Python pyparsing.Group() Examples

def parse(input_string):
        word = Word(alphanums)
        command = Group(OneOrMore(word))
        token = Suppress("->")
        device = Group(OneOrMore(word))
        argument = Group(OneOrMore(word))
        event = command + token + device + Optional(token + argument)
        parse_results = event.parseString(input_string)
        cmd = parse_results[0]
        dev = parse_results[1]
        cmd_str = "_".join(cmd)
        dev_str = "_".join(dev)
        try:
            val = parse_results[2]
        except IndexError:
            val_str = None
        else:
            val_str = "_".join(val)
        return cmd_str, dev_str, val_str 

def _parse_filter():
        op = pyparsing.oneOf('! & |')
        lpar  = pyparsing.Literal('(').suppress()
        rpar  = pyparsing.Literal(')').suppress()

        k = pyparsing.Word(pyparsing.alphanums)
        # NOTE: We may need to expand on this list, but as this is not a real
        # LDAP server we should be OK.
        # Value to contain:
        #   numbers, upper/lower case letters, astrisk, at symbol, minus, full
        #   stop, backslash or a space
        v = pyparsing.Word(pyparsing.alphanums + "-*@.\\ äöü")
        rel = pyparsing.oneOf("= ~= >= <=")

        expr = pyparsing.Forward()
        atom = pyparsing.Group(lpar + op + expr + rpar) \
                            | pyparsing.Combine(lpar + k + rel + v + rpar)
        expr << atom + pyparsing.ZeroOrMore( expr )

        return expr 

def getLogLineBNF():
    global logLineBNF
    
    if logLineBNF is None:
        integer = Word( nums )
        ipAddress = delimitedList( integer, ".", combine=True )
        
        timeZoneOffset = Word("+-",nums)
        month = Word(string.uppercase, string.lowercase, exact=3)
        serverDateTime = Group( Suppress("[") + 
                                Combine( integer + "/" + month + "/" + integer +
                                        ":" + integer + ":" + integer + ":" + integer ) +
                                timeZoneOffset + 
                                Suppress("]") )
                         
        logLineBNF = ( ipAddress.setResultsName("ipAddr") + 
                       Suppress("-") +
                       ("-" | Word( alphas+nums+"@._" )).setResultsName("auth") +
                       serverDateTime.setResultsName("timestamp") + 
                       dblQuotedString.setResultsName("cmd").setParseAction(getCmdFields) +
                       (integer | "-").setResultsName("statusCode") + 
                       (integer | "-").setResultsName("numBytesSent")  + 
                       dblQuotedString.setResultsName("referrer").setParseAction(removeQuotes) +
                       dblQuotedString.setResultsName("clientSfw").setParseAction(removeQuotes) )
    return logLineBNF 

def expression(self):
        expression = pyparsing.Forward()

        # (1 + (2 + 3))
        nested_expression = pyparsing.nestedExpr(
            "(", ")", expression).setParseAction(self._combine_lists)

        # FOO(2 , 3)
        function_call = (
            _TOKEN().setResultsName("function")
            + _OPEN_PARENTHESIS()
            + pyparsing.delimitedList(
                pyparsing.Combine(expression, adjacent=False, joinString=" "),
                delim=",").setResultsName("func_args")
            + _CLOSE_PARENTHESIS()
        )

        expression << pyparsing.OneOrMore(
            function_call.setParseAction(self._is_known_function)
            | pyparsing.Group(nested_expression)
            | _TOKEN()
            | _NOT_TOKEN()
        )

        return pyparsing.Combine(expression, adjacent=False, joinString=" ") 

def _parse(mystr):

    LBRACE, RBRACE, EQUAL = map(pp.Suppress, "{}=")
    field = pp.Word(pp.printables + ' ', excludeChars='[]=')
    field.addParseAction(pp.tokenMap(str.rstrip))
    string = pp.dblQuotedString().setParseAction(pp.removeQuotes)
    number = pp.pyparsing_common.number()
    date_expr = pp.Regex(r'\d\d\d\d-\d\d-\d\d')
    time_expr = pp.Regex(r'\d\d:\d\d:\d\d\.\d\d\d')
    nan = pp.Keyword('nan')
    scalar_value = (string | date_expr | time_expr | number | nan)

    list_marker = pp.Suppress("[]")
    value_list = pp.Forward()
    jobject = pp.Forward()

