--- /dev/null
+# module pyparsing.py\r
+#\r
+# Copyright (c) 2003-2016 Paul T. McGuire\r
+#\r
+# Permission is hereby granted, free of charge, to any person obtaining\r
+# a copy of this software and associated documentation files (the\r
+# "Software"), to deal in the Software without restriction, including\r
+# without limitation the rights to use, copy, modify, merge, publish,\r
+# distribute, sublicense, and/or sell copies of the Software, and to\r
+# permit persons to whom the Software is furnished to do so, subject to\r
+# the following conditions:\r
+#\r
+# The above copyright notice and this permission notice shall be\r
+# included in all copies or substantial portions of the Software.\r
+#\r
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,\r
+# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF\r
+# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.\r
+# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY\r
+# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,\r
+# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE\r
+# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.\r
+#\r
+\r
+__doc__ = \\r
+"""\r
+pyparsing module - Classes and methods to define and execute parsing grammars\r
+\r
+The pyparsing module is an alternative approach to creating and executing simple grammars,\r
+vs. the traditional lex/yacc approach, or the use of regular expressions. With pyparsing, you\r
+don't need to learn a new syntax for defining grammars or matching expressions - the parsing module\r
+provides a library of classes that you use to construct the grammar directly in Python.\r
+\r
+Here is a program to parse "Hello, World!" (or any greeting of the form \r
+C{"<salutation>, <addressee>!"}), built up using L{Word}, L{Literal}, and L{And} elements \r
+(L{'+'<ParserElement.__add__>} operator gives L{And} expressions, strings are auto-converted to\r
+L{Literal} expressions)::\r
+\r
+ from pyparsing import Word, alphas\r
+\r
+ # define grammar of a greeting\r
+ greet = Word(alphas) + "," + Word(alphas) + "!"\r
+\r
+ hello = "Hello, World!"\r
+ print (hello, "->", greet.parseString(hello))\r
+\r
+The program outputs the following::\r
+\r
+ Hello, World! -> ['Hello', ',', 'World', '!']\r
+\r
+The Python representation of the grammar is quite readable, owing to the self-explanatory\r
+class names, and the use of '+', '|' and '^' operators.\r
+\r
+The L{ParseResults} object returned from L{ParserElement.parseString<ParserElement.parseString>} can be accessed as a nested list, a dictionary, or an\r
+object with named attributes.\r
+\r
+The pyparsing module handles some of the problems that are typically vexing when writing text parsers:\r
+ - extra or missing whitespace (the above program will also handle "Hello,World!", "Hello , World !", etc.)\r
+ - quoted strings\r
+ - embedded comments\r
+"""\r
+\r
+__version__ = "2.1.10"\r
+__versionTime__ = "07 Oct 2016 01:31 UTC"\r
+__author__ = "Paul McGuire <ptmcg@users.sourceforge.net>"\r
+\r
+import string\r
+from weakref import ref as wkref\r
+import copy\r
+import sys\r
+import warnings\r
+import re\r
+import sre_constants\r
+import collections\r
+import pprint\r
+import traceback\r
+import types\r
+from datetime import datetime\r
+\r
+try:\r
+ from _thread import RLock\r
+except ImportError:\r
+ from threading import RLock\r
+\r
+try:\r
+ from collections import OrderedDict as _OrderedDict\r
+except ImportError:\r
+ try:\r
+ from ordereddict import OrderedDict as _OrderedDict\r
+ except ImportError:\r
+ _OrderedDict = None\r
+\r
+#~ sys.stderr.write( "testing pyparsing module, version %s, %s\n" % (__version__,__versionTime__ ) )\r
+\r
+__all__ = [\r
+'And', 'CaselessKeyword', 'CaselessLiteral', 'CharsNotIn', 'Combine', 'Dict', 'Each', 'Empty',\r
+'FollowedBy', 'Forward', 'GoToColumn', 'Group', 'Keyword', 'LineEnd', 'LineStart', 'Literal',\r
+'MatchFirst', 'NoMatch', 'NotAny', 'OneOrMore', 'OnlyOnce', 'Optional', 'Or',\r
+'ParseBaseException', 'ParseElementEnhance', 'ParseException', 'ParseExpression', 'ParseFatalException',\r
+'ParseResults', 'ParseSyntaxException', 'ParserElement', 'QuotedString', 'RecursiveGrammarException',\r
+'Regex', 'SkipTo', 'StringEnd', 'StringStart', 'Suppress', 'Token', 'TokenConverter', \r
+'White', 'Word', 'WordEnd', 'WordStart', 'ZeroOrMore',\r
+'alphanums', 'alphas', 'alphas8bit', 'anyCloseTag', 'anyOpenTag', 'cStyleComment', 'col',\r
+'commaSeparatedList', 'commonHTMLEntity', 'countedArray', 'cppStyleComment', 'dblQuotedString',\r
+'dblSlashComment', 'delimitedList', 'dictOf', 'downcaseTokens', 'empty', 'hexnums',\r
+'htmlComment', 'javaStyleComment', 'line', 'lineEnd', 'lineStart', 'lineno',\r
+'makeHTMLTags', 'makeXMLTags', 'matchOnlyAtCol', 'matchPreviousExpr', 'matchPreviousLiteral',\r
+'nestedExpr', 'nullDebugAction', 'nums', 'oneOf', 'opAssoc', 'operatorPrecedence', 'printables',\r
+'punc8bit', 'pythonStyleComment', 'quotedString', 'removeQuotes', 'replaceHTMLEntity', \r
+'replaceWith', 'restOfLine', 'sglQuotedString', 'srange', 'stringEnd',\r
+'stringStart', 'traceParseAction', 'unicodeString', 'upcaseTokens', 'withAttribute',\r
+'indentedBlock', 'originalTextFor', 'ungroup', 'infixNotation','locatedExpr', 'withClass',\r
+'CloseMatch', 'tokenMap', 'pyparsing_common',\r
+]\r
+\r
+system_version = tuple(sys.version_info)[:3]\r
+PY_3 = system_version[0] == 3\r
+if PY_3:\r
+ _MAX_INT = sys.maxsize\r
+ basestring = str\r
+ unichr = chr\r
+ _ustr = str\r
+\r
+ # build list of single arg builtins, that can be used as parse actions\r
+ singleArgBuiltins = [sum, len, sorted, reversed, list, tuple, set, any, all, min, max]\r
+\r
+else:\r
+ _MAX_INT = sys.maxint\r
+ range = xrange\r
+\r
+ def _ustr(obj):\r
+ """Drop-in replacement for str(obj) that tries to be Unicode friendly. It first tries\r
+ str(obj). If that fails with a UnicodeEncodeError, then it tries unicode(obj). It\r
+ then < returns the unicode object | encodes it with the default encoding | ... >.\r
+ """\r
+ if isinstance(obj,unicode):\r
+ return obj\r
+\r
+ try:\r
+ # If this works, then _ustr(obj) has the same behaviour as str(obj), so\r
+ # it won't break any existing code.\r
+ return str(obj)\r
+\r
+ except UnicodeEncodeError:\r
+ # Else encode it\r
+ ret = unicode(obj).encode(sys.getdefaultencoding(), 'xmlcharrefreplace')\r
+ xmlcharref = Regex('&#\d+;')\r
+ xmlcharref.setParseAction(lambda t: '\\u' + hex(int(t[0][2:-1]))[2:])\r
+ return xmlcharref.transformString(ret)\r
+\r
+ # build list of single arg builtins, tolerant of Python version, that can be used as parse actions\r
+ singleArgBuiltins = []\r
+ import __builtin__\r
+ for fname in "sum len sorted reversed list tuple set any all min max".split():\r
+ try:\r
+ singleArgBuiltins.append(getattr(__builtin__,fname))\r
+ except AttributeError:\r
+ continue\r
+ \r
+_generatorType = type((y for y in range(1)))\r
+ \r
+def _xml_escape(data):\r
+ """Escape &, <, >, ", ', etc. in a string of data."""\r
+\r
+ # ampersand must be replaced first\r
+ from_symbols = '&><"\''\r
+ to_symbols = ('&'+s+';' for s in "amp gt lt quot apos".split())\r
+ for from_,to_ in zip(from_symbols, to_symbols):\r
+ data = data.replace(from_, to_)\r
+ return data\r
+\r
+class _Constants(object):\r
+ pass\r
+\r
+alphas = string.ascii_uppercase + string.ascii_lowercase\r
+nums = "0123456789"\r
+hexnums = nums + "ABCDEFabcdef"\r
+alphanums = alphas + nums\r
+_bslash = chr(92)\r
+printables = "".join(c for c in string.printable if c not in string.whitespace)\r
+\r
+class ParseBaseException(Exception):\r
+ """base exception class for all parsing runtime exceptions"""\r
+ # Performance tuning: we construct a *lot* of these, so keep this\r
+ # constructor as small and fast as possible\r
+ def __init__( self, pstr, loc=0, msg=None, elem=None ):\r
+ self.loc = loc\r
+ if msg is None:\r
+ self.msg = pstr\r
+ self.pstr = ""\r
+ else:\r
+ self.msg = msg\r
+ self.pstr = pstr\r
+ self.parserElement = elem\r
+ self.args = (pstr, loc, msg)\r
+\r
+ @classmethod\r
+ def _from_exception(cls, pe):\r
+ """\r
+ internal factory method to simplify creating one type of ParseException \r
+ from another - avoids having __init__ signature conflicts among subclasses\r
+ """\r
+ return cls(pe.pstr, pe.loc, pe.msg, pe.parserElement)\r
+\r
+ def __getattr__( self, aname ):\r
+ """supported attributes by name are:\r
+ - lineno - returns the line number of the exception text\r
+ - col - returns the column number of the exception text\r
+ - line - returns the line containing the exception text\r
+ """\r
+ if( aname == "lineno" ):\r
+ return lineno( self.loc, self.pstr )\r
+ elif( aname in ("col", "column") ):\r
+ return col( self.loc, self.pstr )\r
+ elif( aname == "line" ):\r
+ return line( self.loc, self.pstr )\r
+ else:\r
+ raise AttributeError(aname)\r
+\r
+ def __str__( self ):\r
+ return "%s (at char %d), (line:%d, col:%d)" % \\r
+ ( self.msg, self.loc, self.lineno, self.column )\r
+ def __repr__( self ):\r
+ return _ustr(self)\r
+ def markInputline( self, markerString = ">!<" ):\r
+ """Extracts the exception line from the input string, and marks\r
+ the location of the exception with a special symbol.\r
+ """\r
+ line_str = self.line\r
+ line_column = self.column - 1\r
+ if markerString:\r
+ line_str = "".join((line_str[:line_column],\r
+ markerString, line_str[line_column:]))\r
+ return line_str.strip()\r
+ def __dir__(self):\r
+ return "lineno col line".split() + dir(type(self))\r
+\r
+class ParseException(ParseBaseException):\r
+ """\r
+ Exception thrown when parse expressions don't match class;\r
+ supported attributes by name are:\r
+ - lineno - returns the line number of the exception text\r
+ - col - returns the column number of the exception text\r
+ - line - returns the line containing the exception text\r
+ \r
+ Example::\r
+ try:\r
+ Word(nums).setName("integer").parseString("ABC")\r
+ except ParseException as pe:\r
+ print(pe)\r
+ print("column: {}".format(pe.col))\r
+ \r
+ prints::\r
+ Expected integer (at char 0), (line:1, col:1)\r
+ column: 1\r
+ """\r
+ pass\r
+\r
+class ParseFatalException(ParseBaseException):\r
+ """user-throwable exception thrown when inconsistent parse content\r
+ is found; stops all parsing immediately"""\r
+ pass\r
+\r
+class ParseSyntaxException(ParseFatalException):\r
+ """just like L{ParseFatalException}, but thrown internally when an\r
+ L{ErrorStop<And._ErrorStop>} ('-' operator) indicates that parsing is to stop \r
+ immediately because an unbacktrackable syntax error has been found"""\r
+ pass\r
+\r
+#~ class ReparseException(ParseBaseException):\r
+ #~ """Experimental class - parse actions can raise this exception to cause\r
+ #~ pyparsing to reparse the input string:\r
+ #~ - with a modified input string, and/or\r
+ #~ - with a modified start location\r
+ #~ Set the values of the ReparseException in the constructor, and raise the\r
+ #~ exception in a parse action to cause pyparsing to use the new string/location.\r
+ #~ Setting the values as None causes no change to be made.\r
+ #~ """\r
+ #~ def __init_( self, newstring, restartLoc ):\r
+ #~ self.newParseText = newstring\r
+ #~ self.reparseLoc = restartLoc\r
+\r
+class RecursiveGrammarException(Exception):\r
+ """exception thrown by L{ParserElement.validate} if the grammar could be improperly recursive"""\r
+ def __init__( self, parseElementList ):\r
+ self.parseElementTrace = parseElementList\r
+\r
+ def __str__( self ):\r
+ return "RecursiveGrammarException: %s" % self.parseElementTrace\r
+\r
+class _ParseResultsWithOffset(object):\r
+ def __init__(self,p1,p2):\r
+ self.tup = (p1,p2)\r
+ def __getitem__(self,i):\r
+ return self.tup[i]\r
+ def __repr__(self):\r
+ return repr(self.tup[0])\r
+ def setOffset(self,i):\r
+ self.tup = (self.tup[0],i)\r
+\r
+class ParseResults(object):\r
+ """\r
+ Structured parse results, to provide multiple means of access to the parsed data:\r
+ - as a list (C{len(results)})\r
+ - by list index (C{results[0], results[1]}, etc.)\r
+ - by attribute (C{results.<resultsName>} - see L{ParserElement.setResultsName})\r
+\r
+ Example::\r
+ integer = Word(nums)\r
+ date_str = (integer.setResultsName("year") + '/' \r
+ + integer.setResultsName("month") + '/' \r
+ + integer.setResultsName("day"))\r
+ # equivalent form:\r
+ # date_str = integer("year") + '/' + integer("month") + '/' + integer("day")\r
+\r
+ # parseString returns a ParseResults object\r
+ result = date_str.parseString("1999/12/31")\r
+\r
+ def test(s, fn=repr):\r
+ print("%s -> %s" % (s, fn(eval(s))))\r
+ test("list(result)")\r
+ test("result[0]")\r
+ test("result['month']")\r
+ test("result.day")\r
+ test("'month' in result")\r
+ test("'minutes' in result")\r
+ test("result.dump()", str)\r
+ prints::\r
+ list(result) -> ['1999', '/', '12', '/', '31']\r
+ result[0] -> '1999'\r
+ result['month'] -> '12'\r
+ result.day -> '31'\r
+ 'month' in result -> True\r
+ 'minutes' in result -> False\r
+ result.dump() -> ['1999', '/', '12', '/', '31']\r
+ - day: 31\r
+ - month: 12\r
+ - year: 1999\r
+ """\r
+ def __new__(cls, toklist=None, name=None, asList=True, modal=True ):\r
+ if isinstance(toklist, cls):\r
+ return toklist\r
+ retobj = object.__new__(cls)\r
+ retobj.__doinit = True\r
+ return retobj\r
+\r
+ # Performance tuning: we construct a *lot* of these, so keep this\r
+ # constructor as small and fast as possible\r
+ def __init__( self, toklist=None, name=None, asList=True, modal=True, isinstance=isinstance ):\r
+ if self.__doinit:\r
+ self.__doinit = False\r
+ self.__name = None\r
+ self.__parent = None\r
+ self.__accumNames = {}\r
+ self.__asList = asList\r
+ self.__modal = modal\r
+ if toklist is None:\r
+ toklist = []\r
+ if isinstance(toklist, list):\r
+ self.__toklist = toklist[:]\r
+ elif isinstance(toklist, _generatorType):\r
+ self.__toklist = list(toklist)\r
+ else:\r
+ self.__toklist = [toklist]\r
+ self.__tokdict = dict()\r
+\r
+ if name is not None and name:\r
+ if not modal:\r
+ self.__accumNames[name] = 0\r
+ if isinstance(name,int):\r
+ name = _ustr(name) # will always return a str, but use _ustr for consistency\r
+ self.__name = name\r
+ if not (isinstance(toklist, (type(None), basestring, list)) and toklist in (None,'',[])):\r
+ if isinstance(toklist,basestring):\r
+ toklist = [ toklist ]\r
+ if asList:\r
+ if isinstance(toklist,ParseResults):\r
+ self[name] = _ParseResultsWithOffset(toklist.copy(),0)\r
+ else:\r
+ self[name] = _ParseResultsWithOffset(ParseResults(toklist[0]),0)\r
+ self[name].__name = name\r
+ else:\r
+ try:\r
+ self[name] = toklist[0]\r
+ except (KeyError,TypeError,IndexError):\r
