Python pyparsing.Regex类代码示例

 def define_identifier(self):
     """
     Return the syntax definition for an identifier.
     
     """
     # --- Defining the individual identifiers:
     # Getting all the Unicode numbers in a single string:
     unicode_numbers = "".join([unichr(n) for n in xrange(0x10000)
                                if unichr(n).isdigit()])
     unicode_number_expr = Regex("[%s]" % unicode_numbers, re.UNICODE)
     space_char = re.escape(self._grammar.get_token("identifier_spacing"))
     identifier0 = Regex("[\w%s]+" % space_char, re.UNICODE)
     # Identifiers cannot start with a number:
     identifier0 = Combine(~unicode_number_expr + identifier0)
     identifier0.setName("individual_identifier")
     
     # --- Defining the namespaces:
     namespace_sep = Suppress(self._grammar.get_token("namespace_separator"))
     namespace = Group(ZeroOrMore(identifier0 + namespace_sep))
     namespace.setName("namespace")
     
     # --- The full identifier, which could have a namespace:
     identifier = Combine(namespace.setResultsName("namespace_parts") +
                          identifier0.setResultsName("identifier"))
     identifier.setName("full_identifier")
     
     return identifier

def nexus_iter(infile):
    import pyparsing
    pyparsing.ParserElement.enablePackrat()
    from pyparsing import Word, Literal, QuotedString, CaselessKeyword, \
         OneOrMore, Group, Optional, Suppress, Regex, Dict
    ## beginblock = Suppress(CaselessKeyword("begin") +
    ##                       CaselessKeyword("trees") + ";")
    ## endblock = Suppress((CaselessKeyword("end") |
    ##                      CaselessKeyword("endblock")) + ";")
    comment = Optional(Suppress("[&") + Regex(r'[^]]+') + Suppress("]"))
    ## translate = CaselessKeyword("translate").suppress()
    name = Word(string.letters+string.digits+"_.") | QuotedString("'")
    ## ttrec = Group(Word(string.digits).setResultsName("number") +
    ##               name.setResultsName("name") +
    ##               Optional(",").suppress())
    ## ttable = Group(translate + OneOrMore(ttrec) + Suppress(";"))
    newick = Regex(r'[^;]+;')
    tree = (CaselessKeyword("tree").suppress() +
            Optional("*").suppress() +
            name.setResultsName("tree_name") +
            comment.setResultsName("tree_comment") +
            Suppress("=") +
            comment.setResultsName("root_comment") +
            newick.setResultsName("newick"))
    ## treesblock = Group(beginblock +
    ##                    Optional(ttable.setResultsName("ttable")) +
    ##                    Group(OneOrMore(tree)) +
    ##                    endblock)

    def not_begin(s): return s.strip().lower() != "begin trees;"
    def not_end(s): return s.strip().lower() not in ("end;", "endblock;")
    def parse_ttable(f):
        ttable = {}
        while True:
            s = f.next().strip()
            if not s: continue
            if s.lower() == ";": break
            if s[-1] == ",": s = s[:-1]
            k, v = s.split()
            ttable[k] = v
            if s[-1] == ";": break
        return ttable
            
    # read lines between "begin trees;" and "end;"
    f = itertools.takewhile(not_end, itertools.dropwhile(not_begin, infile))
    s = f.next().strip().lower()
    if s != "begin trees;":
        print sys.stderr, "Expecting 'begin trees;', got %s" % s
        raise StopIteration
    ttable = {}
    while True:
        try: s = f.next().strip()
        except StopIteration: break
        if not s: continue
        if s.lower() == "translate":
            ttable = parse_ttable(f)
            print "ttable: %s" % len(ttable)
        elif s.split()[0].lower()=='tree':
            match = tree.parseString(s)
            yield nexus.Newick(match, ttable)

 def parseDate(self, dateString):
   dateGrammar = Regex("\d{4}")("year") + Regex("\d{2}")("month") +\
     Regex("\d{2}")("day") + Regex("\d{2}")("hours") +\
     Suppress(":") + Regex("\d{2}")("minutes")
   results = dateGrammar.parseString(dateString)
   return {
     "year" : results["year"],
     "month" : results["month"],
     "day" : results["day"],
     "hours" : results["hours"],
     "minutes" : results["minutes"],
   }

    def _make_grammar(self):
        from pyparsing import (QuotedString, ZeroOrMore, Combine,
                               Literal, Optional, OneOrMore,
                               Regex, CaselessKeyword)

