone(re.match('^(\()?([^()]+)(?(1)\))$', '(a'))
self.assertEqual(re.match('^(?:(a)|c)((?(1)b|d))$', '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))$', 'a').groups(),
('a', ''))
# Tests for bug #1177831: exercise groups other than the first group
p = re.compile('(?P<g1>a)(?P<g2>b)?((?(g2)c|d))')
self.assertEqual(p.match('abc').groups(),
('a', 'b', 'c'))
self.assertEqual(p.match('ad').groups(),
('a', None, 'd'))
self.assertIsNone(p.match('abd'))
self.assertIsNone(p.match('ac'))
Example #2
| Source File: test_re.py From BinderFilter with MIT License | 6 votes |
|
def test_symbolic_groups(self):
re.compile('(?P<a>x)(?P=a)(?(a)y)')
re.compile('(?P<a1>x)(?P=a1)(?(a1)y)')
self.assertRaises(re.error, re.compile, '(?P<a>)(?P<a>)')
self.assertRaises(re.error, re.compile, '(?Px)')
self.assertRaises(re.error, re.compile, '(?P=)')
self.assertRaises(re.error, re.compile, '(?P=1)')
self.assertRaises(re.error, re.compile, '(?P=a)')
self.assertRaises(re.error, re.compile, '(?P=a1)')
self.assertRaises(re.error, re.compile, '(?P=a.)')
self.assertRaises(re.error, re.compile, '(?P<)')
self.assertRaises(re.error, re.compile, '(?P<>)')
self.assertRaises(re.error, re.compile, '(?P<1>)')
self.assertRaises(re.error, re.compile, '(?P<a.>)')
self.assertRaises(re.error, re.compile, '(?())')
self.assertRaises(re.error, re.compile, '(?(a))')
self.assertRaises(re.error, re.compile, '(?(1a))')
self.assertRaises(re.error, re.compile, '(?(a.))')
Example #3
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 6 votes |
|
def test_debug_flag(self):
pat = r'(\.)(?:[ch]|py)(?(1)$|: )'
with captured_stdout() as out:
re.compile(pat, re.DEBUG)
dump = '''\
subpattern 1
literal 46
subpattern None
branch
in
literal 99
literal 104
or
literal 112
literal 121
subpattern None
groupref_exists 1
at at_end
else
literal 58
literal 32
'''
self.assertEqual(out.getvalue(), dump)
# Debug output is output again even a second time (bypassing
# the cache -- issue #20426).
with captured_stdout() as out:
re.compile(pat, re.DEBUG)
self.assertEqual(out.getvalue(), dump)
Example #4
| Source File: test_re.py From BinderFilter with MIT License | 6 votes |
|
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('(?P<g1>a)(?P<g2>b)?((?(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)
Example #5
| Source File: test_re.py From jawfish with MIT License | 6 votes |
|
def test_bug_6509(self):
# Replacement strings of both types must parse properly.
# all strings
pat = re.compile('a(\w)')
self.assertEqual(pat.sub('b\\1', 'ac'), 'bc')
pat = re.compile('a(.)')
self.assertEqual(pat.sub('b\\1', 'a\u1234'), 'b\u1234')
pat = re.compile('..')
self.assertEqual(pat.sub(lambda m: 'str', 'a5'), 'str')
# all bytes
pat = re.compile(b'a(\w)')
self.assertEqual(pat.sub(b'b\\1', b'ac'), b'bc')
pat = re.compile(b'a(.)')
self.assertEqual(pat.sub(b'b\\1', b'a\xCD'), b'b\xCD')
pat = re.compile(b'..')