+ self[name] = toklist\r
+\r
+ def __getitem__( self, i ):\r
+ if isinstance( i, (int,slice) ):\r
+ return self.__toklist[i]\r
+ else:\r
+ if i not in self.__accumNames:\r
+ return self.__tokdict[i][-1][0]\r
+ else:\r
+ return ParseResults([ v[0] for v in self.__tokdict[i] ])\r
+\r
+ def __setitem__( self, k, v, isinstance=isinstance ):\r
+ if isinstance(v,_ParseResultsWithOffset):\r
+ self.__tokdict[k] = self.__tokdict.get(k,list()) + [v]\r
+ sub = v[0]\r
+ elif isinstance(k,(int,slice)):\r
+ self.__toklist[k] = v\r
+ sub = v\r
+ else:\r
+ self.__tokdict[k] = self.__tokdict.get(k,list()) + [_ParseResultsWithOffset(v,0)]\r
+ sub = v\r
+ if isinstance(sub,ParseResults):\r
+ sub.__parent = wkref(self)\r
+\r
+ def __delitem__( self, i ):\r
+ if isinstance(i,(int,slice)):\r
+ mylen = len( self.__toklist )\r
+ del self.__toklist[i]\r
+\r
+ # convert int to slice\r
+ if isinstance(i, int):\r
+ if i < 0:\r
+ i += mylen\r
+ i = slice(i, i+1)\r
+ # get removed indices\r
+ removed = list(range(*i.indices(mylen)))\r
+ removed.reverse()\r
+ # fixup indices in token dictionary\r
+ for name,occurrences in self.__tokdict.items():\r
+ for j in removed:\r
+ for k, (value, position) in enumerate(occurrences):\r
+ occurrences[k] = _ParseResultsWithOffset(value, position - (position > j))\r
+ else:\r
+ del self.__tokdict[i]\r
+\r
+ def __contains__( self, k ):\r
+ return k in self.__tokdict\r
+\r
+ def __len__( self ): return len( self.__toklist )\r
+ def __bool__(self): return ( not not self.__toklist )\r
+ __nonzero__ = __bool__\r
+ def __iter__( self ): return iter( self.__toklist )\r
+ def __reversed__( self ): return iter( self.__toklist[::-1] )\r
+ def _iterkeys( self ):\r
+ if hasattr(self.__tokdict, "iterkeys"):\r
+ return self.__tokdict.iterkeys()\r
+ else:\r
+ return iter(self.__tokdict)\r
+\r
+ def _itervalues( self ):\r
+ return (self[k] for k in self._iterkeys())\r
+ \r
+ def _iteritems( self ):\r
+ return ((k, self[k]) for k in self._iterkeys())\r
+\r
+ if PY_3:\r
+ keys = _iterkeys \r
+ """Returns an iterator of all named result keys (Python 3.x only)."""\r
+\r
+ values = _itervalues\r
+ """Returns an iterator of all named result values (Python 3.x only)."""\r
+\r
+ items = _iteritems\r
+ """Returns an iterator of all named result key-value tuples (Python 3.x only)."""\r
+\r
+ else:\r
+ iterkeys = _iterkeys\r
+ """Returns an iterator of all named result keys (Python 2.x only)."""\r
+\r
+ itervalues = _itervalues\r
+ """Returns an iterator of all named result values (Python 2.x only)."""\r
+\r
+ iteritems = _iteritems\r
+ """Returns an iterator of all named result key-value tuples (Python 2.x only)."""\r
+\r
+ def keys( self ):\r
+ """Returns all named result keys (as a list in Python 2.x, as an iterator in Python 3.x)."""\r
+ return list(self.iterkeys())\r
+\r
+ def values( self ):\r
+ """Returns all named result values (as a list in Python 2.x, as an iterator in Python 3.x)."""\r
+ return list(self.itervalues())\r
+ \r
+ def items( self ):\r
+ """Returns all named result key-values (as a list of tuples in Python 2.x, as an iterator in Python 3.x)."""\r
+ return list(self.iteritems())\r
+\r
+ def haskeys( self ):\r
+ """Since keys() returns an iterator, this method is helpful in bypassing\r
+ code that looks for the existence of any defined results names."""\r
+ return bool(self.__tokdict)\r
+ \r
+ def pop( self, *args, **kwargs):\r
+ """\r
+ Removes and returns item at specified index (default=C{last}).\r
+ Supports both C{list} and C{dict} semantics for C{pop()}. If passed no\r
+ argument or an integer argument, it will use C{list} semantics\r
+ and pop tokens from the list of parsed tokens. If passed a \r
+ non-integer argument (most likely a string), it will use C{dict}\r
+ semantics and pop the corresponding value from any defined \r
+ results names. A second default return value argument is \r
+ supported, just as in C{dict.pop()}.\r
+\r
+ Example::\r
+ def remove_first(tokens):\r
+ tokens.pop(0)\r
+ print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321']\r
+ print(OneOrMore(Word(nums)).addParseAction(remove_first).parseString("0 123 321")) # -> ['123', '321']\r
+\r
+ label = Word(alphas)\r
+ patt = label("LABEL") + OneOrMore(Word(nums))\r
+ print(patt.parseString("AAB 123 321").dump())\r
+\r
+ # Use pop() in a parse action to remove named result (note that corresponding value is not\r
+ # removed from list form of results)\r
+ def remove_LABEL(tokens):\r
+ tokens.pop("LABEL")\r
+ return tokens\r
+ patt.addParseAction(remove_LABEL)\r
+ print(patt.parseString("AAB 123 321").dump())\r
+ prints::\r
+ ['AAB', '123', '321']\r
+ - LABEL: AAB\r
+\r
+ ['AAB', '123', '321']\r
+ """\r
+ if not args:\r
+ args = [-1]\r
+ for k,v in kwargs.items():\r
+ if k == 'default':\r
+ args = (args[0], v)\r
+ else:\r
+ raise TypeError("pop() got an unexpected keyword argument '%s'" % k)\r
+ if (isinstance(args[0], int) or \r
+ len(args) == 1 or \r
+ args[0] in self):\r
+ index = args[0]\r
+ ret = self[index]\r
+ del self[index]\r
+ return ret\r
+ else:\r
+ defaultvalue = args[1]\r
+ return defaultvalue\r
+\r
+ def get(self, key, defaultValue=None):\r
+ """\r
+ Returns named result matching the given key, or if there is no\r
+ such name, then returns the given C{defaultValue} or C{None} if no\r
+ C{defaultValue} is specified.\r
+\r
+ Similar to C{dict.get()}.\r
+ \r
+ Example::\r
+ integer = Word(nums)\r
+ date_str = integer("year") + '/' + integer("month") + '/' + integer("day") \r
+\r
+ result = date_str.parseString("1999/12/31")\r
+ print(result.get("year")) # -> '1999'\r
+ print(result.get("hour", "not specified")) # -> 'not specified'\r
+ print(result.get("hour")) # -> None\r
+ """\r
+ if key in self:\r
+ return self[key]\r
+ else:\r
+ return defaultValue\r
+\r
+ def insert( self, index, insStr ):\r
+ """\r
+ Inserts new element at location index in the list of parsed tokens.\r
+ \r
+ Similar to C{list.insert()}.\r
+\r
+ Example::\r
+ print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321']\r
+\r
+ # use a parse action to insert the parse location in the front of the parsed results\r
+ def insert_locn(locn, tokens):\r
+ tokens.insert(0, locn)\r
+ print(OneOrMore(Word(nums)).addParseAction(insert_locn).parseString("0 123 321")) # -> [0, '0', '123', '321']\r
+ """\r
+ self.__toklist.insert(index, insStr)\r
+ # fixup indices in token dictionary\r
+ for name,occurrences in self.__tokdict.items():\r
+ for k, (value, position) in enumerate(occurrences):\r
+ occurrences[k] = _ParseResultsWithOffset(value, position + (position > index))\r
+\r
+ def append( self, item ):\r
+ """\r
+ Add single element to end of ParseResults list of elements.\r
+\r
+ Example::\r
+ print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321']\r
+ \r
+ # use a parse action to compute the sum of the parsed integers, and add it to the end\r
+ def append_sum(tokens):\r
+ tokens.append(sum(map(int, tokens)))\r
+ print(OneOrMore(Word(nums)).addParseAction(append_sum).parseString("0 123 321")) # -> ['0', '123', '321', 444]\r
+ """\r
+ self.__toklist.append(item)\r
+\r
+ def extend( self, itemseq ):\r
+ """\r
+ Add sequence of elements to end of ParseResults list of elements.\r
+\r
+ Example::\r
+ patt = OneOrMore(Word(alphas))\r
+ \r
+ # use a parse action to append the reverse of the matched strings, to make a palindrome\r
+ def make_palindrome(tokens):\r
+ tokens.extend(reversed([t[::-1] for t in tokens]))\r
+ return ''.join(tokens)\r
+ print(patt.addParseAction(make_palindrome).parseString("lskdj sdlkjf lksd")) # -> 'lskdjsdlkjflksddsklfjkldsjdksl'\r
+ """\r
+ if isinstance(itemseq, ParseResults):\r
+ self += itemseq\r
+ else:\r
+ self.__toklist.extend(itemseq)\r
+\r
+ def clear( self ):\r
+ """\r
+ Clear all elements and results names.\r
+ """\r
+ del self.__toklist[:]\r
+ self.__tokdict.clear()\r
+\r
+ def __getattr__( self, name ):\r
+ try:\r
+ return self[name]\r
+ except KeyError:\r
+ return ""\r
+ \r
+ if name in self.__tokdict:\r
+ if name not in self.__accumNames:\r
+ return self.__tokdict[name][-1][0]\r
+ else:\r
+ return ParseResults([ v[0] for v in self.__tokdict[name] ])\r
+ else:\r
+ return ""\r
+\r
+ def __add__( self, other ):\r
+ ret = self.copy()\r
+ ret += other\r
+ return ret\r
+\r
+ def __iadd__( self, other ):\r
+ if other.__tokdict:\r
+ offset = len(self.__toklist)\r
+ addoffset = lambda a: offset if a<0 else a+offset\r
+ otheritems = other.__tokdict.items()\r
+ otherdictitems = [(k, _ParseResultsWithOffset(v[0],addoffset(v[1])) )\r
+ for (k,vlist) in otheritems for v in vlist]\r
+ for k,v in otherdictitems:\r
+ self[k] = v\r
+ if isinstance(v[0],ParseResults):\r
+ v[0].__parent = wkref(self)\r
+ \r
+ self.__toklist += other.__toklist\r
+ self.__accumNames.update( other.__accumNames )\r
+ return self\r
+\r
+ def __radd__(self, other):\r
+ if isinstance(other,int) and other == 0:\r
+ # useful for merging many ParseResults using sum() builtin\r
+ return self.copy()\r
+ else:\r
+ # this may raise a TypeError - so be it\r
+ return other + self\r
+ \r
+ def __repr__( self ):\r
+ return "(%s, %s)" % ( repr( self.__toklist ), repr( self.__tokdict ) )\r
+\r
+ def __str__( self ):\r
+ return '[' + ', '.join(_ustr(i) if isinstance(i, ParseResults) else repr(i) for i in self.__toklist) + ']'\r
+\r
+ def _asStringList( self, sep='' ):\r
+ out = []\r
+ for item in self.__toklist:\r
+ if out and sep:\r
+ out.append(sep)\r
+ if isinstance( item, ParseResults ):\r
+ out += item._asStringList()\r
+ else:\r
+ out.append( _ustr(item) )\r
+ return out\r
+\r
+ def asList( self ):\r
+ """\r
+ Returns the parse results as a nested list of matching tokens, all converted to strings.\r
+\r
+ Example::\r
+ patt = OneOrMore(Word(alphas))\r
+ result = patt.parseString("sldkj lsdkj sldkj")\r
+ # even though the result prints in string-like form, it is actually a pyparsing ParseResults\r
+ print(type(result), result) # -> <class 'pyparsing.ParseResults'> ['sldkj', 'lsdkj', 'sldkj']\r
+ \r
+ # Use asList() to create an actual list\r
+ result_list = result.asList()\r
+ print(type(result_list), result_list) # -> <class 'list'> ['sldkj', 'lsdkj', 'sldkj']\r
+ """\r
+ return [res.asList() if isinstance(res,ParseResults) else res for res in self.__toklist]\r
+\r
+ def asDict( self ):\r
+ """\r
+ Returns the named parse results as a nested dictionary.\r
+\r
+ Example::\r
+ integer = Word(nums)\r
+ date_str = integer("year") + '/' + integer("month") + '/' + integer("day")\r
+ \r
+ result = date_str.parseString('12/31/1999')\r
+ print(type(result), repr(result)) # -> <class 'pyparsing.ParseResults'> (['12', '/', '31', '/', '1999'], {'day': [('1999', 4)], 'year': [('12', 0)], 'month': [('31', 2)]})\r
+ \r
+ result_dict = result.asDict()\r
+ print(type(result_dict), repr(result_dict)) # -> <class 'dict'> {'day': '1999', 'year': '12', 'month': '31'}\r
+\r
+ # even though a ParseResults supports dict-like access, sometime you just need to have a dict\r
+ import json\r
+ print(json.dumps(result)) # -> Exception: TypeError: ... is not JSON serializable\r
+ print(json.dumps(result.asDict())) # -> {"month": "31", "day": "1999", "year": "12"}\r
+ """\r
+ if PY_3:\r
+ item_fn = self.items\r
+ else:\r
+ item_fn = self.iteritems\r
+ \r
+ def toItem(obj):\r
+ if isinstance(obj, ParseResults):\r
+ if obj.haskeys():\r
+ return obj.asDict()\r
+ else:\r
+ return [toItem(v) for v in obj]\r
+ else:\r
+ return obj\r
+ \r
+ return dict((k,toItem(v)) for k,v in item_fn())\r
+\r
+ def copy( self ):\r
+ """\r
+ Returns a new copy of a C{ParseResults} object.\r
+ """\r
+ ret = ParseResults( self.__toklist )\r
+ ret.__tokdict = self.__tokdict.copy()\r
+ ret.__parent = self.__parent\r
+ ret.__accumNames.update( self.__accumNames )\r
+ ret.__name = self.__name\r
+ return ret\r
+\r
+ def asXML( self, doctag=None, namedItemsOnly=False, indent="", formatted=True ):\r
+ """\r
+ (Deprecated) Returns the parse results as XML. Tags are created for tokens and lists that have defined results names.\r
+ """\r
+ nl = "\n"\r
+ out = []\r
+ namedItems = dict((v[1],k) for (k,vlist) in self.__tokdict.items()\r
+ for v in vlist)\r
+ nextLevelIndent = indent + " "\r
+\r
+ # collapse out indents if formatting is not desired\r
+ if not formatted:\r
+ indent = ""\r
+ nextLevelIndent = ""\r
+ nl = ""\r
+\r
+ selfTag = None\r
+ if doctag is not None:\r
+ selfTag = doctag\r
+ else:\r
+ if self.__name:\r
+ selfTag = self.__name\r
+\r
+ if not selfTag:\r
+ if namedItemsOnly:\r
+ return ""\r
+ else:\r
+ selfTag = "ITEM"\r
+\r
+ out += [ nl, indent, "<", selfTag, ">" ]\r
+\r
+ for i,res in enumerate(self.__toklist):\r
+ if isinstance(res,ParseResults):\r
+ if i in namedItems:\r
+ out += [ res.asXML(namedItems[i],\r
+ namedItemsOnly and doctag is None,\r
+ nextLevelIndent,\r
+ formatted)]\r
+ else:\r
+ out += [ res.asXML(None,\r
+ namedItemsOnly and doctag is None,\r
+ nextLevelIndent,\r
+ formatted)]\r
+ else:\r
+ # individual token, see if there is a name for it\r
+ resTag = None\r
+ if i in namedItems:\r
+ resTag = namedItems[i]\r
+ if not resTag:\r
+ if namedItemsOnly:\r
+ continue\r
+ else:\r
+ resTag = "ITEM"\r
+ xmlBodyText = _xml_escape(_ustr(res))\r
+ out += [ nl, nextLevelIndent, "<", resTag, ">",\r
+ xmlBodyText,\r
+ "</", resTag, ">" ]\r
+\r
+ out += [ nl, indent, "</", selfTag, ">" ]\r
+ return "".join(out)\r
+\r
+ def __lookup(self,sub):\r
+ for k,vlist in self.__tokdict.items():\r
+ for v,loc in vlist:\r
+ if sub is v:\r
+ return k\r
+ return None\r
+\r
+ def getName(self):\r
+ """\r
+ Returns the results name for this token expression. Useful when several \r
+ different expressions might match at a particular location.\r
+\r
+ Example::\r
+ integer = Word(nums)\r
+ ssn_expr = Regex(r"\d\d\d-\d\d-\d\d\d\d")\r
+ house_number_expr = Suppress('#') + Word(nums, alphanums)\r
+ user_data = (Group(house_number_expr)("house_number") \r
+ | Group(ssn_expr)("ssn")\r
+ | Group(integer)("age"))\r
+ user_info = OneOrMore(user_data)\r
+ \r
+ result = user_info.parseString("22 111-22-3333 #221B")\r
+ for item in result:\r
+ print(item.getName(), ':', item[0])\r
+ prints::\r
+ age : 22\r
+ ssn : 111-22-3333\r
+ house_number : 221B\r
+ """\r
+ if self.__name:\r
+ return self.__name\r
+ elif self.__parent:\r
+ par = self.__parent()\r
+ if par:\r
+ return par.__lookup(self)\r
+ else:\r
+ return None\r
+ elif (len(self) == 1 and\r
+ len(self.__tokdict) == 1 and\r
+ next(iter(self.__tokdict.values()))[0][1] in (0,-1)):\r
+ return next(iter(self.__tokdict.keys()))\r
+ else:\r
+ return None\r
+\r
+ def dump(self, indent='', depth=0, full=True):\r