        def escape_handler(s, loc, toks):
            if toks[0] == '\\\\':
                return "\\"
            elif toks[0] == '\\\'':
                return "'"
            elif toks[0] == '\\"':
                return '"'
            elif toks[0] == '\\f':
                return "\f"
            elif toks[0] == '\\n':
                return "\n"
            elif toks[0] == '\\r':
                return "\r"
            elif toks[0] == '\\t':
                return "\t"
            elif toks[0] == '\\ ':
                return " "
            else:
                return toks[0][1:]

        escape = Combine(Regex(r'\\.')).setParseAction(escape_handler)
        word = Combine(OneOrMore(escape | Regex(r'[^\s\\]+')))
        whitespace = Regex(r'\s+').suppress()
        quotedstring = Combine(OneOrMore(QuotedString('"', escChar='\\') | QuotedString("'", escChar='\\')))
        command = Regex(r'[^\s:]+') + Literal(":").suppress() + (quotedstring | word)
        include = quotedstring | command | word
        exclude = (Literal("-") | Literal("^")).suppress() + (quotedstring | command | word)
        or_keyword = CaselessKeyword("or")
        and_keyword = CaselessKeyword("and")
        keyword = or_keyword | and_keyword

        argument = (keyword | exclude | include)
        expr = ZeroOrMore(Optional(whitespace) + argument)

        # arguments.leaveWhitespace()

        command.setParseAction(CommandExpr)
        include.setParseAction(IncludeExpr)
        exclude.setParseAction(ExcludeExpr)
        or_keyword.setParseAction(OrKeywordExpr)
        and_keyword.setParseAction(AndKeywordExpr)

        # or_expr.setParseAction(lambda s, loc, toks: OrOperator(toks[0], toks[2]))
        # and_expr.setParseAction(lambda s, loc, toks: AndOperator(toks[0], toks[2]))
        # no_expr.setParseAction(lambda s, loc, toks: AndOperator(toks[0], toks[1]))
        # expr.setParseAction(Operator)

        return expr

    def __init__(self, EvaluateVariableChild=None, EvaluateNumberChild=None):
        EvaluateVariableChild = EvaluateVariableChild or EvaluateVariable
        EvaluateNumberChild = EvaluateNumberChild or EvaluateNumber
        # what is a float number
        floatNumber = Regex(r'[-]?\d+(\.\d*)?([eE][-+]?\d+)?')
        # a variable is a combination of letters, numbers, and underscor
        variable = Word(alphanums + "_")
        # a sign is plus or minus
        signOp = oneOf('+ -')
        # an operand is a variable or a floating point number
        operand = floatNumber ^ variable
        # when a floatNumber is found, parse it with evaluate number
        floatNumber.setParseAction(EvaluateNumberChild)
        # when a variable is found, parse it with the EvaluateVariableChild
        # or EvaluateVariable
        variable.setParseAction(EvaluateVariableChild)
        # comparisons include lt,le,gt,ge,eq,ne
        comparisonOp = oneOf("< <= > >= == !=")
        # negation of the boolean is !
        notOp = oneOf("!")
        # an expression is a either a comparison or
        # a NOT operation (where NOT a is essentially (a == False))
        comparisonExpression = operatorPrecedence(operand,
                                                  [
                                                   (comparisonOp,
                                                    2,
                                                    opAssoc.LEFT,
                                                    EvaluateComparison
                                                    ),
                                                   (notOp,
                                                    1,
                                                    opAssoc.RIGHT,
                                                    EvaluateNot
                                                    ),
                                                  ])

        # boolean logic of AND or OR
        boolOp = oneOf("& |")

        # a bool expression contains a nested bool expression or a comparison,
        # joined with a boolean operation
        boolExpression = Forward()
        boolPossible = boolExpression | comparisonExpression
        self.boolExpression = operatorPrecedence(boolPossible,
                                                 [
                                                  (boolOp,
                                                   2,
                                                   opAssoc.RIGHT,
                                                   EvaluateOrAnd
                                                   ),
                                                 ])
        return

def parse_treesblock(infile):
    import string
    from pyparsing import Optional, Word, Regex, CaselessKeyword, Suppress
    from pyparsing import QuotedString
    comment = Optional(Suppress("[&") + Regex(r'[^]]+') + Suppress("]"))
    name = Word(alphanums+"_") | QuotedString("'")
    newick = Regex(r'[^;]+;')
    tree = (CaselessKeyword("tree").suppress() +
            Optional("*").suppress() +
            name.setResultsName("tree_name") +
            comment.setResultsName("tree_comment") +
            Suppress("=") +
            comment.setResultsName("root_comment") +
            newick.setResultsName("newick"))
    ## treesblock = Group(beginblock +
    ##                    Optional(ttable.setResultsName("ttable")) +
    ##                    Group(OneOrMore(tree)) +
    ##                    endblock)

