From 84aec138dc4311c3f23ff78e9916e00b457a1896 Mon Sep 17 00:00:00 2001 From: insiv Date: Wed, 17 Aug 2016 23:05:38 +0700 Subject: Upgrade pyparsing to version 2.1.8 --- pkg_resources/_vendor/pyparsing.py | 3358 +++++++++++++++++++++++++++--------- 1 file changed, 2564 insertions(+), 794 deletions(-) (limited to 'pkg_resources/_vendor/pyparsing.py') diff --git a/pkg_resources/_vendor/pyparsing.py b/pkg_resources/_vendor/pyparsing.py index 2284cadc..89cffc10 100644 --- a/pkg_resources/_vendor/pyparsing.py +++ b/pkg_resources/_vendor/pyparsing.py @@ -48,7 +48,7 @@ The program outputs the following:: The Python representation of the grammar is quite readable, owing to the self-explanatory class names, and the use of '+', '|' and '^' operators. -The parsed results returned from C{parseString()} can be accessed as a nested list, a dictionary, or an +The parsed results returned from L{I{ParserElement.parseString}} can be accessed as a nested list, a dictionary, or an object with named attributes. The pyparsing module handles some of the problems that are typically vexing when writing text parsers: @@ -57,8 +57,8 @@ The pyparsing module handles some of the problems that are typically vexing when - embedded comments """ -__version__ = "2.0.6" -__versionTime__ = "9 Nov 2015 19:03" +__version__ = "2.1.8" +__versionTime__ = "14 Aug 2016 08:43 UTC" __author__ = "Paul McGuire " import string @@ -70,8 +70,22 @@ import re import sre_constants import collections import pprint -import functools -import itertools +import traceback +import types +from datetime import datetime + +try: + from _thread import RLock +except ImportError: + from threading import RLock + +try: + from collections import OrderedDict as _OrderedDict +except ImportError: + try: + from ordereddict import OrderedDict as _OrderedDict + except ImportError: + _OrderedDict = None #~ sys.stderr.write( "testing pyparsing module, version %s, %s\n" % (__version__,__versionTime__ ) ) @@ -81,21 +95,23 @@ __all__ = [ 'MatchFirst', 'NoMatch', 'NotAny', 'OneOrMore', 'OnlyOnce', 'Optional', 'Or', 'ParseBaseException', 'ParseElementEnhance', 'ParseException', 'ParseExpression', 'ParseFatalException', 'ParseResults', 'ParseSyntaxException', 'ParserElement', 'QuotedString', 'RecursiveGrammarException', -'Regex', 'SkipTo', 'StringEnd', 'StringStart', 'Suppress', 'Token', 'TokenConverter', 'Upcase', +'Regex', 'SkipTo', 'StringEnd', 'StringStart', 'Suppress', 'Token', 'TokenConverter', 'White', 'Word', 'WordEnd', 'WordStart', 'ZeroOrMore', 'alphanums', 'alphas', 'alphas8bit', 'anyCloseTag', 'anyOpenTag', 'cStyleComment', 'col', 'commaSeparatedList', 'commonHTMLEntity', 'countedArray', 'cppStyleComment', 'dblQuotedString', 'dblSlashComment', 'delimitedList', 'dictOf', 'downcaseTokens', 'empty', 'hexnums', -'htmlComment', 'javaStyleComment', 'keepOriginalText', 'line', 'lineEnd', 'lineStart', 'lineno', +'htmlComment', 'javaStyleComment', 'line', 'lineEnd', 'lineStart', 'lineno', 'makeHTMLTags', 'makeXMLTags', 'matchOnlyAtCol', 'matchPreviousExpr', 'matchPreviousLiteral', 'nestedExpr', 'nullDebugAction', 'nums', 'oneOf', 'opAssoc', 'operatorPrecedence', 'printables', 'punc8bit', 'pythonStyleComment', 'quotedString', 'removeQuotes', 'replaceHTMLEntity', 'replaceWith', 'restOfLine', 'sglQuotedString', 'srange', 'stringEnd', 'stringStart', 'traceParseAction', 'unicodeString', 'upcaseTokens', 'withAttribute', 'indentedBlock', 'originalTextFor', 'ungroup', 'infixNotation','locatedExpr', 'withClass', +'tokenMap', 'pyparsing_common', ] -PY_3 = sys.version.startswith('3') +system_version = tuple(sys.version_info)[:3] +PY_3 = system_version[0] == 3 if PY_3: _MAX_INT = sys.maxsize basestring = str @@ -123,18 +139,11 @@ else: return str(obj) except UnicodeEncodeError: - # The Python docs (http://docs.python.org/ref/customization.html#l2h-182) - # state that "The return value must be a string object". However, does a - # unicode object (being a subclass of basestring) count as a "string - # object"? - # If so, then return a unicode object: - return unicode(obj) - # Else encode it... but how? There are many choices... :) - # Replace unprintables with escape codes? - #return unicode(obj).encode(sys.getdefaultencoding(), 'backslashreplace_errors') - # Replace unprintables with question marks? - #return unicode(obj).encode(sys.getdefaultencoding(), 'replace') - # ... + # Else encode it + ret = unicode(obj).encode(sys.getdefaultencoding(), 'xmlcharrefreplace') + xmlcharref = Regex('&#\d+;') + xmlcharref.setParseAction(lambda t: '\\u' + hex(int(t[0][2:-1]))[2:]) + return xmlcharref.transformString(ret) # build list of single arg builtins, tolerant of Python version, that can be used as parse actions singleArgBuiltins = [] @@ -160,7 +169,7 @@ def _xml_escape(data): class _Constants(object): pass -alphas = string.ascii_lowercase + string.ascii_uppercase +alphas = string.ascii_uppercase + string.ascii_lowercase nums = "0123456789" hexnums = nums + "ABCDEFabcdef" alphanums = alphas + nums @@ -180,6 +189,15 @@ class ParseBaseException(Exception): self.msg = msg self.pstr = pstr self.parserElement = elem + self.args = (pstr, loc, msg) + + @classmethod + def _from_exception(cls, pe): + """ + internal factory method to simplify creating one type of ParseException + from another - avoids having __init__ signature conflicts among subclasses + """ + return cls(pe.pstr, pe.loc, pe.msg, pe.parserElement) def __getattr__( self, aname ): """supported attributes by name are: @@ -212,15 +230,26 @@ class ParseBaseException(Exception): markerString, line_str[line_column:])) return line_str.strip() def __dir__(self): - return "loc msg pstr parserElement lineno col line " \ - "markInputline __str__ __repr__".split() + return "lineno col line".split() + dir(type(self)) class ParseException(ParseBaseException): - """exception thrown when parse expressions don't match class; - supported attributes by name are: - - lineno - returns the line number of the exception text - - col - returns the column number of the exception text - - line - returns the line containing the exception text + """ + Exception thrown when parse expressions don't match class; + supported attributes by name are: + - lineno - returns the line number of the exception text + - col - returns the column number of the exception text + - line - returns the line containing the exception text + + Example:: + try: + Word(nums).setName("integer").parseString("ABC") + except ParseException as pe: + print(pe) + print("column: {}".format(pe.col)) + + prints:: + Expected integer (at char 0), (line:1, col:1) + column: 1 """ pass @@ -230,12 +259,10 @@ class ParseFatalException(ParseBaseException): pass class ParseSyntaxException(ParseFatalException): - """just like C{L{ParseFatalException}}, but thrown internally when an - C{L{ErrorStop}} ('-' operator) indicates that parsing is to stop immediately because - an unbacktrackable syntax error has been found""" - def __init__(self, pe): - super(ParseSyntaxException, self).__init__( - pe.pstr, pe.loc, pe.msg, pe.parserElement) + """just like L{ParseFatalException}, but thrown internally when an + L{ErrorStop} ('-' operator) indicates that parsing is to stop + immediately because an unbacktrackable syntax error has been found""" + pass #~ class ReparseException(ParseBaseException): #~ """Experimental class - parse actions can raise this exception to cause @@ -251,7 +278,7 @@ class ParseSyntaxException(ParseFatalException): #~ self.reparseLoc = restartLoc class RecursiveGrammarException(Exception): - """exception thrown by C{validate()} if the grammar could be improperly recursive""" + """exception thrown by L{ParserElement.validate} if the grammar could be improperly recursive""" def __init__( self, parseElementList ): self.parseElementTrace = parseElementList @@ -269,12 +296,44 @@ class _ParseResultsWithOffset(object): self.tup = (self.tup[0],i) class ParseResults(object): - """Structured parse results, to provide multiple means of access to the parsed data: + """ + Structured parse results, to provide multiple means of access to the parsed data: - as a list (C{len(results)}) - by list index (C{results[0], results[1]}, etc.) - - by attribute (C{results.}) - """ - def __new__(cls, toklist, name=None, asList=True, modal=True ): + - by attribute (C{results.} - see L{ParserElement.setResultsName}) + + Example:: + integer = Word(nums) + date_str = (integer.setResultsName("year") + '/' + + integer.setResultsName("month") + '/' + + integer.setResultsName("day")) + # equivalent form: + # date_str = integer("year") + '/' + integer("month") + '/' + integer("day") + + result = date_str.parseString("1999/12/31") + + def test(s, fn=repr): + print("%s -> %s" % (s, fn(eval(s)))) + test("list(result)") + test("result[0]") + test("result['month']") + test("result.day") + test("'month' in result") + test("'minutes' in result") + test("result.dump()", str) + prints:: + list(result) -> ['1999', '/', '12', '/', '31'] + result[0] -> '1999' + result['month'] -> '12' + result.day -> '31' + 'month' in result -> True + 'minutes' in result -> False + result.dump() -> ['1999', '/', '12', '/', '31'] + - day: 31 + - month: 12 + - year: 1999 + """ + def __new__(cls, toklist=None, name=None, asList=True, modal=True ): if isinstance(toklist, cls): return toklist retobj = object.__new__(cls) @@ -283,12 +342,16 @@ class ParseResults(object): # Performance tuning: we construct a *lot* of these, so keep this # constructor as small and fast as possible - def __init__( self, toklist, name=None, asList=True, modal=True, isinstance=isinstance ): + def __init__( self, toklist=None, name=None, asList=True, modal=True, isinstance=isinstance ): if self.__doinit: self.__doinit = False self.__name = None self.__parent = None self.__accumNames = {} + self.__asList = asList + self.__modal = modal + if toklist is None: + toklist = [] if isinstance(toklist, list): self.__toklist = toklist[:] elif isinstance(toklist, _generatorType): @@ -331,7 +394,7 @@ class ParseResults(object): if isinstance(v,_ParseResultsWithOffset): self.__tokdict[k] = self.__tokdict.get(k,list()) + [v] sub = v[0] - elif isinstance(k,int): + elif isinstance(k,(int,slice)): self.__toklist[k] = v sub = v else: @@ -354,11 +417,6 @@ class ParseResults(object): removed = list(range(*i.indices(mylen))) removed.reverse() # fixup indices in token dictionary - #~ for name in self.__tokdict: - #~ occurrences = self.__tokdict[name] - #~ for j in removed: - #~ for k, (value, position) in enumerate(occurrences): - #~ occurrences[k] = _ParseResultsWithOffset(value, position - (position > j)) for name,occurrences in self.__tokdict.items(): for j in removed: for k, (value, position) in enumerate(occurrences): @@ -370,39 +428,52 @@ class ParseResults(object): return k in self.__tokdict def __len__( self ): return len( self.__toklist ) - def __bool__(self): return len( self.__toklist ) > 0 + def __bool__(self): return ( not not self.__toklist ) __nonzero__ = __bool__ def __iter__( self ): return iter( self.__toklist ) def __reversed__( self ): return iter( self.__toklist[::-1] ) - def iterkeys( self ): - """Returns all named result keys.""" + def _iterkeys( self ): if hasattr(self.__tokdict, "iterkeys"): return self.__tokdict.iterkeys() else: return iter(self.__tokdict) - def itervalues( self ): - """Returns all named result values.""" - return (self[k] for k in self.iterkeys()) + def _itervalues( self ): + return (self[k] for k in self._iterkeys()) - def iteritems( self ): - return ((k, self[k]) for k in self.iterkeys()) + def _iteritems( self ): + return ((k, self[k]) for k in self._iterkeys()) if PY_3: - keys = iterkeys - values = itervalues - items = iteritems + keys = _iterkeys + """Returns an iterator of all named result keys (Python 3.x only).""" + + values = _itervalues + """Returns an iterator of all named result values (Python 3.x only).""" + + items = _iteritems + """Returns an iterator of all named result key-value tuples (Python 3.x only).""" + else: + iterkeys = _iterkeys + """Returns an iterator of all named result keys (Python 2.x only).""" + + itervalues = _itervalues + """Returns an iterator of all named result values (Python 2.x only).""" + + iteritems = _iteritems + """Returns an iterator of all named result key-value tuples (Python 2.x only).""" + def keys( self ): - """Returns all named result keys.""" + """Returns all named result keys (as a list in Python 2.x, as an iterator in Python 3.x).""" return list(self.iterkeys()) def values( self ): - """Returns all named result values.""" + """Returns all named result values (as a list in Python 2.x, as an iterator in Python 3.x).""" return list(self.itervalues()) def items( self ): - """Returns all named result keys and values as a list of tuples.""" + """Returns all named result key-values (as a list of tuples in Python 2.x, as an iterator in Python 3.x).""" return list(self.iteritems()) def haskeys( self ): @@ -411,14 +482,39 @@ class ParseResults(object): return bool(self.__tokdict) def pop( self, *args, **kwargs): - """Removes and returns item at specified index (default=last). - Supports both list and dict semantics for pop(). If passed no - argument or an integer argument, it will use list semantics - and pop tokens from the list of parsed tokens. If passed a - non-integer argument (most likely a string), it will use dict - semantics and pop the corresponding value from any defined - results names. A second default return value argument is - supported, just as in dict.pop().""" + """ + Removes and returns item at specified index (default=C{last}). + Supports both C{list} and C{dict} semantics for C{pop()}. If passed no + argument or an integer argument, it will use C{list} semantics + and pop tokens from the list of parsed tokens. If passed a + non-integer argument (most likely a string), it will use C{dict} + semantics and pop the corresponding value from any defined + results names. A second default return value argument is + supported, just as in C{dict.pop()}. + + Example:: + def remove_first(tokens): + tokens.pop(0) + print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321'] + print(OneOrMore(Word(nums)).addParseAction(remove_first).parseString("0 123 321")) # -> ['123', '321'] + + label = Word(alphas) + patt = label("LABEL") + OneOrMore(Word(nums)) + print(patt.parseString("AAB 123 321").dump()) + + # Use pop() in a parse action to remove named result (note that corresponding value is not + # removed from list form of results) + def remove_LABEL(tokens): + tokens.pop("LABEL") + return tokens + patt.addParseAction(remove_LABEL) + print(patt.parseString("AAB 123 321").dump()) + prints:: + ['AAB', '123', '321'] + - LABEL: AAB + + ['AAB', '123', '321'] + """ if not args: args = [-1] for k,v in kwargs.items(): @@ -438,39 +534,83 @@ class ParseResults(object): return defaultvalue def get(self, key, defaultValue=None): - """Returns named result matching the given key, or if there is no - such name, then returns the given C{defaultValue} or C{None} if no - C{defaultValue} is specified.""" + """ + Returns named result matching the given key, or if there is no + such name, then returns the given C{defaultValue} or C{None} if no + C{defaultValue} is specified. + + Similar to C{dict.get()}. + + Example:: + integer = Word(nums) + date_str = integer("year") + '/' + integer("month") + '/' + integer("day") + + result = date_str.parseString("1999/12/31") + print(result.get("year")) # -> '1999' + print(result.get("hour", "not specified")) # -> 'not specified' + print(result.get("hour")) # -> None + """ if key in self: return self[key] else: return defaultValue def insert( self, index, insStr ): - """Inserts new element at location index in the list of parsed tokens.""" + """ + Inserts new element at location index in the list of parsed tokens. + + Similar to C{list.insert()}. + + Example:: + print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321'] + + # use a parse action to insert the parse location in the front of the parsed results + def insert_locn(locn, tokens): + tokens.insert(0, locn) + print(OneOrMore(Word(nums)).addParseAction(insert_locn).parseString("0 123 321")) # -> [0, '0', '123', '321'] + """ self.__toklist.insert(index, insStr) # fixup indices in token dictionary - #~ for name in self.__tokdict: - #~ occurrences = self.__tokdict[name] - #~ for k, (value, position) in enumerate(occurrences): - #~ occurrences[k] = _ParseResultsWithOffset(value, position + (position > index)) for name,occurrences in self.__tokdict.items(): for k, (value, position) in enumerate(occurrences): occurrences[k] = _ParseResultsWithOffset(value, position + (position > index)) def append( self, item ): - """Add single element to end of ParseResults list of elements.""" + """ + Add single element to end of ParseResults list of elements. + + Example:: + print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321'] + + # use a parse action to compute the sum of the parsed integers, and add it to the end + def append_sum(tokens): + tokens.append(sum(map(int, tokens))) + print(OneOrMore(Word(nums)).addParseAction(append_sum).parseString("0 123 321")) # -> ['0', '123', '321', 444] + """ self.__toklist.append(item) def extend( self, itemseq ): - """Add sequence of elements to end of ParseResults list of elements.""" + """ + Add sequence of elements to end of ParseResults list of elements. + + Example:: + patt = OneOrMore(Word(alphas)) + + # use a parse action to append the reverse of the matched strings, to make a palindrome + def make_palindrome(tokens): + tokens.extend(reversed([t[::-1] for t in tokens])) + return ''.join(tokens) + print(patt.addParseAction(make_palindrome).parseString("lskdj sdlkjf lksd")) # -> 'lskdjsdlkjflksddsklfjkldsjdksl' + """ if isinstance(itemseq, ParseResults): self += itemseq else: self.__toklist.extend(itemseq) def clear( self ): - """Clear all elements and results names.""" + """ + Clear all elements and results names. + """ del self.__toklist[:] self.__tokdict.clear() @@ -511,7 +651,11 @@ class ParseResults(object): def __radd__(self, other): if isinstance(other,int) and other == 0: + # useful for merging many ParseResults using sum() builtin return self.copy() + else: + # this may raise a TypeError - so be it + return other + self def __repr__( self ): return "(%s, %s)" % ( repr( self.__toklist ), repr( self.__tokdict ) ) @@ -531,18 +675,60 @@ class ParseResults(object): return out def asList( self ): - """Returns the parse results as a nested list of matching tokens, all converted to strings.""" + """ + Returns the parse results as a nested list of matching tokens, all converted to strings. + + Example:: + patt = OneOrMore(Word(alphas)) + result = patt.parseString("sldkj lsdkj sldkj") + # even though the result prints in string-like form, it is actually a pyparsing ParseResults + print(type(result), result) # -> ['sldkj', 'lsdkj', 'sldkj'] + + # Use asList() to create an actual list + result_list = result.asList() + print(type(result_list), result_list) # -> ['sldkj', 'lsdkj', 'sldkj'] + """ return [res.asList() if isinstance(res,ParseResults) else res for res in self.__toklist] def asDict( self ): - """Returns the named parse results as dictionary.""" + """ + Returns the named parse results as a nested dictionary. + + Example:: + integer = Word(nums) + date_str = integer("year") + '/' + integer("month") + '/' + integer("day") + + result = date_str.parseString('12/31/1999') + print(type(result), repr(result)) # -> (['12', '/', '31', '/', '1999'], {'day': [('1999', 4)], 'year': [('12', 0)], 'month': [('31', 2)]}) + + result_dict = result.asDict() + print(type(result_dict), repr(result_dict)) # -> {'day': '1999', 'year': '12', 'month': '31'} + + # even though a ParseResults supports dict-like access, sometime you just need to have a dict + import json + print(json.dumps(result)) # -> Exception: TypeError: ... is not JSON serializable + print(json.dumps(result.asDict())) # -> {"month": "31", "day": "1999", "year": "12"} + """ if PY_3: - return dict( self.items() ) + item_fn = self.items else: - return dict( self.iteritems() ) + item_fn = self.iteritems + + def toItem(obj): + if isinstance(obj, ParseResults): + if obj.haskeys(): + return obj.asDict() + else: + return [toItem(v) for v in obj] + else: + return obj + + return dict((k,toItem(v)) for k,v in item_fn()) def copy( self ): - """Returns a new copy of a C{ParseResults} object.""" + """ + Returns a new copy of a C{ParseResults} object. + """ ret = ParseResults( self.__toklist ) ret.__tokdict = self.__tokdict.copy() ret.__parent = self.