+ """\r
+ Diagnostic method for listing out the contents of a C{ParseResults}.\r
+ Accepts an optional C{indent} argument so that this string can be embedded\r
+ in a nested display of other data.\r
+\r
+ Example::\r
+ integer = Word(nums)\r
+ date_str = integer("year") + '/' + integer("month") + '/' + integer("day")\r
+ \r
+ result = date_str.parseString('12/31/1999')\r
+ print(result.dump())\r
+ prints::\r
+ ['12', '/', '31', '/', '1999']\r
+ - day: 1999\r
+ - month: 31\r
+ - year: 12\r
+ """\r
+ out = []\r
+ NL = '\n'\r
+ out.append( indent+_ustr(self.asList()) )\r
+ if full:\r
+ if self.haskeys():\r
+ items = sorted((str(k), v) for k,v in self.items())\r
+ for k,v in items:\r
+ if out:\r
+ out.append(NL)\r
+ out.append( "%s%s- %s: " % (indent,(' '*depth), k) )\r
+ if isinstance(v,ParseResults):\r
+ if v:\r
+ out.append( v.dump(indent,depth+1) )\r
+ else:\r
+ out.append(_ustr(v))\r
+ else:\r
+ out.append(repr(v))\r
+ elif any(isinstance(vv,ParseResults) for vv in self):\r
+ v = self\r
+ for i,vv in enumerate(v):\r
+ if isinstance(vv,ParseResults):\r
+ out.append("\n%s%s[%d]:\n%s%s%s" % (indent,(' '*(depth)),i,indent,(' '*(depth+1)),vv.dump(indent,depth+1) ))\r
+ else:\r
+ out.append("\n%s%s[%d]:\n%s%s%s" % (indent,(' '*(depth)),i,indent,(' '*(depth+1)),_ustr(vv)))\r
+ \r
+ return "".join(out)\r
+\r
+ def pprint(self, *args, **kwargs):\r
+ """\r
+ Pretty-printer for parsed results as a list, using the C{pprint} module.\r
+ Accepts additional positional or keyword args as defined for the \r
+ C{pprint.pprint} method. (U{http://docs.python.org/3/library/pprint.html#pprint.pprint})\r
+\r
+ Example::\r
+ ident = Word(alphas, alphanums)\r
+ num = Word(nums)\r
+ func = Forward()\r
+ term = ident | num | Group('(' + func + ')')\r
+ func <<= ident + Group(Optional(delimitedList(term)))\r
+ result = func.parseString("fna a,b,(fnb c,d,200),100")\r
+ result.pprint(width=40)\r
+ prints::\r
+ ['fna',\r
+ ['a',\r
+ 'b',\r
+ ['(', 'fnb', ['c', 'd', '200'], ')'],\r
+ '100']]\r
+ """\r
+ pprint.pprint(self.asList(), *args, **kwargs)\r
+\r
+ # add support for pickle protocol\r
+ def __getstate__(self):\r
+ return ( self.__toklist,\r
+ ( self.__tokdict.copy(),\r
+ self.__parent is not None and self.__parent() or None,\r
+ self.__accumNames,\r
+ self.__name ) )\r
+\r
+ def __setstate__(self,state):\r
+ self.__toklist = state[0]\r
+ (self.__tokdict,\r
+ par,\r
+ inAccumNames,\r
+ self.__name) = state[1]\r
+ self.__accumNames = {}\r
+ self.__accumNames.update(inAccumNames)\r
+ if par is not None:\r
+ self.__parent = wkref(par)\r
+ else:\r
+ self.__parent = None\r
+\r
+ def __getnewargs__(self):\r
+ return self.__toklist, self.__name, self.__asList, self.__modal\r
+\r
+ def __dir__(self):\r
+ return (dir(type(self)) + list(self.keys()))\r
+\r
+collections.MutableMapping.register(ParseResults)\r
+\r
+def col (loc,strg):\r
+ """Returns current column within a string, counting newlines as line separators.\r
+ The first column is number 1.\r
+\r
+ Note: the default parsing behavior is to expand tabs in the input string\r
+ before starting the parsing process. See L{I{ParserElement.parseString}<ParserElement.parseString>} for more information\r
+ on parsing strings containing C{<TAB>}s, and suggested methods to maintain a\r
+ consistent view of the parsed string, the parse location, and line and column\r
+ positions within the parsed string.\r
+ """\r
+ s = strg\r
+ return 1 if 0<loc<len(s) and s[loc-1] == '\n' else loc - s.rfind("\n", 0, loc)\r
+\r
+def lineno(loc,strg):\r
+ """Returns current line number within a string, counting newlines as line separators.\r
+ The first line is number 1.\r
+\r
+ Note: the default parsing behavior is to expand tabs in the input string\r
+ before starting the parsing process. See L{I{ParserElement.parseString}<ParserElement.parseString>} for more information\r
+ on parsing strings containing C{<TAB>}s, and suggested methods to maintain a\r
+ consistent view of the parsed string, the parse location, and line and column\r
+ positions within the parsed string.\r
+ """\r
+ return strg.count("\n",0,loc) + 1\r
+\r
+def line( loc, strg ):\r
+ """Returns the line of text containing loc within a string, counting newlines as line separators.\r
+ """\r
+ lastCR = strg.rfind("\n", 0, loc)\r
+ nextCR = strg.find("\n", loc)\r
+ if nextCR >= 0:\r
+ return strg[lastCR+1:nextCR]\r
+ else:\r
+ return strg[lastCR+1:]\r
+\r
+def _defaultStartDebugAction( instring, loc, expr ):\r
+ print (("Match " + _ustr(expr) + " at loc " + _ustr(loc) + "(%d,%d)" % ( lineno(loc,instring), col(loc,instring) )))\r
+\r
+def _defaultSuccessDebugAction( instring, startloc, endloc, expr, toks ):\r
+ print ("Matched " + _ustr(expr) + " -> " + str(toks.asList()))\r
+\r
+def _defaultExceptionDebugAction( instring, loc, expr, exc ):\r
+ print ("Exception raised:" + _ustr(exc))\r
+\r
+def nullDebugAction(*args):\r
+ """'Do-nothing' debug action, to suppress debugging output during parsing."""\r
+ pass\r
+\r
+# Only works on Python 3.x - nonlocal is toxic to Python 2 installs\r
+#~ 'decorator to trim function calls to match the arity of the target'\r
+#~ def _trim_arity(func, maxargs=3):\r
+ #~ if func in singleArgBuiltins:\r
+ #~ return lambda s,l,t: func(t)\r
+ #~ limit = 0\r
+ #~ foundArity = False\r
+ #~ def wrapper(*args):\r
+ #~ nonlocal limit,foundArity\r
+ #~ while 1:\r
+ #~ try:\r
+ #~ ret = func(*args[limit:])\r
+ #~ foundArity = True\r
+ #~ return ret\r
+ #~ except TypeError:\r
+ #~ if limit == maxargs or foundArity:\r
+ #~ raise\r
+ #~ limit += 1\r
+ #~ continue\r
+ #~ return wrapper\r
+\r
+# this version is Python 2.x-3.x cross-compatible\r
+'decorator to trim function calls to match the arity of the target'\r
+def _trim_arity(func, maxargs=2):\r
+ if func in singleArgBuiltins:\r
+ return lambda s,l,t: func(t)\r
+ limit = [0]\r
+ foundArity = [False]\r
+ \r
+ # traceback return data structure changed in Py3.5 - normalize back to plain tuples\r
+ if system_version[:2] >= (3,5):\r
+ def extract_stack(limit=0):\r
+ # special handling for Python 3.5.0 - extra deep call stack by 1\r
+ offset = -3 if system_version == (3,5,0) else -2\r
+ frame_summary = traceback.extract_stack(limit=-offset+limit-1)[offset]\r
+ return [(frame_summary.filename, frame_summary.lineno)]\r
+ def extract_tb(tb, limit=0):\r
+ frames = traceback.extract_tb(tb, limit=limit)\r
+ frame_summary = frames[-1]\r
+ return [(frame_summary.filename, frame_summary.lineno)]\r
+ else:\r
+ extract_stack = traceback.extract_stack\r
+ extract_tb = traceback.extract_tb\r
+ \r
+ # synthesize what would be returned by traceback.extract_stack at the call to \r
+ # user's parse action 'func', so that we don't incur call penalty at parse time\r
+ \r
+ LINE_DIFF = 6\r
+ # IF ANY CODE CHANGES, EVEN JUST COMMENTS OR BLANK LINES, BETWEEN THE NEXT LINE AND \r
+ # THE CALL TO FUNC INSIDE WRAPPER, LINE_DIFF MUST BE MODIFIED!!!!\r
+ this_line = extract_stack(limit=2)[-1]\r
+ pa_call_line_synth = (this_line[0], this_line[1]+LINE_DIFF)\r
+\r
+ def wrapper(*args):\r
+ while 1:\r
+ try:\r
+ ret = func(*args[limit[0]:])\r
+ foundArity[0] = True\r
+ return ret\r
+ except TypeError:\r
+ # re-raise TypeErrors if they did not come from our arity testing\r
+ if foundArity[0]:\r
+ raise\r
+ else:\r
+ try:\r
+ tb = sys.exc_info()[-1]\r
+ if not extract_tb(tb, limit=2)[-1][:2] == pa_call_line_synth:\r
+ raise\r
+ finally:\r
+ del tb\r
+\r
+ if limit[0] <= maxargs:\r
+ limit[0] += 1\r
+ continue\r
+ raise\r
+\r
+ # copy func name to wrapper for sensible debug output\r
+ func_name = "<parse action>"\r
+ try:\r
+ func_name = getattr(func, '__name__', \r
+ getattr(func, '__class__').__name__)\r
+ except Exception:\r
+ func_name = str(func)\r
+ wrapper.__name__ = func_name\r
+\r
+ return wrapper\r
+\r
+class ParserElement(object):\r
+ """Abstract base level parser element class."""\r
+ DEFAULT_WHITE_CHARS = " \n\t\r"\r
+ verbose_stacktrace = False\r
+\r
+ @staticmethod\r
+ def setDefaultWhitespaceChars( chars ):\r
+ r"""\r
+ Overrides the default whitespace chars\r
+\r
+ Example::\r
+ # default whitespace chars are space, <TAB> and newline\r
+ OneOrMore(Word(alphas)).parseString("abc def\nghi jkl") # -> ['abc', 'def', 'ghi', 'jkl']\r
+ \r
+ # change to just treat newline as significant\r
+ ParserElement.setDefaultWhitespaceChars(" \t")\r
+ OneOrMore(Word(alphas)).parseString("abc def\nghi jkl") # -> ['abc', 'def']\r
+ """\r
+ ParserElement.DEFAULT_WHITE_CHARS = chars\r
+\r
+ @staticmethod\r
+ def inlineLiteralsUsing(cls):\r
+ """\r
+ Set class to be used for inclusion of string literals into a parser.\r
+ \r
+ Example::\r
+ # default literal class used is Literal\r
+ integer = Word(nums)\r
+ date_str = integer("year") + '/' + integer("month") + '/' + integer("day") \r
+\r
+ date_str.parseString("1999/12/31") # -> ['1999', '/', '12', '/', '31']\r
+\r
+\r
+ # change to Suppress\r
+ ParserElement.inlineLiteralsUsing(Suppress)\r
+ date_str = integer("year") + '/' + integer("month") + '/' + integer("day") \r
+\r
+ date_str.parseString("1999/12/31") # -> ['1999', '12', '31']\r
+ """\r
+ ParserElement._literalStringClass = cls\r
+\r
+ def __init__( self, savelist=False ):\r
+ self.parseAction = list()\r
+ self.failAction = None\r
+ #~ self.name = "<unknown>" # don't define self.name, let subclasses try/except upcall\r
+ self.strRepr = None\r
+ self.resultsName = None\r
+ self.saveAsList = savelist\r
+ self.skipWhitespace = True\r
+ self.whiteChars = ParserElement.DEFAULT_WHITE_CHARS\r
+ self.copyDefaultWhiteChars = True\r
+ self.mayReturnEmpty = False # used when checking for left-recursion\r
+ self.keepTabs = False\r
+ self.ignoreExprs = list()\r
+ self.debug = False\r
+ self.streamlined = False\r
+ self.mayIndexError = True # used to optimize exception handling for subclasses that don't advance parse index\r
+ self.errmsg = ""\r
+ self.modalResults = True # used to mark results names as modal (report only last) or cumulative (list all)\r
+ self.debugActions = ( None, None, None ) #custom debug actions\r
+ self.re = None\r
+ self.callPreparse = True # used to avoid redundant calls to preParse\r
+ self.callDuringTry = False\r
+\r
+ def copy( self ):\r
+ """\r
+ Make a copy of this C{ParserElement}. Useful for defining different parse actions\r
+ for the same parsing pattern, using copies of the original parse element.\r
+ \r
+ Example::\r
+ integer = Word(nums).setParseAction(lambda toks: int(toks[0]))\r
+ integerK = integer.copy().addParseAction(lambda toks: toks[0]*1024) + Suppress("K")\r
+ integerM = integer.copy().addParseAction(lambda toks: toks[0]*1024*1024) + Suppress("M")\r
+ \r
+ print(OneOrMore(integerK | integerM | integer).parseString("5K 100 640K 256M"))\r
+ prints::\r
+ [5120, 100, 655360, 268435456]\r
+ Equivalent form of C{expr.copy()} is just C{expr()}::\r
+ integerM = integer().addParseAction(lambda toks: toks[0]*1024*1024) + Suppress("M")\r
+ """\r
+ cpy = copy.copy( self )\r
+ cpy.parseAction = self.parseAction[:]\r
+ cpy.ignoreExprs = self.ignoreExprs[:]\r
+ if self.copyDefaultWhiteChars:\r
+ cpy.whiteChars = ParserElement.DEFAULT_WHITE_CHARS\r
+ return cpy\r
+\r
+ def setName( self, name ):\r
+ """\r
+ Define name for this expression, makes debugging and exception messages clearer.\r
+ \r
+ Example::\r
+ Word(nums).parseString("ABC") # -> Exception: Expected W:(0123...) (at char 0), (line:1, col:1)\r
+ Word(nums).setName("integer").parseString("ABC") # -> Exception: Expected integer (at char 0), (line:1, col:1)\r
+ """\r
+ self.name = name\r
+ self.errmsg = "Expected " + self.name\r
+ if hasattr(self,"exception"):\r
+ self.exception.msg = self.errmsg\r
+ return self\r
+\r
+ def setResultsName( self, name, listAllMatches=False ):\r
+ """\r
+ Define name for referencing matching tokens as a nested attribute\r
+ of the returned parse results.\r
+ NOTE: this returns a *copy* of the original C{ParserElement} object;\r
+ this is so that the client can define a basic element, such as an\r
+ integer, and reference it in multiple places with different names.\r
+\r
+ You can also set results names using the abbreviated syntax,\r
+ C{expr("name")} in place of C{expr.setResultsName("name")} - \r
+ see L{I{__call__}<__call__>}.\r
+\r
+ Example::\r
+ date_str = (integer.setResultsName("year") + '/' \r
+ + integer.setResultsName("month") + '/' \r
+ + integer.setResultsName("day"))\r
+\r
+ # equivalent form:\r
+ date_str = integer("year") + '/' + integer("month") + '/' + integer("day")\r
+ """\r
+ newself = self.copy()\r
+ if name.endswith("*"):\r
+ name = name[:-1]\r
+ listAllMatches=True\r
+ newself.resultsName = name\r
+ newself.modalResults = not listAllMatches\r
+ return newself\r
+\r
+ def setBreak(self,breakFlag = True):\r
+ """Method to invoke the Python pdb debugger when this element is\r
+ about to be parsed. Set C{breakFlag} to True to enable, False to\r
+ disable.\r
+ """\r
+ if breakFlag:\r
+ _parseMethod = self._parse\r
+ def breaker(instring, loc, doActions=True, callPreParse=True):\r
+ import pdb\r
+ pdb.set_trace()\r
+ return _parseMethod( instring, loc, doActions, callPreParse )\r
+ breaker._originalParseMethod = _parseMethod\r
+ self._parse = breaker\r
+ else:\r
+ if hasattr(self._parse,"_originalParseMethod"):\r
+ self._parse = self._parse._originalParseMethod\r
+ return self\r
+\r
+ def setParseAction( self, *fns, **kwargs ):\r
+ """\r
+ Define action to perform when successfully matching parse element definition.\r
+ Parse action fn is a callable method with 0-3 arguments, called as C{fn(s,loc,toks)},\r
+ C{fn(loc,toks)}, C{fn(toks)}, or just C{fn()}, where:\r
+ - s = the original string being parsed (see note below)\r
+ - loc = the location of the matching substring\r
+ - toks = a list of the matched tokens, packaged as a C{L{ParseResults}} object\r
+ If the functions in fns modify the tokens, they can return them as the return\r
+ value from fn, and the modified list of tokens will replace the original.\r
+ Otherwise, fn does not need to return any value.\r
+\r
+ Optional keyword arguments:\r
+ - callDuringTry = (default=C{False}) indicate if parse action should be run during lookaheads and alternate testing\r
+\r
+ Note: the default parsing behavior is to expand tabs in the input string\r
+ before starting the parsing process. See L{I{parseString}<parseString>} for more information\r
+ on parsing strings containing C{<TAB>}s, and suggested methods to maintain a\r
+ consistent view of the parsed string, the parse location, and line and column\r
+ positions within the parsed string.\r
+ \r
+ Example::\r