    def parse_ttable(f):
        ttable = {}
        while True:
            s = f.next().strip()
            if s.lower() == ";":
                break
            if s[-1] in ",;":
                s = s[:-1]
            k, v = s.split()
            ttable[k] = v
            if s[-1] == ";":
                break
        return ttable

    ttable = {}
    while True:
        try:
            s = infile.next().strip()
        except StopIteration:
            break
        if s.lower() == "translate":
            ttable = parse_ttable(infile)
            # print("ttable: %s" % len(ttable))
        else:
            match = tree.parseString(s)
            yield Newick(match, ttable)

	def getkw_bnf(self):
		sect_begin   = Literal("{").suppress()
		sect_end   = Literal("}").suppress()
		array_begin   = Literal("[").suppress()
		array_end   = Literal("]").suppress()
		tag_begin   = Literal("<").suppress()
		tag_end   = Literal(">").suppress()
		eql   = Literal("=").suppress()
		dmark = Literal('$').suppress()
		end_data=Literal('$end').suppress()
		prtable = alphanums+r'!$%&*+-./<>[email protected]^_|~'
		ival=Regex('[-]?\d+')
		dval=Regex('-?\d+\.\d*([eE]?[+-]?\d+)?')
		lval=Regex('([Yy]es|[Nn]o|[Tt]rue|[Ff]alse|[Oo]n|[Oo]ff)')
	
		# Helper definitions

		kstr= quotedString.setParseAction(removeQuotes) ^ \
				dval ^ ival ^ lval ^ Word(prtable)
		name = Word(alphas+"_",alphanums+"_")
		vec=array_begin+delimitedList(dval ^ ival ^ lval ^ Word(prtable) ^ \
				Literal("\n").suppress() ^ \
				quotedString.setParseAction(removeQuotes))+array_end
		sect=name+sect_begin
		tag_sect=name+Group(tag_begin+name+tag_end)+sect_begin

		# Grammar
		keyword = name + eql + kstr
		vector = name + eql + vec
		data=Combine(dmark+name)+SkipTo(end_data)+end_data
		section=Forward()
		sect_def=(sect | tag_sect ) #| vec_sect)
		input=section | data | vector | keyword 
		section << sect_def+ZeroOrMore(input) + sect_end

		# Parsing actions	
		ival.setParseAction(self.conv_ival)
		dval.setParseAction(self.conv_dval)
		lval.setParseAction(self.conv_lval)
		keyword.setParseAction(self.store_key)
		vector.setParseAction(self.store_vector)
		data.setParseAction(self.store_data)
		sect.setParseAction(self.add_sect)
		tag_sect.setParseAction(self.add_sect)
		sect_end.setParseAction(self.pop_sect)

		bnf=ZeroOrMore(input) + StringEnd().setFailAction(parse_error)
		bnf.ignore(pythonStyleComment)
		return bnf

def make_sexp_parser ():
    """
    Returns a  simple parser for  nested lists of real  numbers. Round
    parens () are assumed as customary in lisps.
    """

    # Punctuation literals (note round parens):
    LPAR, RPAR = map (Suppress, "()")

    # Real numbers:
    real_string = Regex (r"[+-]?\d+\.\d*([eE][+-]?\d+)?")
    real = real_string.setParseAction (lambda tokens: float (tokens[0]))

    # Voodoo:
    sexp = Forward ()
    sexp_list = Group (LPAR + ZeroOrMore (sexp) + RPAR)
    sexp << (real | sexp_list)

    return lambda s: sexp.parseString (s)[0]

def ListParser():
	"""
	A parser for list columns, where each list is composed of pairs of values.
	"""

	value = Regex(r'[-+]?[0-9]+(?:\.[0-9]*)?(?:e[-+]?[0-9]+)?', IGNORECASE)
	value.setParseAction(lambda toks: float(toks[0]))

	item = Suppress('(') + value + Suppress(',') + value + Suppress(')')
	item.setParseAction(tuple)

	lst = Suppress('[') + delimitedList(item) + Suppress(']')
	lst.setParseAction(list)