__parent @@ -551,7 +737,9 @@ class ParseResults(object): return ret def asXML( self, doctag=None, namedItemsOnly=False, indent="", formatted=True ): - """Returns the parse results as XML. Tags are created for tokens and lists that have defined results names.""" + """ + (Deprecated) Returns the parse results as XML. Tags are created for tokens and lists that have defined results names. + """ nl = "\n" out = [] namedItems = dict((v[1],k) for (k,vlist) in self.__tokdict.items() @@ -617,7 +805,27 @@ class ParseResults(object): return None def getName(self): - """Returns the results name for this token expression.""" + """ + Returns the results name for this token expression. Useful when several + different expressions might match at a particular location. + + Example:: + integer = Word(nums) + ssn_expr = Regex(r"\d\d\d-\d\d-\d\d\d\d") + house_number_expr = Suppress('#') + Word(nums, alphanums) + user_data = (Group(house_number_expr)("house_number") + | Group(ssn_expr)("ssn") + | Group(integer)("age")) + user_info = OneOrMore(user_data) + + result = user_info.parseString("22 111-22-3333 #221B") + for item in result: + print(item.getName(), ':', item[0]) + prints:: + age : 22 + ssn : 111-22-3333 + house_number : 221B + """ if self.__name: return self.__name elif self.__parent: @@ -633,40 +841,72 @@ class ParseResults(object): else: return None - def dump(self,indent='',depth=0): - """Diagnostic method for listing out the contents of a C{ParseResults}. - Accepts an optional C{indent} argument so that this string can be embedded - in a nested display of other data.""" + def dump(self, indent='', depth=0, full=True): + """ + Diagnostic method for listing out the contents of a C{ParseResults}. + Accepts an optional C{indent} argument so that this string can be embedded + in a nested display of other data. + + Example:: + integer = Word(nums) + date_str = integer("year") + '/' + integer("month") + '/' + integer("day") + + result = date_str.parseString('12/31/1999') + print(result.dump()) + prints:: + ['12', '/', '31', '/', '1999'] + - day: 1999 + - month: 31 + - year: 12 + """ out = [] NL = '\n' out.append( indent+_ustr(self.asList()) ) - if self.haskeys(): - items = sorted(self.items()) - for k,v in items: - if out: - out.append(NL) - out.append( "%s%s- %s: " % (indent,(' '*depth), k) ) - if isinstance(v,ParseResults): - if v: - out.append( v.dump(indent,depth+1) ) + if full: + if self.haskeys(): + items = sorted(self.items()) + for k,v in items: + if out: + out.append(NL) + out.append( "%s%s- %s: " % (indent,(' '*depth), k) ) + if isinstance(v,ParseResults): + if v: + out.append( v.dump(indent,depth+1) ) + else: + out.append(_ustr(v)) else: out.append(_ustr(v)) - else: - out.append(_ustr(v)) - elif any(isinstance(vv,ParseResults) for vv in self): - v = self - for i,vv in enumerate(v): - if isinstance(vv,ParseResults): - out.append("\n%s%s[%d]:\n%s%s%s" % (indent,(' '*(depth)),i,indent,(' '*(depth+1)),vv.dump(indent,depth+1) )) - else: - out.append("\n%s%s[%d]:\n%s%s%s" % (indent,(' '*(depth)),i,indent,(' '*(depth+1)),_ustr(vv))) + elif any(isinstance(vv,ParseResults) for vv in self): + v = self + for i,vv in enumerate(v): + if isinstance(vv,ParseResults): + out.append("\n%s%s[%d]:\n%s%s%s" % (indent,(' '*(depth)),i,indent,(' '*(depth+1)),vv.dump(indent,depth+1) )) + else: + out.append("\n%s%s[%d]:\n%s%s%s" % (indent,(' '*(depth)),i,indent,(' '*(depth+1)),_ustr(vv))) return "".join(out) def pprint(self, *args, **kwargs): - """Pretty-printer for parsed results as a list, using the C{pprint} module. - Accepts additional positional or keyword args as defined for the - C{pprint.pprint} method. (U{http://docs.python.org/3/library/pprint.html#pprint.pprint})""" + """ + Pretty-printer for parsed results as a list, using the C{pprint} module. + Accepts additional positional or keyword args as defined for the + C{pprint.pprint} method. (U{http://docs.python.org/3/library/pprint.html#pprint.pprint}) + + Example:: + ident = Word(alphas, alphanums) + num = Word(nums) + func = Forward() + term = ident | num | Group('(' + func + ')') + func <<= ident + Group(Optional(delimitedList(term))) + result = func.parseString("fna a,b,(fnb c,d,200),100") + result.pprint(width=40) + prints:: + ['fna', + ['a', + 'b', + ['(', 'fnb', ['c', 'd', '200'], ')'], + '100']] + """ pprint.pprint(self.asList(), *args, **kwargs) # add support for pickle protocol @@ -690,8 +930,11 @@ class ParseResults(object): else: self.__parent = None + def __getnewargs__(self): + return self.__toklist, self.__name, self.__asList, self.__modal + def __dir__(self): - return dir(super(ParseResults,self)) + list(self.keys()) + return (dir(type(self)) + list(self.keys())) collections.MutableMapping.register(ParseResults) @@ -771,6 +1014,31 @@ def _trim_arity(func, maxargs=2): return lambda s,l,t: func(t) limit = [0] foundArity = [False] + + # traceback return data structure changed in Py3.5 - normalize back to plain tuples + if system_version[:2] >= (3,5): + def extract_stack(limit=0): + # special handling for Python 3.5.0 - extra deep call stack by 1 + offset = -3 if system_version == (3,5,0) else -2 + frame_summary = traceback.extract_stack(limit=-offset+limit-1)[offset] + return [(frame_summary.filename, frame_summary.lineno)] + def extract_tb(tb, limit=0): + frames = traceback.extract_tb(tb, limit=limit) + frame_summary = frames[-1] + return [(frame_summary.filename, frame_summary.lineno)] + else: + extract_stack = traceback.extract_stack + extract_tb = traceback.extract_tb + + # synthesize what would be returned by traceback.extract_stack at the call to + # user's parse action 'func', so that we don't incur call penalty at parse time + + LINE_DIFF = 6 + # IF ANY CODE CHANGES, EVEN JUST COMMENTS OR BLANK LINES, BETWEEN THE NEXT LINE AND + # THE CALL TO FUNC INSIDE WRAPPER, LINE_DIFF MUST BE MODIFIED!!!! + this_line = extract_stack(limit=2)[-1] + pa_call_line_synth = (this_line[0], this_line[1]+LINE_DIFF) + def wrapper(*args): while 1: try: @@ -778,12 +1046,33 @@ def _trim_arity(func, maxargs=2): foundArity[0] = True return ret except TypeError: - if limit[0] <= maxargs and not foundArity[0]: + # re-raise TypeErrors if they did not come from our arity testing + if foundArity[0]: + raise + else: + try: + tb = sys.exc_info()[-1] + if not extract_tb(tb, limit=2)[-1][:2] == pa_call_line_synth: + raise + finally: + del tb + + if limit[0] <= maxargs: limit[0] += 1 continue raise + + # copy func name to wrapper for sensible debug output + func_name = "" + try: + func_name = getattr(func, '__name__', + getattr(func, '__class__').__name__) + except Exception: + func_name = str(func) + wrapper.__name__ = func_name + return wrapper - + class ParserElement(object): """Abstract base level parser element class.""" DEFAULT_WHITE_CHARS = " \n\t\r" @@ -791,7 +1080,16 @@ class ParserElement(object): @staticmethod def setDefaultWhitespaceChars( chars ): - """Overrides the default whitespace chars + r""" + Overrides the default whitespace chars + + Example:: + # default whitespace chars are space, and newline + OneOrMore(Word(alphas)).parseString("abc def\nghi jkl") # -> ['abc', 'def', 'ghi', 'jkl'] + + # change to just treat newline as significant + ParserElement.setDefaultWhitespaceChars(" \t") + OneOrMore(Word(alphas)).parseString("abc def\nghi jkl") # -> ['abc', 'def'] """ ParserElement.DEFAULT_WHITE_CHARS = chars @@ -799,8 +1097,22 @@ class ParserElement(object): def inlineLiteralsUsing(cls): """ Set class to be used for inclusion of string literals into a parser. + + Example:: + # default literal class used is Literal + integer = Word(nums) + date_str = integer("year") + '/' + integer("month") + '/' + integer("day") + + date_str.parseString("1999/12/31") # -> ['1999', '/', '12', '/', '31'] + + + # change to Suppress + ParserElement.inlineLiteralsUsing(Suppress) + date_str = integer("year") + '/' + integer("month") + '/' + integer("day") + + date_str.parseString("1999/12/31") # -> ['1999', '12', '31'] """ - ParserElement.literalStringClass = cls + ParserElement._literalStringClass = cls def __init__( self, savelist=False ): self.parseAction = list() @@ -826,8 +1138,21 @@ class ParserElement(object): self.callDuringTry = False def copy( self ): - """Make a copy of this C{ParserElement}. Useful for defining different parse actions - for the same parsing pattern, using copies of the original parse element.""" + """ + Make a copy of this C{ParserElement}. Useful for defining different parse actions + for the same parsing pattern, using copies of the original parse element. + + Example:: + integer = Word(nums).setParseAction(lambda toks: int(toks[0])) + integerK = integer.copy().addParseAction(lambda toks: toks[0]*1024) + Suppress("K") + integerM = integer.copy().addParseAction(lambda toks: toks[0]*1024*1024) + Suppress("M") + + print(OneOrMore(integerK | integerM | integer).parseString("5K 100 640K 256M")) + prints:: + [5120, 100, 655360, 268435456] + Equivalent form of C{expr.copy()} is just C{expr()}:: + integerM = integer().addParseAction(lambda toks: toks[0]*1024*1024) + Suppress("M") + """ cpy = copy.copy( self ) cpy.parseAction = self.parseAction[:] cpy.ignoreExprs = self.ignoreExprs[:] @@ -836,7 +1161,13 @@ class ParserElement(object): return cpy def setName( self, name ): - """Define name for this expression, for use in debugging.""" + """ + Define name for this expression, makes debugging and exception messages clearer. + + Example:: + Word(nums).parseString("ABC") # -> Exception: Expected W:(0123...) (at char 0), (line:1, col:1) + Word(nums).setName("integer").parseString("ABC") # -> Exception: Expected integer (at char 0), (line:1, col:1) + """ self.name = name self.errmsg = "Expected " + self.name if hasattr(self,"exception"): @@ -844,15 +1175,24 @@ class ParserElement(object): return self def setResultsName( self, name, listAllMatches=False ): - """Define name for referencing matching tokens as a nested attribute - of the returned parse results. - NOTE: this returns a *copy* of the original C{ParserElement} object; - this is so that the client can define a basic element, such as an - integer, and reference it in multiple places with different names. - - You can also set results names using the abbreviated syntax, - C{expr("name")} in place of C{expr.setResultsName("name")} - - see L{I{__call__}<__call__>}. + """ + Define name for referencing matching tokens as a nested attribute + of the returned parse results. + NOTE: this returns a *copy* of the original C{ParserElement} object; + this is so that the client can define a basic element, such as an + integer, and reference it in multiple places with different names. + + You can also set results names using the abbreviated syntax, + C{expr("name")} in place of C{expr.setResultsName("name")} - + see L{I{__call__}<__call__>}. + + Example:: + date_str = (integer.setResultsName("year") + '/' + + integer.setResultsName("month") + '/' + + integer.setResultsName("day")) + + # equivalent form: + date_str = integer("year") + '/' + integer("month") + '/' + integer("day") """ newself = self.copy() if name.endswith("*"): @@ -881,42 +1221,76 @@ class ParserElement(object): return self def setParseAction( self, *fns, **kwargs ): - """Define action to perform when successfully matching parse element definition. - Parse action fn is a callable method with 0-3 arguments, called as C{fn(s,loc,toks)}, - C{fn(loc,toks)}, C{fn(toks)}, or just C{fn()}, where: - - s = the original string being parsed (see note below) - - loc = the location of the matching substring - - toks = a list of the matched tokens, packaged as a C{L{ParseResults}} object - If the functions in fns modify the tokens, they can return them as the return - value from fn, and the modified list of tokens will replace the original. - Otherwise, fn does not need to return any value. - - Note: the default parsing behavior is to expand tabs in the input string - before starting the parsing process. See L{I{parseString}} for more information - on parsing strings containing C{}s, and suggested methods to maintain a - consistent view of the parsed string, the parse location, and line and column - positions within the parsed string. - """ + """ + Define action to perform when successfully matching parse element definition. + Parse action fn is a callable method with 0-3 arguments, called as C{fn(s,loc,toks)}, + C{fn(loc,toks)}, C{fn(toks)}, or just C{fn()}, where: + - s = the original string being parsed (see note below) + - loc = the location of the matching substring + - toks = a list of the matched tokens, packaged as a C{L{ParseResults}} object + If the functions in fns modify the tokens, they can return them as the return + value from fn, and the modified list of tokens will replace the original. + Otherwise, fn does not need to return any value. + + Optional keyword arguments: + - callDuringTry = (default=C{False}) indicate if parse action should be run during lookaheads and alternate testing + + Note: the default parsing behavior is to expand tabs in the input string + before starting the parsing process. See L{I{parseString}} for more information + on parsing strings containing C{}s, and suggested methods to maintain a + consistent view of the parsed string, the parse location, and line and column + positions within the parsed string. + + Example:: + integer = Word(nums) + date_str = integer + '/' + integer + '/' + integer + + date_str.parseString("1999/12/31") # -> ['1999', '/', '12', '/', '31'] + + # use parse action to convert to ints at parse time + integer = Word(nums).setParseAction(lambda toks: int(toks[0])) + date_str = integer + '/' + integer + '/' + integer + + # note that integer fields are now ints, not strings + date_str.parseString("1999/12/31") # -> [1999, '/', 12, '/', 31] + """ self.parseAction = list(map(_trim_arity, list(fns))) self.callDuringTry = kwargs.get("callDuringTry", False) return self def addParseAction( self, *fns, **kwargs ): - """Add parse action to expression's list of parse actions. See L{I{setParseAction}}.""" + """ + Add parse action to expression's list of parse actions. See L{I{setParseAction}}. + + See examples in L{I{copy}}. + """ self.parseAction += list(map(_trim_arity, list(fns))) self.callDuringTry = self.callDuringTry or kwargs.get("callDuringTry", False) return self def addCondition(self, *fns, **kwargs): """Add a boolean predicate function to expression's list of parse actions. See - L{I{setParseAction}}. Optional keyword argument C{message} can - be used to define a custom message to be used in the raised exception.""" - msg = kwargs.get("message") or "failed user-defined condition" + L{I{setParseAction}} for function call signatures. Unlike C{setParseAction}, + functions passed to C{addCondition} need to return boolean success/fail of the condition. + + Optional keyword arguments: + - message = define a custom message to be used in the raised exception + - fatal = if True, will raise ParseFatalException to stop parsing immediately; otherwise will raise ParseException + + Example:: + integer = Word(nums).setParseAction(lambda toks: int(toks[0])) + year_int = integer.copy() + year_int.addCondition(lambda toks: toks[0] >= 2000, message="Only support years 2000 and later") + date_str = year_int + '/' + integer + '/' + integer + + result = date_str.parseString("1999/12/31") # -> Exception: Only support years 2000 and later (at char 0), (line:1, col:1) + """ + msg = kwargs.get("message", "failed user-defined condition") + exc_type = ParseFatalException if kwargs.get("fatal", False) else ParseException for fn in fns: def pa(s,l,t): if not bool(_trim_arity(fn)(s,l,t)): - raise ParseException(s,l,msg) - return t + raise exc_type(s,l,msg) self.parseAction.append(pa) self.callDuringTry = self.callDuringTry or kwargs.get("callDuringTry", False) return self @@ -1043,43 +1417,132 @@ class ParserElement(object): return self._parse( instring, loc, doActions=False )[0] except ParseFatalException: raise ParseException( instring, loc, self.errmsg, self) + + def canParseNext(self, instring, loc): + try: + self.tryParse(instring, loc) + except (ParseException, IndexError): + return False + else: + return True + + class _UnboundedCache(object): + def __init__(self): + cache = {} + self.not_in_cache = not_in_cache = object() + + def get(self, key): + return cache.get(key, not_in_cache) + + def set(self, key, value): + cache[key] = value + + def clear(self): + cache.clear() + + self.get = types.MethodType(get, self) + self.set = types.MethodType(set, self) + self.clear = types.MethodType(clear, self) + + if _OrderedDict is not None: + class _FifoCache(object): + def __init__(self, size): + self.not_in_cache = not_in_cache = object() + + cache = _OrderedDict() + + def get(self, key): + return cache.get(key, not_in_cache) + + def set(self, key, value): + cache[key] = value + if len(cache) > size: + cache.popitem(False) + + def clear(self): + cache.clear() + + self.get = types.MethodType(get, self) + self.set = types.MethodType(set, self) + self.clear = types.MethodType(clear, self) + + else: + class _FifoCache(object): + def __init__(self, size): + self.not_in_cache = not_in_cache = object() + + cache = {} + key_fifo = collections.deque([], size) + + def get(self, key): + return cache.get(key, not_in_cache) + + def set(self, key, value): + cache[key] = value + if len(cache) > size: + cache.pop(key_fifo.popleft(), None) + key_fifo.append(key) + + def clear(self): + cache.clear() + key_fifo.clear() + + self.get = types.MethodType(get, self) + self.set = types.MethodType(set, self) + self.clear = types.MethodType(clear, self) + + # argument cache for optimizing repeated calls when backtracking through recursive expressions + packrat_cache = {} # this is set later by enabledPackrat(); this is here so that resetCache() doesn't fail + packrat_cache_lock = RLock() + packrat_cache_stats = [0, 0] # this method gets repeatedly called during backtracking with the same arguments - # we can cache these arguments and save ourselves the trouble of re-parsing the contained expression def _parseCache( self, instring, loc, doActions=True, callPreParse=True ): - lookup = (self,instring,loc,callPreParse,doActions) - if lookup in ParserElement._exprArgCache: - value = ParserElement._exprArgCache[ lookup ] - if isinstance(value, Exception): - raise value - return (value[0],value[1].copy()) - else: - try: - value = self._parseNoCache( instring, loc, doActions, callPreParse ) - ParserElement._exprArgCache[ lookup ] = (value[0],value[1].copy()) - return value - except ParseBaseException as pe: - pe.__traceback__ = None - ParserElement._exprArgCache[ lookup ] = pe - raise + HIT, MISS = 0, 1 + lookup = (self, instring, loc, callPreParse, doActions) + with ParserElement.packrat_cache_lock: + cache = ParserElement.packrat_cache + value = cache.get(lookup) + if value is cache.not_in_cache: + ParserElement.packrat_cache_stats[MISS] += 1 + try: + value = self._parseNoCache(instring, loc, doActions, callPreParse) + except ParseBaseException as pe: + # cache a copy of the exception, without the traceback + cache.set(lookup, pe.