    memberDef1 = pp.Group(field + EQUAL + scalar_value)
    memberDef2 = pp.Group(field + EQUAL + jobject)
    memberDef3 = pp.Group(field + list_marker + EQUAL + LBRACE + value_list +
                          RBRACE)
    memberDef = memberDef1 | memberDef2 | memberDef3

    value_list <<= (pp.delimitedList(scalar_value, ",") |
                    pp.ZeroOrMore(pp.Group(pp.Dict(memberDef2))))
    value_list.setParseAction(lambda t: [pp.ParseResults(t[:])])

    members = pp.OneOrMore(memberDef)
    jobject <<= pp.Dict(LBRACE + pp.ZeroOrMore(memberDef) + RBRACE)
    # force empty jobject to be a dict
    jobject.setParseAction(lambda t: t or {})

    parser = members
    parser = pp.OneOrMore(pp.Group(pp.Dict(memberDef)))

    return parser.parseString(mystr) 

def _apply_modifiers(rule, modifiers):
        if 'grouped' in modifiers:
            rule = pp.Group(rule)

        if 'optional' in modifiers:
            rule = pp.Optional(rule)
        else:
            for modifier in modifiers:
                if modifier.startswith('at_least'):
                    times = rule_at_least.parseString(modifier)[0]
                    if times > 0:
                        rule_multiple = rule
                        for _ in range(1, times):
                            rule_multiple = rule_multiple + rule
                        rule = rule_multiple + pp.ZeroOrMore(rule)
                    else:
                        rule = pp.Optional(pp.ZeroOrMore(rule))

        return rule 

def audio_visual_key(name=None):
    """
    Creates the grammar for an Audio Visual Key code.

    This is a variation on the ISAN (International Standard Audiovisual Number)

    :param name: name for the field
    :return: grammar for an ISRC field
    """

    if name is None:
        name = 'AVI Field'

    society_code = basic.numeric(3)
    society_code = society_code.setName('Society Code') \
        .setResultsName('society_code')

    av_number = basic.alphanum(15, extended=True, isLast=True)
    field_empty = pp.Regex('[ ]{15}')
    field_empty.setParseAction(pp.replaceWith(''))
    av_number = av_number | field_empty
    av_number = av_number.setName('Audio-Visual Number') \
        .setResultsName('av_number')

    field = pp.Group(society_code + pp.Optional(av_number))

    field.setParseAction(lambda v: _to_avi(v[0]))

    field = field.setName(name)

    return field.setResultsName('audio_visual_key') 

def __init__(self):
        if PYPARSING:
            category = Word(alphas + "_-*", alphanums + "_-*")
            operator = oneOf("and or ,")
            neg_operator = "not"
            elementRef = category
            definition = elementRef + ZeroOrMore(operator + elementRef)
            nestedformula = Group(Suppress(Optional(Literal("("))) + definition + Suppress(Optional(Literal(")"))))
            neg_nestedformula = Optional(neg_operator) + nestedformula
            self.finalformula = neg_nestedformula + ZeroOrMore(operator + neg_nestedformula)

            elementRef.setParseAction(self.__compute_element)
            neg_nestedformula.setParseAction(self.__compute_neg_formula)
            nestedformula.setParseAction(self.__compute_formula)
            self.finalformula.setParseAction(self.__myreduce) 

def todict(cls, parseresult, mode='parens', ns={}):
        """
        Helper function to return dictionary given the parse results
        from a pyparsing.nestedExpr object (containing keywords).

        The ns is a dynamic namespace (typically the IPython Notebook
        namespace) used to update the class-level namespace.
        """
        grouped, kwargs = [], {}
        tokens = cls.collect_tokens(parseresult, mode)
        # Group tokens without '=' and append to last token containing '='
        for group in groupby(tokens, lambda el: '=' in el):
            (val, items) = group
            if val is True:
                grouped += list(items)
            if val is False:
                elements =list(items)
                # Assume anything before ) or } can be joined with commas
                # (e.g tuples with spaces in them)
                joiner=',' if any(((')' in el) or ('}' in el))
                                  for el in elements) else ''
                grouped[-1] += joiner + joiner.join(elements)

        for keyword in grouped:
            # Tuple ('a', 3) becomes (,'a',3) and '(,' is never valid
            # Same for some of the other joining errors corrected here
            for (fst,snd) in [('(,', '('), ('{,', '{'), ('=,','='),
                              (',:',':'), (':,', ':'), (',,', ','),
                              (',.', '.')]:
                keyword = keyword.replace(fst, snd)
            try:
                kwargs.update(eval('dict(%s)' % keyword,
                                   dict(cls.namespace, **ns)))
            except:
                if cls.abort_on_eval_failure:
                    raise SyntaxError("Could not evaluate keyword: %r"
                                      % keyword)
                msg = "Ignoring keyword pair that fails to evaluate: '%s'"
                parsewarning.warning(msg % keyword)

        return kwargs 

def extract_into_diagram(self, el_id: int):
        """
        Used when we encounter the same token twice in the same tree. When this happens, we replace all instances of that
        token with a terminal, and create a new subdiagram for the token
        """
        position = self.first[el_id]