+ integer = Word(nums)\r
+ date_str = integer + '/' + integer + '/' + integer\r
+\r
+ date_str.parseString("1999/12/31") # -> ['1999', '/', '12', '/', '31']\r
+\r
+ # use parse action to convert to ints at parse time\r
+ integer = Word(nums).setParseAction(lambda toks: int(toks[0]))\r
+ date_str = integer + '/' + integer + '/' + integer\r
+\r
+ # note that integer fields are now ints, not strings\r
+ date_str.parseString("1999/12/31") # -> [1999, '/', 12, '/', 31]\r
+ """\r
+ self.parseAction = list(map(_trim_arity, list(fns)))\r
+ self.callDuringTry = kwargs.get("callDuringTry", False)\r
+ return self\r
+\r
+ def addParseAction( self, *fns, **kwargs ):\r
+ """\r
+ Add parse action to expression's list of parse actions. See L{I{setParseAction}<setParseAction>}.\r
+ \r
+ See examples in L{I{copy}<copy>}.\r
+ """\r
+ self.parseAction += list(map(_trim_arity, list(fns)))\r
+ self.callDuringTry = self.callDuringTry or kwargs.get("callDuringTry", False)\r
+ return self\r
+\r
+ def addCondition(self, *fns, **kwargs):\r
+ """Add a boolean predicate function to expression's list of parse actions. See \r
+ L{I{setParseAction}<setParseAction>} for function call signatures. Unlike C{setParseAction}, \r
+ functions passed to C{addCondition} need to return boolean success/fail of the condition.\r
+\r
+ Optional keyword arguments:\r
+ - message = define a custom message to be used in the raised exception\r
+ - fatal = if True, will raise ParseFatalException to stop parsing immediately; otherwise will raise ParseException\r
+ \r
+ Example::\r
+ integer = Word(nums).setParseAction(lambda toks: int(toks[0]))\r
+ year_int = integer.copy()\r
+ year_int.addCondition(lambda toks: toks[0] >= 2000, message="Only support years 2000 and later")\r
+ date_str = year_int + '/' + integer + '/' + integer\r
+\r
+ result = date_str.parseString("1999/12/31") # -> Exception: Only support years 2000 and later (at char 0), (line:1, col:1)\r
+ """\r
+ msg = kwargs.get("message", "failed user-defined condition")\r
+ exc_type = ParseFatalException if kwargs.get("fatal", False) else ParseException\r
+ for fn in fns:\r
+ def pa(s,l,t):\r
+ if not bool(_trim_arity(fn)(s,l,t)):\r
+ raise exc_type(s,l,msg)\r
+ self.parseAction.append(pa)\r
+ self.callDuringTry = self.callDuringTry or kwargs.get("callDuringTry", False)\r
+ return self\r
+\r
+ def setFailAction( self, fn ):\r
+ """Define action to perform if parsing fails at this expression.\r
+ Fail acton fn is a callable function that takes the arguments\r
+ C{fn(s,loc,expr,err)} where:\r
+ - s = string being parsed\r
+ - loc = location where expression match was attempted and failed\r
+ - expr = the parse expression that failed\r
+ - err = the exception thrown\r
+ The function returns no value. It may throw C{L{ParseFatalException}}\r
+ if it is desired to stop parsing immediately."""\r
+ self.failAction = fn\r
+ return self\r
+\r
+ def _skipIgnorables( self, instring, loc ):\r
+ exprsFound = True\r
+ while exprsFound:\r
+ exprsFound = False\r
+ for e in self.ignoreExprs:\r
+ try:\r
+ while 1:\r
+ loc,dummy = e._parse( instring, loc )\r
+ exprsFound = True\r
+ except ParseException:\r
+ pass\r
+ return loc\r
+\r
+ def preParse( self, instring, loc ):\r
+ if self.ignoreExprs:\r
+ loc = self._skipIgnorables( instring, loc )\r
+\r
+ if self.skipWhitespace:\r
+ wt = self.whiteChars\r
+ instrlen = len(instring)\r
+ while loc < instrlen and instring[loc] in wt:\r
+ loc += 1\r
+\r
+ return loc\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ return loc, []\r
+\r
+ def postParse( self, instring, loc, tokenlist ):\r
+ return tokenlist\r
+\r
+ #~ @profile\r
+ def _parseNoCache( self, instring, loc, doActions=True, callPreParse=True ):\r
+ debugging = ( self.debug ) #and doActions )\r
+\r
+ if debugging or self.failAction:\r
+ #~ print ("Match",self,"at loc",loc,"(%d,%d)" % ( lineno(loc,instring), col(loc,instring) ))\r
+ if (self.debugActions[0] ):\r
+ self.debugActions[0]( instring, loc, self )\r
+ if callPreParse and self.callPreparse:\r
+ preloc = self.preParse( instring, loc )\r
+ else:\r
+ preloc = loc\r
+ tokensStart = preloc\r
+ try:\r
+ try:\r
+ loc,tokens = self.parseImpl( instring, preloc, doActions )\r
+ except IndexError:\r
+ raise ParseException( instring, len(instring), self.errmsg, self )\r
+ except ParseBaseException as err:\r
+ #~ print ("Exception raised:", err)\r
+ if self.debugActions[2]:\r
+ self.debugActions[2]( instring, tokensStart, self, err )\r
+ if self.failAction:\r
+ self.failAction( instring, tokensStart, self, err )\r
+ raise\r
+ else:\r
+ if callPreParse and self.callPreparse:\r
+ preloc = self.preParse( instring, loc )\r
+ else:\r
+ preloc = loc\r
+ tokensStart = preloc\r
+ if self.mayIndexError or loc >= len(instring):\r
+ try:\r
+ loc,tokens = self.parseImpl( instring, preloc, doActions )\r
+ except IndexError:\r
+ raise ParseException( instring, len(instring), self.errmsg, self )\r
+ else:\r
+ loc,tokens = self.parseImpl( instring, preloc, doActions )\r
+\r
+ tokens = self.postParse( instring, loc, tokens )\r
+\r
+ retTokens = ParseResults( tokens, self.resultsName, asList=self.saveAsList, modal=self.modalResults )\r
+ if self.parseAction and (doActions or self.callDuringTry):\r
+ if debugging:\r
+ try:\r
+ for fn in self.parseAction:\r
+ tokens = fn( instring, tokensStart, retTokens )\r
+ if tokens is not None:\r
+ retTokens = ParseResults( tokens,\r
+ self.resultsName,\r
+ asList=self.saveAsList and isinstance(tokens,(ParseResults,list)),\r
+ modal=self.modalResults )\r
+ except ParseBaseException as err:\r
+ #~ print "Exception raised in user parse action:", err\r
+ if (self.debugActions[2] ):\r
+ self.debugActions[2]( instring, tokensStart, self, err )\r
+ raise\r
+ else:\r
+ for fn in self.parseAction:\r
+ tokens = fn( instring, tokensStart, retTokens )\r
+ if tokens is not None:\r
+ retTokens = ParseResults( tokens,\r
+ self.resultsName,\r
+ asList=self.saveAsList and isinstance(tokens,(ParseResults,list)),\r
+ modal=self.modalResults )\r
+\r
+ if debugging:\r
+ #~ print ("Matched",self,"->",retTokens.asList())\r
+ if (self.debugActions[1] ):\r
+ self.debugActions[1]( instring, tokensStart, loc, self, retTokens )\r
+\r
+ return loc, retTokens\r
+\r
+ def tryParse( self, instring, loc ):\r
+ try:\r
+ return self._parse( instring, loc, doActions=False )[0]\r
+ except ParseFatalException:\r
+ raise ParseException( instring, loc, self.errmsg, self)\r
+ \r
+ def canParseNext(self, instring, loc):\r
+ try:\r
+ self.tryParse(instring, loc)\r
+ except (ParseException, IndexError):\r
+ return False\r
+ else:\r
+ return True\r
+\r
+ class _UnboundedCache(object):\r
+ def __init__(self):\r
+ cache = {}\r
+ self.not_in_cache = not_in_cache = object()\r
+\r
+ def get(self, key):\r
+ return cache.get(key, not_in_cache)\r
+\r
+ def set(self, key, value):\r
+ cache[key] = value\r
+\r
+ def clear(self):\r
+ cache.clear()\r
+\r
+ self.get = types.MethodType(get, self)\r
+ self.set = types.MethodType(set, self)\r
+ self.clear = types.MethodType(clear, self)\r
+\r
+ if _OrderedDict is not None:\r
+ class _FifoCache(object):\r
+ def __init__(self, size):\r
+ self.not_in_cache = not_in_cache = object()\r
+\r
+ cache = _OrderedDict()\r
+\r
+ def get(self, key):\r
+ return cache.get(key, not_in_cache)\r
+\r
+ def set(self, key, value):\r
+ cache[key] = value\r
+ if len(cache) > size:\r
+ cache.popitem(False)\r
+\r
+ def clear(self):\r
+ cache.clear()\r
+\r
+ self.get = types.MethodType(get, self)\r
+ self.set = types.MethodType(set, self)\r
+ self.clear = types.MethodType(clear, self)\r
+\r
+ else:\r
+ class _FifoCache(object):\r
+ def __init__(self, size):\r
+ self.not_in_cache = not_in_cache = object()\r
+\r
+ cache = {}\r
+ key_fifo = collections.deque([], size)\r
+\r
+ def get(self, key):\r
+ return cache.get(key, not_in_cache)\r
+\r
+ def set(self, key, value):\r
+ cache[key] = value\r
+ if len(cache) > size:\r
+ cache.pop(key_fifo.popleft(), None)\r
+ key_fifo.append(key)\r
+\r
+ def clear(self):\r
+ cache.clear()\r
+ key_fifo.clear()\r
+\r
+ self.get = types.MethodType(get, self)\r
+ self.set = types.MethodType(set, self)\r
+ self.clear = types.MethodType(clear, self)\r
+\r
+ # argument cache for optimizing repeated calls when backtracking through recursive expressions\r
+ packrat_cache = {} # this is set later by enabledPackrat(); this is here so that resetCache() doesn't fail\r
+ packrat_cache_lock = RLock()\r
+ packrat_cache_stats = [0, 0]\r
+\r
+ # this method gets repeatedly called during backtracking with the same arguments -\r
+ # we can cache these arguments and save ourselves the trouble of re-parsing the contained expression\r
+ def _parseCache( self, instring, loc, doActions=True, callPreParse=True ):\r
+ HIT, MISS = 0, 1\r
+ lookup = (self, instring, loc, callPreParse, doActions)\r
+ with ParserElement.packrat_cache_lock:\r
+ cache = ParserElement.packrat_cache\r
+ value = cache.get(lookup)\r
+ if value is cache.not_in_cache:\r
+ ParserElement.packrat_cache_stats[MISS] += 1\r
+ try:\r
+ value = self._parseNoCache(instring, loc, doActions, callPreParse)\r
+ except ParseBaseException as pe:\r
+ # cache a copy of the exception, without the traceback\r
+ cache.set(lookup, pe.__class__(*pe.args))\r
+ raise\r
+ else:\r
+ cache.set(lookup, (value[0], value[1].copy()))\r
+ return value\r
+ else:\r
+ ParserElement.packrat_cache_stats[HIT] += 1\r
+ if isinstance(value, Exception):\r
+ raise value\r
+ return (value[0], value[1].copy())\r
+\r
+ _parse = _parseNoCache\r
+\r
+ @staticmethod\r
+ def resetCache():\r
+ ParserElement.packrat_cache.clear()\r
+ ParserElement.packrat_cache_stats[:] = [0] * len(ParserElement.packrat_cache_stats)\r
+\r
+ _packratEnabled = False\r
+ @staticmethod\r
+ def enablePackrat(cache_size_limit=128):\r
+ """Enables "packrat" parsing, which adds memoizing to the parsing logic.\r
+ Repeated parse attempts at the same string location (which happens\r
+ often in many complex grammars) can immediately return a cached value,\r
+ instead of re-executing parsing/validating code. Memoizing is done of\r
+ both valid results and parsing exceptions.\r
+ \r
+ Parameters:\r
+ - cache_size_limit - (default=C{128}) - if an integer value is provided\r
+ will limit the size of the packrat cache; if None is passed, then\r
+ the cache size will be unbounded; if 0 is passed, the cache will\r
+ be effectively disabled.\r
+ \r
+ This speedup may break existing programs that use parse actions that\r
+ have side-effects. For this reason, packrat parsing is disabled when\r
+ you first import pyparsing. To activate the packrat feature, your\r
+ program must call the class method C{ParserElement.enablePackrat()}. If\r
+ your program uses C{psyco} to "compile as you go", you must call\r
+ C{enablePackrat} before calling C{psyco.full()}. If you do not do this,\r
+ Python will crash. For best results, call C{enablePackrat()} immediately\r
+ after importing pyparsing.\r
+ \r
+ Example::\r
+ import pyparsing\r
+ pyparsing.ParserElement.enablePackrat()\r
+ """\r
+ if not ParserElement._packratEnabled:\r
+ ParserElement._packratEnabled = True\r
+ if cache_size_limit is None:\r
+ ParserElement.packrat_cache = ParserElement._UnboundedCache()\r
+ else:\r
+ ParserElement.packrat_cache = ParserElement._FifoCache(cache_size_limit)\r
+ ParserElement._parse = ParserElement._parseCache\r
+\r
+ def parseString( self, instring, parseAll=False ):\r
+ """\r
+ Execute the parse expression with the given string.\r
+ This is the main interface to the client code, once the complete\r
+ expression has been built.\r
+\r
+ If you want the grammar to require that the entire input string be\r
+ successfully parsed, then set C{parseAll} to True (equivalent to ending\r
+ the grammar with C{L{StringEnd()}}).\r
+\r
+ Note: C{parseString} implicitly calls C{expandtabs()} on the input string,\r
+ in order to report proper column numbers in parse actions.\r
+ If the input string contains tabs and\r
+ the grammar uses parse actions that use the C{loc} argument to index into the\r
+ string being parsed, you can ensure you have a consistent view of the input\r
+ string by:\r
+ - calling C{parseWithTabs} on your grammar before calling C{parseString}\r
+ (see L{I{parseWithTabs}<parseWithTabs>})\r
+ - define your parse action using the full C{(s,loc,toks)} signature, and\r
+ reference the input string using the parse action's C{s} argument\r
+ - explictly expand the tabs in your input string before calling\r
+ C{parseString}\r
+ \r
+ Example::\r
+ Word('a').parseString('aaaaabaaa') # -> ['aaaaa']\r
+ Word('a').parseString('aaaaabaaa', parseAll=True) # -> Exception: Expected end of text\r
+ """\r
+ ParserElement.resetCache()\r
+ if not self.streamlined:\r
+ self.streamline()\r
+ #~ self.saveAsList = True\r
+ for e in self.ignoreExprs:\r
+ e.streamline()\r
+ if not self.keepTabs:\r
+ instring = instring.expandtabs()\r
+ try:\r
+ loc, tokens = self._parse( instring, 0 )\r
+ if parseAll:\r
+ loc = self.preParse( instring, loc )\r
+ se = Empty() + StringEnd()\r
+ se._parse( instring, loc )\r
+ except ParseBaseException as exc:\r
+ if ParserElement.verbose_stacktrace:\r
+ raise\r
+ else:\r
+ # catch and re-raise exception from here, clears out pyparsing internal stack trace\r
+ raise exc\r
+ else:\r
+ return tokens\r
+\r
+ def scanString( self, instring, maxMatches=_MAX_INT, overlap=False ):\r
+ """\r
+ Scan the input string for expression matches. Each match will return the\r
+ matching tokens, start location, and end location. May be called with optional\r
+ C{maxMatches} argument, to clip scanning after 'n' matches are found. If\r
+ C{overlap} is specified, then overlapping matches will be reported.\r
+\r
+ Note that the start and end locations are reported relative to the string\r
+ being parsed. See L{I{parseString}<parseString>} for more information on parsing\r
+ strings with embedded tabs.\r
+\r
+ Example::\r
+ source = "sldjf123lsdjjkf345sldkjf879lkjsfd987"\r
+ print(source)\r
+ for tokens,start,end in Word(alphas).scanString(source):\r
+ print(' '*start + '^'*(end-start))\r
+ print(' '*start + tokens[0])\r
+ \r
+ prints::\r
+ \r
+ sldjf123lsdjjkf345sldkjf879lkjsfd987\r
+ ^^^^^\r
+ sldjf\r
+ ^^^^^^^\r
+ lsdjjkf\r
+ ^^^^^^\r
+ sldkjf\r
+ ^^^^^^\r
+ lkjsfd\r
+ """\r
+ if not self.streamlined:\r
+ self.streamline()\r
+ for e in self.ignoreExprs:\r
+ e.streamline()\r
+\r
+ if not self.keepTabs:\r
+ instring = _ustr(instring).expandtabs()\r
+ instrlen = len(instring)\r
+ loc = 0\r
+ preparseFn = self.preParse\r
+ parseFn = self._parse\r