	def parse(s):
		try:
			return lst.parseString(s).asList()
		except ParseBaseException as e:
			raise ValueError(e)

	return parse

    def translate(self, text, filename):
        self.source = text
        self.super = None
        self.inheritance = 0
        self.declaration_lines = ['inheritance = 0']
        self.block_lines = []
        self.body_lines = ['def body():']

        self.target_lines = self.body_lines
        self.indent = 1
        
        template_close = Literal('%>')
        white = White()
        
        attribute = Word(alphanums + '_') + Literal('=') + QuotedString('"') + Optional(white)
        
        directive = "<%@" +  Optional(white) + Word(alphanums + '_') + white + ZeroOrMore(attribute) + template_close
        declaration = "<%!" + SkipTo(template_close) + template_close
        expression = "<%=" + SkipTo(template_close) + template_close
        scriptlet = '<%' + SkipTo(template_close) + template_close
        
        template_text = directive | declaration | expression | scriptlet
        plain_text = Regex(r'((?!<%).|\s)+', re.MULTILINE)
        
        body = template_text | plain_text
        lit = OneOrMore(body)
        
        directive.setParseAction(self.compile_directive)
        declaration.setParseAction(self.compile_declaration)
        expression.setParseAction(self.compile_expression)
        scriptlet.setParseAction(self.compile_scriptlet)
        plain_text.setParseAction(self.compile_plain_text)
        
        lit.leaveWhitespace()
        lit.parseString(self.source)
        
        translated =  '\n' + '\n'.join(self.declaration_lines + ['\n'] + self.block_lines + ['\n'] + self.body_lines)
        if self.super:
            translated = self.super.module_source + translated
        return translated

        return [(t[0][0].port, t[0][1].port)]

def normalize_ip(t):
    # returns a normalized ip
    return t.ip + "/" + (str(t.mask.mask) if t.mask else "32")


port = Group(Word(nums).setParseAction(to_int)('port'))
port_range = Group((port + Word("-").suppress() + port)('range'))

normalized_port_range = (port ^ port_range).setParseAction(to_port_range)

ports  = delimitedList(normalized_port_range)('ports')

# IP addresses, name of another group, or sg-*
security_group = Regex("sg-[\w\d]+")
group_name = Regex("[\w\d\-]+")

mask = Word("/") + Word(nums).setParseAction(to_int)('mask')
ip= (Combine(Word(nums) + ('.' + Word(nums))*3)('ip') + Optional(mask)('mask')).setParseAction(normalize_ip)

parser = Optional(protocol)('protocol') + \
         Optional(port_) + \
         ports + \
         (ip.setResultsName('ip_and_mask') ^ security_group.setResultsName('security_group') ^ group_name('group_name'))


class Rule(object):

    def __init__(self, protocol, from_port, to_port, address=None, group=None, group_name=None):
        """constructs a new rule

        # val = val.replace(".", "\\.")
    elif val.startswith('`') and val.endswith('`'):
          val = "'" + val[1:-1].replace("``","`") + "'"
    elif val.startswith("+"):
        val = val[1:]
    un = ast.literal_eval(val)
    return un


def to_string(instring, tokensStart, retTokens):
    val = retTokens[0]
    val = "'"+val[1:-1].replace("''", "\\'")+"'"
    return {"literal": ast.literal_eval(val)}

# NUMBERS
realNum = Regex(r"[+-]?(\d+\.\d*|\.\d+)([eE][+-]?\d+)?").addParseAction(unquote)
intNum = Regex(r"[+-]?\d+([eE]\+?\d+)?").addParseAction(unquote)

# STRINGS, NUMBERS, VARIABLES
sqlString = Regex(r"\'(\'\'|\\.|[^'])*\'").addParseAction(to_string)
identString = Regex(r'\"(\"\"|\\.|[^"])*\"').addParseAction(unquote)
mysqlidentString = Regex(r'\`(\`\`|\\.|[^`])*\`').addParseAction(unquote)
ident = Combine(~RESERVED + (delimitedList(Literal("*") | Word(alphas + "_", alphanums + "_$") | identString | mysqlidentString, delim=".", combine=True))).setName("identifier")