__class__(*pe.args)) + raise + else: + cache.set(lookup, (value[0], value[1].copy())) + return value + else: + ParserElement.packrat_cache_stats[HIT] += 1 + if isinstance(value, Exception): + raise value + return (value[0], value[1].copy()) _parse = _parseNoCache - # argument cache for optimizing repeated calls when backtracking through recursive expressions - _exprArgCache = {} @staticmethod def resetCache(): - ParserElement._exprArgCache.clear() + ParserElement.packrat_cache.clear() + ParserElement.packrat_cache_stats[:] = [0] * len(ParserElement.packrat_cache_stats) _packratEnabled = False @staticmethod - def enablePackrat(): + def enablePackrat(cache_size_limit=128): """Enables "packrat" parsing, which adds memoizing to the parsing logic. Repeated parse attempts at the same string location (which happens often in many complex grammars) can immediately return a cached value, instead of re-executing parsing/validating code. Memoizing is done of both valid results and parsing exceptions. - + + Parameters: + - cache_size_limit - (default=C{128}) - if an integer value is provided + will limit the size of the packrat cache; if None is passed, then + the cache size will be unbounded; if 0 is passed, the cache will + be effectively disabled. + This speedup may break existing programs that use parse actions that have side-effects. For this reason, packrat parsing is disabled when you first import pyparsing. To activate the packrat feature, your @@ -1088,32 +1551,45 @@ class ParserElement(object): C{enablePackrat} before calling C{psyco.full()}. If you do not do this, Python will crash. For best results, call C{enablePackrat()} immediately after importing pyparsing. + + Example:: + import pyparsing + pyparsing.ParserElement.enablePackrat() """ if not ParserElement._packratEnabled: ParserElement._packratEnabled = True + if cache_size_limit is None: + ParserElement.packrat_cache = ParserElement._UnboundedCache() + else: + ParserElement.packrat_cache = ParserElement._FifoCache(cache_size_limit) ParserElement._parse = ParserElement._parseCache def parseString( self, instring, parseAll=False ): - """Execute the parse expression with the given string. - This is the main interface to the client code, once the complete - expression has been built. - - If you want the grammar to require that the entire input string be - successfully parsed, then set C{parseAll} to True (equivalent to ending - the grammar with C{L{StringEnd()}}). - - Note: C{parseString} implicitly calls C{expandtabs()} on the input string, - in order to report proper column numbers in parse actions. - If the input string contains tabs and - the grammar uses parse actions that use the C{loc} argument to index into the - string being parsed, you can ensure you have a consistent view of the input - string by: - - calling C{parseWithTabs} on your grammar before calling C{parseString} - (see L{I{parseWithTabs}}) - - define your parse action using the full C{(s,loc,toks)} signature, and - reference the input string using the parse action's C{s} argument - - explicitly expand the tabs in your input string before calling - C{parseString} + """ + Execute the parse expression with the given string. + This is the main interface to the client code, once the complete + expression has been built. + + If you want the grammar to require that the entire input string be + successfully parsed, then set C{parseAll} to True (equivalent to ending + the grammar with C{L{StringEnd()}}). + + Note: C{parseString} implicitly calls C{expandtabs()} on the input string, + in order to report proper column numbers in parse actions. + If the input string contains tabs and + the grammar uses parse actions that use the C{loc} argument to index into the + string being parsed, you can ensure you have a consistent view of the input + string by: + - calling C{parseWithTabs} on your grammar before calling C{parseString} + (see L{I{parseWithTabs}}) + - define your parse action using the full C{(s,loc,toks)} signature, and + reference the input string using the parse action's C{s} argument + - explictly expand the tabs in your input string before calling + C{parseString} + + Example:: + Word('a').parseString('aaaaabaaa') # -> ['aaaaa'] + Word('a').parseString('aaaaabaaa', parseAll=True) # -> Exception: Expected end of text """ ParserElement.resetCache() if not self.streamlined: @@ -1139,14 +1615,35 @@ class ParserElement(object): return tokens def scanString( self, instring, maxMatches=_MAX_INT, overlap=False ): - """Scan the input string for expression matches. Each match will return the - matching tokens, start location, and end location. May be called with optional - C{maxMatches} argument, to clip scanning after 'n' matches are found. If - C{overlap} is specified, then overlapping matches will be reported. - - Note that the start and end locations are reported relative to the string - being parsed. See L{I{parseString}} for more information on parsing - strings with embedded tabs.""" + """ + Scan the input string for expression matches. Each match will return the + matching tokens, start location, and end location. May be called with optional + C{maxMatches} argument, to clip scanning after 'n' matches are found. If + C{overlap} is specified, then overlapping matches will be reported. + + Note that the start and end locations are reported relative to the string + being parsed. See L{I{parseString}} for more information on parsing + strings with embedded tabs. + + Example:: + source = "sldjf123lsdjjkf345sldkjf879lkjsfd987" + print(source) + for tokens,start,end in Word(alphas).scanString(source): + print(' '*start + '^'*(end-start)) + print(' '*start + tokens[0]) + + prints:: + + sldjf123lsdjjkf345sldkjf879lkjsfd987 + ^^^^^ + sldjf + ^^^^^^^ + lsdjjkf + ^^^^^^ + sldkjf + ^^^^^^ + lkjsfd + """ if not self.streamlined: self.streamline() for e in self.ignoreExprs: @@ -1189,12 +1686,22 @@ class ParserElement(object): raise exc def transformString( self, instring ): - """Extension to C{L{scanString}}, to modify matching text with modified tokens that may - be returned from a parse action. To use C{transformString}, define a grammar and - attach a parse action to it that modifies the returned token list. - Invoking C{transformString()} on a target string will then scan for matches, - and replace the matched text patterns according to the logic in the parse - action. C{transformString()} returns the resulting transformed string.""" + """ + Extension to C{L{scanString}}, to modify matching text with modified tokens that may + be returned from a parse action. To use C{transformString}, define a grammar and + attach a parse action to it that modifies the returned token list. + Invoking C{transformString()} on a target string will then scan for matches, + and replace the matched text patterns according to the logic in the parse + action. C{transformString()} returns the resulting transformed string. + + Example:: + wd = Word(alphas) + wd.setParseAction(lambda toks: toks[0].title()) + + print(wd.transformString("now is the winter of our discontent made glorious summer by this sun of york.")) + Prints:: + Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York. + """ out = [] lastE = 0 # force preservation of s, to minimize unwanted transformation of string, and to @@ -1222,9 +1729,18 @@ class ParserElement(object): raise exc def searchString( self, instring, maxMatches=_MAX_INT ): - """Another extension to C{L{scanString}}, simplifying the access to the tokens found - to match the given parse expression. May be called with optional - C{maxMatches} argument, to clip searching after 'n' matches are found. + """ + Another extension to C{L{scanString}}, simplifying the access to the tokens found + to match the given parse expression. May be called with optional + C{maxMatches} argument, to clip searching after 'n' matches are found. + + Example:: + # a capitalized word starts with an uppercase letter, followed by zero or more lowercase letters + cap_word = Word(alphas.upper(), alphas.lower()) + + print(cap_word.searchString("More than Iron, more than Lead, more than Gold I need Electricity")) + prints:: + ['More', 'Iron', 'Lead', 'Gold', 'I'] """ try: return ParseResults([ t for t,s,e in self.scanString( instring, maxMatches ) ]) @@ -1235,10 +1751,34 @@ class ParserElement(object): # catch and re-raise exception from here, clears out pyparsing internal stack trace raise exc + def split(self, instring, maxsplit=_MAX_INT, includeSeparators=False): + """ + Generator method to split a string using the given expression as a separator. + May be called with optional C{maxsplit} argument, to limit the number of splits; + and the optional C{includeSeparators} argument (default=C{False}), if the separating + matching text should be included in the split results. + + Example:: + punc = oneOf(list(".,;:/-!?")) + print(list(punc.split("This, this?, this sentence, is badly punctuated!"))) + prints:: + ['This', ' this', '', ' this sentence', ' is badly punctuated', ''] + """ + splits = 0 + last = 0 + for t,s,e in self.scanString(instring, maxMatches=maxsplit): + yield instring[last:s] + if includeSeparators: + yield t[0] + last = e + yield instring[last:] + def __add__(self, other ): - """Implementation of + operator - returns C{L{And}}""" + """ + Implementation of + operator - returns C{L{And}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1246,9 +1786,11 @@ class ParserElement(object): return And( [ self, other ] ) def __radd__(self, other ): - """Implementation of + operator when left operand is not a C{L{ParserElement}}""" + """ + Implementation of + operator when left operand is not a C{L{ParserElement}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1256,9 +1798,11 @@ class ParserElement(object): return other + self def __sub__(self, other): - """Implementation of - operator, returns C{L{And}} with error stop""" + """ + Implementation of - operator, returns C{L{And}} with error stop + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1266,9 +1810,11 @@ class ParserElement(object): return And( [ self, And._ErrorStop(), other ] ) def __rsub__(self, other ): - """Implementation of - operator when left operand is not a C{L{ParserElement}}""" + """ + Implementation of - operator when left operand is not a C{L{ParserElement}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1276,24 +1822,24 @@ class ParserElement(object): return other - self def __mul__(self,other): - """Implementation of * operator, allows use of C{expr * 3} in place of - C{expr + expr + expr}. Expressions may also me multiplied by a 2-integer - tuple, similar to C{{min,max}} multipliers in regular expressions. Tuples - may also include C{None} as in: - - C{expr*(n,None)} or C{expr*(n,)} is equivalent + """ + Implementation of * operator, allows use of C{expr * 3} in place of + C{expr + expr + expr}. Expressions may also me multiplied by a 2-integer + tuple, similar to C{{min,max}} multipliers in regular expressions. Tuples + may also include C{None} as in: + - C{expr*(n,None)} or C{expr*(n,)} is equivalent to C{expr*n + L{ZeroOrMore}(expr)} (read as "at least n instances of C{expr}") - - C{expr*(None,n)} is equivalent to C{expr*(0,n)} + - C{expr*(None,n)} is equivalent to C{expr*(0,n)} (read as "0 to n instances of C{expr}") - - C{expr*(None,None)} is equivalent to C{L{ZeroOrMore}(expr)} - - C{expr*(1,None)} is equivalent to C{L{OneOrMore}(expr)} - - Note that C{expr*(None,n)} does not raise an exception if - more than n exprs exist in the input stream; that is, - C{expr*(None,n)} does not enforce a maximum number of expr - occurrences. If this behavior is desired, then write - C{expr*(None,n) + ~expr} - + - C{expr*(None,None)} is equivalent to C{L{ZeroOrMore}(expr)} + - C{expr*(1,None)} is equivalent to C{L{OneOrMore}(expr)} + + Note that C{expr*(None,n)} does not raise an exception if + more than n exprs exist in the input stream; that is, + C{expr*(None,n)} does not enforce a maximum number of expr + occurrences. If this behavior is desired, then write + C{expr*(None,n) + ~expr} """ if isinstance(other,int): minElements, optElements = other,0 @@ -1347,9 +1893,11 @@ class ParserElement(object): return self.__mul__(other) def __or__(self, other ): - """Implementation of | operator - returns C{L{MatchFirst}}""" + """ + Implementation of | operator - returns C{L{MatchFirst}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1357,9 +1905,11 @@ class ParserElement(object): return MatchFirst( [ self, other ] ) def __ror__(self, other ): - """Implementation of | operator when left operand is not a C{L{ParserElement}}""" + """ + Implementation of | operator when left operand is not a C{L{ParserElement}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1367,9 +1917,11 @@ class ParserElement(object): return other | self def __xor__(self, other ): - """Implementation of ^ operator - returns C{L{Or}}""" + """ + Implementation of ^ operator - returns C{L{Or}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1377,9 +1929,11 @@ class ParserElement(object): return Or( [ self, other ] ) def __rxor__(self, other ): - """Implementation of ^ operator when left operand is not a C{L{ParserElement}}""" + """ + Implementation of ^ operator when left operand is not a C{L{ParserElement}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1387,9 +1941,11 @@ class ParserElement(object): return other ^ self def __and__(self, other ): - """Implementation of & operator - returns C{L{Each}}""" + """ + Implementation of & operator - returns C{L{Each}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1397,9 +1953,11 @@ class ParserElement(object): return Each( [ self, other ] ) def __rand__(self, other ): - """Implementation of & operator when left operand is not a C{L{ParserElement}}""" + """ + Implementation of & operator when left operand is not a C{L{ParserElement}} + """ if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) if not isinstance( other, ParserElement ): warnings.warn("Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2) @@ -1407,41 +1965,49 @@ class ParserElement(object): return other & self def __invert__( self ): - """Implementation of ~ operator - returns C{L{NotAny}}""" + """ + Implementation of ~ operator - returns C{L{NotAny}} + """ return NotAny( self ) def __call__(self, name=None): - """Shortcut for C{L{setResultsName}}, with C{listAllMatches=default}:: - userdata = Word(alphas).setResultsName("name") + Word(nums+"-").setResultsName("socsecno") - could be written as:: - userdata = Word(alphas)("name") + Word(nums+"-")("socsecno") - - If C{name} is given with a trailing C{'*'} character, then C{listAllMatches} will be - passed as C{True}. + """ + Shortcut for C{L{setResultsName}}, with C{listAllMatches=default}. + + If C{name} is given with a trailing C{'*'} character, then C{listAllMatches} will be + passed as C{True}. - If C{name} is omitted, same as calling C{L{copy}}. - """ + If C{name} is omitted, same as calling C{L{copy}}. + + Example:: + # these are equivalent + userdata = Word(alphas).setResultsName("name") + Word(nums+"-").setResultsName("socsecno") + userdata = Word(alphas)("name") + Word(nums+"-")("socsecno") + """ if name is not None: return self.setResultsName(name) else: return self.copy() def suppress( self ): - """Suppresses the output of this C{ParserElement}; useful to keep punctuation from - cluttering up returned output. + """ + Suppresses the output of this C{ParserElement}; useful to keep punctuation from + cluttering up returned output. """ return Suppress( self ) def leaveWhitespace( self ): - """Disables the skipping of whitespace before matching the characters in the - C{ParserElement}'s defined pattern. This is normally only used internally by - the pyparsing module, but may be needed in some whitespace-sensitive grammars. + """ + Disables the skipping of whitespace before matching the characters in the + C{ParserElement}'s defined pattern. This is normally only used internally by + the pyparsing module, but may be needed in some whitespace-sensitive grammars. """ self.skipWhitespace = False return self def setWhitespaceChars( self, chars ): - """Overrides the default whitespace chars + """ + Overrides the default whitespace chars """ self.skipWhitespace = True self.whiteChars = chars @@ -1449,26 +2015,41 @@ class ParserElement(object): return self def parseWithTabs( self ): - """Overrides default behavior to expand C{}s to spaces before parsing the input string. - Must be called before C{parseString} when the input grammar contains elements that - match C{} characters.""" + """ + Overrides default behavior to expand C{}s to spaces before parsing the input string. + Must be called before C{parseString} when the input grammar contains elements that + match C{} characters. + """ self.keepTabs = True return self def ignore( self, other ): - """Define expression to be ignored (e.g., comments) while doing pattern - matching; may be called repeatedly, to define multiple comment or other - ignorable patterns. """ + Define expression to be ignored (e.g., comments) while doing pattern + matching; may be called repeatedly, to define multiple comment or other + ignorable patterns. + + Example:: + patt = OneOrMore(Word(alphas)) + patt.parseString('ablaj /* comment */ lskjd') # -> ['ablaj'] + + patt.ignore(cStyleComment) + patt.parseString('ablaj /* comment */ lskjd') # -> ['ablaj', 'lskjd'] + """ + if isinstance(other, basestring): + other = Suppress(other) + if isinstance( other, Suppress ): if other not in self.ignoreExprs: - self.ignoreExprs.append( other.copy() ) + self.ignoreExprs.append(other) else: self.ignoreExprs.append( Suppress( other.copy() ) ) return self def setDebugActions( self, startAction, successAction, exceptionAction ): - """Enable display of debugging messages while doing pattern matching.""" + """ + Enable display of debugging messages while doing pattern matching. + """ self.debugActions = (startAction or _defaultStartDebugAction, successAction or _defaultSuccessDebugAction, exceptionAction or _defaultExceptionDebugAction) @@ -1476,8 +2057,39 @@ class ParserElement(object): return self def setDebug( self, flag=True ): - """Enable display of debugging messages while doing pattern matching. - Set C{flag} to True to enable, False to disable.""" + """ + Enable display of debugging messages while doing pattern matching. + Set C{flag} to True to enable, False to disable. + + Example:: + wd = Word(alphas).setName("alphaword") + integer = Word(nums).setName("numword") + term = wd | integer + + # turn on debugging for wd + wd.setDebug() + + OneOrMore(term).parseString("abc 123 xyz 890") + + prints:: + Match alphaword at loc 0(1,1) + Matched alphaword -> ['abc'] + Match alphaword at loc 3(1,4) + Exception raised:Expected alphaword (at char 4), (line:1, col:5) + Match alphaword at loc 7(1,8) + Matched alphaword -> ['xyz'] + Match alphaword at loc 11(1,12) + Exception raised:Expected alphaword (at char 12), (line:1, col:13) + Match alphaword at loc 15(1,16) + Exception raised:Expected alphaword (at char 15), (line:1, col:16) + + The output shown is that produced by the default debug actions. Prior to attempting + to match the C{wd} expression, the debugging message C{"Match at loc (,)"} + is shown. Then if the parse succeeds, a C{"Matched"} message is shown, or an C{"Exception raised"} + message is shown. Also note the use of L{setName} to assign a human-readable name to the expression, + which makes debugging and exception messages easier to understand - for instance, the default + name created for the C{Word} expression without calling C{setName} is C{"W:(ABCD...)"}. + """ if flag: self.setDebugActions( _defaultStartDebugAction, _defaultSuccessDebugAction, _defaultExceptionDebugAction ) else: @@ -1499,20 +2111,22 @@ class ParserElement(object): pass def validate( self, validateTrace=[] ): - """Check defined expressions for valid structure, check for infinite recursive definitions.""" + """ + Check defined expressions for valid structure, check for infinite recursive definitions. + """ self.checkRecursion( [] ) def parseFile( self, file_or_filename, parseAll=False ): - """Execute the parse expression on the given file or filename. - If a filename is specified (instead of a file object), - the entire file is opened, read, and closed before parsing. + """ + Execute the parse expression on the given file or filename. + If a filename is specified (instead of a file object), + the entire file is opened, read, and closed before parsing. """ try: file_contents = file_or_filename.read() except AttributeError: - f = open(file_or_filename, "r") - file_contents = f.read() - f.close() + with open(file_or_filename, "r") as f: + file_contents = f.read() try: return self.parseString(file_contents, parseAll) except ParseBaseException as exc: @@ -1524,13 +2138,9 @@ class ParserElement(object): def __eq__(self,other): if isinstance(other, ParserElement): - return self is other or self.__dict__ == other.