        # Replace the original definition of this element with a regular block
        if position.parent:
            ret = EditablePartial.from_call(railroad.NonTerminal, text=position.name)
            if "item" in position.parent.kwargs:
                position.parent.kwargs["item"] = ret
            else:
                position.parent.kwargs["items"][position.parent_index] = ret

        # If the element we're extracting is a group, skip to its content but keep the title
        if position.converted.func == railroad.Group:
            content = position.converted.kwargs["item"]
        else:
            content = position.converted

        self.diagrams[el_id] = EditablePartial.from_call(
            NamedDiagram,
            name=position.name,
            diagram=EditablePartial.from_call(
                railroad.Diagram, content, **self.diagram_kwargs
            ),
            index=position.number,
        )
        del self.first[el_id] 

def table_row(start_tag, end_tag):
    body = start_tag.tag_body
    body.addParseAction(pp.tokenMap(str.strip), pp.tokenMap(strip_html))
    row = pp.Group(
        tr.suppress()
        + pp.ZeroOrMore(start_tag.suppress() + body + end_tag.suppress())
        + tr_end.suppress()
    )
    return row 

def _base_or_array_expression(self):
        array_indices = pyparsing.ZeroOrMore(
            _OPEN_BRACKETS
            + self.expression
            + _CLOSE_BRACKETS
        )
        return (
            self._base_expression()
            + pyparsing.Group(array_indices)
        ).setParseAction(self._create_base_or_array_expression) 

def date_time(name=None):
    """
    Creates the grammar for a date and time field, which is a combination of
    the Date (D) and Time or Duration field (T).

    This field requires first a Date, and then a Time, without any space in
    between.

    :param name: name for the field
    :return: grammar for a Date and Time field
    """
    if name is None:
        name = 'Date and Time Field'

    date = basic.date('Date')
    time = basic.time('Time')

    date = date.setResultsName('date')
    time = time.setResultsName('time')

    field = pp.Group(date + time)

    field.setParseAction(lambda d: _combine_date_time(d[0]))

    field.setName(name)

    return field.setResultsName('date_time') 

def _nested_scopes(opening, closing, body):
  """Constructs a parser for (possibly nested) scopes."""
  scope = pp.Forward()
  scope << pp.Group(  # pylint: disable=expression-not-assigned
      opening +
      pp.ZeroOrMore(body | scope)("members") +
      closing)
  return scope 

def _nested_if_else(if_, pred, else_, endif, match_if_true, match_if_false):
  """Constructs a parser for (possibly nested) if...(else)...endif blocks."""
  ifelse = pp.Forward()
  ifelse << pp.Group(  # pylint: disable=expression-not-assigned
      if_ +
      pred("predicate") +
      pp.ZeroOrMore(match_if_true | ifelse)("if_true") +
      pp.Optional(else_ +
                  pp.ZeroOrMore(match_if_false | ifelse)("if_false")) +
      endif)
  return ifelse


# Some common string patterns to suppress.
# ------------------------------------------------------------------------------ 

def getchunk():
    """
    Using pyparsing, create chunk reader for chunk strings.
    """
    slot = pp.Word("".join([pp.alphas, "_"]), "".join([pp.alphanums, "_"]))
    special_value = pp.Group(pp.oneOf([ACTRVARIABLE, "".join([ACTRNEG, ACTRVARIABLE]), ACTRNEG, VISIONGREATER, VISIONSMALLER, "".join([VISIONGREATER, ACTRVARIABLE]), "".join([VISIONSMALLER, ACTRVARIABLE])])\
            + pp.Word("".join([pp.alphanums, "_", '"', "'"])))
    strvalue = pp.QuotedString('"', unquoteResults=False)
    strvalue2 = pp.QuotedString("'", unquoteResults=False)
    varvalue = pp.Word("".join([pp.alphanums, "_"]))
    value = varvalue | special_value | strvalue | strvalue2
    chunk_reader = pp.OneOrMore(pp.Group(slot + value))
    return chunk_reader 