# EXPRESSIONS
expr = Forward()

# CASE
case = (
    CASE +
    Group(ZeroOrMore((WHEN + expr("when") + THEN + expr("then")).addParseAction(to_when_call)))("case") +

    def define_dot_parser(self):
        """Define dot grammar

        Based on the grammar http://www.graphviz.org/doc/info/lang.html
        """
        # punctuation
        colon  = Literal(":")
        lbrace = Suppress("{")
        rbrace = Suppress("}")
        lbrack = Suppress("[")
        rbrack = Suppress("]")
        lparen = Literal("(")
        rparen = Literal(")")
        equals = Suppress("=")
        comma  = Literal(",")
        dot    = Literal(".")
        slash  = Literal("/")
        bslash = Literal("\\")
        star   = Literal("*")
        semi   = Suppress(";")
        at     = Literal("@")
        minus  = Literal("-")
        pluss  = Suppress("+")

        # keywords
        strict_    = CaselessLiteral("strict")
        graph_     = CaselessLiteral("graph")
        digraph_   = CaselessLiteral("digraph")
        subgraph_  = CaselessLiteral("subgraph")
        node_      = CaselessLiteral("node")
        edge_      = CaselessLiteral("edge")

        punctuation_ = "".join( [ c for c in string.punctuation if c not in '_' ] ) +string.whitespace
        # token definitions

        identifier = Word(alphanums + "_" ).setName("identifier")

        #double_quoted_string = QuotedString('"', multiline=True,escChar='\\',
        #    unquoteResults=True) # dblQuotedString
        double_quoted_string = Regex(r'\"(?:\\\"|\\\\|[^"])*\"', re.MULTILINE)
        double_quoted_string.setParseAction(removeQuotes)
        quoted_string = Combine(double_quoted_string+
            Optional(OneOrMore(pluss+double_quoted_string)),adjacent=False)
        alphastring_ = OneOrMore(CharsNotIn(punctuation_))

        def parse_html(s, loc, toks):
            return '<<%s>>' % ''.join(toks[0])


        opener = '<'
        closer = '>'
        try:
            html_text = pyparsing.nestedExpr( opener, closer,
                (( CharsNotIn(
                    opener + closer ).setParseAction( lambda t:t[0] ))
                )).setParseAction(parse_html)
        except:
            log.debug('nestedExpr not available.')
            log.warning('Old version of pyparsing detected. Version 1.4.8 or '
                        'later is recommended. Parsing of html labels may not '
                        'work properly.')
            html_text = Combine(Literal("<<") + OneOrMore(CharsNotIn(",]")))


        ID = ( alphastring_ | html_text |
            quoted_string | #.setParseAction(strip_quotes) |
            identifier ).setName("ID")


        float_number = Combine(Optional(minus) +
            OneOrMore(Word(nums + "."))).setName("float_number")

        righthand_id =  (float_number | ID ).setName("righthand_id")

        port_angle = (at + ID).setName("port_angle")

        port_location = ((OneOrMore(Group(colon + ID)) |
            Group(colon + lparen + ID + comma + ID + rparen))).setName("port_location")

        port = Combine((Group(port_location + Optional(port_angle)) |
            Group(port_angle + Optional(port_location)))).setName("port")

        node_id = (ID + Optional(port))
        a_list = OneOrMore(ID + Optional(equals + righthand_id) +
            Optional(comma.suppress())).setName("a_list")

        attr_list = OneOrMore(lbrack + Optional(a_list) +
            rbrack).setName("attr_list").setResultsName('attrlist')

        attr_stmt = ((graph_ | node_ | edge_) + attr_list).setName("attr_stmt")

        edgeop = (Literal("--") | Literal("->")).setName("edgeop")

        stmt_list = Forward()
        graph_stmt = (lbrace + Optional(stmt_list) +
            rbrace + Optional(semi) ).setName("graph_stmt")


        edge_point = Forward()

#.........这里部分代码省略.........

def formula_grammar(table):
    """
    Construct a parser for molecular formulas.

    :Parameters:

        *table* = None : PeriodicTable
             If table is specified, then elements and their associated fields
             will be chosen from that periodic table rather than the default.

    :Returns:
        *parser* : pyparsing.ParserElement.
            The ``parser.parseString()`` method returns a list of
            pairs (*count,fragment*), where fragment is an *isotope*,
            an *element* or a list of pairs (*count,fragment*).