__dict__ + return self is other or vars(self) == vars(other) elif isinstance(other, basestring): - try: - self.parseString(_ustr(other), parseAll=True) - return True - except ParseBaseException: - return False + return self.matches(other) else: return super(ParserElement,self)==other @@ -1546,40 +2156,161 @@ class ParserElement(object): def __rne__(self,other): return not (self == other) - def runTests(self, tests, parseAll=False): - """Execute the parse expression on a series of test strings, showing each - test, the parsed results or where the parse failed. Quick and easy way to - run a parse expression against a list of sample strings. + def matches(self, testString, parseAll=True): + """ + Method for quick testing of a parser against a test string. Good for simple + inline microtests of sub expressions while building up larger parser.0 - Parameters: - - tests - a list of separate test strings, or a multiline string of test strings - - parseAll - (default=False) - flag to pass to C{L{parseString}} when running tests + Parameters: + - testString - to test against this expression for a match + - parseAll - (default=C{True}) - flag to pass to C{L{parseString}} when running tests + + Example:: + expr = Word(nums) + assert expr.matches("100") + """ + try: + self.parseString(_ustr(testString), parseAll=parseAll) + return True + except ParseBaseException: + return False + + def runTests(self, tests, parseAll=True, comment='#', fullDump=True, printResults=True, failureTests=False): + """ + Execute the parse expression on a series of test strings, showing each + test, the parsed results or where the parse failed. Quick and easy way to + run a parse expression against a list of sample strings. + + Parameters: + - tests - a list of separate test strings, or a multiline string of test strings + - parseAll - (default=C{True}) - flag to pass to C{L{parseString}} when running tests + - comment - (default=C{'#'}) - expression for indicating embedded comments in the test + string; pass None to disable comment filtering + - fullDump - (default=C{True}) - dump results as list followed by results names in nested outline; + if False, only dump nested list + - printResults - (default=C{True}) prints test output to stdout + - failureTests - (default=C{False}) indicates if these tests are expected to fail parsing + + Returns: a (success, results) tuple, where success indicates that all tests succeeded + (or failed if C{failureTests} is True), and the results contain a list of lines of each + test's output + + Example:: + number_expr = pyparsing_common.number.copy() + + result = number_expr.runTests(''' + # unsigned integer + 100 + # negative integer + -100 + # float with scientific notation + 6.02e23 + # integer with scientific notation + 1e-12 + ''') + print("Success" if result[0] else "Failed!") + + result = number_expr.runTests(''' + # stray character + 100Z + # missing leading digit before '.' + -.100 + # too many '.' + 3.14.159 + ''', failureTests=True) + print("Success" if result[0] else "Failed!") + prints:: + # unsigned integer + 100 + [100] + + # negative integer + -100 + [-100] + + # float with scientific notation + 6.02e23 + [6.02e+23] + + # integer with scientific notation + 1e-12 + [1e-12] + + Success + + # stray character + 100Z + ^ + FAIL: Expected end of text (at char 3), (line:1, col:4) + + # missing leading digit before '.' + -.100 + ^ + FAIL: Expected {real number with scientific notation | real number | signed integer} (at char 0), (line:1, col:1) + + # too many '.' + 3.14.159 + ^ + FAIL: Expected end of text (at char 4), (line:1, col:5) + + Success """ if isinstance(tests, basestring): - tests = map(str.strip, tests.splitlines()) + tests = list(map(str.strip, tests.rstrip().splitlines())) + if isinstance(comment, basestring): + comment = Literal(comment) + allResults = [] + comments = [] + success = True for t in tests: - out = [t] + if comment is not None and comment.matches(t, False) or comments and not t: + comments.append(t) + continue + if not t: + continue + out = ['\n'.join(comments), t] + comments = [] try: - out.append(self.parseString(t, parseAll=parseAll).dump()) - except ParseException as pe: + result = self.parseString(t, parseAll=parseAll) + out.append(result.dump(full=fullDump)) + success = success and not failureTests + except ParseBaseException as pe: + fatal = "(FATAL)" if isinstance(pe, ParseFatalException) else "" if '\n' in t: out.append(line(pe.loc, t)) - out.append(' '*(col(pe.loc,t)-1) + '^') + out.append(' '*(col(pe.loc,t)-1) + '^' + fatal) else: - out.append(' '*pe.loc + '^') - out.append(str(pe)) - out.append('') - print('\n'.join(out)) + out.append(' '*pe.loc + '^' + fatal) + out.append("FAIL: " + str(pe)) + success = success and failureTests + result = pe + except Exception as exc: + out.append("FAIL-EXCEPTION: " + str(exc)) + success = success and failureTests + result = exc + + if printResults: + if fullDump: + out.append('') + print('\n'.join(out)) + + allResults.append((t, result)) + + return success, allResults class Token(ParserElement): - """Abstract C{ParserElement} subclass, for defining atomic matching patterns.""" + """ + Abstract C{ParserElement} subclass, for defining atomic matching patterns. + """ def __init__( self ): super(Token,self).__init__( savelist=False ) class Empty(Token): - """An empty token, will always match.""" + """ + An empty token, will always match. + """ def __init__( self ): super(Empty,self).__init__() self.name = "Empty" @@ -1588,7 +2319,9 @@ class Empty(Token): class NoMatch(Token): - """A token that will never match.""" + """ + A token that will never match. + """ def __init__( self ): super(NoMatch,self).__init__() self.name = "NoMatch" @@ -1601,7 +2334,19 @@ class NoMatch(Token): class Literal(Token): - """Token to exactly match a specified string.""" + """ + Token to exactly match a specified string. + + Example:: + Literal('blah').parseString('blah') # -> ['blah'] + Literal('blah').parseString('blahfooblah') # -> ['blah'] + Literal('blah').parseString('bla') # -> Exception: Expected "blah" + + For case-insensitive matching, use L{CaselessLiteral}. + + For keyword matching (force word break before and after the matched string), + use L{Keyword} or L{CaselessKeyword}. + """ def __init__( self, matchString ): super(Literal,self).__init__() self.match = matchString @@ -1627,17 +2372,24 @@ class Literal(Token): return loc+self.matchLen, self.match raise ParseException(instring, loc, self.errmsg, self) _L = Literal -ParserElement.literalStringClass = Literal +ParserElement._literalStringClass = Literal class Keyword(Token): - """Token to exactly match a specified string as a keyword, that is, it must be - immediately followed by a non-keyword character. Compare with C{L{Literal}}:: - Literal("if") will match the leading C{'if'} in C{'ifAndOnlyIf'}. - Keyword("if") will not; it will only match the leading C{'if'} in C{'if x=1'}, or C{'if(y==2)'} - Accepts two optional constructor arguments in addition to the keyword string: - C{identChars} is a string of characters that would be valid identifier characters, - defaulting to all alphanumerics + "_" and "$"; C{caseless} allows case-insensitive - matching, default is C{False}. + """ + Token to exactly match a specified string as a keyword, that is, it must be + immediately followed by a non-keyword character. Compare with C{L{Literal}}: + - C{Literal("if")} will match the leading C{'if'} in C{'ifAndOnlyIf'}. + - C{Keyword("if")} will not; it will only match the leading C{'if'} in C{'if x=1'}, or C{'if(y==2)'} + Accepts two optional constructor arguments in addition to the keyword string: + - C{identChars} is a string of characters that would be valid identifier characters, + defaulting to all alphanumerics + "_" and "$" + - C{caseless} allows case-insensitive matching, default is C{False}. + + Example:: + Keyword("start").parseString("start") # -> ['start'] + Keyword("start").parseString("starting") # -> Exception + + For case-insensitive matching, use L{CaselessKeyword}. """ DEFAULT_KEYWORD_CHARS = alphanums+"_$" @@ -1686,9 +2438,15 @@ class Keyword(Token): Keyword.DEFAULT_KEYWORD_CHARS = chars class CaselessLiteral(Literal): - """Token to match a specified string, ignoring case of letters. - Note: the matched results will always be in the case of the given - match string, NOT the case of the input text. + """ + Token to match a specified string, ignoring case of letters. + Note: the matched results will always be in the case of the given + match string, NOT the case of the input text. + + Example:: + OneOrMore(CaselessLiteral("CMD")).parseString("cmd CMD Cmd10") # -> ['CMD', 'CMD', 'CMD'] + + (Contrast with example for L{CaselessKeyword}.) """ def __init__( self, matchString ): super(CaselessLiteral,self).__init__( matchString.upper() ) @@ -1703,6 +2461,14 @@ class CaselessLiteral(Literal): raise ParseException(instring, loc, self.errmsg, self) class CaselessKeyword(Keyword): + """ + Caseless version of L{Keyword}. + + Example:: + OneOrMore(CaselessKeyword("CMD")).parseString("cmd CMD Cmd10") # -> ['CMD', 'CMD'] + + (Contrast with example for L{CaselessLiteral}.) + """ def __init__( self, matchString, identChars=Keyword.DEFAULT_KEYWORD_CHARS ): super(CaselessKeyword,self).__init__( matchString, identChars, caseless=True ) @@ -1713,16 +2479,51 @@ class CaselessKeyword(Keyword): raise ParseException(instring, loc, self.errmsg, self) class Word(Token): - """Token for matching words composed of allowed character sets. - Defined with string containing all allowed initial characters, - an optional string containing allowed body characters (if omitted, - defaults to the initial character set), and an optional minimum, - maximum, and/or exact length. The default value for C{min} is 1 (a - minimum value < 1 is not valid); the default values for C{max} and C{exact} - are 0, meaning no maximum or exact length restriction. An optional - C{exclude} parameter can list characters that might be found in - the input C{bodyChars} string; useful to define a word of all printables - except for one or two characters, for instance. + """ + Token for matching words composed of allowed character sets. + Defined with string containing all allowed initial characters, + an optional string containing allowed body characters (if omitted, + defaults to the initial character set), and an optional minimum, + maximum, and/or exact length. The default value for C{min} is 1 (a + minimum value < 1 is not valid); the default values for C{max} and C{exact} + are 0, meaning no maximum or exact length restriction. An optional + C{excludeChars} parameter can list characters that might be found in + the input C{bodyChars} string; useful to define a word of all printables + except for one or two characters, for instance. + + L{srange} is useful for defining custom character set strings for defining + C{Word} expressions, using range notation from regular expression character sets. + + A common mistake is to use C{Word} to match a specific literal string, as in + C{Word("Address")}. Remember that C{Word} uses the string argument to define + I{sets} of matchable characters. This expression would match "Add", "AAA", + "dAred", or any other word made up of the characters 'A', 'd', 'r', 'e', and 's'. + To match an exact literal string, use L{Literal} or L{Keyword}. + + pyparsing includes helper strings for building Words: + - L{alphas} + - L{nums} + - L{alphanums} + - L{hexnums} + - L{alphas8bit} (alphabetic characters in ASCII range 128-255 - accented, tilded, umlauted, etc.) + - L{punc8bit} (non-alphabetic characters in ASCII range 128-255 - currency, symbols, superscripts, diacriticals, etc.) + - L{printables} (any non-whitespace character) + + Example:: + # a word composed of digits + integer = Word(nums) # equivalent to Word("0123456789") or Word(srange("0-9")) + + # a word with a leading capital, and zero or more lowercase + capital_word = Word(alphas.upper(), alphas.lower()) + + # hostnames are alphanumeric, with leading alpha, and '-' + hostname = Word(alphas, alphanums+'-') + + # roman numeral (not a strict parser, accepts invalid mix of characters) + roman = Word("IVXLCDM") + + # any string of non-whitespace characters, except for ',' + csv_value = Word(printables, excludeChars=",") """ def __init__( self, initChars, bodyChars=None, min=1, max=0, exact=0, asKeyword=False, excludeChars=None ): super(Word,self).__init__() @@ -1837,8 +2638,17 @@ class Word(Token): class Regex(Token): - """Token for matching strings that match a given regular expression. - Defined with string specifying the regular expression in a form recognized by the inbuilt Python re module. + """ + Token for matching strings that match a given regular expression. + Defined with string specifying the regular expression in a form recognized by the inbuilt Python re module. + If the given regex contains named groups (defined using C{(?P...)}), these will be preserved as + named parse results. + + Example:: + realnum = Regex(r"[+-]?\d+\.\d*") + date = Regex(r'(?P\d{4})-(?P\d\d)-(?P\d\d)') + # ref: http://stackoverflow.com/questions/267399/how-do-you-match-only-valid-roman-numerals-with-a-regular-expression + roman = Regex(r"M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})") """ compiledREtype = type(re.compile("[A-Z]")) def __init__( self, pattern, flags=0): @@ -1846,7 +2656,7 @@ class Regex(Token): super(Regex,self).__init__() if isinstance(pattern, basestring): - if len(pattern) == 0: + if not pattern: warnings.warn("null string passed to Regex; use Empty() instead", SyntaxWarning, stacklevel=2) @@ -1901,23 +2711,36 @@ class Regex(Token): class QuotedString(Token): - """Token for matching strings that are delimited by quoting characters. + r""" + Token for matching strings that are delimited by quoting characters. + + Defined with the following parameters: + - quoteChar - string of one or more characters defining the quote delimiting string + - escChar - character to escape quotes, typically backslash (default=C{None}) + - escQuote - special quote sequence to escape an embedded quote string (such as SQL's "" to escape an embedded ") (default=C{None}) + - multiline - boolean indicating whether quotes can span multiple lines (default=C{False}) + - unquoteResults - boolean indicating whether the matched text should be unquoted (default=C{True}) + - endQuoteChar - string of one or more characters defining the end of the quote delimited string (default=C{None} => same as quoteChar) + - convertWhitespaceEscapes - convert escaped whitespace (C{'\t'}, C{'\n'}, etc.) to actual whitespace (default=C{True}) + + Example:: + qs = QuotedString('"') + print(qs.searchString('lsjdf "This is the quote" sldjf')) + complex_qs = QuotedString('{{', endQuoteChar='}}') + print(complex_qs.searchString('lsjdf {{This is the "quote"}} sldjf')) + sql_qs = QuotedString('"', escQuote='""') + print(sql_qs.searchString('lsjdf "This is the quote with ""embedded"" quotes" sldjf')) + prints:: + [['This is the quote']] + [['This is the "quote"']] + [['This is the quote with "embedded" quotes']] """ - def __init__( self, quoteChar, escChar=None, escQuote=None, multiline=False, unquoteResults=True, endQuoteChar=None): - """ - Defined with the following parameters: - - quoteChar - string of one or more characters defining the quote delimiting string - - escChar - character to escape quotes, typically backslash (default=None) - - escQuote - special quote sequence to escape an embedded quote string (such as SQL's "" to escape an embedded ") (default=None) - - multiline - boolean indicating whether quotes can span multiple lines (default=C{False}) - - unquoteResults - boolean indicating whether the matched text should be unquoted (default=C{True}) - - endQuoteChar - string of one or more characters defining the end of the quote delimited string (default=C{None} => same as quoteChar) - """ + def __init__( self, quoteChar, escChar=None, escQuote=None, multiline=False, unquoteResults=True, endQuoteChar=None, convertWhitespaceEscapes=True): super(QuotedString,self).__init__() # remove white space from quote chars - wont work anyway quoteChar = quoteChar.strip() - if len(quoteChar) == 0: + if not quoteChar: warnings.warn("quoteChar cannot be the empty string",SyntaxWarning,stacklevel=2) raise SyntaxError() @@ -1925,7 +2748,7 @@ class QuotedString(Token): endQuoteChar = quoteChar else: endQuoteChar = endQuoteChar.strip() - if len(endQuoteChar) == 0: + if not endQuoteChar: warnings.warn("endQuoteChar cannot be the empty string",SyntaxWarning,stacklevel=2) raise SyntaxError() @@ -1937,6 +2760,7 @@ class QuotedString(Token): self.escChar = escChar self.escQuote = escQuote self.unquoteResults = unquoteResults + self.convertWhitespaceEscapes = convertWhitespaceEscapes if multiline: self.flags = re.MULTILINE | re.DOTALL @@ -1990,6 +2814,17 @@ class QuotedString(Token): ret = ret[self.quoteCharLen:-self.endQuoteCharLen] if isinstance(ret,basestring): + # replace escaped whitespace + if '\\' in ret and self.convertWhitespaceEscapes: + ws_map = { + r'\t' : '\t', + r'\n' : '\n', + r'\f' : '\f', + r'\r' : '\r', + } + for wslit,wschar in ws_map.items(): + ret = ret.replace(wslit, wschar) + # replace escaped characters if self.escChar: ret = re.sub(self.escCharReplacePattern,"\g<1>",ret) @@ -2013,11 +2848,20 @@ class QuotedString(Token): class CharsNotIn(Token): - """Token for matching words composed of characters *not* in a given set. - Defined with string containing all disallowed characters, and an optional - minimum, maximum, and/or exact length. The default value for C{min} is 1 (a - minimum value < 1 is not valid); the default values for C{max} and C{exact} - are 0, meaning no maximum or exact length restriction. + """ + Token for matching words composed of characters I{not} in a given set (will + include whitespace in matched characters if not listed in the provided exclusion set - see example). + Defined with string containing all disallowed characters, and an optional + minimum, maximum, and/or exact length. The default value for C{min} is 1 (a + minimum value < 1 is not valid); the default values for C{max} and C{exact} + are 0, meaning no maximum or exact length restriction. + + Example:: + # define a comma-separated-value as anything that is not a ',' + csv_value = CharsNotIn(',') + print(delimitedList(csv_value).parseString("dkls,lsdkjf,s12 34,@!#,213")) + prints:: + ['dkls', 'lsdkjf', 's12 34', '@!#', '213'] """ def __init__( self, notChars, min=1, max=0, exact=0 ): super(CharsNotIn,self).__init__() @@ -2075,11 +2919,13 @@ class CharsNotIn(Token): return self.strRepr class White(Token): - """Special matching class for matching whitespace. Normally, whitespace is ignored - by pyparsing grammars. This class is included when some whitespace structures - are significant. Define with a string containing the whitespace characters to be - matched; default is C{" \\t\\r\\n"}. Also takes optional C{min}, C{max}, and C{exact} arguments, - as defined for the C{L{Word}} class.""" + """ + Special matching class for matching whitespace. Normally, whitespace is ignored + by pyparsing grammars. This class is included when some whitespace structures + are significant. Define with a string containing the whitespace characters to be + matched; default is C{" \\t\\r\\n"}. Also takes optional C{min}, C{max}, and C{exact} arguments, + as defined for the C{L{Word}} class. + """ whiteStrs = { " " : "", "\t": "", @@ -2131,7 +2977,9 @@ class _PositionToken(Token): self.mayIndexError = False class GoToColumn(_PositionToken): - """Token to advance to a specific column of input text; useful for tabular report scraping.""" + """ + Token to advance to a specific column of input text; useful for tabular report scraping. + """ def __init__( self, colno ): super(GoToColumn,self).__init__() self.col = colno @@ -2154,7 +3002,9 @@ class GoToColumn(_PositionToken): return newloc, ret class LineStart(_PositionToken): - """Matches if current position is at the beginning of a line within the parse string""" + """ + Matches if current position is at the beginning of a line within the parse string + """ def __init__( self ): super(LineStart,self).__init__() self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n","") ) @@ -2174,7 +3024,9 @@ class LineStart(_PositionToken): return loc, [] class LineEnd(_PositionToken): - """Matches if current position is at the end of a line within the parse string""" + """ + Matches if current position is at the end of a line within the parse string + """ def __init__( self ): super(LineEnd,self).__init__() self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n","") ) @@ -2192,7 +3044,9 @@ class LineEnd(_PositionToken): raise ParseException(instring, loc, self.errmsg, self) class StringStart(_PositionToken): - """Matches if current position is at the beginning of the parse string""" + """ + Matches if current position is at the beginning of the parse string + """ def __init__( self ): super(StringStart,self).__init__() self.errmsg = "Expected start of text" @@ -2205,7 +3059,9 @@ class StringStart(_PositionToken): return loc, [] class StringEnd(_PositionToken): - """Matches if current position is at the end of the parse string""" + """ + Matches if current position is at the end of the parse string + """ def __init__( self ): super(StringEnd,self).__init__() self.errmsg = "Expected end of text" @@ -2221,11 +3077,12 @@ class StringEnd(_PositionToken): raise ParseException(instring, loc, self.errmsg, self) class WordStart(_PositionToken): - """Matches if the current position is at the beginning of a Word, and - is not preceded by any character in a given set of C{wordChars} - (default=C{printables}). To emulate the C{\b} behavior of regular expressions, - use C{WordStart(alphanums)}. C{WordStart} will also match at the beginning of - the string being parsed, or at the beginning of a line. + """ + Matches if the current position is at the beginning of a Word, and + is not preceded by any character in a given set of C{wordChars} + (default=C{printables}). To emulate the C{\b} behavior of regular expressions, + use C{WordStart(alphanums)}. C{WordStart} will also match at the beginning of + the string being parsed, or at the beginning of a line. """ def __init__(self, wordChars = printables): super(WordStart,self).