+ ParserElement.resetCache()\r
+ matches = 0\r
+ try:\r
+ while loc <= instrlen and matches < maxMatches:\r
+ try:\r
+ preloc = preparseFn( instring, loc )\r
+ nextLoc,tokens = parseFn( instring, preloc, callPreParse=False )\r
+ except ParseException:\r
+ loc = preloc+1\r
+ else:\r
+ if nextLoc > loc:\r
+ matches += 1\r
+ yield tokens, preloc, nextLoc\r
+ if overlap:\r
+ nextloc = preparseFn( instring, loc )\r
+ if nextloc > loc:\r
+ loc = nextLoc\r
+ else:\r
+ loc += 1\r
+ else:\r
+ loc = nextLoc\r
+ else:\r
+ loc = preloc+1\r
+ except ParseBaseException as exc:\r
+ if ParserElement.verbose_stacktrace:\r
+ raise\r
+ else:\r
+ # catch and re-raise exception from here, clears out pyparsing internal stack trace\r
+ raise exc\r
+\r
+ def transformString( self, instring ):\r
+ """\r
+ Extension to C{L{scanString}}, to modify matching text with modified tokens that may\r
+ be returned from a parse action. To use C{transformString}, define a grammar and\r
+ attach a parse action to it that modifies the returned token list.\r
+ Invoking C{transformString()} on a target string will then scan for matches,\r
+ and replace the matched text patterns according to the logic in the parse\r
+ action. C{transformString()} returns the resulting transformed string.\r
+ \r
+ Example::\r
+ wd = Word(alphas)\r
+ wd.setParseAction(lambda toks: toks[0].title())\r
+ \r
+ print(wd.transformString("now is the winter of our discontent made glorious summer by this sun of york."))\r
+ Prints::\r
+ Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York.\r
+ """\r
+ out = []\r
+ lastE = 0\r
+ # force preservation of <TAB>s, to minimize unwanted transformation of string, and to\r
+ # keep string locs straight between transformString and scanString\r
+ self.keepTabs = True\r
+ try:\r
+ for t,s,e in self.scanString( instring ):\r
+ out.append( instring[lastE:s] )\r
+ if t:\r
+ if isinstance(t,ParseResults):\r
+ out += t.asList()\r
+ elif isinstance(t,list):\r
+ out += t\r
+ else:\r
+ out.append(t)\r
+ lastE = e\r
+ out.append(instring[lastE:])\r
+ out = [o for o in out if o]\r
+ return "".join(map(_ustr,_flatten(out)))\r
+ except ParseBaseException as exc:\r
+ if ParserElement.verbose_stacktrace:\r
+ raise\r
+ else:\r
+ # catch and re-raise exception from here, clears out pyparsing internal stack trace\r
+ raise exc\r
+\r
+ def searchString( self, instring, maxMatches=_MAX_INT ):\r
+ """\r
+ Another extension to C{L{scanString}}, simplifying the access to the tokens found\r
+ to match the given parse expression. May be called with optional\r
+ C{maxMatches} argument, to clip searching after 'n' matches are found.\r
+ \r
+ Example::\r
+ # a capitalized word starts with an uppercase letter, followed by zero or more lowercase letters\r
+ cap_word = Word(alphas.upper(), alphas.lower())\r
+ \r
+ print(cap_word.searchString("More than Iron, more than Lead, more than Gold I need Electricity"))\r
+ prints::\r
+ ['More', 'Iron', 'Lead', 'Gold', 'I']\r
+ """\r
+ try:\r
+ return ParseResults([ t for t,s,e in self.scanString( instring, maxMatches ) ])\r
+ except ParseBaseException as exc:\r
+ if ParserElement.verbose_stacktrace:\r
+ raise\r
+ else:\r
+ # catch and re-raise exception from here, clears out pyparsing internal stack trace\r
+ raise exc\r
+\r
+ def split(self, instring, maxsplit=_MAX_INT, includeSeparators=False):\r
+ """\r
+ Generator method to split a string using the given expression as a separator.\r
+ May be called with optional C{maxsplit} argument, to limit the number of splits;\r
+ and the optional C{includeSeparators} argument (default=C{False}), if the separating\r
+ matching text should be included in the split results.\r
+ \r
+ Example:: \r
+ punc = oneOf(list(".,;:/-!?"))\r
+ print(list(punc.split("This, this?, this sentence, is badly punctuated!")))\r
+ prints::\r
+ ['This', ' this', '', ' this sentence', ' is badly punctuated', '']\r
+ """\r
+ splits = 0\r
+ last = 0\r
+ for t,s,e in self.scanString(instring, maxMatches=maxsplit):\r
+ yield instring[last:s]\r
+ if includeSeparators:\r
+ yield t[0]\r
+ last = e\r
+ yield instring[last:]\r
+\r
+ def __add__(self, other ):\r
+ """\r
+ Implementation of + operator - returns C{L{And}}. Adding strings to a ParserElement\r
+ converts them to L{Literal}s by default.\r
+ \r
+ Example::\r
+ greet = Word(alphas) + "," + Word(alphas) + "!"\r
+ hello = "Hello, World!"\r
+ print (hello, "->", greet.parseString(hello))\r
+ Prints::\r
+ Hello, World! -> ['Hello', ',', 'World', '!']\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return And( [ self, other ] )\r
+\r
+ def __radd__(self, other ):\r
+ """\r
+ Implementation of + operator when left operand is not a C{L{ParserElement}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return other + self\r
+\r
+ def __sub__(self, other):\r
+ """\r
+ Implementation of - operator, returns C{L{And}} with error stop\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return And( [ self, And._ErrorStop(), other ] )\r
+\r
+ def __rsub__(self, other ):\r
+ """\r
+ Implementation of - operator when left operand is not a C{L{ParserElement}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return other - self\r
+\r
+ def __mul__(self,other):\r
+ """\r
+ Implementation of * operator, allows use of C{expr * 3} in place of\r
+ C{expr + expr + expr}. Expressions may also me multiplied by a 2-integer\r
+ tuple, similar to C{{min,max}} multipliers in regular expressions. Tuples\r
+ may also include C{None} as in:\r
+ - C{expr*(n,None)} or C{expr*(n,)} is equivalent\r
+ to C{expr*n + L{ZeroOrMore}(expr)}\r
+ (read as "at least n instances of C{expr}")\r
+ - C{expr*(None,n)} is equivalent to C{expr*(0,n)}\r
+ (read as "0 to n instances of C{expr}")\r
+ - C{expr*(None,None)} is equivalent to C{L{ZeroOrMore}(expr)}\r
+ - C{expr*(1,None)} is equivalent to C{L{OneOrMore}(expr)}\r
+\r
+ Note that C{expr*(None,n)} does not raise an exception if\r
+ more than n exprs exist in the input stream; that is,\r
+ C{expr*(None,n)} does not enforce a maximum number of expr\r
+ occurrences. If this behavior is desired, then write\r
+ C{expr*(None,n) + ~expr}\r
+ """\r
+ if isinstance(other,int):\r
+ minElements, optElements = other,0\r
+ elif isinstance(other,tuple):\r
+ other = (other + (None, None))[:2]\r
+ if other[0] is None:\r
+ other = (0, other[1])\r
+ if isinstance(other[0],int) and other[1] is None:\r
+ if other[0] == 0:\r
+ return ZeroOrMore(self)\r
+ if other[0] == 1:\r
+ return OneOrMore(self)\r
+ else:\r
+ return self*other[0] + ZeroOrMore(self)\r
+ elif isinstance(other[0],int) and isinstance(other[1],int):\r
+ minElements, optElements = other\r
+ optElements -= minElements\r
+ else:\r
+ raise TypeError("cannot multiply 'ParserElement' and ('%s','%s') objects", type(other[0]),type(other[1]))\r
+ else:\r
+ raise TypeError("cannot multiply 'ParserElement' and '%s' objects", type(other))\r
+\r
+ if minElements < 0:\r
+ raise ValueError("cannot multiply ParserElement by negative value")\r
+ if optElements < 0:\r
+ raise ValueError("second tuple value must be greater or equal to first tuple value")\r
+ if minElements == optElements == 0:\r
+ raise ValueError("cannot multiply ParserElement by 0 or (0,0)")\r
+\r
+ if (optElements):\r
+ def makeOptionalList(n):\r
+ if n>1:\r
+ return Optional(self + makeOptionalList(n-1))\r
+ else:\r
+ return Optional(self)\r
+ if minElements:\r
+ if minElements == 1:\r
+ ret = self + makeOptionalList(optElements)\r
+ else:\r
+ ret = And([self]*minElements) + makeOptionalList(optElements)\r
+ else:\r
+ ret = makeOptionalList(optElements)\r
+ else:\r
+ if minElements == 1:\r
+ ret = self\r
+ else:\r
+ ret = And([self]*minElements)\r
+ return ret\r
+\r
+ def __rmul__(self, other):\r
+ return self.__mul__(other)\r
+\r
+ def __or__(self, other ):\r
+ """\r
+ Implementation of | operator - returns C{L{MatchFirst}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return MatchFirst( [ self, other ] )\r
+\r
+ def __ror__(self, other ):\r
+ """\r
+ Implementation of | operator when left operand is not a C{L{ParserElement}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return other | self\r
+\r
+ def __xor__(self, other ):\r
+ """\r
+ Implementation of ^ operator - returns C{L{Or}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return Or( [ self, other ] )\r
+\r
+ def __rxor__(self, other ):\r
+ """\r
+ Implementation of ^ operator when left operand is not a C{L{ParserElement}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return other ^ self\r
+\r
+ def __and__(self, other ):\r
+ """\r
+ Implementation of & operator - returns C{L{Each}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return Each( [ self, other ] )\r
+\r
+ def __rand__(self, other ):\r
+ """\r
+ Implementation of & operator when left operand is not a C{L{ParserElement}}\r
+ """\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ if not isinstance( other, ParserElement ):\r
+ warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),\r
+ SyntaxWarning, stacklevel=2)\r
+ return None\r
+ return other & self\r
+\r
+ def __invert__( self ):\r
+ """\r
+ Implementation of ~ operator - returns C{L{NotAny}}\r
+ """\r
+ return NotAny( self )\r
+\r
+ def __call__(self, name=None):\r
+ """\r
+ Shortcut for C{L{setResultsName}}, with C{listAllMatches=False}.\r
+ \r
+ If C{name} is given with a trailing C{'*'} character, then C{listAllMatches} will be\r
+ passed as C{True}.\r
+ \r
+ If C{name} is omitted, same as calling C{L{copy}}.\r
+\r
+ Example::\r
+ # these are equivalent\r
+ userdata = Word(alphas).setResultsName("name") + Word(nums+"-").setResultsName("socsecno")\r
+ userdata = Word(alphas)("name") + Word(nums+"-")("socsecno") \r
+ """\r
+ if name is not None:\r
+ return self.setResultsName(name)\r
+ else:\r
+ return self.copy()\r
+\r
+ def suppress( self ):\r
+ """\r
+ Suppresses the output of this C{ParserElement}; useful to keep punctuation from\r
+ cluttering up returned output.\r
+ """\r
+ return Suppress( self )\r
+\r
+ def leaveWhitespace( self ):\r
+ """\r
+ Disables the skipping of whitespace before matching the characters in the\r
+ C{ParserElement}'s defined pattern. This is normally only used internally by\r
+ the pyparsing module, but may be needed in some whitespace-sensitive grammars.\r
+ """\r
+ self.skipWhitespace = False\r
+ return self\r
+\r
+ def setWhitespaceChars( self, chars ):\r
+ """\r
+ Overrides the default whitespace chars\r
+ """\r
+ self.skipWhitespace = True\r
+ self.whiteChars = chars\r
+ self.copyDefaultWhiteChars = False\r
+ return self\r
+\r
+ def parseWithTabs( self ):\r
+ """\r
+ Overrides default behavior to expand C{<TAB>}s to spaces before parsing the input string.\r
+ Must be called before C{parseString} when the input grammar contains elements that\r
+ match C{<TAB>} characters.\r
+ """\r
+ self.keepTabs = True\r
+ return self\r
+\r
+ def ignore( self, other ):\r
+ """\r
+ Define expression to be ignored (e.g., comments) while doing pattern\r
+ matching; may be called repeatedly, to define multiple comment or other\r
+ ignorable patterns.\r
+ \r
+ Example::\r
+ patt = OneOrMore(Word(alphas))\r
+ patt.parseString('ablaj /* comment */ lskjd') # -> ['ablaj']\r
+ \r
+ patt.ignore(cStyleComment)\r
+ patt.parseString('ablaj /* comment */ lskjd') # -> ['ablaj', 'lskjd']\r
+ """\r
+ if isinstance(other, basestring):\r
+ other = Suppress(other)\r
+\r
+ if isinstance( other, Suppress ):\r
+ if other not in self.ignoreExprs:\r
+ self.ignoreExprs.append(other)\r
+ else:\r
+ self.ignoreExprs.append( Suppress( other.copy() ) )\r
+ return self\r
+\r
+ def setDebugActions( self, startAction, successAction, exceptionAction ):\r
+ """\r
+ Enable display of debugging messages while doing pattern matching.\r
+ """\r
+ self.debugActions = (startAction or _defaultStartDebugAction,\r
+ successAction or _defaultSuccessDebugAction,\r
+ exceptionAction or _defaultExceptionDebugAction)\r
+ self.debug = True\r
+ return self\r
+\r
+ def setDebug( self, flag=True ):\r
+ """\r
+ Enable display of debugging messages while doing pattern matching.\r
+ Set C{flag} to True to enable, False to disable.\r
+\r
+ Example::\r
+ wd = Word(alphas).setName("alphaword")\r
+ integer = Word(nums).setName("numword")\r
+ term = wd | integer\r
+ \r
+ # turn on debugging for wd\r
+ wd.setDebug()\r
+\r
+ OneOrMore(term).parseString("abc 123 xyz 890")\r
+ \r
+ prints::\r
+ Match alphaword at loc 0(1,1)\r
+ Matched alphaword -> ['abc']\r
+ Match alphaword at loc 3(1,4)\r
+ Exception raised:Expected alphaword (at char 4), (line:1, col:5)\r
+ Match alphaword at loc 7(1,8)\r
+ Matched alphaword -> ['xyz']\r
+ Match alphaword at loc 11(1,12)\r
+ Exception raised:Expected alphaword (at char 12), (line:1, col:13)\r
+ Match alphaword at loc 15(1,16)\r
+ Exception raised:Expected alphaword (at char 15), (line:1, col:16)\r
+\r
+ The output shown is that produced by the default debug actions - custom debug actions can be\r
+ specified using L{setDebugActions}. Prior to attempting\r
+ to match the C{wd} expression, the debugging message C{"Match <exprname> at loc <n>(<line>,<col>)"}\r
+ is shown. Then if the parse succeeds, a C{"Matched"} message is shown, or an C{"Exception raised"}\r
+ message is shown. Also note the use of L{setName} to assign a human-readable name to the expression,\r
+ which makes debugging and exception messages easier to understand - for instance, the default\r
+ name created for the C{Word} expression without calling C{setName} is C{"W:(ABCD...)"}.\r
+ """\r
+ if flag:\r
+ self.setDebugActions( _defaultStartDebugAction, _defaultSuccessDebugAction, _defaultExceptionDebugAction )\r
+ else:\r
+ self.debug = False\r
+ return self\r
+\r
+ def __str__( self ):\r
+ return self.name\r
+\r
+ def __repr__( self ):\r
+ return _ustr(self)\r
+\r
+ def streamline( self ):\r
+ self.streamlined = True\r
+ self.strRepr = None\r
+ return self\r
+\r
+ def checkRecursion( self, parseElementList ):\r
+ pass\r
+\r
+ def validate( self, validateTrace=[] ):\r
+ """\r
+ Check defined expressions for valid structure, check for infinite recursive definitions.\r
+ """\r
+ self.checkRecursion( [] )\r
+\r
+ def parseFile( self, file_or_filename, parseAll=False ):\r
+ """\r
+ Execute the parse expression on the given file or filename.\r
+ If a filename is specified (instead of a file object),\r
+ the entire file is opened, read, and closed before parsing.\r
+ """\r
+ try:\r
+ file_contents = file_or_filename.read()\r
+ except AttributeError:\r
+ with open(file_or_filename, "r") as f:\r
+ file_contents = f.read()\r
+ try:\r