    """
    # Recursive
    composite = Forward()
    mixture = Forward()

    # whitespace and separators
    space = Optional(White().suppress())
    separator = space+Literal('+').suppress()+space

    # Lookup the element in the element table
    symbol = Regex("[A-Z][a-z]*")
    symbol = symbol.setParseAction(lambda s,l,t: table.symbol(t[0]))

    # Translate isotope
    openiso = Literal('[').suppress()
    closeiso = Literal(']').suppress()
    isotope = Optional(~White()+openiso+Regex("[1-9][0-9]*")+closeiso,
                       default='0')
    isotope = isotope.setParseAction(lambda s,l,t: int(t[0]) if t[0] else 0)

    # Translate ion
    openion = Literal('{').suppress()
    closeion = Literal('}').suppress()
    ion = Optional(~White()
                     +openion
                     +Regex("([1-9][0-9]*)?[+-]")
                     +closeion,
                   default='0+')
    ion = ion.setParseAction(lambda s,l,t: int(t[0][-1]+(t[0][:-1] if len(t[0])>1 else '1')))

    # Translate counts
    fract = Regex("(0|[1-9][0-9]*|)([.][0-9]*)")
    fract = fract.setParseAction(lambda s,l,t: float(t[0]) if t[0] else 1)
    whole = Regex("[1-9][0-9]*")
    whole = whole.setParseAction(lambda s,l,t: int(t[0]) if t[0] else 1)
    count = Optional(~White()+(fract|whole),default=1)

    # Convert symbol,isotope,ion,count to (count,isotope)
    element = symbol+isotope+ion+count
    def convert_element(string,location,tokens):
        #print "convert_element received",tokens
        symbol,isotope,ion,count = tokens[0:4]
        if isotope != 0: symbol = symbol[isotope]
        if ion != 0: symbol = symbol.ion[ion]
        return (count,symbol)
    element = element.setParseAction(convert_element)

    # Convert "count elements" to a pair
    implicit_group = count+OneOrMore(element)
    def convert_implicit(string,location,tokens):
        #print "implicit",tokens
        count = tokens[0]
        fragment = tokens[1:]
        return fragment if count==1 else (count,fragment)
    implicit_group = implicit_group.setParseAction(convert_implicit)

    # Convert "(composite) count" to a pair
    opengrp = space + Literal('(').suppress() + space
    closegrp = space + Literal(')').suppress() + space
    explicit_group = opengrp + composite + closegrp + count
    def convert_explicit(string,location,tokens):
        #print "explicit",tokens
        count = tokens[-1]
        fragment = tokens[:-1]
        return fragment if count == 1 else (count,fragment)
    explicit_group = explicit_group.setParseAction(convert_explicit)

    # Build composite from a set of groups
    group = implicit_group | explicit_group
    implicit_separator = separator | space
    composite << group + ZeroOrMore(implicit_separator + group)

    density = Literal('@').suppress() + count + Optional(Regex("[ni]"),default='i')
    compound = composite + Optional(density,default=None)
    def convert_compound(string,location,tokens):
        #print "compound",tokens
        if tokens[-1] is None:
            return Formula(structure=_immutable(tokens[:-1]))
        elif tokens[-1] == 'n':
            return Formula(structure=_immutable(tokens[:-2]), natural_density=tokens[-2])
        else:
            return Formula(structure=_immutable(tokens[:-2]), density=tokens[-2])
    compound = compound.setParseAction(convert_compound)

#.........这里部分代码省略.........

    ParserElement,
)

ParserElement.enablePackrat()

COLON, LBRACK, RBRACK, LBRACE, RBRACE, TILDE, CARAT = map(Literal, ":[]{}~^")
LPAR, RPAR = map(Suppress, "()")
and_ = CaselessKeyword("AND")
or_ = CaselessKeyword("OR")
not_ = CaselessKeyword("NOT")
to_ = CaselessKeyword("TO")
keyword = and_ | or_ | not_

expression = Forward()

valid_word = Regex(r'([a-zA-Z0-9*_+.-]|\\[!(){}\[\]^"~*?\\:])+').setName("word")
valid_word.setParseAction(lambda t: t[0].replace("\\\\", chr(127)).replace("\\", "").replace(chr(127), "\\"))

string = QuotedString('"')

required_modifier = Literal("+")("required")
prohibit_modifier = Literal("-")("prohibit")
integer = Regex(r"\d+").setParseAction(lambda t: int(t[0]))
proximity_modifier = Group(TILDE + integer("proximity"))
number = Regex(r"\d+(\.\d+)?").setParseAction(lambda t: float(t[0]))
fuzzy_modifier = TILDE + Optional(number, default=0.5)("fuzzy")

term = Forward()
field_name = valid_word.copy().setName("fieldname")
incl_range_search = Group(LBRACK + term("lower") + to_ + term("upper") + RBRACK)
excl_range_search = Group(LBRACE + term("lower") + to_ + term("upper") + RBRACE)