self.assertEqual(pat.sub(lambda m: b'bytes', b'a5'), b'bytes')
Example #6
| Source File: test_re.py From jawfish with MIT License | 6 votes |
|
def test_finditer(self):
iter = re.finditer(r":+", "a:b::c:::d")
self.assertEqual([item.group(0) for item in iter],
[":", "::", ":::"])
pat = re.compile(r":+")
iter = pat.finditer("a:b::c:::d", 1, 10)
self.assertEqual([item.group(0) for item in iter],
[":", "::", ":::"])
pat = re.compile(r":+")
iter = pat.finditer("a:b::c:::d", pos=1, endpos=10)
self.assertEqual([item.group(0) for item in iter],
[":", "::", ":::"])
pat = re.compile(r":+")
iter = pat.finditer("a:b::c:::d", endpos=10, pos=1)
self.assertEqual([item.group(0) for item in iter],
[":", "::", ":::"])
pat = re.compile(r":+")
iter = pat.finditer("a:b::c:::d", pos=3, endpos=8)
self.assertEqual([item.group(0) for item in iter],
["::", "::"])
Example #7
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 6 votes |
|
def test_symbolic_groups(self):
re.compile('(?P<a>x)(?P=a)(?(a)y)')
re.compile('(?P<a1>x)(?P=a1)(?(a1)y)')
self.assertRaises(re.error, re.compile, '(?P<a>)(?P<a>)')
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<a.>)')
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.))')
Example #8
| Source File: sre_compile.py From kobo-predict with BSD 2-Clause "Simplified" License | 6 votes |
|
def compile(p, flags=0):
# internal: convert pattern list to internal format
if isstring(p):
pattern = p
p = sre_parse.parse(p, flags)
else:
pattern = None
code = _code(p, flags)
# print(code)
# map in either direction
groupindex = p.pattern.groupdict
indexgroup = [None] * p.pattern.groups
for k, i in groupindex.items():
indexgroup[i] = k
return _sre.compile(
pattern, flags | p.pattern.flags, code,
p.pattern.groups-1,
groupindex, indexgroup
)
Example #9
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 6 votes |
|
def test_other_escapes(self):
self.assertRaises(re.error, re.compile, "\\")
self.assertEqual(re.match(r"\(", '(').group(), '(')
self.assertIsNone(re.match(r"\(", ')'))
self.assertEqual(re.match(r"\\", '\\').group(), '\\')
self.assertEqual(re.match(r"[\]]", ']').group(), ']')
self.assertIsNone(re.match(r"[\]]", '['))
self.assertEqual(re.match(r"[a\-c]", '-').group(), '-')
self.assertIsNone(re.match(r"[a\-c]", 'b'))
self.assertEqual(re.match(r"[\^a]+", 'a^').group(), 'a^')
self.assertIsNone(re.match(r"[\^a]+", 'b'))
re.purge() # for warnings
for c in 'ceghijklmopquyzCEFGHIJKLMNOPQRTUVXY':
warn = FutureWarning if c in 'Uu' else DeprecationWarning
with check_py3k_warnings(('', warn)):
self.assertEqual(re.match('\\%c$' % c, c).group(), c)
self.assertIsNone(re.match('\\%c' % c, 'a'))
for c in 'ceghijklmopquyzABCEFGHIJKLMNOPQRTUVXYZ':
warn = FutureWarning if c in 'Uu' else DeprecationWarning
with check_py3k_warnings(('', warn)):
self.assertEqual(re.match('[\\%c]$' % c, c).group(), c)
self.assertIsNone(re.match('[\\%c]' % c, 'a'))
Example #10
| Source File: test_re.py From jawfish with MIT License | 6 votes |
|
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('(?P<g1>a)(?P<g2>b)?((?(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)
Example #11
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
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")
Example #12
| Source File: test_re.py From BinderFilter with MIT License | 5 votes |
|
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)
p = re.compile('(?iu)' + upper_char)
q = p.match(lower_char)
self.assertNotEqual(q, None)
p = re.compile('(?iu)' + lower_char)
q = p.match(upper_char)
self.assertNotEqual(q, None)
Example #13
| Source File: sre_compile.py From kobo-predict with BSD 2-Clause "Simplified" License | 5 votes |
|
def _code(p, flags):
flags = p.pattern.flags | flags
code = []
# compile info block
_compile_info(code, p, flags)
# compile the pattern
_compile(code, p.data, flags)
code.append(SUCCESS)
return code
Example #14
| Source File: test_re.py From BinderFilter with MIT License | 5 votes |
|
def test_bug_764548(self):
# bug 764548, re.compile() barfs on str/unicode subclasses
try:
unicode
except NameError:
return # no problem if we have no unicode