+ return self.parseString(file_contents, parseAll)\r
+ except ParseBaseException as exc:\r
+ if ParserElement.verbose_stacktrace:\r
+ raise\r
+ else:\r
+ # catch and re-raise exception from here, clears out pyparsing internal stack trace\r
+ raise exc\r
+\r
+ def __eq__(self,other):\r
+ if isinstance(other, ParserElement):\r
+ return self is other or vars(self) == vars(other)\r
+ elif isinstance(other, basestring):\r
+ return self.matches(other)\r
+ else:\r
+ return super(ParserElement,self)==other\r
+\r
+ def __ne__(self,other):\r
+ return not (self == other)\r
+\r
+ def __hash__(self):\r
+ return hash(id(self))\r
+\r
+ def __req__(self,other):\r
+ return self == other\r
+\r
+ def __rne__(self,other):\r
+ return not (self == other)\r
+\r
+ def matches(self, testString, parseAll=True):\r
+ """\r
+ Method for quick testing of a parser against a test string. Good for simple \r
+ inline microtests of sub expressions while building up larger parser.\r
+ \r
+ Parameters:\r
+ - testString - to test against this expression for a match\r
+ - parseAll - (default=C{True}) - flag to pass to C{L{parseString}} when running tests\r
+ \r
+ Example::\r
+ expr = Word(nums)\r
+ assert expr.matches("100")\r
+ """\r
+ try:\r
+ self.parseString(_ustr(testString), parseAll=parseAll)\r
+ return True\r
+ except ParseBaseException:\r
+ return False\r
+ \r
+ def runTests(self, tests, parseAll=True, comment='#', fullDump=True, printResults=True, failureTests=False):\r
+ """\r
+ Execute the parse expression on a series of test strings, showing each\r
+ test, the parsed results or where the parse failed. Quick and easy way to\r
+ run a parse expression against a list of sample strings.\r
+ \r
+ Parameters:\r
+ - tests - a list of separate test strings, or a multiline string of test strings\r
+ - parseAll - (default=C{True}) - flag to pass to C{L{parseString}} when running tests \r
+ - comment - (default=C{'#'}) - expression for indicating embedded comments in the test \r
+ string; pass None to disable comment filtering\r
+ - fullDump - (default=C{True}) - dump results as list followed by results names in nested outline;\r
+ if False, only dump nested list\r
+ - printResults - (default=C{True}) prints test output to stdout\r
+ - failureTests - (default=C{False}) indicates if these tests are expected to fail parsing\r
+\r
+ Returns: a (success, results) tuple, where success indicates that all tests succeeded\r
+ (or failed if C{failureTests} is True), and the results contain a list of lines of each \r
+ test's output\r
+ \r
+ Example::\r
+ number_expr = pyparsing_common.number.copy()\r
+\r
+ result = number_expr.runTests('''\r
+ # unsigned integer\r
+ 100\r
+ # negative integer\r
+ -100\r
+ # float with scientific notation\r
+ 6.02e23\r
+ # integer with scientific notation\r
+ 1e-12\r
+ ''')\r
+ print("Success" if result[0] else "Failed!")\r
+\r
+ result = number_expr.runTests('''\r
+ # stray character\r
+ 100Z\r
+ # missing leading digit before '.'\r
+ -.100\r
+ # too many '.'\r
+ 3.14.159\r
+ ''', failureTests=True)\r
+ print("Success" if result[0] else "Failed!")\r
+ prints::\r
+ # unsigned integer\r
+ 100\r
+ [100]\r
+\r
+ # negative integer\r
+ -100\r
+ [-100]\r
+\r
+ # float with scientific notation\r
+ 6.02e23\r
+ [6.02e+23]\r
+\r
+ # integer with scientific notation\r
+ 1e-12\r
+ [1e-12]\r
+\r
+ Success\r
+ \r
+ # stray character\r
+ 100Z\r
+ ^\r
+ FAIL: Expected end of text (at char 3), (line:1, col:4)\r
+\r
+ # missing leading digit before '.'\r
+ -.100\r
+ ^\r
+ FAIL: Expected {real number with scientific notation | real number | signed integer} (at char 0), (line:1, col:1)\r
+\r
+ # too many '.'\r
+ 3.14.159\r
+ ^\r
+ FAIL: Expected end of text (at char 4), (line:1, col:5)\r
+\r
+ Success\r
+\r
+ Each test string must be on a single line. If you want to test a string that spans multiple\r
+ lines, create a test like this::\r
+\r
+ expr.runTest(r"this is a test\\n of strings that spans \\n 3 lines")\r
+ \r
+ (Note that this is a raw string literal, you must include the leading 'r'.)\r
+ """\r
+ if isinstance(tests, basestring):\r
+ tests = list(map(str.strip, tests.rstrip().splitlines()))\r
+ if isinstance(comment, basestring):\r
+ comment = Literal(comment)\r
+ allResults = []\r
+ comments = []\r
+ success = True\r
+ for t in tests:\r
+ if comment is not None and comment.matches(t, False) or comments and not t:\r
+ comments.append(t)\r
+ continue\r
+ if not t:\r
+ continue\r
+ out = ['\n'.join(comments), t]\r
+ comments = []\r
+ try:\r
+ t = t.replace(r'\n','\n')\r
+ result = self.parseString(t, parseAll=parseAll)\r
+ out.append(result.dump(full=fullDump))\r
+ success = success and not failureTests\r
+ except ParseBaseException as pe:\r
+ fatal = "(FATAL)" if isinstance(pe, ParseFatalException) else ""\r
+ if '\n' in t:\r
+ out.append(line(pe.loc, t))\r
+ out.append(' '*(col(pe.loc,t)-1) + '^' + fatal)\r
+ else:\r
+ out.append(' '*pe.loc + '^' + fatal)\r
+ out.append("FAIL: " + str(pe))\r
+ success = success and failureTests\r
+ result = pe\r
+ except Exception as exc:\r
+ out.append("FAIL-EXCEPTION: " + str(exc))\r
+ success = success and failureTests\r
+ result = exc\r
+\r
+ if printResults:\r
+ if fullDump:\r
+ out.append('')\r
+ print('\n'.join(out))\r
+\r
+ allResults.append((t, result))\r
+ \r
+ return success, allResults\r
+\r
+ \r
+class Token(ParserElement):\r
+ """\r
+ Abstract C{ParserElement} subclass, for defining atomic matching patterns.\r
+ """\r
+ def __init__( self ):\r
+ super(Token,self).__init__( savelist=False )\r
+\r
+\r
+class Empty(Token):\r
+ """\r
+ An empty token, will always match.\r
+ """\r
+ def __init__( self ):\r
+ super(Empty,self).__init__()\r
+ self.name = "Empty"\r
+ self.mayReturnEmpty = True\r
+ self.mayIndexError = False\r
+\r
+\r
+class NoMatch(Token):\r
+ """\r
+ A token that will never match.\r
+ """\r
+ def __init__( self ):\r
+ super(NoMatch,self).__init__()\r
+ self.name = "NoMatch"\r
+ self.mayReturnEmpty = True\r
+ self.mayIndexError = False\r
+ self.errmsg = "Unmatchable token"\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+\r
+class Literal(Token):\r
+ """\r
+ Token to exactly match a specified string.\r
+ \r
+ Example::\r
+ Literal('blah').parseString('blah') # -> ['blah']\r
+ Literal('blah').parseString('blahfooblah') # -> ['blah']\r
+ Literal('blah').parseString('bla') # -> Exception: Expected "blah"\r
+ \r
+ For case-insensitive matching, use L{CaselessLiteral}.\r
+ \r
+ For keyword matching (force word break before and after the matched string),\r
+ use L{Keyword} or L{CaselessKeyword}.\r
+ """\r
+ def __init__( self, matchString ):\r
+ super(Literal,self).__init__()\r
+ self.match = matchString\r
+ self.matchLen = len(matchString)\r
+ try:\r
+ self.firstMatchChar = matchString[0]\r
+ except IndexError:\r
+ warnings.warn("null string passed to Literal; use Empty() instead",\r
+ SyntaxWarning, stacklevel=2)\r
+ self.__class__ = Empty\r
+ self.name = '"%s"' % _ustr(self.match)\r
+ self.errmsg = "Expected " + self.name\r
+ self.mayReturnEmpty = False\r
+ self.mayIndexError = False\r
+\r
+ # Performance tuning: this routine gets called a *lot*\r
+ # if this is a single character match string and the first character matches,\r
+ # short-circuit as quickly as possible, and avoid calling startswith\r
+ #~ @profile\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if (instring[loc] == self.firstMatchChar and\r
+ (self.matchLen==1 or instring.startswith(self.match,loc)) ):\r
+ return loc+self.matchLen, self.match\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+_L = Literal\r
+ParserElement._literalStringClass = Literal\r
+\r
+class Keyword(Token):\r
+ """\r
+ Token to exactly match a specified string as a keyword, that is, it must be\r
+ immediately followed by a non-keyword character. Compare with C{L{Literal}}:\r
+ - C{Literal("if")} will match the leading C{'if'} in C{'ifAndOnlyIf'}.\r
+ - C{Keyword("if")} will not; it will only match the leading C{'if'} in C{'if x=1'}, or C{'if(y==2)'}\r
+ Accepts two optional constructor arguments in addition to the keyword string:\r
+ - C{identChars} is a string of characters that would be valid identifier characters,\r
+ defaulting to all alphanumerics + "_" and "$"\r
+ - C{caseless} allows case-insensitive matching, default is C{False}.\r
+ \r
+ Example::\r
+ Keyword("start").parseString("start") # -> ['start']\r
+ Keyword("start").parseString("starting") # -> Exception\r
+\r
+ For case-insensitive matching, use L{CaselessKeyword}.\r
+ """\r
+ DEFAULT_KEYWORD_CHARS = alphanums+"_$"\r
+\r
+ def __init__( self, matchString, identChars=None, caseless=False ):\r
+ super(Keyword,self).__init__()\r
+ if identChars is None:\r
+ identChars = Keyword.DEFAULT_KEYWORD_CHARS\r
+ self.match = matchString\r
+ self.matchLen = len(matchString)\r
+ try:\r
+ self.firstMatchChar = matchString[0]\r
+ except IndexError:\r
+ warnings.warn("null string passed to Keyword; use Empty() instead",\r
+ SyntaxWarning, stacklevel=2)\r
+ self.name = '"%s"' % self.match\r
+ self.errmsg = "Expected " + self.name\r
+ self.mayReturnEmpty = False\r
+ self.mayIndexError = False\r
+ self.caseless = caseless\r
+ if caseless:\r
+ self.caselessmatch = matchString.upper()\r
+ identChars = identChars.upper()\r
+ self.identChars = set(identChars)\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if self.caseless:\r
+ if ( (instring[ loc:loc+self.matchLen ].upper() == self.caselessmatch) and\r
+ (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen].upper() not in self.identChars) and\r
+ (loc == 0 or instring[loc-1].upper() not in self.identChars) ):\r
+ return loc+self.matchLen, self.match\r
+ else:\r
+ if (instring[loc] == self.firstMatchChar and\r
+ (self.matchLen==1 or instring.startswith(self.match,loc)) and\r
+ (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen] not in self.identChars) and\r
+ (loc == 0 or instring[loc-1] not in self.identChars) ):\r
+ return loc+self.matchLen, self.match\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ def copy(self):\r
+ c = super(Keyword,self).copy()\r
+ c.identChars = Keyword.DEFAULT_KEYWORD_CHARS\r
+ return c\r
+\r
+ @staticmethod\r
+ def setDefaultKeywordChars( chars ):\r
+ """Overrides the default Keyword chars\r
+ """\r
+ Keyword.DEFAULT_KEYWORD_CHARS = chars\r
+\r
+class CaselessLiteral(Literal):\r
+ """\r
+ Token to match a specified string, ignoring case of letters.\r
+ Note: the matched results will always be in the case of the given\r
+ match string, NOT the case of the input text.\r
+\r
+ Example::\r
+ OneOrMore(CaselessLiteral("CMD")).parseString("cmd CMD Cmd10") # -> ['CMD', 'CMD', 'CMD']\r
+ \r
+ (Contrast with example for L{CaselessKeyword}.)\r
+ """\r
+ def __init__( self, matchString ):\r
+ super(CaselessLiteral,self).__init__( matchString.upper() )\r
+ # Preserve the defining literal.\r
+ self.returnString = matchString\r
+ self.name = "'%s'" % self.returnString\r
+ self.errmsg = "Expected " + self.name\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if instring[ loc:loc+self.matchLen ].upper() == self.match:\r
+ return loc+self.matchLen, self.returnString\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+class CaselessKeyword(Keyword):\r
+ """\r
+ Caseless version of L{Keyword}.\r
+\r
+ Example::\r
+ OneOrMore(CaselessKeyword("CMD")).parseString("cmd CMD Cmd10") # -> ['CMD', 'CMD']\r
+ \r
+ (Contrast with example for L{CaselessLiteral}.)\r
+ """\r
+ def __init__( self, matchString, identChars=None ):\r
+ super(CaselessKeyword,self).__init__( matchString, identChars, caseless=True )\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if ( (instring[ loc:loc+self.matchLen ].upper() == self.caselessmatch) and\r
+ (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen].upper() not in self.identChars) ):\r
+ return loc+self.matchLen, self.match\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+class CloseMatch(Token):\r
+ """\r
+ A variation on L{Literal} which matches "close" matches, that is, \r
+ strings with at most 'n' mismatching characters. C{CloseMatch} takes parameters:\r
+ - C{match_string} - string to be matched\r
+ - C{maxMismatches} - (C{default=1}) maximum number of mismatches allowed to count as a match\r
+ \r
+ The results from a successful parse will contain the matched text from the input string and the following named results:\r
+ - C{mismatches} - a list of the positions within the match_string where mismatches were found\r
+ - C{original} - the original match_string used to compare against the input string\r
+ \r
+ If C{mismatches} is an empty list, then the match was an exact match.\r
+ \r
+ Example::\r
+ patt = CloseMatch("ATCATCGAATGGA")\r
+ patt.parseString("ATCATCGAAXGGA") # -> (['ATCATCGAAXGGA'], {'mismatches': [[9]], 'original': ['ATCATCGAATGGA']})\r
+ patt.parseString("ATCAXCGAAXGGA") # -> Exception: Expected 'ATCATCGAATGGA' (with up to 1 mismatches) (at char 0), (line:1, col:1)\r
+\r
+ # exact match\r
+ patt.parseString("ATCATCGAATGGA") # -> (['ATCATCGAATGGA'], {'mismatches': [[]], 'original': ['ATCATCGAATGGA']})\r
+\r
+ # close match allowing up to 2 mismatches\r
+ patt = CloseMatch("ATCATCGAATGGA", maxMismatches=2)\r
+ patt.parseString("ATCAXCGAAXGGA") # -> (['ATCAXCGAAXGGA'], {'mismatches': [[4, 9]], 'original': ['ATCATCGAATGGA']})\r
+ """\r
+ def __init__(self, match_string, maxMismatches=1):\r
+ super(CloseMatch,self).__init__()\r
+ self.name = match_string\r
+ self.match_string = match_string\r
+ self.maxMismatches = maxMismatches\r
+ self.errmsg = "Expected %r (with up to %d mismatches)" % (self.match_string, self.maxMismatches)\r
+ self.mayIndexError = False\r
+ self.mayReturnEmpty = False\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ start = loc\r
+ instrlen = len(instring)\r
+ maxloc = start + len(self.match_string)\r
+\r
+ if maxloc <= instrlen:\r
+ match_string = self.match_string\r
+ match_stringloc = 0\r
+ mismatches = []\r
+ maxMismatches = self.maxMismatches\r
+\r
+ for match_stringloc,s_m in enumerate(zip(instring[loc:maxloc], self.match_string)):\r
+ src,mat = s_m\r
+ if src != mat:\r
+ mismatches.append(match_stringloc)\r
+ if len(mismatches) > maxMismatches:\r
+ break\r
+ else:\r
+ loc = match_stringloc + 1\r
+ results = ParseResults([instring[start:loc]])\r
+ results['original'] = self.match_string\r
+ results['mismatches'] = mismatches\r
+ return loc, results\r
+\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+\r
+class Word(Token):\r
+ """\r
+ Token for matching words composed of allowed character sets.\r
+ Defined with string containing all allowed initial characters,\r