__init__() @@ -2240,11 +3097,12 @@ class WordStart(_PositionToken): return loc, [] class WordEnd(_PositionToken): - """Matches if the current position is at the end of a Word, and - is not followed by any character in a given set of C{wordChars} - (default=C{printables}). To emulate the C{\b} behavior of regular expressions, - use C{WordEnd(alphanums)}. C{WordEnd} will also match at the end of - the string being parsed, or at the end of a line. + """ + Matches if the current position is at the end of a Word, and + is not followed by any character in a given set of C{wordChars} + (default=C{printables}). To emulate the C{\b} behavior of regular expressions, + use C{WordEnd(alphanums)}. C{WordEnd} will also match at the end of + the string being parsed, or at the end of a line. """ def __init__(self, wordChars = printables): super(WordEnd,self).__init__() @@ -2262,18 +3120,21 @@ class WordEnd(_PositionToken): class ParseExpression(ParserElement): - """Abstract subclass of ParserElement, for combining and post-processing parsed tokens.""" + """ + Abstract subclass of ParserElement, for combining and post-processing parsed tokens. + """ def __init__( self, exprs, savelist = False ): super(ParseExpression,self).__init__(savelist) if isinstance( exprs, _generatorType ): exprs = list(exprs) if isinstance( exprs, basestring ): - self.exprs = [ Literal( exprs ) ] - elif isinstance( exprs, collections.Sequence ): + self.exprs = [ ParserElement._literalStringClass( exprs ) ] + elif isinstance( exprs, collections.Iterable ): + exprs = list(exprs) # if sequence of strings provided, wrap with Literal if all(isinstance(expr, basestring) for expr in exprs): - exprs = map(Literal, exprs) + exprs = map(ParserElement._literalStringClass, exprs) self.exprs = list(exprs) else: try: @@ -2351,7 +3212,7 @@ class ParseExpression(ParserElement): self.mayReturnEmpty |= other.mayReturnEmpty self.mayIndexError |= other.mayIndexError - self.errmsg = "Expected " + str(self) + self.errmsg = "Expected " + _ustr(self) return self @@ -2371,9 +3232,19 @@ class ParseExpression(ParserElement): return ret class And(ParseExpression): - """Requires all given C{ParseExpression}s to be found in the given order. - Expressions may be separated by whitespace. - May be constructed using the C{'+'} operator. + """ + Requires all given C{ParseExpression}s to be found in the given order. + Expressions may be separated by whitespace. + May be constructed using the C{'+'} operator. + May also be constructed using the C{'-'} operator, which will suppress backtracking. + + Example:: + integer = Word(nums) + name_expr = OneOrMore(Word(alphas)) + + expr = And([integer("id"),name_expr("name"),integer("age")]) + # more easily written as: + expr = integer("id") + name_expr("name") + integer("age") """ class _ErrorStop(Empty): @@ -2405,9 +3276,9 @@ class And(ParseExpression): raise except ParseBaseException as pe: pe.__traceback__ = None - raise ParseSyntaxException(pe) + raise ParseSyntaxException._from_exception(pe) except IndexError: - raise ParseSyntaxException( ParseException(instring, len(instring), self.errmsg, self) ) + raise ParseSyntaxException(instring, len(instring), self.errmsg, self) else: loc, exprtokens = e._parse( instring, loc, doActions ) if exprtokens or exprtokens.haskeys(): @@ -2416,7 +3287,7 @@ class And(ParseExpression): def __iadd__(self, other ): if isinstance( other, basestring ): - other = Literal( other ) + other = ParserElement._literalStringClass( other ) return self.append( other ) #And( [ self, other ] ) def checkRecursion( self, parseElementList ): @@ -2437,9 +3308,18 @@ class And(ParseExpression): class Or(ParseExpression): - """Requires that at least one C{ParseExpression} is found. - If two expressions match, the expression that matches the longest string will be used. - May be constructed using the C{'^'} operator. + """ + Requires that at least one C{ParseExpression} is found. + If two expressions match, the expression that matches the longest string will be used. + May be constructed using the C{'^'} operator. + + Example:: + # construct Or using '^' operator + + number = Word(nums) ^ Combine(Word(nums) + '.' + Word(nums)) + print(number.searchString("123 3.1416 789")) + prints:: + [['123'], ['3.1416'], ['789']] """ def __init__( self, exprs, savelist = False ): super(Or,self).__init__(exprs, savelist) @@ -2488,7 +3368,7 @@ class Or(ParseExpression): def __ixor__(self, other ): if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) return self.append( other ) #Or( [ self, other ] ) def __str__( self ): @@ -2507,9 +3387,21 @@ class Or(ParseExpression): class MatchFirst(ParseExpression): - """Requires that at least one C{ParseExpression} is found. - If two expressions match, the first one listed is the one that will match. - May be constructed using the C{'|'} operator. + """ + Requires that at least one C{ParseExpression} is found. + If two expressions match, the first one listed is the one that will match. + May be constructed using the C{'|'} operator. + + Example:: + # construct MatchFirst using '|' operator + + # watch the order of expressions to match + number = Word(nums) | Combine(Word(nums) + '.' + Word(nums)) + print(number.searchString("123 3.1416 789")) # Fail! -> [['123'], ['3'], ['1416'], ['789']] + + # put more selective expression first + number = Combine(Word(nums) + '.' + Word(nums)) | Word(nums) + print(number.searchString("123 3.1416 789")) # Better -> [['123'], ['3.1416'], ['789']] """ def __init__( self, exprs, savelist = False ): super(MatchFirst,self).__init__(exprs, savelist) @@ -2544,7 +3436,7 @@ class MatchFirst(ParseExpression): def __ior__(self, other ): if isinstance( other, basestring ): - other = ParserElement.literalStringClass( other ) + other = ParserElement._literalStringClass( other ) return self.append( other ) #MatchFirst( [ self, other ] ) def __str__( self ): @@ -2563,9 +3455,58 @@ class MatchFirst(ParseExpression): class Each(ParseExpression): - """Requires all given C{ParseExpression}s to be found, but in any order. - Expressions may be separated by whitespace. - May be constructed using the C{'&'} operator. + """ + Requires all given C{ParseExpression}s to be found, but in any order. + Expressions may be separated by whitespace. + May be constructed using the C{'&'} operator. + + Example:: + color = oneOf("RED ORANGE YELLOW GREEN BLUE PURPLE BLACK WHITE BROWN") + shape_type = oneOf("SQUARE CIRCLE TRIANGLE STAR HEXAGON OCTAGON") + integer = Word(nums) + shape_attr = "shape:" + shape_type("shape") + posn_attr = "posn:" + Group(integer("x") + ',' + integer("y"))("posn") + color_attr = "color:" + color("color") + size_attr = "size:" + integer("size") + + # use Each (using operator '&') to accept attributes in any order + # (shape and posn are required, color and size are optional) + shape_spec = shape_attr & posn_attr & Optional(color_attr) & Optional(size_attr) + + shape_spec.runTests(''' + shape: SQUARE color: BLACK posn: 100, 120 + shape: CIRCLE size: 50 color: BLUE posn: 50,80 + color:GREEN size:20 shape:TRIANGLE posn:20,40 + ''' + ) + prints:: + shape: SQUARE color: BLACK posn: 100, 120 + ['shape:', 'SQUARE', 'color:', 'BLACK', 'posn:', ['100', ',', '120']] + - color: BLACK + - posn: ['100', ',', '120'] + - x: 100 + - y: 120 + - shape: SQUARE + + + shape: CIRCLE size: 50 color: BLUE posn: 50,80 + ['shape:', 'CIRCLE', 'size:', '50', 'color:', 'BLUE', 'posn:', ['50', ',', '80']] + - color: BLUE + - posn: ['50', ',', '80'] + - x: 50 + - y: 80 + - shape: CIRCLE + - size: 50 + + + color: GREEN size: 20 shape: TRIANGLE posn: 20,40 + ['color:', 'GREEN', 'size:', '20', 'shape:', 'TRIANGLE', 'posn:', ['20', ',', '40']] + - color: GREEN + - posn: ['20', ',', '40'] + - x: 20 + - y: 40 + - shape: TRIANGLE + - size: 20 """ def __init__( self, exprs, savelist = True ): super(Each,self).__init__(exprs, savelist) @@ -2619,17 +3560,7 @@ class Each(ParseExpression): loc,results = e._parse(instring,loc,doActions) resultlist.append(results) - finalResults = ParseResults([]) - for r in resultlist: - dups = {} - for k in r.keys(): - if k in finalResults: - tmp = ParseResults(finalResults[k]) - tmp += ParseResults(r[k]) - dups[k] = tmp - finalResults += ParseResults(r) - for k,v in dups.items(): - finalResults[k] = v + finalResults = sum(resultlist, ParseResults([])) return loc, finalResults def __str__( self ): @@ -2648,11 +3579,16 @@ class Each(ParseExpression): class ParseElementEnhance(ParserElement): - """Abstract subclass of C{ParserElement}, for combining and post-processing parsed tokens.""" + """ + Abstract subclass of C{ParserElement}, for combining and post-processing parsed tokens. + """ def __init__( self, expr, savelist=False ): super(ParseElementEnhance,self).__init__(savelist) if isinstance( expr, basestring ): - expr = Literal(expr) + if issubclass(ParserElement._literalStringClass, Token): + expr = ParserElement._literalStringClass(expr) + else: + expr = ParserElement._literalStringClass(Literal(expr)) self.expr = expr self.strRepr = None if expr is not None: @@ -2720,10 +3656,22 @@ class ParseElementEnhance(ParserElement): class FollowedBy(ParseElementEnhance): - """Lookahead matching of the given parse expression. C{FollowedBy} - does *not* advance the parsing position within the input string, it only + """ + Lookahead matching of the given parse expression. C{FollowedBy} + does I{not} advance the parsing position within the input string, it only verifies that the specified parse expression matches at the current - position. C{FollowedBy} always returns a null token list.""" + position. C{FollowedBy} always returns a null token list. + + Example:: + # use FollowedBy to match a label only if it is followed by a ':' + data_word = Word(alphas) + label = data_word + FollowedBy(':') + attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) + + OneOrMore(attr_expr).parseString("shape: SQUARE color: BLACK posn: upper left").pprint() + prints:: + [['shape', 'SQUARE'], ['color', 'BLACK'], ['posn', 'upper left']] + """ def __init__( self, expr ): super(FollowedBy,self).__init__(expr) self.mayReturnEmpty = True @@ -2734,11 +3682,16 @@ class FollowedBy(ParseElementEnhance): class NotAny(ParseElementEnhance): - """Lookahead to disallow matching with the given parse expression. C{NotAny} - does *not* advance the parsing position within the input string, it only - verifies that the specified parse expression does *not* match at the current - position. Also, C{NotAny} does *not* skip over leading whitespace. C{NotAny} - always returns a null token list. May be constructed using the '~' operator.""" + """ + Lookahead to disallow matching with the given parse expression. C{NotAny} + does I{not} advance the parsing position within the input string, it only + verifies that the specified parse expression does I{not} match at the current + position. Also, C{NotAny} does I{not} skip over leading whitespace. C{NotAny} + always returns a null token list. May be constructed using the '~' operator. + + Example:: + + """ def __init__( self, expr ): super(NotAny,self).__init__(expr) #~ self.leaveWhitespace() @@ -2747,11 +3700,7 @@ class NotAny(ParseElementEnhance): self.errmsg = "Found unwanted token, "+_ustr(self.expr) def parseImpl( self, instring, loc, doActions=True ): - try: - self.expr.tryParse( instring, loc ) - except (ParseException,IndexError): - pass - else: + if self.expr.canParseNext(instring, loc): raise ParseException(instring, loc, self.errmsg, self) return loc, [] @@ -2764,80 +3713,114 @@ class NotAny(ParseElementEnhance): return self.strRepr - -class ZeroOrMore(ParseElementEnhance): - """Optional repetition of zero or more of the given expression.""" - def __init__( self, expr ): - super(ZeroOrMore,self).__init__(expr) - self.mayReturnEmpty = True +class _MultipleMatch(ParseElementEnhance): + def __init__( self, expr, stopOn=None): + super(_MultipleMatch, self).__init__(expr) + ender = stopOn + if isinstance(ender, basestring): + ender = ParserElement._literalStringClass(ender) + self.not_ender = ~ender if ender is not None else None def parseImpl( self, instring, loc, doActions=True ): - tokens = [] + self_expr_parse = self.expr._parse + self_skip_ignorables = self._skipIgnorables + check_ender = self.not_ender is not None + if check_ender: + try_not_ender = self.not_ender.tryParse + + # must be at least one (but first see if we are the stopOn sentinel; + # if so, fail) + if check_ender: + try_not_ender(instring, loc) + loc, tokens = self_expr_parse( instring, loc, doActions, callPreParse=False ) try: - loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False ) - hasIgnoreExprs = ( len(self.ignoreExprs) > 0 ) + hasIgnoreExprs = (not not self.ignoreExprs) while 1: + if check_ender: + try_not_ender(instring, loc) if hasIgnoreExprs: - preloc = self._skipIgnorables( instring, loc ) + preloc = self_skip_ignorables( instring, loc ) else: preloc = loc - loc, tmptokens = self.expr._parse( instring, preloc, doActions ) + loc, tmptokens = self_expr_parse( instring, preloc, doActions ) if tmptokens or tmptokens.haskeys(): tokens += tmptokens except (ParseException,IndexError): pass return loc, tokens + +class OneOrMore(_MultipleMatch): + """ + Repetition of one or more of the given expression. + + Parameters: + - expr - expression that must match one or more times + - stopOn - (default=C{None}) - expression for a terminating sentinel + (only required if the sentinel would ordinarily match the repetition + expression) + + Example:: + data_word = Word(alphas) + label = data_word + FollowedBy(':') + attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).setParseAction(' '.join)) + + text = "shape: SQUARE posn: upper left color: BLACK" + OneOrMore(attr_expr).parseString(text).pprint() # Fail! read 'color' as data instead of next label -> [['shape', 'SQUARE color']] + + # use stopOn attribute for OneOrMore to avoid reading label string as part of the data + attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) + OneOrMore(attr_expr).parseString(text).pprint() # Better -> [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'BLACK']] + + # could also be written as + (attr_expr * (1,)).parseString(text).pprint() + """ def __str__( self ): if hasattr(self,"name"): return self.name if self.strRepr is None: - self.strRepr = "[" + _ustr(self.expr) + "]..." + self.strRepr = "{" + _ustr(self.expr) + "}..." return self.strRepr def setResultsName( self, name, listAllMatches=False ): - ret = super(ZeroOrMore,self).setResultsName(name,listAllMatches) + ret = super(OneOrMore,self).setResultsName(name,listAllMatches) ret.saveAsList = True return ret +class ZeroOrMore(_MultipleMatch): + """ + Optional repetition of zero or more of the given expression. + + Parameters: + - expr - expression that must match zero or more times + - stopOn - (default=C{None}) - expression for a terminating sentinel + (only required if the sentinel would ordinarily match the repetition + expression) -class OneOrMore(ParseElementEnhance): - """Repetition of one or more of the given expression.""" + Example: similar to L{OneOrMore} + """ + def __init__( self, expr, stopOn=None): + super(ZeroOrMore,self).__init__(expr, stopOn=stopOn) + self.mayReturnEmpty = True + def parseImpl( self, instring, loc, doActions=True ): - # must be at least one - loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False ) try: - hasIgnoreExprs = ( len(self.ignoreExprs) > 0 ) - while 1: - if hasIgnoreExprs: - preloc = self._skipIgnorables( instring, loc ) - else: - preloc = loc - loc, tmptokens = self.expr._parse( instring, preloc, doActions ) - if tmptokens or tmptokens.haskeys(): - tokens += tmptokens + return super(ZeroOrMore, self).parseImpl(instring, loc, doActions) except (ParseException,IndexError): - pass - - return loc, tokens + return loc, [] def __str__( self ): if hasattr(self,"name"): return self.name if self.strRepr is None: - self.strRepr = "{" + _ustr(self.expr) + "}..." + self.strRepr = "[" + _ustr(self.expr) + "]..." return self.strRepr - def setResultsName( self, name, listAllMatches=False ): - ret = super(OneOrMore,self).setResultsName(name,listAllMatches) - ret.saveAsList = True - return ret - class _NullToken(object): def __bool__(self): return False @@ -2847,9 +3830,39 @@ class _NullToken(object): _optionalNotMatched = _NullToken() class Optional(ParseElementEnhance): - """Optional matching of the given expression. - A default return string can also be specified, if the optional expression - is not found. + """ + Optional matching of the given expression. + + Parameters: + - expr - expression that must match zero or more times + - default (optional) - value to be returned if the optional expression is not found. + + Example:: + # US postal code can be a 5-digit zip, plus optional 4-digit qualifier + zip = Combine(Word(nums, exact=5) + Optional('-' + Word(nums, exact=4))) + zip.runTests(''' + # traditional ZIP code + 12345 + + # ZIP+4 form + 12101-0001 + + # invalid ZIP + 98765- + ''') + prints:: + # traditional ZIP code + 12345 + ['12345'] + + # ZIP+4 form + 12101-0001 + ['12101-0001'] + + # invalid ZIP + 98765- + ^ + FAIL: Expected end of text (at char 5), (line:1, col:6) """ def __init__( self, expr, default=_optionalNotMatched ): super(Optional,self).__init__( expr, savelist=False ) @@ -2879,13 +3892,60 @@ class Optional(ParseElementEnhance): return self.strRepr - class SkipTo(ParseElementEnhance): - """Token for skipping over all undefined text until the matched expression is found. - If C{include} is set to true, the matched expression is also parsed (the skipped text - and matched expression are returned as a 2-element list). The C{ignore} - argument is used to define grammars (typically quoted strings and comments) that - might contain false matches. + """ + Token for skipping over all undefined text until the matched expression is found. + + Parameters: + - expr - target expression marking the end of the data to be skipped + - include - (default=C{False}) if True, the target expression is also parsed + (the skipped text and target expression are returned as a 2-element list). + - ignore - (default=C{None}) used to define grammars (typically quoted strings and + comments) that might contain false matches to the target expression + - failOn - (default=C{None}) define expressions that are not allowed to be + included in the skipped test; if found before the target expression is found, + the SkipTo is not a match + + Example:: + report = ''' + Outstanding Issues Report - 1 Jan 2000 + + # | Severity | Description | Days Open + -----+----------+-------------------------------------------+----------- + 101 | Critical | Intermittent system crash | 6 + 94 | Cosmetic | Spelling error on Login ('log|n') | 14 + 79 | Minor | System slow when running too many reports | 47 + ''' + integer = Word(nums) + SEP = Suppress('|') + # use SkipTo to simply match everything up until the next SEP + # - ignore quoted strings, so that a '|' character inside a quoted string does not match + # - parse action will call token.strip() for each matched token, i.e., the description body + string_data = SkipTo(SEP, ignore=quotedString) + string_data.setParseAction(tokenMap(str.strip)) + ticket_expr = (integer("issue_num") + SEP + + string_data("sev") + SEP + + string_data("desc") + SEP + + integer("days_open")) + + for tkt in ticket_expr.searchString(report): + print tkt.dump() + prints:: + ['101', 'Critical', 'Intermittent system crash', '6'] + - days_open: 6 + - desc: Intermittent system crash + - issue_num: 101 + - sev: Critical + ['94', 'Cosmetic', "Spelling error on Login ('log|n')", '14'] + - days_open: 14 + - desc: Spelling error on Login ('log|n') + - issue_num: 94 + - sev: Cosmetic + ['79', 'Minor', 'System slow when running too many reports', '47'] + - days_open: 47 + - desc: System slow when running too many reports + - issue_num: 79 + - sev: Minor """ def __init__( self, other, include=False, ignore=None, failOn=None ): super( SkipTo, self ).