def getrule():
    """
    Using pyparsing, get rule out of a string.
    """
    arrow = pp.Literal("==>")
    buff = pp.Word(pp.alphas, "".join([pp.alphanums, "_"]))
    special_valueLHS = pp.oneOf([x for x in _LHSCONVENTIONS.keys()])
    end_buffer = pp.Literal(">")
    special_valueRHS = pp.oneOf([x for x in _RHSCONVENTIONS.keys()])
    chunk = getchunk()
    rule_reader = pp.Group(pp.OneOrMore(pp.Group(special_valueLHS + buff + end_buffer + pp.Group(pp.Optional(chunk))))) + arrow + pp.Group(pp.OneOrMore(pp.Group(special_valueRHS + buff + end_buffer + pp.Group(pp.Optional(chunk)))))
    return rule_reader 

def __init__(self, cfg_mgr, pre_defined_actions, scripts_dir):
        """init
        Args:
            cfg_mgr: Instance of ibis.utilities.config_manager.ConfigManager
            pre_defined_actions: list of default action ids
        """
        self.cfg_mgr = cfg_mgr
        self.logger = get_logger(self.cfg_mgr)
        self.pre_defined_actions = pre_defined_actions
        header_pattern = pp.Keyword('action').setResultsName('action_header')
        body_pattern = pp.Word(pp.alphanums + '._/')
        body_pattern = body_pattern.setResultsName('action_id')
        self.pattern = header_pattern + \
            pp.Group(body_pattern).setResultsName('action_body')
        self.scripts_dir = scripts_dir 

def parse_table(attribute, string):
    Line = OneOrMore(Float)('data') + Literal(';') + Optional(Comments, default='')('name')
    Grammar = Suppress(Keyword('mpc.{}'.format(attribute)) + Keyword('=') + Keyword('[') + Optional(Comments)) + OneOrMore(Group(Line)) + Suppress(Keyword(']') + Optional(Comments))

    result, i, j = Grammar.scanString(string).next()

    _list = list()
    for r in result:
        _list.append([int_else_float_except_string(s) for s in r['data'].asList()])

    return _list 

def _arguments(self):
        return pyparsing.Group(
            pyparsing.Optional(pyparsing.delimitedList(self._argument()))
        ) 

def _multiword_argument(self):
        return pyparsing.Group(
            self._variable()
            + pyparsing.OneOrMore(self._variable())
        ).setParseAction(util.action(pre_ast.CompositeBlock)) 

def get_location_language(identifier: ParserElement) -> ParserElement:
    """Build a location parser."""
    return Group(location_tag + nest(identifier))(LOCATION) 

def _enum_definition(self):
        """Detect an enum definition.

        e.g.
             enum foo {
                OPTION_1: 1 + 2,
                OPTION_2
             }
        """
        return (
            _ENUM
            + pyparsing.Optional(self._identifier())("enum_name")
            + _OPEN_CURLY
            + pyparsing.ZeroOrMore(
                pyparsing.Group(
                    self._identifier()("name")
                    + pyparsing.Optional(
                        _EQUALS
                        # This allows us to get even invalid expressions.
                        + pyparsing.SkipTo(pyparsing.Word(",}"))("expression")
                    )
                    + pyparsing.Optional(_COMMA)
                )
            )("fields")
            + _CLOSE_CURLY
            + self._maybe_attributes()("attributes")
        ).setParseAction(self._process_enum_definition) 

def _arguments(self):
        return pyparsing.Group(
            pyparsing.Optional(
                pyparsing.delimitedList(self.expression()))) 

def get_gene_modification_language(concept_qualified: ParserElement) -> ParserElement:
    """Build a gene modification parser."""
    concept = MatchFirst([
        concept_qualified,
        gmod_default_ns,
    ])
    return gmod_tag + nest(Group(concept)(CONCEPT)) 

def meta_section():
    return pyparsing.Group(
        pyparsing.Literal("meta") +
        _COLON +
        pyparsing.OneOrMore(
            statement()
        ).setResultsName("statements")
    ).setResultsName("meta") 

def statement():
    return pyparsing.Group(
        _IDENTIFIER.setResultsName("lhs") + _EQUALS +
        pyparsing.Combine(
            (anything_in_curly() |
             pyparsing.QuotedString("'", escChar="\\", unquoteResults=False) |
             pyparsing.QuotedString("\"", escChar="\\", unquoteResults=False) |
             _REGEX) +
            pyparsing.ZeroOrMore(_KEYWORD),
            adjacent=False,
            joinString=" ",
        ).setResultsName("rhs")
    ) 

def strings_section():
    return pyparsing.Group(
        pyparsing.Literal("strings") +
        _COLON +
        pyparsing.OneOrMore(statement()).setResultsName("statements")
    ).setResultsName("strings") 

def jsParse(inStr):
	# This disaster is a context-free grammar parser for parsing javascript object literals.
	# It needs to be able to handle a lot of the definitional messes you find in in-the-wild
	# javascript object literals.
	# Unfortunately, Javascript is /way/ more tolerant then JSON when it comes to object literals
	# so we can't just parse objects using python's `json` library.