+ an optional string containing allowed body characters (if omitted,\r
+ defaults to the initial character set), and an optional minimum,\r
+ maximum, and/or exact length. The default value for C{min} is 1 (a\r
+ minimum value < 1 is not valid); the default values for C{max} and C{exact}\r
+ are 0, meaning no maximum or exact length restriction. An optional\r
+ C{excludeChars} parameter can list characters that might be found in \r
+ the input C{bodyChars} string; useful to define a word of all printables\r
+ except for one or two characters, for instance.\r
+ \r
+ L{srange} is useful for defining custom character set strings for defining \r
+ C{Word} expressions, using range notation from regular expression character sets.\r
+ \r
+ A common mistake is to use C{Word} to match a specific literal string, as in \r
+ C{Word("Address")}. Remember that C{Word} uses the string argument to define\r
+ I{sets} of matchable characters. This expression would match "Add", "AAA",\r
+ "dAred", or any other word made up of the characters 'A', 'd', 'r', 'e', and 's'.\r
+ To match an exact literal string, use L{Literal} or L{Keyword}.\r
+\r
+ pyparsing includes helper strings for building Words:\r
+ - L{alphas}\r
+ - L{nums}\r
+ - L{alphanums}\r
+ - L{hexnums}\r
+ - L{alphas8bit} (alphabetic characters in ASCII range 128-255 - accented, tilded, umlauted, etc.)\r
+ - L{punc8bit} (non-alphabetic characters in ASCII range 128-255 - currency, symbols, superscripts, diacriticals, etc.)\r
+ - L{printables} (any non-whitespace character)\r
+\r
+ Example::\r
+ # a word composed of digits\r
+ integer = Word(nums) # equivalent to Word("0123456789") or Word(srange("0-9"))\r
+ \r
+ # a word with a leading capital, and zero or more lowercase\r
+ capital_word = Word(alphas.upper(), alphas.lower())\r
+\r
+ # hostnames are alphanumeric, with leading alpha, and '-'\r
+ hostname = Word(alphas, alphanums+'-')\r
+ \r
+ # roman numeral (not a strict parser, accepts invalid mix of characters)\r
+ roman = Word("IVXLCDM")\r
+ \r
+ # any string of non-whitespace characters, except for ','\r
+ csv_value = Word(printables, excludeChars=",")\r
+ """\r
+ def __init__( self, initChars, bodyChars=None, min=1, max=0, exact=0, asKeyword=False, excludeChars=None ):\r
+ super(Word,self).__init__()\r
+ if excludeChars:\r
+ initChars = ''.join(c for c in initChars if c not in excludeChars)\r
+ if bodyChars:\r
+ bodyChars = ''.join(c for c in bodyChars if c not in excludeChars)\r
+ self.initCharsOrig = initChars\r
+ self.initChars = set(initChars)\r
+ if bodyChars :\r
+ self.bodyCharsOrig = bodyChars\r
+ self.bodyChars = set(bodyChars)\r
+ else:\r
+ self.bodyCharsOrig = initChars\r
+ self.bodyChars = set(initChars)\r
+\r
+ self.maxSpecified = max > 0\r
+\r
+ if min < 1:\r
+ raise ValueError("cannot specify a minimum length < 1; use Optional(Word()) if zero-length word is permitted")\r
+\r
+ self.minLen = min\r
+\r
+ if max > 0:\r
+ self.maxLen = max\r
+ else:\r
+ self.maxLen = _MAX_INT\r
+\r
+ if exact > 0:\r
+ self.maxLen = exact\r
+ self.minLen = exact\r
+\r
+ self.name = _ustr(self)\r
+ self.errmsg = "Expected " + self.name\r
+ self.mayIndexError = False\r
+ self.asKeyword = asKeyword\r
+\r
+ if ' ' not in self.initCharsOrig+self.bodyCharsOrig and (min==1 and max==0 and exact==0):\r
+ if self.bodyCharsOrig == self.initCharsOrig:\r
+ self.reString = "[%s]+" % _escapeRegexRangeChars(self.initCharsOrig)\r
+ elif len(self.initCharsOrig) == 1:\r
+ self.reString = "%s[%s]*" % \\r
+ (re.escape(self.initCharsOrig),\r
+ _escapeRegexRangeChars(self.bodyCharsOrig),)\r
+ else:\r
+ self.reString = "[%s][%s]*" % \\r
+ (_escapeRegexRangeChars(self.initCharsOrig),\r
+ _escapeRegexRangeChars(self.bodyCharsOrig),)\r
+ if self.asKeyword:\r
+ self.reString = r"\b"+self.reString+r"\b"\r
+ try:\r
+ self.re = re.compile( self.reString )\r
+ except Exception:\r
+ self.re = None\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if self.re:\r
+ result = self.re.match(instring,loc)\r
+ if not result:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ loc = result.end()\r
+ return loc, result.group()\r
+\r
+ if not(instring[ loc ] in self.initChars):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ start = loc\r
+ loc += 1\r
+ instrlen = len(instring)\r
+ bodychars = self.bodyChars\r
+ maxloc = start + self.maxLen\r
+ maxloc = min( maxloc, instrlen )\r
+ while loc < maxloc and instring[loc] in bodychars:\r
+ loc += 1\r
+\r
+ throwException = False\r
+ if loc - start < self.minLen:\r
+ throwException = True\r
+ if self.maxSpecified and loc < instrlen and instring[loc] in bodychars:\r
+ throwException = True\r
+ if self.asKeyword:\r
+ if (start>0 and instring[start-1] in bodychars) or (loc<instrlen and instring[loc] in bodychars):\r
+ throwException = True\r
+\r
+ if throwException:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ return loc, instring[start:loc]\r
+\r
+ def __str__( self ):\r
+ try:\r
+ return super(Word,self).__str__()\r
+ except Exception:\r
+ pass\r
+\r
+\r
+ if self.strRepr is None:\r
+\r
+ def charsAsStr(s):\r
+ if len(s)>4:\r
+ return s[:4]+"..."\r
+ else:\r
+ return s\r
+\r
+ if ( self.initCharsOrig != self.bodyCharsOrig ):\r
+ self.strRepr = "W:(%s,%s)" % ( charsAsStr(self.initCharsOrig), charsAsStr(self.bodyCharsOrig) )\r
+ else:\r
+ self.strRepr = "W:(%s)" % charsAsStr(self.initCharsOrig)\r
+\r
+ return self.strRepr\r
+\r
+\r
+class Regex(Token):\r
+ """\r
+ Token for matching strings that match a given regular expression.\r
+ Defined with string specifying the regular expression in a form recognized by the inbuilt Python re module.\r
+ If the given regex contains named groups (defined using C{(?P<name>...)}), these will be preserved as \r
+ named parse results.\r
+\r
+ Example::\r
+ realnum = Regex(r"[+-]?\d+\.\d*")\r
+ date = Regex(r'(?P<year>\d{4})-(?P<month>\d\d?)-(?P<day>\d\d?)')\r
+ # ref: http://stackoverflow.com/questions/267399/how-do-you-match-only-valid-roman-numerals-with-a-regular-expression\r
+ roman = Regex(r"M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})")\r
+ """\r
+ compiledREtype = type(re.compile("[A-Z]"))\r
+ def __init__( self, pattern, flags=0):\r
+ """The parameters C{pattern} and C{flags} are passed to the C{re.compile()} function as-is. See the Python C{re} module for an explanation of the acceptable patterns and flags."""\r
+ super(Regex,self).__init__()\r
+\r
+ if isinstance(pattern, basestring):\r
+ if not pattern:\r
+ warnings.warn("null string passed to Regex; use Empty() instead",\r
+ SyntaxWarning, stacklevel=2)\r
+\r
+ self.pattern = pattern\r
+ self.flags = flags\r
+\r
+ try:\r
+ self.re = re.compile(self.pattern, self.flags)\r
+ self.reString = self.pattern\r
+ except sre_constants.error:\r
+ warnings.warn("invalid pattern (%s) passed to Regex" % pattern,\r
+ SyntaxWarning, stacklevel=2)\r
+ raise\r
+\r
+ elif isinstance(pattern, Regex.compiledREtype):\r
+ self.re = pattern\r
+ self.pattern = \\r
+ self.reString = str(pattern)\r
+ self.flags = flags\r
+ \r
+ else:\r
+ raise ValueError("Regex may only be constructed with a string or a compiled RE object")\r
+\r
+ self.name = _ustr(self)\r
+ self.errmsg = "Expected " + self.name\r
+ self.mayIndexError = False\r
+ self.mayReturnEmpty = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ result = self.re.match(instring,loc)\r
+ if not result:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ loc = result.end()\r
+ d = result.groupdict()\r
+ ret = ParseResults(result.group())\r
+ if d:\r
+ for k in d:\r
+ ret[k] = d[k]\r
+ return loc,ret\r
+\r
+ def __str__( self ):\r
+ try:\r
+ return super(Regex,self).__str__()\r
+ except Exception:\r
+ pass\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "Re:(%s)" % repr(self.pattern)\r
+\r
+ return self.strRepr\r
+\r
+\r
+class QuotedString(Token):\r
+ r"""\r
+ Token for matching strings that are delimited by quoting characters.\r
+ \r
+ Defined with the following parameters:\r
+ - quoteChar - string of one or more characters defining the quote delimiting string\r
+ - escChar - character to escape quotes, typically backslash (default=C{None})\r
+ - escQuote - special quote sequence to escape an embedded quote string (such as SQL's "" to escape an embedded ") (default=C{None})\r
+ - multiline - boolean indicating whether quotes can span multiple lines (default=C{False})\r
+ - unquoteResults - boolean indicating whether the matched text should be unquoted (default=C{True})\r
+ - endQuoteChar - string of one or more characters defining the end of the quote delimited string (default=C{None} => same as quoteChar)\r
+ - convertWhitespaceEscapes - convert escaped whitespace (C{'\t'}, C{'\n'}, etc.) to actual whitespace (default=C{True})\r
+\r
+ Example::\r
+ qs = QuotedString('"')\r
+ print(qs.searchString('lsjdf "This is the quote" sldjf'))\r
+ complex_qs = QuotedString('{{', endQuoteChar='}}')\r
+ print(complex_qs.searchString('lsjdf {{This is the "quote"}} sldjf'))\r
+ sql_qs = QuotedString('"', escQuote='""')\r
+ print(sql_qs.searchString('lsjdf "This is the quote with ""embedded"" quotes" sldjf'))\r
+ prints::\r
+ [['This is the quote']]\r
+ [['This is the "quote"']]\r
+ [['This is the quote with "embedded" quotes']]\r
+ """\r
+ def __init__( self, quoteChar, escChar=None, escQuote=None, multiline=False, unquoteResults=True, endQuoteChar=None, convertWhitespaceEscapes=True):\r
+ super(QuotedString,self).__init__()\r
+\r
+ # remove white space from quote chars - wont work anyway\r
+ quoteChar = quoteChar.strip()\r
+ if not quoteChar:\r
+ warnings.warn("quoteChar cannot be the empty string",SyntaxWarning,stacklevel=2)\r
+ raise SyntaxError()\r
+\r
+ if endQuoteChar is None:\r
+ endQuoteChar = quoteChar\r
+ else:\r
+ endQuoteChar = endQuoteChar.strip()\r
+ if not endQuoteChar:\r
+ warnings.warn("endQuoteChar cannot be the empty string",SyntaxWarning,stacklevel=2)\r
+ raise SyntaxError()\r
+\r
+ self.quoteChar = quoteChar\r
+ self.quoteCharLen = len(quoteChar)\r
+ self.firstQuoteChar = quoteChar[0]\r
+ self.endQuoteChar = endQuoteChar\r
+ self.endQuoteCharLen = len(endQuoteChar)\r
+ self.escChar = escChar\r
+ self.escQuote = escQuote\r
+ self.unquoteResults = unquoteResults\r
+ self.convertWhitespaceEscapes = convertWhitespaceEscapes\r
+\r
+ if multiline:\r
+ self.flags = re.MULTILINE | re.DOTALL\r
+ self.pattern = r'%s(?:[^%s%s]' % \\r
+ ( re.escape(self.quoteChar),\r
+ _escapeRegexRangeChars(self.endQuoteChar[0]),\r
+ (escChar is not None and _escapeRegexRangeChars(escChar) or '') )\r
+ else:\r
+ self.flags = 0\r
+ self.pattern = r'%s(?:[^%s\n\r%s]' % \\r
+ ( re.escape(self.quoteChar),\r
+ _escapeRegexRangeChars(self.endQuoteChar[0]),\r
+ (escChar is not None and _escapeRegexRangeChars(escChar) or '') )\r
+ if len(self.endQuoteChar) > 1:\r
+ self.pattern += (\r
+ '|(?:' + ')|(?:'.join("%s[^%s]" % (re.escape(self.endQuoteChar[:i]),\r
+ _escapeRegexRangeChars(self.endQuoteChar[i]))\r
+ for i in range(len(self.endQuoteChar)-1,0,-1)) + ')'\r
+ )\r
+ if escQuote:\r
+ self.pattern += (r'|(?:%s)' % re.escape(escQuote))\r
+ if escChar:\r
+ self.pattern += (r'|(?:%s.)' % re.escape(escChar))\r
+ self.escCharReplacePattern = re.escape(self.escChar)+"(.)"\r
+ self.pattern += (r')*%s' % re.escape(self.endQuoteChar))\r
+\r
+ try:\r
+ self.re = re.compile(self.pattern, self.flags)\r
+ self.reString = self.pattern\r
+ except sre_constants.error:\r
+ warnings.warn("invalid pattern (%s) passed to Regex" % self.pattern,\r
+ SyntaxWarning, stacklevel=2)\r
+ raise\r
+\r
+ self.name = _ustr(self)\r
+ self.errmsg = "Expected " + self.name\r
+ self.mayIndexError = False\r
+ self.mayReturnEmpty = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ result = instring[loc] == self.firstQuoteChar and self.re.match(instring,loc) or None\r
+ if not result:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ loc = result.end()\r
+ ret = result.group()\r
+\r
+ if self.unquoteResults:\r
+\r
+ # strip off quotes\r
+ ret = ret[self.quoteCharLen:-self.endQuoteCharLen]\r
+\r
+ if isinstance(ret,basestring):\r
+ # replace escaped whitespace\r
+ if '\\' in ret and self.convertWhitespaceEscapes:\r
+ ws_map = {\r
+ r'\t' : '\t',\r
+ r'\n' : '\n',\r
+ r'\f' : '\f',\r
+ r'\r' : '\r',\r
+ }\r
+ for wslit,wschar in ws_map.items():\r
+ ret = ret.replace(wslit, wschar)\r
+\r
+ # replace escaped characters\r
+ if self.escChar:\r
+ ret = re.sub(self.escCharReplacePattern,"\g<1>",ret)\r
+\r
+ # replace escaped quotes\r
+ if self.escQuote:\r
+ ret = ret.replace(self.escQuote, self.endQuoteChar)\r
+\r
+ return loc, ret\r
+\r
+ def __str__( self ):\r
+ try:\r
+ return super(QuotedString,self).__str__()\r
+ except Exception:\r
+ pass\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "quoted string, starting with %s ending with %s" % (self.quoteChar, self.endQuoteChar)\r
+\r
+ return self.strRepr\r
+\r
+\r
+class CharsNotIn(Token):\r
+ """\r
+ Token for matching words composed of characters I{not} in a given set (will\r
+ include whitespace in matched characters if not listed in the provided exclusion set - see example).\r
+ Defined with string containing all disallowed characters, and an optional\r
+ minimum, maximum, and/or exact length. The default value for C{min} is 1 (a\r
+ minimum value < 1 is not valid); the default values for C{max} and C{exact}\r
+ are 0, meaning no maximum or exact length restriction.\r
+\r
+ Example::\r
+ # define a comma-separated-value as anything that is not a ','\r
+ csv_value = CharsNotIn(',')\r
+ print(delimitedList(csv_value).parseString("dkls,lsdkjf,s12 34,@!#,213"))\r
+ prints::\r
+ ['dkls', 'lsdkjf', 's12 34', '@!#', '213']\r
+ """\r
+ def __init__( self, notChars, min=1, max=0, exact=0 ):\r
+ super(CharsNotIn,self).__init__()\r
+ self.skipWhitespace = False\r
+ self.notChars = notChars\r
+\r
+ if min < 1:\r
+ raise ValueError("cannot specify a minimum length < 1; use Optional(CharsNotIn()) if zero-length char group is permitted")\r
+\r
+ self.minLen = min\r
+\r
+ if max > 0:\r
+ self.maxLen = max\r
+ else:\r
+ self.maxLen = _MAX_INT\r
+\r
+ if exact > 0:\r
+ self.maxLen = exact\r
+ self.minLen = exact\r
+\r
+ self.name = _ustr(self)\r
+ self.errmsg = "Expected " + self.name\r
+ self.mayReturnEmpty = ( self.minLen == 0 )\r
+ self.mayIndexError = False\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if instring[loc] in self.notChars:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ start = loc\r
+ loc += 1\r
+ notchars = self.notChars\r
+ maxlen = min( start+self.maxLen, len(instring) )\r
+ while loc < maxlen and \\r