class my_unicode(unicode): pass
pat = re.compile(my_unicode("abc"))
self.assertEqual(pat.match("xyz"), None)
Example #15
| Source File: sre_compile.py From kobo-predict with BSD 2-Clause "Simplified" License | 5 votes |
|
def _compile_charset(charset, flags, code, fixup=None, fixes=None):
# compile charset subprogram
emit = code.append
for op, av in _optimize_charset(charset, fixup, fixes):
emit(op)
if op is NEGATE:
pass
elif op is LITERAL:
emit(av)
elif op is RANGE or op is RANGE_IGNORE:
emit(av[0])
emit(av[1])
elif op is CHARSET:
code.extend(av)
elif op is BIGCHARSET:
code.extend(av)
elif op is CATEGORY:
if flags & SRE_FLAG_LOCALE:
emit(CH_LOCALE[av])
elif flags & SRE_FLAG_UNICODE:
emit(CH_UNICODE[av])
else:
emit(av)
else:
raise error("internal: unsupported set operator %r" % (op,))
emit(FAILURE)
Example #16
| Source File: test_re.py From BinderFilter with MIT License | 5 votes |
|
def test_bug_926075(self):
try:
unicode
except NameError:
return # no problem if we have no unicode
self.assertTrue(re.compile('bug_926075') is not
re.compile(eval("u'bug_926075'")))
Example #17
| Source File: test_re.py From BinderFilter with MIT License | 5 votes |
|
def test_bug_931848(self):
try:
unicode
except NameError:
pass
pattern = eval('u"[\u002E\u3002\uFF0E\uFF61]"')
self.assertEqual(re.compile(pattern).split("a.b.c"),
['a','b','c'])
Example #18
| Source File: test_re.py From BinderFilter with MIT License | 5 votes |
|
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(), ())
Example #19
| Source File: test_re.py From BinderFilter with MIT License | 5 votes |
|
def test_dealloc(self):
# 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
self.assertRaises(TypeError, re.finditer, "a", {})
self.assertRaises(OverflowError, _sre.compile, "abc", 0, [long_overflow])
Example #20
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
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'])
Example #21
| Source File: test_re.py From BinderFilter with MIT License | 5 votes |
|
def test_weakref(self):
s = 'QabbbcR'
x = re.compile('ab+c')
y = proxy(x)
self.assertEqual(x.findall('QabbbcR'), y.findall('QabbbcR'))
Example #22
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
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<?foo>)')
Example #23
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
def test_backref_group_name_in_exception(self):
# Issue 17341: Poor error message when compiling invalid regex
with self.assertRaisesRegexp(sre_constants.error, '<foo>'):
re.compile('(?P=<foo>)')
Example #24
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
def test_repeat_minmax_overflow_maxrepeat(self):
try:
from _sre import MAXREPEAT
except ImportError:
self.skipTest('requires _sre.MAXREPEAT constant')
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)
Example #25
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
def test_repeat_minmax_overflow(self):
# Issue #13169
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))
Example #26
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
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)
Example #27
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
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#')
Example #28
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
def test_empty_array(self):
# SF buf 1647541
import array
typecodes = 'cbBhHiIlLfd'
if have_unicode:
typecodes += 'u'
for typecode in typecodes:
a = array.array(typecode)
self.assertIsNone(re.compile("bla").match(a))
self.assertEqual(re.compile("").match(a).groups(), ())
Example #29
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
def test_bug_581080(self):
iter = re.finditer(r"\s", "a b")
self.assertEqual(iter.next().span(), (1,2))
self.assertRaises(StopIteration, iter.next)
scanner = re.compile(r"\s").scanner("a b")
self.assertEqual(scanner.search().span(), (1, 2))
self.assertIsNone(scanner.search())
Example #30
| Source File: test_re.py From ironpython2 with Apache License 2.0 | 5 votes |
|
def test_bug_931848(self):
pattern = u(r"[\u002E\u3002\uFF0E\uFF61]")
self.assertEqual(re.compile(pattern).split("a.b.c"),
['a','b','c'])