__init__( other ) @@ -2894,77 +3954,85 @@ class SkipTo(ParseElementEnhance): self.mayIndexError = False self.includeMatch = include self.asList = False - if failOn is not None and isinstance(failOn, basestring): - self.failOn = Literal(failOn) + if isinstance(failOn, basestring): + self.failOn = ParserElement._literalStringClass(failOn) else: self.failOn = failOn self.errmsg = "No match found for "+_ustr(self.expr) def parseImpl( self, instring, loc, doActions=True ): - startLoc = loc + startloc = loc instrlen = len(instring) expr = self.expr - failParse = False - while loc <= instrlen: - try: - if self.failOn: + expr_parse = self.expr._parse + self_failOn_canParseNext = self.failOn.canParseNext if self.failOn is not None else None + self_ignoreExpr_tryParse = self.ignoreExpr.tryParse if self.ignoreExpr is not None else None + + tmploc = loc + while tmploc <= instrlen: + if self_failOn_canParseNext is not None: + # break if failOn expression matches + if self_failOn_canParseNext(instring, tmploc): + break + + if self_ignoreExpr_tryParse is not None: + # advance past ignore expressions + while 1: try: - self.failOn.tryParse(instring, loc) + tmploc = self_ignoreExpr_tryParse(instring, tmploc) except ParseBaseException: - pass - else: - failParse = True - raise ParseException(instring, loc, "Found expression " + str(self.failOn)) - failParse = False - if self.ignoreExpr is not None: - while 1: - try: - loc = self.ignoreExpr.tryParse(instring,loc) - # print("found ignoreExpr, advance to", loc) - except ParseBaseException: - break - expr._parse( instring, loc, doActions=False, callPreParse=False ) - skipText = instring[startLoc:loc] - if self.includeMatch: - loc,mat = expr._parse(instring,loc,doActions,callPreParse=False) - if mat: - skipRes = ParseResults( skipText ) - skipRes += mat - return loc, [ skipRes ] - else: - return loc, [ skipText ] - else: - return loc, [ skipText ] - except (ParseException,IndexError): - if failParse: - raise - else: - loc += 1 - raise ParseException(instring, loc, self.errmsg, self) + break + + try: + expr_parse(instring, tmploc, doActions=False, callPreParse=False) + except (ParseException, IndexError): + # no match, advance loc in string + tmploc += 1 + else: + # matched skipto expr, done + break + + else: + # ran off the end of the input string without matching skipto expr, fail + raise ParseException(instring, loc, self.errmsg, self) + + # build up return values + loc = tmploc + skiptext = instring[startloc:loc] + skipresult = ParseResults(skiptext) + + if self.includeMatch: + loc, mat = expr_parse(instring,loc,doActions,callPreParse=False) + skipresult += mat + + return loc, skipresult class Forward(ParseElementEnhance): - """Forward declaration of an expression to be defined later - - used for recursive grammars, such as algebraic infix notation. - When the expression is known, it is assigned to the C{Forward} variable using the '<<' operator. - - Note: take care when assigning to C{Forward} not to overlook precedence of operators. - Specifically, '|' has a lower precedence than '<<', so that:: - fwdExpr << a | b | c - will actually be evaluated as:: - (fwdExpr << a) | b | c - thereby leaving b and c out as parseable alternatives. It is recommended that you - explicitly group the values inserted into the C{Forward}:: - fwdExpr << (a | b | c) - Converting to use the '<<=' operator instead will avoid this problem. + """ + Forward declaration of an expression to be defined later - + used for recursive grammars, such as algebraic infix notation. + When the expression is known, it is assigned to the C{Forward} variable using the '<<' operator. + + Note: take care when assigning to C{Forward} not to overlook precedence of operators. + Specifically, '|' has a lower precedence than '<<', so that:: + fwdExpr << a | b | c + will actually be evaluated as:: + (fwdExpr << a) | b | c + thereby leaving b and c out as parseable alternatives. It is recommended that you + explicitly group the values inserted into the C{Forward}:: + fwdExpr << (a | b | c) + Converting to use the '<<=' operator instead will avoid this problem. + + See L{ParseResults.pprint} for an example of a recursive parser created using + C{Forward}. """ def __init__( self, other=None ): super(Forward,self).__init__( other, savelist=False ) def __lshift__( self, other ): if isinstance( other, basestring ): - other = ParserElement.literalStringClass(other) + other = ParserElement._literalStringClass(other) self.expr = other - self.mayReturnEmpty = other.mayReturnEmpty self.strRepr = None self.mayIndexError = self.expr.mayIndexError self.mayReturnEmpty = self.expr.mayReturnEmpty @@ -2998,7 +4066,9 @@ class Forward(ParseElementEnhance): def __str__( self ): if hasattr(self,"name"): return self.name + return self.__class__.__name__ + ": ..." + # stubbed out for now - creates awful memory and perf issues self._revertClass = self.__class__ self.__class__ = _ForwardNoRecurse try: @@ -3023,26 +4093,29 @@ class _ForwardNoRecurse(Forward): return "..." class TokenConverter(ParseElementEnhance): - """Abstract subclass of C{ParseExpression}, for converting parsed results.""" + """ + Abstract subclass of C{ParseExpression}, for converting parsed results. + """ def __init__( self, expr, savelist=False ): super(TokenConverter,self).__init__( expr )#, savelist ) self.saveAsList = False -class Upcase(TokenConverter): - """Converter to upper case all matching tokens.""" - def __init__(self, *args): - super(Upcase,self).__init__(*args) - warnings.warn("Upcase class is deprecated, use upcaseTokens parse action instead", - DeprecationWarning,stacklevel=2) - - def postParse( self, instring, loc, tokenlist ): - return list(map( str.upper, tokenlist )) - - class Combine(TokenConverter): - """Converter to concatenate all matching tokens to a single string. - By default, the matching patterns must also be contiguous in the input string; - this can be disabled by specifying C{'adjacent=False'} in the constructor. + """ + Converter to concatenate all matching tokens to a single string. + By default, the matching patterns must also be contiguous in the input string; + this can be disabled by specifying C{'adjacent=False'} in the constructor. + + Example:: + real = Word(nums) + '.' + Word(nums) + print(real.parseString('3.1416')) # -> ['3', '.', '1416'] + # will also erroneously match the following + print(real.parseString('3. 1416')) # -> ['3', '.', '1416'] + + real = Combine(Word(nums) + '.' + Word(nums)) + print(real.parseString('3.1416')) # -> ['3.1416'] + # no match when there are internal spaces + print(real.parseString('3. 1416')) # -> Exception: Expected W:(0123...) """ def __init__( self, expr, joinString="", adjacent=True ): super(Combine,self).__init__( expr ) @@ -3072,7 +4145,19 @@ class Combine(TokenConverter): return retToks class Group(TokenConverter): - """Converter to return the matched tokens as a list - useful for returning tokens of C{L{ZeroOrMore}} and C{L{OneOrMore}} expressions.""" + """ + Converter to return the matched tokens as a list - useful for returning tokens of C{L{ZeroOrMore}} and C{L{OneOrMore}} expressions. + + Example:: + ident = Word(alphas) + num = Word(nums) + term = ident | num + func = ident + Optional(delimitedList(term)) + print(func.parseString("fn a,b,100")) # -> ['fn', 'a', 'b', '100'] + + func = ident + Group(Optional(delimitedList(term))) + print(func.parseString("fn a,b,100")) # -> ['fn', ['a', 'b', '100']] + """ def __init__( self, expr ): super(Group,self).__init__( expr ) self.saveAsList = True @@ -3081,9 +4166,40 @@ class Group(TokenConverter): return [ tokenlist ] class Dict(TokenConverter): - """Converter to return a repetitive expression as a list, but also as a dictionary. - Each element can also be referenced using the first token in the expression as its key. - Useful for tabular report scraping when the first column can be used as a item key. + """ + Converter to return a repetitive expression as a list, but also as a dictionary. + Each element can also be referenced using the first token in the expression as its key. + Useful for tabular report scraping when the first column can be used as a item key. + + Example:: + data_word = Word(alphas) + label = data_word + FollowedBy(':') + attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).setParseAction(' '.join)) + + text = "shape: SQUARE posn: upper left color: light blue texture: burlap" + attr_expr = (label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) + + # print attributes as plain groups + print(OneOrMore(attr_expr).parseString(text).dump()) + + # instead of OneOrMore(expr), parse using Dict(OneOrMore(Group(expr))) - Dict will auto-assign names + result = Dict(OneOrMore(Group(attr_expr))).parseString(text) + print(result.dump()) + + # access named fields as dict entries, or output as dict + print(result['shape']) + print(result.asDict()) + prints:: + ['shape', 'SQUARE', 'posn', 'upper left', 'color', 'light blue', 'texture', 'burlap'] + + [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']] + - color: light blue + - posn: upper left + - shape: SQUARE + - texture: burlap + SQUARE + {'color': 'light blue', 'posn': 'upper left', 'texture': 'burlap', 'shape': 'SQUARE'} + See more examples at L{ParseResults} of accessing fields by results name. """ def __init__( self, expr ): super(Dict,self).__init__( expr ) @@ -3115,7 +4231,24 @@ class Dict(TokenConverter): class Suppress(TokenConverter): - """Converter for ignoring the results of a parsed expression.""" + """ + Converter for ignoring the results of a parsed expression. + + Example:: + source = "a, b, c,d" + wd = Word(alphas) + wd_list1 = wd + ZeroOrMore(',' + wd) + print(wd_list1.parseString(source)) + + # often, delimiters that are useful during parsing are just in the + # way afterward - use Suppress to keep them out of the parsed output + wd_list2 = wd + ZeroOrMore(Suppress(',') + wd) + print(wd_list2.parseString(source)) + prints:: + ['a', ',', 'b', ',', 'c', ',', 'd'] + ['a', 'b', 'c', 'd'] + (See also L{delimitedList}.) + """ def postParse( self, instring, loc, tokenlist ): return [] @@ -3124,7 +4257,9 @@ class Suppress(TokenConverter): class OnlyOnce(object): - """Wrapper for parse actions, to ensure they are only called once.""" + """ + Wrapper for parse actions, to ensure they are only called once. + """ def __init__(self, methodCall): self.callable = _trim_arity(methodCall) self.called = False @@ -3138,20 +4273,36 @@ class OnlyOnce(object): self.called = False def traceParseAction(f): - """Decorator for debugging parse actions.""" + """ + Decorator for debugging parse actions. + + Example:: + wd = Word(alphas) + + @traceParseAction + def remove_duplicate_chars(tokens): + return ''.join(sorted(set(''.join(tokens))) + + wds = OneOrMore(wd).setParseAction(remove_duplicate_chars) + print(wds.parseString("slkdjs sld sldd sdlf sdljf")) + prints:: + >>entering remove_duplicate_chars(line: 'slkdjs sld sldd sdlf sdljf', 0, (['slkdjs', 'sld', 'sldd', 'sdlf', 'sdljf'], {})) + <3: thisFunc = paArgs[0].__class__.__name__ + '.' + thisFunc - sys.stderr.write( ">>entering %s(line: '%s', %d, %s)\n" % (thisFunc,line(l,s),l,t) ) + sys.stderr.write( ">>entering %s(line: '%s', %d, %r)\n" % (thisFunc,line(l,s),l,t) ) try: ret = f(*paArgs) except Exception as exc: sys.stderr.write( "< ['aa', 'bb', 'cc'] + delimitedList(Word(hexnums), delim=':', combine=True).parseString("AA:BB:CC:DD:EE") # -> ['AA:BB:CC:DD:EE'] """ dlName = _ustr(expr)+" ["+_ustr(delim)+" "+_ustr(expr)+"]..." if combine: @@ -3177,11 +4333,15 @@ def delimitedList( expr, delim=",", combine=False ): return ( expr + ZeroOrMore( Suppress( delim ) + expr ) ).setName(dlName) def countedArray( expr, intExpr=None ): - """Helper to define a counted list of expressions. - This helper defines a pattern of the form:: - integer expr expr expr... - where the leading integer tells how many expr expressions follow. - The matched tokens returns the array of expr tokens as a list - the leading count token is suppressed. + """ + Helper to define a counted list of expressions. + This helper defines a pattern of the form:: + integer expr expr expr... + where the leading integer tells how many expr expressions follow. + The matched tokens returns the array of expr tokens as a list - the leading count token is suppressed. + + Example:: + countedArray(Word(alphas)).parseString('2 ab cd ef') # -> ['ab', 'cd'] """ arrayExpr = Forward() def countFieldParseAction(s,l,t): @@ -3194,7 +4354,7 @@ def countedArray( expr, intExpr=None ): intExpr = intExpr.copy() intExpr.setName("arrayLen") intExpr.addParseAction(countFieldParseAction, callDuringTry=True) - return ( intExpr + arrayExpr ) + return ( intExpr + arrayExpr ).setName('(len) ' + _ustr(expr) + '...') def _flatten(L): ret = [] @@ -3206,16 +4366,17 @@ def _flatten(L): return ret def matchPreviousLiteral(expr): - """Helper to define an expression that is indirectly defined from - the tokens matched in a previous expression, that is, it looks - for a 'repeat' of a previous expression. For example:: - first = Word(nums) - second = matchPreviousLiteral(first) - matchExpr = first + ":" + second - will match C{"1:1"}, but not C{"1:2"}. Because this matches a - previous literal, will also match the leading C{"1:1"} in C{"1:10"}. - If this is not desired, use C{matchPreviousExpr}. - Do *not* use with packrat parsing enabled. + """ + Helper to define an expression that is indirectly defined from + the tokens matched in a previous expression, that is, it looks + for a 'repeat' of a previous expression. For example:: + first = Word(nums) + second = matchPreviousLiteral(first) + matchExpr = first + ":" + second + will match C{"1:1"}, but not C{"1:2"}. Because this matches a + previous literal, will also match the leading C{"1:1"} in C{"1:10"}. + If this is not desired, use C{matchPreviousExpr}. + Do I{not} use with packrat parsing enabled. """ rep = Forward() def copyTokenToRepeater(s,l,t): @@ -3225,24 +4386,26 @@ def matchPreviousLiteral(expr): else: # flatten t tokens tflat = _flatten(t.asList()) - rep << And( [ Literal(tt) for tt in tflat ] ) + rep << And(Literal(tt) for tt in tflat) else: rep << Empty() expr.addParseAction(copyTokenToRepeater, callDuringTry=True) + rep.setName('(prev) ' + _ustr(expr)) return rep def matchPreviousExpr(expr): - """Helper to define an expression that is indirectly defined from - the tokens matched in a previous expression, that is, it looks - for a 'repeat' of a previous expression. For example:: - first = Word(nums) - second = matchPreviousExpr(first) - matchExpr = first + ":" + second - will match C{"1:1"}, but not C{"1:2"}. Because this matches by - expressions, will *not* match the leading C{"1:1"} in C{"1:10"}; - the expressions are evaluated first, and then compared, so - C{"1"} is compared with C{"10"}. - Do *not* use with packrat parsing enabled. + """ + Helper to define an expression that is indirectly defined from + the tokens matched in a previous expression, that is, it looks + for a 'repeat' of a previous expression. For example:: + first = Word(nums) + second = matchPreviousExpr(first) + matchExpr = first + ":" + second + will match C{"1:1"}, but not C{"1:2"}. Because this matches by + expressions, will I{not} match the leading C{"1:1"} in C{"1:10"}; + the expressions are evaluated first, and then compared, so + C{"1"} is compared with C{"10"}. + Do I{not} use with packrat parsing enabled. """ rep = Forward() e2 = expr.copy() @@ -3255,6 +4418,7 @@ def matchPreviousExpr(expr): raise ParseException("",0,"") rep.setParseAction( mustMatchTheseTokens, callDuringTry=True ) expr.addParseAction(copyTokenToRepeater, callDuringTry=True) + rep.setName('(prev) ' + _ustr(expr)) return rep def _escapeRegexRangeChars(s): @@ -3266,16 +4430,27 @@ def _escapeRegexRangeChars(s): return _ustr(s) def oneOf( strs, caseless=False, useRegex=True ): - """Helper to quickly define a set of alternative Literals, and makes sure to do - longest-first testing when there is a conflict, regardless of the input order, - but returns a C{L{MatchFirst}} for best performance. - - Parameters: - - strs - a string of space-delimited literals, or a list of string literals - - caseless - (default=False) - treat all literals as caseless - - useRegex - (default=True) - as an optimization, will generate a Regex + """ + Helper to quickly define a set of alternative Literals, and makes sure to do + longest-first testing when there is a conflict, regardless of the input order, + but returns a C{L{MatchFirst}} for best performance. + + Parameters: + - strs - a string of space-delimited literals, or a collection of string literals + - caseless - (default=C{False}) - treat all literals as caseless + - useRegex - (default=C{True}) - as an optimization, will generate a Regex object; otherwise, will generate a C{MatchFirst} object (if C{caseless=True}, or if creating a C{Regex} raises an exception) + + Example:: + comp_oper = oneOf("< = > <= >= !=") + var = Word(alphas) + number = Word(nums) + term = var | number + comparison_expr = term + comp_oper + term + print(comparison_expr.searchString("B = 12 AA=23 B<=AA AA>12")) + prints:: + [['B', '=', '12'], ['AA', '=', '23'], ['B', '<=', 'AA'], ['AA', '>', '12']] """ if caseless: isequal = ( lambda a,b: a.upper() == b.upper() ) @@ -3289,12 +4464,10 @@ def oneOf( strs, caseless=False, useRegex=True ): symbols = [] if isinstance(strs,basestring): symbols = strs.split() - elif isinstance(strs, collections.Sequence): - symbols = list(strs[:]) - elif isinstance(strs, _generatorType): + elif isinstance(strs, collections.Iterable): symbols = list(strs) else: - warnings.warn("Invalid argument to oneOf, expected string or list", + warnings.warn("Invalid argument to oneOf, expected string or iterable", SyntaxWarning, stacklevel=2) if not symbols: return NoMatch() @@ -3318,41 +4491,76 @@ def oneOf( strs, caseless=False, useRegex=True ): #~ print (strs,"->", "|".join( [ _escapeRegexChars(sym) for sym in symbols] )) try: if len(symbols)==len("".join(symbols)): - return Regex( "[%s]" % "".join(_escapeRegexRangeChars(sym) for sym in symbols) ) + return Regex( "[%s]" % "".join(_escapeRegexRangeChars(sym) for sym in symbols) ).setName(' | '.join(symbols)) else: - return Regex( "|".join(re.escape(sym) for sym in symbols) ) + return Regex( "|".join(re.escape(sym) for sym in symbols) ).setName(' | '.join(symbols)) except: warnings.warn("Exception creating Regex for oneOf, building MatchFirst", SyntaxWarning, stacklevel=2) # last resort, just use MatchFirst - return MatchFirst( [ parseElementClass(sym) for sym in symbols ] ) + return MatchFirst(parseElementClass(sym) for sym in symbols).setName(' | '.join(symbols)) def dictOf( key, value ): - """Helper to easily and clearly define a dictionary by specifying the respective patterns - for the key and value. Takes care of defining the C{L{Dict}}, C{L{ZeroOrMore}}, and C{L{Group}} tokens - in the proper order. The key pattern can include delimiting markers or punctuation, - as long as they are suppressed, thereby leaving the significant key text. The value - pattern can include named results, so that the C{Dict} results can include named token - fields. + """ + Helper to easily and clearly define a dictionary by specifying the respective patterns + for the key and value. Takes care of defining the C{L{Dict}}, C{L{ZeroOrMore}}, and C{L{Group}} tokens + in the proper order. The key pattern can include delimiting markers or punctuation, + as long as they are suppressed, thereby leaving the significant key text. The value + pattern can include named results, so that the C{Dict} results can include named token + fields. + + Example:: + text = "shape: SQUARE posn: upper left color: light blue texture: burlap" + attr_expr = (label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) + print(OneOrMore(attr_expr).parseString(text).dump()) + + attr_label = label + attr_value = Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join) + + # similar to Dict, but simpler call format + result = dictOf(attr_label, attr_value).parseString(text) + print(result.dump()) + print(result['shape']) + print(result.shape) # object attribute access works too + print(result.asDict()) + prints:: + [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']] + - color: light blue + - posn: upper left + - shape: SQUARE + - texture: burlap + SQUARE + SQUARE + {'color': 'light blue', 'shape': 'SQUARE', 'posn': 'upper left', 'texture': 'burlap'} """ return Dict( ZeroOrMore( Group ( key + value ) ) ) def originalTextFor(expr, asString=True): - """Helper to return the original, untokenized text for a given expression. Useful to - restore the parsed fields of an HTML start tag into the raw tag text itself, or to - revert separate tokens with intervening whitespace back to the original matching - input text. Simpler to use than the parse action C{L{keepOriginalText}}, and does not - require the inspect module to chase up the call stack. By default, returns a - string containing the original parsed text. + """ + Helper to return the original, untokenized text for a given expression. Useful to + restore the parsed fields of an HTML start tag into the raw tag text itself, or to + revert separate tokens with intervening whitespace back to the original matching + input text. By default, returns astring containing