test_sre.py

mallet是自然语言处理、机器学习领域的一个开源项目。

test(r"""pat.match('a').group(1, 2, 3)""", ('a', None, None))
test(r"""pat.match('b').group('a1', 'b2', 'c3')""", (None, 'b', None))
test(r"""pat.match('ac').group(1, 'b2', 3)""", ('a', None, 'c'))

# bug 448951 (similar to 429357, but with single char match)
# (Also test greedy matches.)
for op in '','?','*':
    test(r"""sre.match(r'((.%s):)?z', 'z').groups()"""%op, (None, None))
    test(r"""sre.match(r'((.%s):)?z', 'a:z').groups()"""%op, ('a:', 'a'))

if verbose:
    print "Running tests on sre.escape"

p = ""
for i in range(0, 256):
    p = p + chr(i)
    test(r"""sre.match(sre.escape(chr(i)), chr(i)) is not None""", 1)
    test(r"""sre.match(sre.escape(chr(i)), chr(i)).span()""", (0,1))

pat = sre.compile(sre.escape(p))
test(r"""pat.match(p) is not None""", 1)
test(r"""pat.match(p).span()""", (0,256))

if verbose:
    print 'Running tests on sre.Scanner'

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 = sre.Scanner([
    (r"[a-zA-Z_]\w*", s_ident),
    (r"\d+\.\d*", s_float),
    (r"\d+", s_int),
    (r"=|\+|-|\*|/", s_operator),
    (r"\s+", None),
    ])

# sanity check
test('scanner.scan("sum = 3*foo + 312.50 + bar")',
     (['sum', 'op=', 3, 'op*', 'foo', 'op+', 312.5, 'op+', 'bar'], ''))

if verbose:
    print 'Pickling a SRE_Pattern instance'

try:
    import pickle
    pat = sre.compile(r'a(?:b|(c|e){1,2}?|d)+?(.)')
    s = pickle.dumps(pat)
    pat = pickle.loads(s)
except:
    print TestFailed, 're module pickle'

try:
    import cPickle
    pat = sre.compile(r'a(?:b|(c|e){1,2}?|d)+?(.)')
    s = cPickle.dumps(pat)
    pat = cPickle.loads(s)
except:
    print TestFailed, 're module cPickle'

# constants
test(r"""sre.I""", sre.IGNORECASE)
test(r"""sre.L""", sre.LOCALE)
test(r"""sre.M""", sre.MULTILINE)
test(r"""sre.S""", sre.DOTALL)
test(r"""sre.X""", sre.VERBOSE)
test(r"""sre.T""", sre.TEMPLATE)
test(r"""sre.U""", sre.UNICODE)

for flags in [sre.I, sre.M, sre.X, sre.S, sre.L, sre.T, sre.U]:
    try:
        r = sre.compile('^pattern$', flags)
    except:
        print 'Exception raised on flag', flags

if verbose:
    print 'Test engine limitations'

# Try nasty case that overflows the straightforward recursive
# implementation of repeated groups.
test("sre.match('(x)*', 50000*'x').span()", (0, 50000), RuntimeError)
test("sre.match(r'(x)*y', 50000*'x'+'y').span()", (0, 50001), RuntimeError)
test("sre.match(r'(x)*?y', 50000*'x'+'y').span()", (0, 50001), RuntimeError)

from re_tests import *

if verbose:
    print 'Running re_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=sre.compile(pattern)
    except sre.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 (sre.error), msg:
            print '=== Unexpected exception', t, repr(msg)
        if outcome==SYNTAX_ERROR:
            print '=== Compiled incorrectly', t
        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
                continue

            # Try the match on a unicode string, and check that it
            # still succeeds.
            try:
                u = unicode(s, "latin-1")
            except NameError:
                pass
            except TypeError:
                continue # skip unicode test strings
            else:
                result=obj.search(u)
                if result==None:
                    print '=== Fails on unicode match', t

            # Try the match on a unicode pattern, and check that it
            # still succeeds.
            try:
                u = unicode(pattern, "latin-1")
            except NameError:
                pass
            else:
                obj=sre.compile(u)
                result=obj.search(s)
                if result==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':
                obj=sre.compile(pattern)
                result=obj.search(s, result.start(0), result.end(0)+1)
                if result==None:
                    print '=== Failed on range-limited match', t

            # Try the match with IGNORECASE enabled, and check that it
            # still succeeds.
            obj=sre.compile(pattern, sre.IGNORECASE)
            result=obj.search(s)
            if result==None:
                print '=== Fails on case-insensitive match', t

            # Try the match with LOCALE enabled, and check that it
            # still succeeds.
            obj=sre.compile(pattern, sre.LOCALE)
            result=obj.search(s)
            if result==None:
                print '=== Fails on locale-sensitive match', t

            # Try the match with UNICODE locale enabled, and check
            # that it still succeeds.
            if have_unicode:
                obj=sre.compile(pattern, sre.UNICODE)
                result=obj.search(s)
                if result==None:
                    print '=== Fails on unicode-sensitive match', t