Python sre_parse.parse函数代码示例

def test():
  import ClearMap.Utils.InverseRegularExpression as ire;
  import sre_parse as sre;

  reload(ire)  
  source = '/test/test_(?P<row>\d{4})_(?P<col>\d{3}).tif';
  p = sre.parse(source);
  ire.patternToExpression(p)

  reload(ire)  
  source = r'/test/test_(?:\d)_(?P<col>\d{3})_[7-9][.](?=col)tif$';
  p = sre.parse(source);
  ire.patternToExpression(p)

def charclass_runner(pat):
    r = Regex().get_parse_tree(pat)
    regexlint_version = r.children[0].matching_character_codes
    sre_parsed = sre_parse.parse(pat)
    print(sre_parsed)
    if isinstance(sre_parsed[0][1], int):
        sre_chars = sre_parsed
    else:
        sre_chars = sre_parsed[0][1]
    print('inner', sre_chars)
    golden = list(expand_sre_in(sre_chars))
    order_matters = True
    try:
        if (sre_parsed[0][0] == sre_constants.NOT_LITERAL or
            sre_parsed[0][1][0][0] == sre_constants.NEGATE):
            golden = [i for i in range(256) if i not in golden]
            order_matters = False

    except TypeError:
        pass

    print('sre_parse', golden)
    print('regexlint', regexlint_version)
    if order_matters:
        assert golden == regexlint_version
    else:
        print('extra:', sorted(set(regexlint_version) - set(golden)))
        print('missing:', sorted(set(golden) - set(regexlint_version)))

        assert sorted(golden) == sorted(regexlint_version)

def reverse_group_map(re_str):
    r = re.compile(re_str)
    ast = sre_parse.parse(re_str)
    group_indices = r.groupindex
    group_index_map = dict((index, group) 
                           for (group, index) in r.groupindex.items())
    return group_index_map

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

    # XXX: <fl> get rid of this limitation!
    if p.pattern.groups > 100:
        raise AssertionError(
            "sorry, but this version only supports 100 named groups"
        )

    # 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
    )

def isFileExpression(source):
    """Checks if filename is a regular expression denoting a file list
    
    Arguments:
        source (str): source file name
        
    Returns:
        bool: True if source is true regular expression with at least one non-literal
        
    Note:
        The any character '.' is not treated as a non-literal because of possible filename extensions
    """    
    
    if not isinstance(source, basestring):
        return False;
    
    if isFile(source):
        return False;
    else:
        #searchRegex = re.compile('.*\\\\d\{(?P<digit>\d)\}.*').search
        #m = searchRegex(source);
        #if m is None:
        #    return False;
        #else:
        #    return True;
        
        #parse regular expression 
        p = sre_parse.parse(source);
        for l in p:
          #note: allow for a filname.ext patterns although this is strictly a regular expression which should be denoted as filename\.ext
          if l[0] != 'literal' and l[0] != 'any':
            return True;
        
        return False;

def compile_regexp_to_noncapturing(pattern, flags=0):
    """
    Convert all grouping parentheses in the given regexp pattern to
    non-capturing groups, and return the result.  E.g.:

        >>> from nltk.internals import compile_regexp_to_noncapturing
        >>> compile_regexp_to_noncapturing('ab(c(x+)(z*))?d')
        'ab(?:c(?:x+)(?:z*))?d'

    :type pattern: str
    :rtype: str
    """
    def convert_regexp_to_noncapturing_parsed(parsed_pattern):
        res_data = []
        for key, value in parsed_pattern.data:
            if key == sre_constants.SUBPATTERN:
                index, subpattern = value
                value = (None, convert_regexp_to_noncapturing(subpattern))
            elif key == sre_constants.GROUPREF:
                raise ValueError('Regular expressions with back-references are not supported: {0}'.format(pattern))
            res_data.append((key, value))
        parsed_pattern.data = res_data
        parsed_pattern.pattern.groups = 1
        parsed_pattern.pattern.groupdict = {}
        return parsed_pattern

    return sre_compile.compile(convert_regexp_to_noncapturing_parsed(sre_parse.parse(pattern)))

def regex_score(regex, search_string):
    """
        Returns a closeness score of how well the regex matches the string.
        Will return -1 if it doesn't match.
    """

    match = re.search(regex, search_string)

    if match:
        # Base score is the longest distance between regex, match,
        # and search_string
        regex_match_dist = levenshtein_distance(
            regex.pattern.lower(), match.group(0).lower())
        match_string_dist = levenshtein_distance(
            match.group(0).lower(), search_string.lower())

        score = max(regex_match_dist, match_string_dist)