the original parsed text. - If the optional C{asString} argument is passed as C{False}, then the return value is a - C{L{ParseResults}} containing any results names that were originally matched, and a - single token containing the original matched text from the input string. So if - the expression passed to C{L{originalTextFor}} contains expressions with defined - results names, you must set C{asString} to C{False} if you want to preserve those - results name values.""" + If the optional C{asString} argument is passed as C{False}, then the return value is a + C{L{ParseResults}} containing any results names that were originally matched, and a + single token containing the original matched text from the input string. So if + the expression passed to C{L{originalTextFor}} contains expressions with defined + results names, you must set C{asString} to C{False} if you want to preserve those + results name values. + + Example:: + src = "this is test bold text normal text " + for tag in ("b","i"): + opener,closer = makeHTMLTags(tag) + patt = originalTextFor(opener + SkipTo(closer) + closer) + print(patt.searchString(src)[0]) + prints:: + [' bold text '] + ['text'] + """ locMarker = Empty().setParseAction(lambda s,loc,t: loc) endlocMarker = locMarker.copy() endlocMarker.callPreparse = False @@ -3361,27 +4569,37 @@ def originalTextFor(expr, asString=True): extractText = lambda s,l,t: s[t._original_start:t._original_end] else: def extractText(s,l,t): - del t[:] - t.insert(0, s[t._original_start:t._original_end]) - del t["_original_start"] - del t["_original_end"] + t[:] = [s[t.pop('_original_start'):t.pop('_original_end')]] matchExpr.setParseAction(extractText) + matchExpr.ignoreExprs = expr.ignoreExprs return matchExpr def ungroup(expr): - """Helper to undo pyparsing's default grouping of And expressions, even - if all but one are non-empty.""" + """ + Helper to undo pyparsing's default grouping of And expressions, even + if all but one are non-empty. + """ return TokenConverter(expr).setParseAction(lambda t:t[0]) def locatedExpr(expr): - """Helper to decorate a returned token with its starting and ending locations in the input string. - This helper adds the following results names: - - locn_start = location where matched expression begins - - locn_end = location where matched expression ends - - value = the actual parsed results - - Be careful if the input text contains C{} characters, you may want to call - C{L{ParserElement.parseWithTabs}} + """ + Helper to decorate a returned token with its starting and ending locations in the input string. + This helper adds the following results names: + - locn_start = location where matched expression begins + - locn_end = location where matched expression ends + - value = the actual parsed results + + Be careful if the input text contains C{} characters, you may want to call + C{L{ParserElement.parseWithTabs}} + + Example:: + wd = Word(alphas) + for match in locatedExpr(wd).searchString("ljsdf123lksdjjf123lkkjj1222"): + print(match) + prints:: + [[0, 'ljsdf', 5]] + [[8, 'lksdjjf', 15]] + [[18, 'lkkjj', 23]] """ locator = Empty().setParseAction(lambda s,l,t: l) return Group(locator("locn_start") + expr("value") + locator.copy().leaveWhitespace()("locn_end")) @@ -3402,21 +4620,22 @@ _charRange = Group(_singleChar + Suppress("-") + _singleChar) _reBracketExpr = Literal("[") + Optional("^").setResultsName("negate") + Group( OneOrMore( _charRange | _singleChar ) ).setResultsName("body") + "]" def srange(s): - r"""Helper to easily define string ranges for use in Word construction. Borrows - syntax from regexp '[]' string range definitions:: - srange("[0-9]") -> "0123456789" - srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz" - srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_" - The input string must be enclosed in []'s, and the returned string is the expanded - character set joined into a single string. - The values enclosed in the []'s may be:: - a single character - an escaped character with a leading backslash (such as \- or \]) - an escaped hex character with a leading '\x' (\x21, which is a '!' character) - (\0x## is also supported for backwards compatibility) - an escaped octal character with a leading '\0' (\041, which is a '!' character) - a range of any of the above, separated by a dash ('a-z', etc.) - any combination of the above ('aeiouy', 'a-zA-Z0-9_$', etc.) + r""" + Helper to easily define string ranges for use in Word construction. Borrows + syntax from regexp '[]' string range definitions:: + srange("[0-9]") -> "0123456789" + srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz" + srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_" + The input string must be enclosed in []'s, and the returned string is the expanded + character set joined into a single string. + The values enclosed in the []'s may be: + - a single character + - an escaped character with a leading backslash (such as C{\-} or C{\]}) + - an escaped hex character with a leading C{'\x'} (C{\x21}, which is a C{'!'} character) + (C{\0x##} is also supported for backwards compatibility) + - an escaped octal character with a leading C{'\0'} (C{\041}, which is a C{'!'} character) + - a range of any of the above, separated by a dash (C{'a-z'}, etc.) + - any combination of the above (C{'aeiouy'}, C{'a-zA-Z0-9_$'}, etc.) """ _expanded = lambda p: p if not isinstance(p,ParseResults) else ''.join(unichr(c) for c in range(ord(p[0]),ord(p[1])+1)) try: @@ -3425,8 +4644,9 @@ def srange(s): return "" def matchOnlyAtCol(n): - """Helper method for defining parse actions that require matching at a specific - column in the input text. + """ + Helper method for defining parse actions that require matching at a specific + column in the input text. """ def verifyCol(strg,locn,toks): if col(locn,strg) != n: @@ -3434,57 +4654,83 @@ def matchOnlyAtCol(n): return verifyCol def replaceWith(replStr): - """Helper method for common parse actions that simply return a literal value. Especially - useful when used with C{L{transformString}()}. """ - #def _replFunc(*args): - # return [replStr] - #return _replFunc - return functools.partial(next, itertools.repeat([replStr])) + Helper method for common parse actions that simply return a literal value. Especially + useful when used with C{L{transformString}()}. + + Example:: + num = Word(nums).setParseAction(lambda toks: int(toks[0])) + na = oneOf("N/A NA").setParseAction(replaceWith(math.nan)) + term = na | num + + OneOrMore(term).parseString("324 234 N/A 234") # -> [324, 234, nan, 234] + """ + return lambda s,l,t: [replStr] def removeQuotes(s,l,t): - """Helper parse action for removing quotation marks from parsed quoted strings. - To use, add this parse action to quoted string using:: - quotedString.setParseAction( removeQuotes ) """ - return t[0][1:-1] + Helper parse action for removing quotation marks from parsed quoted strings. -def upcaseTokens(s,l,t): - """Helper parse action to convert tokens to upper case.""" - return [ tt.upper() for tt in map(_ustr,t) ] + Example:: + # by default, quotation marks are included in parsed results + quotedString.parseString("'Now is the Winter of our Discontent'") # -> ["'Now is the Winter of our Discontent'"] -def downcaseTokens(s,l,t): - """Helper parse action to convert tokens to lower case.""" - return [ tt.lower() for tt in map(_ustr,t) ] + # use removeQuotes to strip quotation marks from parsed results + quotedString.setParseAction(removeQuotes) + quotedString.parseString("'Now is the Winter of our Discontent'") # -> ["Now is the Winter of our Discontent"] + """ + return t[0][1:-1] + +def tokenMap(func, *args): + """ + Helper to define a parse action by mapping a function to all elements of a ParseResults list.If any additional + args are passed, they are forwarded to the given function as additional arguments after + the token, as in C{hex_integer = Word(hexnums).setParseAction(tokenMap(int, 16))}, which will convert the + parsed data to an integer using base 16. + + Example (compare the last to example in L{ParserElement.transformString}:: + hex_ints = OneOrMore(Word(hexnums)).setParseAction(tokenMap(int, 16)) + hex_ints.runTests(''' + 00 11 22 aa FF 0a 0d 1a + ''') + + upperword = Word(alphas).setParseAction(tokenMap(str.upper)) + OneOrMore(upperword).runTests(''' + my kingdom for a horse + ''') + + wd = Word(alphas).setParseAction(tokenMap(str.title)) + OneOrMore(wd).setParseAction(' '.join).runTests(''' + now is the winter of our discontent made glorious summer by this sun of york + ''') + prints:: + 00 11 22 aa FF 0a 0d 1a + [0, 17, 34, 170, 255, 10, 13, 26] + + my kingdom for a horse + ['MY', 'KINGDOM', 'FOR', 'A', 'HORSE'] + + now is the winter of our discontent made glorious summer by this sun of york + ['Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York'] + """ + def pa(s,l,t): + return [func(tokn, *args) for tokn in t] -def keepOriginalText(s,startLoc,t): - """DEPRECATED - use new helper method C{L{originalTextFor}}. - Helper parse action to preserve original parsed text, - overriding any nested parse actions.""" - try: - endloc = getTokensEndLoc() - except ParseException: - raise ParseFatalException("incorrect usage of keepOriginalText - may only be called as a parse action") - del t[:] - t += ParseResults(s[startLoc:endloc]) - return t - -def getTokensEndLoc(): - """Method to be called from within a parse action to determine the end - location of the parsed tokens.""" - import inspect - fstack = inspect.stack() try: - # search up the stack (through intervening argument normalizers) for correct calling routine - for f in fstack[2:]: - if f[3] == "_parseNoCache": - endloc = f[0].f_locals["loc"] - return endloc - else: - raise ParseFatalException("incorrect usage of getTokensEndLoc - may only be called from within a parse action") - finally: - del fstack + func_name = getattr(func, '__name__', + getattr(func, '__class__').__name__) + except Exception: + func_name = str(func) + pa.__name__ = func_name + return pa + +upcaseTokens = tokenMap(lambda t: _ustr(t).upper()) +"""Helper parse action to convert tokens to upper case.""" + +downcaseTokens = tokenMap(lambda t: _ustr(t).lower()) +"""Helper parse action to convert tokens to lower case.""" + def _makeTags(tagStr, xml): """Internal helper to construct opening and closing tag expressions, given a tag name""" if isinstance(tagStr,basestring): @@ -3508,40 +4754,90 @@ def _makeTags(tagStr, xml): Optional("/",default=[False]).setResultsName("empty").setParseAction(lambda s,l,t:t[0]=='/') + Suppress(">") closeTag = Combine(_L("") - openTag = openTag.setResultsName("start"+"".join(resname.replace(":"," ").title().split())).setName("<%s>" % tagStr) - closeTag = closeTag.setResultsName("end"+"".join(resname.replace(":"," ").title().split())).setName("" % tagStr) + openTag = openTag.setResultsName("start"+"".join(resname.replace(":"," ").title().split())).setName("<%s>" % resname) + closeTag = closeTag.setResultsName("end"+"".join(resname.replace(":"," ").title().split())).setName("" % resname) openTag.tag = resname closeTag.tag = resname return openTag, closeTag def makeHTMLTags(tagStr): - """Helper to construct opening and closing tag expressions for HTML, given a tag name""" + """ + Helper to construct opening and closing tag expressions for HTML, given a tag name. Matches + tags in either upper or lower case, attributes with namespaces and with quoted or unquoted values. + + Example:: + text = 'More info at the pyparsing wiki page' + # makeHTMLTags returns pyparsing expressions for the opening and closing tags as a 2-tuple + a,a_end = makeHTMLTags("A") + link_expr = a + SkipTo(a_end)("link_text") + a_end + + for link in link_expr.searchString(text): + # attributes in the tag (like "href" shown here) are also accessible as named results + print(link.link_text, '->', link.href) + prints:: + pyparsing -> http://pyparsing.wikispaces.com + """ return _makeTags( tagStr, False ) def makeXMLTags(tagStr): - """Helper to construct opening and closing tag expressions for XML, given a tag name""" + """ + Helper to construct opening and closing tag expressions for XML, given a tag name. Matches + tags only in the given upper/lower case. + + Example: similar to L{makeHTMLTags} + """ return _makeTags( tagStr, True ) def withAttribute(*args,**attrDict): - """Helper to create a validating parse action to be used with start tags created - with C{L{makeXMLTags}} or C{L{makeHTMLTags}}. Use C{withAttribute} to qualify a starting tag - with a required attribute value, to avoid false matches on common tags such as - C{} or C{
}. - - Call C{withAttribute} with a series of attribute names and values. Specify the list - of filter attributes names and values as: - - keyword arguments, as in C{(align="right")}, or - - as an explicit dict with C{**} operator, when an attribute name is also a Python + """ + Helper to create a validating parse action to be used with start tags created + with C{L{makeXMLTags}} or C{L{makeHTMLTags}}. Use C{withAttribute} to qualify a starting tag + with a required attribute value, to avoid false matches on common tags such as + C{} or C{
}. + + Call C{withAttribute} with a series of attribute names and values. Specify the list + of filter attributes names and values as: + - keyword arguments, as in C{(align="right")}, or + - as an explicit dict with C{**} operator, when an attribute name is also a Python reserved word, as in C{**{"class":"Customer", "align":"right"}} - - a list of name-value tuples, as in ( ("ns1:class", "Customer"), ("ns2:align","right") ) - For attribute names with a namespace prefix, you must use the second form. Attribute - names are matched insensitive to upper/lower case. + - a list of name-value tuples, as in ( ("ns1:class", "Customer"), ("ns2:align","right") ) + For attribute names with a namespace prefix, you must use the second form. Attribute + names are matched insensitive to upper/lower case. - If just testing for C{class} (with or without a namespace), use C{L{withClass}}. - - To verify that the attribute exists, but without specifying a value, pass - C{withAttribute.ANY_VALUE} as the value. - """ + If just testing for C{class} (with or without a namespace), use C{L{withClass}}. + + To verify that the attribute exists, but without specifying a value, pass + C{withAttribute.ANY_VALUE} as the value. + + Example:: + html = ''' +
+ Some text +
1 4 0 1 0
+
1,3 2,3 1,1
+
this has no type
+
+ + ''' + div,div_end = makeHTMLTags("div") + + # only match div tag having a type attribute with value "grid" + div_grid = div().setParseAction(withAttribute(type="grid")) + grid_expr = div_grid + SkipTo(div | div_end)("body") + for grid_header in grid_expr.searchString(html): + print(grid_header.body) + + # construct a match with any div tag having a type attribute, regardless of the value + div_any_type = div().setParseAction(withAttribute(type=withAttribute.ANY_VALUE)) + div_expr = div_any_type + SkipTo(div | div_end)("body") + for div_header in div_expr.searchString(html): + print(div_header.body) + prints:: + 1 4 0 1 0 + + 1 4 0 1 0 + 1,3 2,3 1,1 + """ if args: attrs = args[:] else: @@ -3558,9 +4854,37 @@ def withAttribute(*args,**attrDict): withAttribute.ANY_VALUE = object() def withClass(classname, namespace=''): - """Simplified version of C{L{withAttribute}} when matching on a div class - made - difficult because C{class} is a reserved word in Python. - """ + """ + Simplified version of C{L{withAttribute}} when matching on a div class - made + difficult because C{class} is a reserved word in Python. + + Example:: + html = ''' +
+ Some text +
1 4 0 1 0
+
1,3 2,3 1,1
+
this <div> has no class
+
+ + ''' + div,div_end = makeHTMLTags("div") + div_grid = div().setParseAction(withClass("grid")) + + grid_expr = div_grid + SkipTo(div | div_end)("body") + for grid_header in grid_expr.searchString(html): + print(grid_header.body) + + div_any_type = div().setParseAction(withClass(withAttribute.ANY_VALUE)) + div_expr = div_any_type + SkipTo(div | div_end)("body") + for div_header in div_expr.searchString(html): + print(div_header.body) + prints:: + 1 4 0 1 0 + + 1 4 0 1 0 + 1,3 2,3 1,1 + """ classattr = "%s:class" % namespace if namespace else "class" return withAttribute(**{classattr : classname}) @@ -3569,40 +4893,69 @@ opAssoc.LEFT = object() opAssoc.RIGHT = object() def infixNotation( baseExpr, opList, lpar=Suppress('('), rpar=Suppress(')') ): - """Helper method for constructing grammars of expressions made up of - operators working in a precedence hierarchy. Operators may be unary or - binary, left- or right-associative. Parse actions can also be attached - to operator expressions. - - Parameters: - - baseExpr - expression representing the most basic element for the nested - - opList - list of tuples, one for each operator precedence level in the - expression grammar; each tuple is of the form - (opExpr, numTerms, rightLeftAssoc, parseAction), where: - - opExpr is the pyparsing expression for the operator; - may also be a string, which will be converted to a Literal; - if numTerms is 3, opExpr is a tuple of two expressions, for the - two operators separating the 3 terms - - numTerms is the number of terms for this operator (must - be 1, 2, or 3) - - rightLeftAssoc is the indicator whether the operator is - right or left associative, using the pyparsing-defined - constants C{opAssoc.RIGHT} and C{opAssoc.LEFT}. - - parseAction is the parse action to be associated with - expressions matching this operator expression (the - parse action tuple member may be omitted) - - lpar - expression for matching left-parentheses (default=Suppress('(')) - - rpar - expression for matching right-parentheses (default=Suppress(')')) + """ + Helper method for constructing grammars of expressions made up of + operators working in a precedence hierarchy. Operators may be unary or + binary, left- or right-associative. Parse actions can also be attached + to operator expressions. + + Parameters: + - baseExpr - expression representing the most basic element for the nested + - opList - list of tuples, one for each operator precedence level in the + expression grammar; each tuple is of the form + (opExpr, numTerms, rightLeftAssoc, parseAction), where: + - opExpr is the pyparsing expression for the operator; + may also be a string, which will be converted to a Literal; + if numTerms is 3, opExpr is a tuple of two expressions, for the + two operators separating the 3 terms + - numTerms is the number of terms for this operator (must + be 1, 2, or 3) + - rightLeftAssoc is the indicator whether the operator is + right or left associative, using the pyparsing-defined + constants C{opAssoc.RIGHT} and C{opAssoc.LEFT}. + - parseAction is the parse action to be associated with + expressions matching this operator expression (the + parse action tuple member may be omitted) + - lpar - expression for matching left-parentheses (default=C{Suppress('(')}) + - rpar - expression for matching right-parentheses (default=C{Suppress(')')}) + + Example:: + # simple example of four-function arithmetic with ints and variable names + integer = pyparsing_common.signedInteger + varname = pyparsing_common.identifier + + arith_expr = infixNotation(integer | varname, + [ + ('-', 1, opAssoc.RIGHT), + (oneOf('* /'), 2, opAssoc.LEFT), + (oneOf('+ -'), 2, opAssoc.LEFT), + ]) + + arith_expr.runTests(''' + 5+3*6 + (5+3)*6 + -2--11 + ''', fullDump=False) + prints:: + 5+3*6 + [[5, '+', [3, '*', 6]]] + + (5+3)*6 + [[[5, '+', 3], '*', 6]] + + -2--11 + [[['-', 2], '-', ['-', 11]]] """ ret = Forward() lastExpr = baseExpr | ( lpar + ret + rpar ) for i,operDef in enumerate(opList): opExpr,arity,rightLeftAssoc,pa = (operDef + (None,))[:4] + termName = "%s term" % opExpr if arity < 3 else "%s%s term" % opExpr if arity == 3: if opExpr is None or len(opExpr) != 2: raise ValueError("if numterms=3, opExpr must be a tuple or list of two expressions") opExpr1, opExpr2 = opExpr - thisExpr = Forward()#.setName("expr%d" % i) + thisExpr = Forward().setName(termName) if rightLeftAssoc == opAssoc.LEFT: if arity == 1: matchExpr = FollowedBy(lastExpr + opExpr) + Group( lastExpr + OneOrMore( opExpr ) ) @@ -3636,37 +4989,77 @@ def infixNotation( baseExpr, opList, lpar=Suppress('('), rpar=Suppress(')') ): raise ValueError("operator must indicate right or left associativity") if pa: matchExpr.setParseAction( pa ) - thisExpr <<= ( matchExpr | lastExpr ) + thisExpr <<= ( matchExpr.setName(termName) | lastExpr ) lastExpr = thisExpr ret <<= lastExpr return ret + operatorPrecedence = infixNotation +"""(Deprecated) Former name of C{L{infixNotation}}, will be dropped in a future release.""" -dblQuotedString = Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*"').setName("string