	TRUE = pp.Keyword("true").setParseAction( pp.replaceWith(True) )
	FALSE = pp.Keyword("false").setParseAction( pp.replaceWith(False) )
	NULL = pp.Keyword("null").setParseAction( pp.replaceWith(None) )

	jsonString = pp.quotedString.setParseAction( pp.removeQuotes )
	jsonNumber = pp.Combine( pp.Optional('-') + ( '0' | pp.Word('123456789',pp.nums) ) +
											pp.Optional( '.' + pp.Word(pp.nums) ) +
											pp.Optional( pp.Word('eE',exact=1) + pp.Word(pp.nums+'+-',pp.nums) ) )

	jsonObject   = pp.Forward()
	jsonValue    = pp.Forward()
	jsonDict     = pp.Forward()
	jsonArray    = pp.Forward()
	jsonElements = pp.Forward()

	rawText      = pp.Regex('[a-zA-Z_$][0-9a-zA-Z_$]*')

	commaToNull = pp.Word(',,', exact=1).setParseAction(pp.replaceWith(None))
	jsonElements << pp.ZeroOrMore(commaToNull) + pp.Optional(jsonObject) + pp.ZeroOrMore((pp.Suppress(',') + jsonObject) | commaToNull)

	jsonValue << ( jsonString | jsonNumber | TRUE | FALSE | NULL )


	dictMembers = pp.delimitedList( pp.Group( (rawText | jsonString) + pp.Suppress(':') + (jsonValue | jsonDict | jsonArray)))
	jsonDict << ( pp.Dict( pp.Suppress('{') + pp.Optional(dictMembers) + pp.ZeroOrMore(pp.Suppress(',')) + pp.Suppress('}') ) )
	jsonArray << ( pp.Group(pp.Suppress('[') + pp.Optional(jsonElements) + pp.Suppress(']') ) )
	jsonObject << (jsonValue | jsonDict | jsonArray)

	jsonComment = pp.cppStyleComment
	jsonObject.ignore( jsonComment )

	def convertDict(s, l, toks):

		return dict(toks.asList())

	def convertNumbers(s,l,toks):
		n = toks[0]
		try:
			return int(n)
		except ValueError:
			return float(n)

	jsonNumber.setParseAction(convertNumbers)
	jsonDict.setParseAction(convertDict)

	# jsonObject.setDebug()
	jsonObject.parseString('"inStr"').pop()
	return jsonObject.parseString(inStr).pop()


# Stolen from http://stackoverflow.com/a/12017573/268006 

def _type_instance(self):
        """A type declaration.

        The modifiers of a typedef:
                struct s *P[];
                         ^^^^<-    The type instance.
        """
        type_instance = (
            # Function pointer     (*f)(int foobar)
            pyparsing.ZeroOrMore(_STAR)
            + _OPEN_PARENTHESIS
            + pyparsing.Optional(_STAR("function_pointer"))
            + self._identifier()("type_instance_name")
            + _CLOSE_PARENTHESIS
            + parsers.anything_in_parentheses()("function_args")
        ) | (
            # Function object  f(foo bar *)
            pyparsing.ZeroOrMore(_STAR)
            + self._identifier()("type_instance_name")
            + parsers.anything_in_parentheses()("function_args")
        ) | (
            # Simple form: *foo[10];
            pyparsing.ZeroOrMore(_STAR)("type_pointer")
            + self._identifier()("type_instance_name")

            # Possibly array: [] , [][]
            + pyparsing.ZeroOrMore(
                _OPEN_BRACKET
                + pyparsing.SkipTo(_CLOSE_BRACKET)(
                    "brackets_with_expression_inside*")
                + _CLOSE_BRACKET)

            # Bitfields:    int x: 7;
            + pyparsing.Optional(
                _COLON
                + pyparsing.SkipTo(
                    _SEMICOLON | _COMMA)("bitfield")
            )
        )

        return pyparsing.Group(
            type_instance
            + self._maybe_attributes()
        )