        # Adjust score: Special anchors slightly reduce distance
        for opcode, argument in sre_parse.parse(regex.pattern):
            if str(opcode) == 'AT':
                if str(argument) == 'AT_BEGINNING' or 'AT_END':
                    # ^ or $, adjust 1 edit
                    score -= 1

                if str(argument) == 'AT_BOUNDARY':
                    # all other anchors reduce 2 edits
                    score -= 2

        return score if score >= 0 else 0
    else:
        return -1

def re_replace_literals(text, mapping):
    """Raises NotImplementedError or re.error"""

    assert isinstance(text, unicode)

    pattern = sre_parse.parse(text)
    return _construct_regexp(pattern, mapping)

    def __init__(self, lexicon, flags=FLAGS):
        self.actions = [None]
        # combine phrases into a compound pattern
        s = sre_parse.Pattern()
        s.flags = flags
        p = []
        # NOTE(kgibbs): These lines must be added to make this file work under
        # Python 2.2, which is commonly used at Google.
        def enumerate(obj):
            i = -1
            for item in obj:
                i += 1
                yield i, item
        # NOTE(kgibbs): End changes.
        for idx, token in enumerate(lexicon):
            phrase = token.pattern
            try:
                subpattern = sre_parse.SubPattern(s,
                    [(SUBPATTERN, (idx + 1, sre_parse.parse(phrase, flags)))])
            except sre_constants.error:
                raise
            p.append(subpattern)
            self.actions.append(token)

        s.groups = len(p)+1  # NOTE(guido): Added to make SRE validation work
        p = sre_parse.SubPattern(s, [(BRANCH, (None, p))])
        self.scanner = sre_compile.compile(p)

def clean_pattern(pattern):
    """
    Cleans URL patterns
     * pattern => token
     * '2'     => ('literal', 50)
     * '2|3'   => ('in', [('literal', 50), ('literal', 51)])
    """
    star = '*'
    parsed = sre_parse.parse(pattern)
    literals = []

    for token in parsed:
        if token[0] == LITERAL:
            character = quote(unichr(token[1]).encode('utf8'))
            literals.append(character)
        elif token[0] == AT:
            pass

        elif literals[-1:] != [star]:
            literals.append(star)

    rule = '/' + ''.join(literals)

    if parsed and not rule.endswith(star):
        if parsed[-1] == (AT, AT_END):
            rule += '$'
        else:
            rule += star

    return rule

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)

    if flags & SRE_FLAG_DEBUG:
        print()
        dis(code)

    # map in either direction
    groupindex = p.state.groupdict
    indexgroup = [None] * p.state.groups
    for k, i in groupindex.items():
        indexgroup[i] = k

    return _sre.compile(
        pattern, flags | p.state.flags, code,
        p.state.groups-1,
        groupindex, tuple(indexgroup)
        )

def charclass_runner(pat):
    r = Regex().get_parse_tree(pat)
    regexlint_version = r.children[0].matching_character_codes
    sre_parsed = sre_parse.parse(pat)
    print sre_parsed
    if isinstance(sre_parsed[0][1], int):
        sre_chars = sre_parsed
    else:
        sre_chars = sre_parsed[0][1]
    golden = list(expand_sre_in(sre_chars))
    order_matters = True
    try:
        if (sre_parsed[0][0] == 'not_literal' or
            sre_parsed[0][1][0][0] == 'negate'):
            golden = [i for i in range(256) if i not in golden]
            order_matters = False
    except TypeError:
        pass

    print golden
    print regexlint_version
    if order_matters:
        assert golden == regexlint_version
    else:
        assert sorted(golden) == sorted(regexlint_version)

def compile(p, flags=0):
    # internal: convert pattern list to internal format

    if type(p) in STRING_TYPES:
        import sre_parse
        pattern = p
        p = sre_parse.parse(p, flags)
    else:
        pattern = None

    code = _code(p, flags)

    # print code

    # FIXME: <fl> get rid of this limitation!
    assert p.pattern.groups <= 100,\
           "sorry, but this version only supports 100 named groups"

    # 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, code,
        p.pattern.groups-1,
        groupindex, indexgroup
        )

def doReversing(p):
#	p = 'ab*de.gh+i{10}'
#	p = re.compile()
	dbg("Pattern:" + p)
	pattern = sre_parse.parse(p, 0) 
	out = reverse(pattern)
	return out

def regex_slice(expr, start, end):
    """
    Get a slice of a regex by calling regex_index on each index.

    Note that this can return expressions that are overly general: for example,
    it can mix characters from both branches of a regex. Being more specific
    than that would take more work.

        >>> regex_slice('test', 0, 1)
        't'
        >>> regex_slice('t?est', 0, 2)
        '[te][es]'
        >>> regex_slice('mo+', 3, 8)
        'ooooo'

    """
    if start < 0 or end < 0:
        raise NotImplementedError("Can't take negative slices of a regex yet")
    result = ''
    for index in range(start, end):
        choices = _regex_index_pattern(parse(expr), index)
        if len(choices) == 0:
            return None
        elif len(choices) == 1:
            regex = unparse(choices[0])
            result += regex
        else:
            regex = round_trip(unparse(('branch', (None, choices))))
            if '|' in regex:
                result += '(%s)' % (regex,)
            else:
                result += regex
    return result