enclosed in double quotes") -sglQuotedString = Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*'").setName("string enclosed in single quotes") -quotedString = Regex(r'''(?:"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*")|(?:'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*')''').setName("quotedString using single or double quotes") -unicodeString = Combine(_L('u') + quotedString.copy()) +dblQuotedString = Combine(Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*')+'"').setName("string enclosed in double quotes") +sglQuotedString = Combine(Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*")+"'").setName("string enclosed in single quotes") +quotedString = Combine(Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*')+'"'| + Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*")+"'").setName("quotedString using single or double quotes") +unicodeString = Combine(_L('u') + quotedString.copy()).setName("unicode string literal") def nestedExpr(opener="(", closer=")", content=None, ignoreExpr=quotedString.copy()): - """Helper method for defining nested lists enclosed in opening and closing - delimiters ("(" and ")" are the default). - - Parameters: - - opener - opening character for a nested list (default="("); can also be a pyparsing expression - - closer - closing character for a nested list (default=")"); can also be a pyparsing expression - - content - expression for items within the nested lists (default=None) - - ignoreExpr - expression for ignoring opening and closing delimiters (default=quotedString) - - If an expression is not provided for the content argument, the nested - expression will capture all whitespace-delimited content between delimiters - as a list of separate values. - - Use the C{ignoreExpr} argument to define expressions that may contain - opening or closing characters that should not be treated as opening - or closing characters for nesting, such as quotedString or a comment - expression. Specify multiple expressions using an C{L{Or}} or C{L{MatchFirst}}. - The default is L{quotedString}, but if no expressions are to be ignored, - then pass C{None} for this argument. + """ + Helper method for defining nested lists enclosed in opening and closing + delimiters ("(" and ")" are the default). + + Parameters: + - opener - opening character for a nested list (default=C{"("}); can also be a pyparsing expression + - closer - closing character for a nested list (default=C{")"}); can also be a pyparsing expression + - content - expression for items within the nested lists (default=C{None}) + - ignoreExpr - expression for ignoring opening and closing delimiters (default=C{quotedString}) + + If an expression is not provided for the content argument, the nested + expression will capture all whitespace-delimited content between delimiters + as a list of separate values. + + Use the C{ignoreExpr} argument to define expressions that may contain + opening or closing characters that should not be treated as opening + or closing characters for nesting, such as quotedString or a comment + expression. Specify multiple expressions using an C{L{Or}} or C{L{MatchFirst}}. + The default is L{quotedString}, but if no expressions are to be ignored, + then pass C{None} for this argument. + + Example:: + data_type = oneOf("void int short long char float double") + decl_data_type = Combine(data_type + Optional(Word('*'))) + ident = Word(alphas+'_', alphanums+'_') + number = pyparsing_common.number + arg = Group(decl_data_type + ident) + LPAR,RPAR = map(Suppress, "()") + + code_body = nestedExpr('{', '}', ignoreExpr=(quotedString | cStyleComment)) + + c_function = (decl_data_type("type") + + ident("name") + + LPAR + Optional(delimitedList(arg), [])("args") + RPAR + + code_body("body")) + c_function.ignore(cStyleComment) + + source_code = ''' + int is_odd(int x) { + return (x%2); + } + + int dec_to_hex(char hchar) { + if (hchar >= '0' && hchar <= '9') { + return (ord(hchar)-ord('0')); + } else { + return (10+ord(hchar)-ord('A')); + } + } + ''' + for func in c_function.searchString(source_code): + print("%(name)s (%(type)s) args: %(args)s" % func) + + prints:: + is_odd (int) args: [['int', 'x']] + dec_to_hex (int) args: [['char', 'hchar']] """ if opener == closer: raise ValueError("opening and closing strings cannot be the same") @@ -3697,23 +5090,86 @@ def nestedExpr(opener="(", closer=")", content=None, ignoreExpr=quotedString.cop ret <<= Group( Suppress(opener) + ZeroOrMore( ignoreExpr | ret | content ) + Suppress(closer) ) else: ret <<= Group( Suppress(opener) + ZeroOrMore( ret | content ) + Suppress(closer) ) + ret.setName('nested %s%s expression' % (opener,closer)) return ret def indentedBlock(blockStatementExpr, indentStack, indent=True): - """Helper method for defining space-delimited indentation blocks, such as - those used to define block statements in Python source code. + """ + Helper method for defining space-delimited indentation blocks, such as + those used to define block statements in Python source code. - Parameters: - - blockStatementExpr - expression defining syntax of statement that + Parameters: + - blockStatementExpr - expression defining syntax of statement that is repeated within the indented block - - indentStack - list created by caller to manage indentation stack + - indentStack - list created by caller to manage indentation stack (multiple statementWithIndentedBlock expressions within a single grammar should share a common indentStack) - - indent - boolean indicating whether block must be indented beyond the + - indent - boolean indicating whether block must be indented beyond the the current level; set to False for block of left-most statements - (default=True) - - A valid block must contain at least one C{blockStatement}. + (default=C{True}) + + A valid block must contain at least one C{blockStatement}. + + Example:: + data = ''' + def A(z): + A1 + B = 100 + G = A2 + A2 + A3 + B + def BB(a,b,c): + BB1 + def BBA(): + bba1 + bba2 + bba3 + C + D + def spam(x,y): + def eggs(z): + pass + ''' + + + indentStack = [1] + stmt = Forward() + + identifier = Word(alphas, alphanums) + funcDecl = ("def" + identifier + Group( "(" + Optional( delimitedList(identifier) ) + ")" ) + ":") + func_body = indentedBlock(stmt, indentStack) + funcDef = Group( funcDecl + func_body ) + + rvalue = Forward() + funcCall = Group(identifier + "(" + Optional(delimitedList(rvalue)) + ")") + rvalue << (funcCall | identifier | Word(nums)) + assignment = Group(identifier + "=" + rvalue) + stmt << ( funcDef | assignment | identifier ) + + module_body = OneOrMore(stmt) + + parseTree = module_body.parseString(data) + parseTree.pprint() + prints:: + [['def', + 'A', + ['(', 'z', ')'], + ':', + [['A1'], [['B', '=', '100']], [['G', '=', 'A2']], ['A2'], ['A3']]], + 'B', + ['def', + 'BB', + ['(', 'a', 'b', 'c', ')'], + ':', + [['BB1'], [['def', 'BBA', ['(', ')'], ':', [['bba1'], ['bba2'], ['bba3']]]]]], + 'C', + 'D', + ['def', + 'spam', + ['(', 'x', 'y', ')'], + ':', + [[['def', 'eggs', ['(', 'z', ')'], ':', [['pass']]]]]]] """ def checkPeerIndent(s,l,t): if l >= len(s): return @@ -3738,9 +5194,9 @@ def indentedBlock(blockStatementExpr, indentStack, indent=True): indentStack.pop() NL = OneOrMore(LineEnd().setWhitespaceChars("\t ").suppress()) - INDENT = Empty() + Empty().setParseAction(checkSubIndent) - PEER = Empty().setParseAction(checkPeerIndent) - UNDENT = Empty().setParseAction(checkUnindent) + INDENT = (Empty() + Empty().setParseAction(checkSubIndent)).setName('INDENT') + PEER = Empty().setParseAction(checkPeerIndent).setName('') + UNDENT = Empty().setParseAction(checkUnindent).setName('UNINDENT') if indent: smExpr = Group( Optional(NL) + #~ FollowedBy(blockStatementExpr) + @@ -3749,57 +5205,371 @@ def indentedBlock(blockStatementExpr, indentStack, indent=True): smExpr = Group( Optional(NL) + (OneOrMore( PEER + Group(blockStatementExpr) + Optional(NL) )) ) blockStatementExpr.ignore(_bslash + LineEnd()) - return smExpr + return smExpr.setName('indented block') alphas8bit = srange(r"[\0xc0-\0xd6\0xd8-\0xf6\0xf8-\0xff]") punc8bit = srange(r"[\0xa1-\0xbf\0xd7\0xf7]") -anyOpenTag,anyCloseTag = makeHTMLTags(Word(alphas,alphanums+"_:")) -commonHTMLEntity = Combine(_L("&") + oneOf("gt lt amp nbsp quot").setResultsName("entity") +";").streamline() -_htmlEntityMap = dict(zip("gt lt amp nbsp quot".split(),'><& "')) -replaceHTMLEntity = lambda t : t.entity in _htmlEntityMap and _htmlEntityMap[t.entity] or None +anyOpenTag,anyCloseTag = makeHTMLTags(Word(alphas,alphanums+"_:").setName('any tag')) +_htmlEntityMap = dict(zip("gt lt amp nbsp quot apos".split(),'><& "\'')) +commonHTMLEntity = Regex('&(?P' + '|'.join(_htmlEntityMap.keys()) +");").setName("common HTML entity") +def replaceHTMLEntity(t): + """Helper parser action to replace common HTML entities with their special characters""" + return _htmlEntityMap.get(t.entity) # it's easy to get these comment structures wrong - they're very common, so may as well make them available -cStyleComment = Regex(r"/\*(?:[^*]*\*+)+?/").setName("C style comment") +cStyleComment = Combine(Regex(r"/\*(?:[^*]|\*(?!/))*") + '*/').setName("C style comment") +"Comment of the form C{/* ... */}" -htmlComment = Regex(r"") -restOfLine = Regex(r".*").leaveWhitespace() -dblSlashComment = Regex(r"\/\/(\\\n|.)*").setName("// comment") -cppStyleComment = Regex(r"/(?:\*(?:[^*]*\*+)+?/|/[^\n]*(?:\n[^\n]*)*?(?:(?").setName("HTML comment") +"Comment of the form C{}" + +restOfLine = Regex(r".*").leaveWhitespace().setName("rest of line") +dblSlashComment = Regex(r"//(?:\\\n|[^\n])*").setName("// comment") +"Comment of the form C{// ... (to end of line)}" + +cppStyleComment = Combine(Regex(r"/\*(?:[^*]|\*(?!/))*") + '*/'| dblSlashComment).setName("C++ style comment") +"Comment of either form C{L{cStyleComment}} or C{L{dblSlashComment}}" javaStyleComment = cppStyleComment +"Same as C{L{cppStyleComment}}" + pythonStyleComment = Regex(r"#.*").setName("Python style comment") +"Comment of the form C{# ... (to end of line)}" + _commasepitem = Combine(OneOrMore(Word(printables, excludeChars=',') + Optional( Word(" \t") + ~Literal(",") + ~LineEnd() ) ) ).streamline().setName("commaItem") commaSeparatedList = delimitedList( Optional( quotedString.copy() | _commasepitem, default="") ).setName("commaSeparatedList") +"""Predefined expression of 1 or more printable words or quoted strings, separated by commas.""" + +# some other useful expressions - using lower-case class name since we are really using this as a namespace +class pyparsing_common: + """ + Here are some common low-level expressions that may be useful in jump-starting parser development: + - numeric forms (L{integers}, L{reals}, L{scientific notation}) + - common L{programming identifiers} + - network addresses (L{MAC}, L{IPv4}, L{IPv6}) + - ISO8601 L{dates} and L{datetime} + - L{UUID} + Parse actions: + - C{L{convertToInteger}} + - C{L{convertToFloat}} + - C{L{convertToDate}} + - C{L{convertToDatetime}} + - C{L{stripHTMLTags}} + + Example:: + pyparsing_common.number.runTests(''' + # any int or real number, returned as the appropriate type + 100 + -100 + +100 + 3.14159 + 6.02e23 + 1e-12 + ''') + + pyparsing_common.fnumber.runTests(''' + # any int or real number, returned as float + 100 + -100 + +100 + 3.14159 + 6.02e23 + 1e-12 + ''') + + pyparsing_common.hex_integer.runTests(''' + # hex numbers + 100 + FF + ''') + + pyparsing_common.fraction.runTests(''' + # fractions + 1/2 + -3/4 + ''') + + pyparsing_common.mixed_integer.runTests(''' + # mixed fractions + 1 + 1/2 + -3/4 + 1-3/4 + ''') + + import uuid + pyparsing_common.uuid.setParseAction(tokenMap(uuid.UUID)) + pyparsing_common.uuid.runTests(''' + # uuid + 12345678-1234-5678-1234-567812345678 + ''') + prints:: + # any int or real number, returned as the appropriate type + 100 + [100] + + -100 + [-100] + + +100 + [100] + + 3.14159 + [3.14159] + + 6.02e23 + [6.02e+23] + + 1e-12 + [1e-12] + + # any int or real number, returned as float + 100 + [100.0] + + -100 + [-100.0] + + +100 + [100.0] + + 3.14159 + [3.14159] + + 6.02e23 + [6.02e+23] + + 1e-12 + [1e-12] + + # hex numbers + 100 + [256] + + FF + [255] + + # fractions + 1/2 + [0.5] + + -3/4 + [-0.75] + + # mixed fractions + 1 + [1] + + 1/2 + [0.5] + + -3/4 + [-0.75] + + 1-3/4 + [1.75] + + # uuid + 12345678-1234-5678-1234-567812345678 + [UUID('12345678-1234-5678-1234-567812345678')] + """ + + convertToInteger = tokenMap(int) + """ + Parse action for converting parsed integers to Python int + """ + + convertToFloat = tokenMap(float) + """ + Parse action for converting parsed numbers to Python float + """ + + integer = Word(nums).setName("integer").setParseAction(convertToInteger) + """expression that parses an unsigned integer, returns an int""" + + hex_integer = Word(hexnums).setName("hex integer").setParseAction(tokenMap(int,16)) + """expression that parses a hexadecimal integer, returns an int""" + + signedInteger = Regex(r'[+-]?\d+').setName("signed integer").setParseAction(convertToInteger) + """expression that parses an integer with optional leading sign, returns an int""" + + fraction = (signedInteger().setParseAction(convertToFloat) + '/' + signedInteger().setParseAction(convertToFloat)).setName("fraction") + """fractional expression of an integer divided by an integer, returns a float""" + fraction.addParseAction(lambda t: t[0]/t[-1]) + + mixed_integer = (fraction | signedInteger + Optional(Optional('-').suppress() + fraction)).setName("fraction or mixed integer-fraction") + """mixed integer of the form 'integer - fraction', with optional leading integer, returns float""" + mixed_integer.addParseAction(sum) + + real = Regex(r'[+-]?\d+\.\d*').setName("real number").setParseAction(convertToFloat) + """expression that parses a floating point number and returns a float""" + sciReal = Regex(r'[+-]?\d+([eE][+-]?\d+|\.\d*([eE][+-]?\d+)?)').setName("real number with scientific notation").setParseAction(convertToFloat) + """expression that parses a floating point number with optional scientific notation and returns a float""" + + # streamlining this expression makes the docs nicer-looking + number = (sciReal | real | signedInteger).streamline() + """any numeric expression, returns the corresponding Python type""" + + fnumber = Regex(r'[+-]?\d+\.?\d*([eE][+-]?\d+)?').setName("fnumber").setParseAction(convertToFloat) + """any int or real number, returned as float""" + + identifier = Word(alphas+'_', alphanums+'_').setName("identifier") + """typical code identifier (leading alpha or '_', followed by 0 or more alphas, nums, or '_')""" + + ipv4_address = Regex(r'(25[0-5]|2[0-4][0-9]|1?[0-9]{1,2})(\.(25[0-5]|2[0-4][0-9]|1?[0-9]{1,2})){3}').setName("IPv4 address") + "IPv4 address (C{0.0.0.0 - 255.255.255.255})" + + _ipv6_part = Regex(r'[0-9a-fA-F]{1,4}').setName("hex_integer") + _full_ipv6_address = (_ipv6_part + (':' + _ipv6_part)*7).setName("full IPv6 address") + _short_ipv6_address = (Optional(_ipv6_part + (':' + _ipv6_part)*(0,6)) + "::" + Optional(_ipv6_part + (':' + _ipv6_part)*(0,6))).setName("short IPv6 address") + _short_ipv6_address.addCondition(lambda t: sum(1 for tt in t if pyparsing_common._ipv6_part.matches(tt)) < 8) + _mixed_ipv6_address = ("::ffff:" + ipv4_address).setName("mixed IPv6 address") + ipv6_address = Combine((_full_ipv6_address | _mixed_ipv6_address | _short_ipv6_address).setName("IPv6 address")).setName("IPv6 address") + "IPv6 address (long, short, or mixed form)" + + mac_address = Regex(r'[0-9a-fA-F]{2}([:.-])[0-9a-fA-F]{2}(?:\1[0-9a-fA-F]{2}){4}').setName("MAC address") + "MAC address xx:xx:xx:xx:xx (may also have '-' or '.' delimiters)" + + @staticmethod + def convertToDate(fmt="%Y-%m-%d"): + """ + Helper to create a parse action for converting parsed date string to Python datetime.date + + Params - + - fmt - format to be passed to datetime.strptime (default=C{"%Y-%m-%d"}) + + Example:: + date_expr = pyparsing_common.iso8601_date.copy() + date_expr.setParseAction(pyparsing_common.convertToDate()) + print(date_expr.parseString("1999-12-31")) + prints:: + [datetime.date(1999, 12, 31)] + """ + def cvt_fn(s,l,t): + try: + return datetime.strptime(t[0], fmt).date() + except ValueError as ve: + raise ParseException(s, l, str(ve)) + return cvt_fn + + @staticmethod + def convertToDatetime(fmt="%Y-%m-%dT%H:%M:%S.%f"): + """ + Helper to create a parse action for converting parsed datetime string to Python datetime.datetime + + Params - + - fmt - format to be passed to datetime.strptime (default=C{"%Y-%m-%dT%H:%M:%S.%f"}) + + Example:: + dt_expr = pyparsing_common.iso8601_datetime.copy() + dt_expr.setParseAction(pyparsing_common.convertToDatetime()) + print(dt_expr.parseString("1999-12-31T23:59:59.999")) + prints:: + [datetime.datetime(1999, 12, 31, 23, 59, 59, 999000)] + """ + def cvt_fn(s,l,t): + try: + return datetime.strptime(t[0], fmt) + except ValueError as ve: + raise ParseException(s, l, str(ve)) + return cvt_fn + + iso8601_date = Regex(r'(?P\d{4})(?:-(?P\d\d)(?:-(?P\d\d))?)?').setName("ISO8601 date") + "ISO8601 date (C{yyyy-mm-dd})" + + iso8601_datetime = Regex(r'(?P\d{4})-(?P\d\d)-(?P\d\d)[T ](?P\d\d):(?P\d\d)(:(?P\d\d(\.\d*)?)?)?(?PZ|[+-]\d\d:?\d\d)?').setName("ISO8601 datetime") + "ISO8601 datetime (C{yyyy-mm-ddThh:mm:ss.s(Z|+-00:00)}) - trailing seconds, milliseconds, and timezone optional; accepts separating C{'T'} or C{' '}" + + uuid = Regex(r'[0-9a-fA-F]{8}(-[0-9a-fA-F]{4}){3}-[0-9a-fA-F]{12}').setName("UUID") + "UUID (C{xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx})" + + _html_stripper = anyOpenTag.suppress() | anyCloseTag.suppress() + @staticmethod + def stripHTMLTags(s, l, tokens): + """ + Parse action to remove HTML tags from web page HTML source + + Example:: + # strip HTML links from normal text + text = 'More info at the pyparsing wiki page' + td,td_end = makeHTMLTags("TD") + table_text = td + SkipTo(td_end).setParseAction(pyparsing_common.stripHTMLTags)("body") + td_end + + print(table_text.parseString(text).body) # -> 'More info at the pyparsing wiki page' + """ + return pyparsing_common._html_stripper.transformString(tokens[0]) if __name__ == "__main__": - selectToken = CaselessLiteral( "select" ) - fromToken = CaselessLiteral( "from" ) - - ident = Word( alphas, alphanums + "_$" ) - columnName = delimitedList( ident, ".", combine=True ).setParseAction( upcaseTokens ) - columnNameList = Group( delimitedList( columnName ) ).setName("columns") - tableName = delimitedList( ident, ".", combine=True ).setParseAction( upcaseTokens ) - tableNameList = Group( delimitedList( tableName ) ).setName("tables") - simpleSQL = ( selectToken + \ - ( '*' | columnNameList ).setResultsName( "columns" ) + \ - fromToken + \ - tableNameList.setResultsName( "tables" ) ) - - simpleSQL.runTests("""\ - SELECT * from XYZZY, ABC - select * from SYS.XYZZY - Select A from Sys.dual - Select AA,BB,CC from Sys.dual - Select A, B, C from Sys.dual - Select A, B, C from Sys.dual - Xelect A, B, C from Sys.dual - Select A, B, C frox Sys.dual - Select - Select ^^^ frox Sys.dual - Select A, B, C from Sys.dual, Table2""") + selectToken = CaselessLiteral("select") + fromToken = CaselessLiteral("from") + + ident = Word(alphas, alphanums + "_$") + + columnName = delimitedList(ident, ".", combine=True).setParseAction(upcaseTokens) + columnNameList = Group(delimitedList(columnName)).setName("columns") + columnSpec = ('*' | columnNameList) + + tableName = delimitedList(ident, ".", combine=True).setParseAction(upcaseTokens) + tableNameList = Group(delimitedList(tableName)).setName("tables") + simpleSQL = selectToken("command") + columnSpec("columns") + fromToken + tableNameList("tables") + + # demo runTests method, including embedded comments in test string + simpleSQL.runTests(""" + # '*' as column list and dotted table name + select * from SYS.XYZZY + + # caseless match on "SELECT", and casts back to "select" + SELECT * from XYZZY, ABC + + # list of column names, and mixed case SELECT keyword + Select AA,BB,CC from Sys.dual + + # multiple tables + Select A, B, C from Sys.dual, Table2 + + # invalid SELECT keyword - should fail + Xelect A, B, C from Sys.dual + + # incomplete command - should fail + Select + + # invalid column name - should fail + Select ^^^ frox Sys.dual + + """) + + pyparsing_common.number.runTests(""" + 100 + -100 + +100 + 3.14159 + 6.02e23 + 1e-12 + """) + + # any int or real number, returned as float + pyparsing_common.fnumber.runTests(""" + 100 + -100 + +100 + 3.14159 + 6.02e23 + 1e-12 + """) + + pyparsing_common.hex_integer.runTests(""" + 100 + FF + """) + + import uuid + pyparsing_common.uuid.setParseAction(tokenMap(uuid.UUID)) + pyparsing_common.uuid.runTests(""" + 12345678-1234-5678-1234-567812345678 + """) -- cgit v1.2.1