+ (instring[loc] not in notchars):\r
+ loc += 1\r
+\r
+ if loc - start < self.minLen:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ return loc, instring[start:loc]\r
+\r
+ def __str__( self ):\r
+ try:\r
+ return super(CharsNotIn, self).__str__()\r
+ except Exception:\r
+ pass\r
+\r
+ if self.strRepr is None:\r
+ if len(self.notChars) > 4:\r
+ self.strRepr = "!W:(%s...)" % self.notChars[:4]\r
+ else:\r
+ self.strRepr = "!W:(%s)" % self.notChars\r
+\r
+ return self.strRepr\r
+\r
+class White(Token):\r
+ """\r
+ Special matching class for matching whitespace. Normally, whitespace is ignored\r
+ by pyparsing grammars. This class is included when some whitespace structures\r
+ are significant. Define with a string containing the whitespace characters to be\r
+ matched; default is C{" \\t\\r\\n"}. Also takes optional C{min}, C{max}, and C{exact} arguments,\r
+ as defined for the C{L{Word}} class.\r
+ """\r
+ whiteStrs = {\r
+ " " : "<SPC>",\r
+ "\t": "<TAB>",\r
+ "\n": "<LF>",\r
+ "\r": "<CR>",\r
+ "\f": "<FF>",\r
+ }\r
+ def __init__(self, ws=" \t\r\n", min=1, max=0, exact=0):\r
+ super(White,self).__init__()\r
+ self.matchWhite = ws\r
+ self.setWhitespaceChars( "".join(c for c in self.whiteChars if c not in self.matchWhite) )\r
+ #~ self.leaveWhitespace()\r
+ self.name = ("".join(White.whiteStrs[c] for c in self.matchWhite))\r
+ self.mayReturnEmpty = True\r
+ self.errmsg = "Expected " + self.name\r
+\r
+ self.minLen = min\r
+\r
+ if max > 0:\r
+ self.maxLen = max\r
+ else:\r
+ self.maxLen = _MAX_INT\r
+\r
+ if exact > 0:\r
+ self.maxLen = exact\r
+ self.minLen = exact\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if not(instring[ loc ] in self.matchWhite):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+ start = loc\r
+ loc += 1\r
+ maxloc = start + self.maxLen\r
+ maxloc = min( maxloc, len(instring) )\r
+ while loc < maxloc and instring[loc] in self.matchWhite:\r
+ loc += 1\r
+\r
+ if loc - start < self.minLen:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ return loc, instring[start:loc]\r
+\r
+\r
+class _PositionToken(Token):\r
+ def __init__( self ):\r
+ super(_PositionToken,self).__init__()\r
+ self.name=self.__class__.__name__\r
+ self.mayReturnEmpty = True\r
+ self.mayIndexError = False\r
+\r
+class GoToColumn(_PositionToken):\r
+ """\r
+ Token to advance to a specific column of input text; useful for tabular report scraping.\r
+ """\r
+ def __init__( self, colno ):\r
+ super(GoToColumn,self).__init__()\r
+ self.col = colno\r
+\r
+ def preParse( self, instring, loc ):\r
+ if col(loc,instring) != self.col:\r
+ instrlen = len(instring)\r
+ if self.ignoreExprs:\r
+ loc = self._skipIgnorables( instring, loc )\r
+ while loc < instrlen and instring[loc].isspace() and col( loc, instring ) != self.col :\r
+ loc += 1\r
+ return loc\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ thiscol = col( loc, instring )\r
+ if thiscol > self.col:\r
+ raise ParseException( instring, loc, "Text not in expected column", self )\r
+ newloc = loc + self.col - thiscol\r
+ ret = instring[ loc: newloc ]\r
+ return newloc, ret\r
+\r
+\r
+class LineStart(_PositionToken):\r
+ """\r
+ Matches if current position is at the beginning of a line within the parse string\r
+ \r
+ Example::\r
+ \r
+ test = '''\\r
+ AAA this line\r
+ AAA and this line\r
+ AAA but not this one\r
+ B AAA and definitely not this one\r
+ '''\r
+\r
+ for t in (LineStart() + 'AAA' + restOfLine).searchString(test):\r
+ print(t)\r
+ \r
+ Prints::\r
+ ['AAA', ' this line']\r
+ ['AAA', ' and this line'] \r
+\r
+ """\r
+ def __init__( self ):\r
+ super(LineStart,self).__init__()\r
+ self.errmsg = "Expected start of line"\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if col(loc, instring) == 1:\r
+ return loc, []\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+class LineEnd(_PositionToken):\r
+ """\r
+ Matches if current position is at the end of a line within the parse string\r
+ """\r
+ def __init__( self ):\r
+ super(LineEnd,self).__init__()\r
+ self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n","") )\r
+ self.errmsg = "Expected end of line"\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if loc<len(instring):\r
+ if instring[loc] == "\n":\r
+ return loc+1, "\n"\r
+ else:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+ elif loc == len(instring):\r
+ return loc+1, []\r
+ else:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+class StringStart(_PositionToken):\r
+ """\r
+ Matches if current position is at the beginning of the parse string\r
+ """\r
+ def __init__( self ):\r
+ super(StringStart,self).__init__()\r
+ self.errmsg = "Expected start of text"\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if loc != 0:\r
+ # see if entire string up to here is just whitespace and ignoreables\r
+ if loc != self.preParse( instring, 0 ):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+ return loc, []\r
+\r
+class StringEnd(_PositionToken):\r
+ """\r
+ Matches if current position is at the end of the parse string\r
+ """\r
+ def __init__( self ):\r
+ super(StringEnd,self).__init__()\r
+ self.errmsg = "Expected end of text"\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if loc < len(instring):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+ elif loc == len(instring):\r
+ return loc+1, []\r
+ elif loc > len(instring):\r
+ return loc, []\r
+ else:\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+class WordStart(_PositionToken):\r
+ """\r
+ Matches if the current position is at the beginning of a Word, and\r
+ is not preceded by any character in a given set of C{wordChars}\r
+ (default=C{printables}). To emulate the C{\b} behavior of regular expressions,\r
+ use C{WordStart(alphanums)}. C{WordStart} will also match at the beginning of\r
+ the string being parsed, or at the beginning of a line.\r
+ """\r
+ def __init__(self, wordChars = printables):\r
+ super(WordStart,self).__init__()\r
+ self.wordChars = set(wordChars)\r
+ self.errmsg = "Not at the start of a word"\r
+\r
+ def parseImpl(self, instring, loc, doActions=True ):\r
+ if loc != 0:\r
+ if (instring[loc-1] in self.wordChars or\r
+ instring[loc] not in self.wordChars):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+ return loc, []\r
+\r
+class WordEnd(_PositionToken):\r
+ """\r
+ Matches if the current position is at the end of a Word, and\r
+ is not followed by any character in a given set of C{wordChars}\r
+ (default=C{printables}). To emulate the C{\b} behavior of regular expressions,\r
+ use C{WordEnd(alphanums)}. C{WordEnd} will also match at the end of\r
+ the string being parsed, or at the end of a line.\r
+ """\r
+ def __init__(self, wordChars = printables):\r
+ super(WordEnd,self).__init__()\r
+ self.wordChars = set(wordChars)\r
+ self.skipWhitespace = False\r
+ self.errmsg = "Not at the end of a word"\r
+\r
+ def parseImpl(self, instring, loc, doActions=True ):\r
+ instrlen = len(instring)\r
+ if instrlen>0 and loc<instrlen:\r
+ if (instring[loc] in self.wordChars or\r
+ instring[loc-1] not in self.wordChars):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+ return loc, []\r
+\r
+\r
+class ParseExpression(ParserElement):\r
+ """\r
+ Abstract subclass of ParserElement, for combining and post-processing parsed tokens.\r
+ """\r
+ def __init__( self, exprs, savelist = False ):\r
+ super(ParseExpression,self).__init__(savelist)\r
+ if isinstance( exprs, _generatorType ):\r
+ exprs = list(exprs)\r
+\r
+ if isinstance( exprs, basestring ):\r
+ self.exprs = [ ParserElement._literalStringClass( exprs ) ]\r
+ elif isinstance( exprs, collections.Iterable ):\r
+ exprs = list(exprs)\r
+ # if sequence of strings provided, wrap with Literal\r
+ if all(isinstance(expr, basestring) for expr in exprs):\r
+ exprs = map(ParserElement._literalStringClass, exprs)\r
+ self.exprs = list(exprs)\r
+ else:\r
+ try:\r
+ self.exprs = list( exprs )\r
+ except TypeError:\r
+ self.exprs = [ exprs ]\r
+ self.callPreparse = False\r
+\r
+ def __getitem__( self, i ):\r
+ return self.exprs[i]\r
+\r
+ def append( self, other ):\r
+ self.exprs.append( other )\r
+ self.strRepr = None\r
+ return self\r
+\r
+ def leaveWhitespace( self ):\r
+ """Extends C{leaveWhitespace} defined in base class, and also invokes C{leaveWhitespace} on\r
+ all contained expressions."""\r
+ self.skipWhitespace = False\r
+ self.exprs = [ e.copy() for e in self.exprs ]\r
+ for e in self.exprs:\r
+ e.leaveWhitespace()\r
+ return self\r
+\r
+ def ignore( self, other ):\r
+ if isinstance( other, Suppress ):\r
+ if other not in self.ignoreExprs:\r
+ super( ParseExpression, self).ignore( other )\r
+ for e in self.exprs:\r
+ e.ignore( self.ignoreExprs[-1] )\r
+ else:\r
+ super( ParseExpression, self).ignore( other )\r
+ for e in self.exprs:\r
+ e.ignore( self.ignoreExprs[-1] )\r
+ return self\r
+\r
+ def __str__( self ):\r
+ try:\r
+ return super(ParseExpression,self).__str__()\r
+ except Exception:\r
+ pass\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "%s:(%s)" % ( self.__class__.__name__, _ustr(self.exprs) )\r
+ return self.strRepr\r
+\r
+ def streamline( self ):\r
+ super(ParseExpression,self).streamline()\r
+\r
+ for e in self.exprs:\r
+ e.streamline()\r
+\r
+ # collapse nested And's of the form And( And( And( a,b), c), d) to And( a,b,c,d )\r
+ # but only if there are no parse actions or resultsNames on the nested And's\r
+ # (likewise for Or's and MatchFirst's)\r
+ if ( len(self.exprs) == 2 ):\r
+ other = self.exprs[0]\r
+ if ( isinstance( other, self.__class__ ) and\r
+ not(other.parseAction) and\r
+ other.resultsName is None and\r
+ not other.debug ):\r
+ self.exprs = other.exprs[:] + [ self.exprs[1] ]\r
+ self.strRepr = None\r
+ self.mayReturnEmpty |= other.mayReturnEmpty\r
+ self.mayIndexError |= other.mayIndexError\r
+\r
+ other = self.exprs[-1]\r
+ if ( isinstance( other, self.__class__ ) and\r
+ not(other.parseAction) and\r
+ other.resultsName is None and\r
+ not other.debug ):\r
+ self.exprs = self.exprs[:-1] + other.exprs[:]\r
+ self.strRepr = None\r
+ self.mayReturnEmpty |= other.mayReturnEmpty\r
+ self.mayIndexError |= other.mayIndexError\r
+\r
+ self.errmsg = "Expected " + _ustr(self)\r
+ \r
+ return self\r
+\r
+ def setResultsName( self, name, listAllMatches=False ):\r
+ ret = super(ParseExpression,self).setResultsName(name,listAllMatches)\r
+ return ret\r
+\r
+ def validate( self, validateTrace=[] ):\r
+ tmp = validateTrace[:]+[self]\r
+ for e in self.exprs:\r
+ e.validate(tmp)\r
+ self.checkRecursion( [] )\r
+ \r
+ def copy(self):\r
+ ret = super(ParseExpression,self).copy()\r
+ ret.exprs = [e.copy() for e in self.exprs]\r
+ return ret\r
+\r
+class And(ParseExpression):\r
+ """\r
+ Requires all given C{ParseExpression}s to be found in the given order.\r
+ Expressions may be separated by whitespace.\r
+ May be constructed using the C{'+'} operator.\r
+ May also be constructed using the C{'-'} operator, which will suppress backtracking.\r
+\r
+ Example::\r
+ integer = Word(nums)\r
+ name_expr = OneOrMore(Word(alphas))\r
+\r
+ expr = And([integer("id"),name_expr("name"),integer("age")])\r
+ # more easily written as:\r
+ expr = integer("id") + name_expr("name") + integer("age")\r
+ """\r
+\r
+ class _ErrorStop(Empty):\r
+ def __init__(self, *args, **kwargs):\r
+ super(And._ErrorStop,self).__init__(*args, **kwargs)\r
+ self.name = '-'\r
+ self.leaveWhitespace()\r
+\r
+ def __init__( self, exprs, savelist = True ):\r
+ super(And,self).__init__(exprs, savelist)\r
+ self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)\r
+ self.setWhitespaceChars( self.exprs[0].whiteChars )\r
+ self.skipWhitespace = self.exprs[0].skipWhitespace\r
+ self.callPreparse = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ # pass False as last arg to _parse for first element, since we already\r
+ # pre-parsed the string as part of our And pre-parsing\r
+ loc, resultlist = self.exprs[0]._parse( instring, loc, doActions, callPreParse=False )\r
+ errorStop = False\r
+ for e in self.exprs[1:]:\r
+ if isinstance(e, And._ErrorStop):\r
+ errorStop = True\r
+ continue\r
+ if errorStop:\r
+ try:\r
+ loc, exprtokens = e._parse( instring, loc, doActions )\r
+ except ParseSyntaxException:\r
+ raise\r
+ except ParseBaseException as pe:\r
+ pe.__traceback__ = None\r
+ raise ParseSyntaxException._from_exception(pe)\r
+ except IndexError:\r
+ raise ParseSyntaxException(instring, len(instring), self.errmsg, self)\r
+ else:\r
+ loc, exprtokens = e._parse( instring, loc, doActions )\r
+ if exprtokens or exprtokens.haskeys():\r
+ resultlist += exprtokens\r
+ return loc, resultlist\r
+\r
+ def __iadd__(self, other ):\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ return self.append( other ) #And( [ self, other ] )\r
+\r
+ def checkRecursion( self, parseElementList ):\r
+ subRecCheckList = parseElementList[:] + [ self ]\r
+ for e in self.exprs:\r
+ e.checkRecursion( subRecCheckList )\r
+ if not e.mayReturnEmpty:\r
+ break\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "{" + " ".join(_ustr(e) for e in self.exprs) + "}"\r
+\r
+ return self.strRepr\r
+\r
+\r
+class Or(ParseExpression):\r
+ """\r
+ Requires that at least one C{ParseExpression} is found.\r
+ If two expressions match, the expression that matches the longest string will be used.\r
+ May be constructed using the C{'^'} operator.\r
+\r
+ Example::\r
+ # construct Or using '^' operator\r
+ \r
+ number = Word(nums) ^ Combine(Word(nums) + '.' + Word(nums))\r
+ print(number.searchString("123 3.1416 789"))\r
+ prints::\r
+ [['123'], ['3.1416'], ['789']]\r
+ """\r
+ def __init__( self, exprs, savelist = False ):\r
+ super(Or,self).__init__(exprs, savelist)\r
+ if self.exprs:\r
+ self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)\r
+ else:\r
+ self.mayReturnEmpty = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ maxExcLoc = -1\r
+ maxException = None\r
+ matches = []\r
+ for e in self.exprs:\r
+ try:\r
+ loc2 = e.tryParse( instring, loc )\r
+ except ParseException as err:\r
+ err.__traceback__ = None\r
+ if err.loc > maxExcLoc:\r
+ maxException = err\r
+ maxExcLoc = err.loc\r
+ except IndexError:\r
+ if len(instring) > maxExcLoc:\r
+ maxException = ParseException(instring,len(instring),e.errmsg,self)\r
+ maxExcLoc = len(instring)\r
+ else:\r
+ # save match among all matches, to retry longest to shortest\r
+ matches.append((loc2, e))\r
+\r
+ if matches:\r
+ matches.sort(key=lambda x: -x[0])\r
+ for _,e in matches:\r
+ try:\r
+ return e._parse( instring, loc, doActions )\r
+ except ParseException as err:\r
+ err.__traceback__ = None\r
+ if err.loc > maxExcLoc:\r
+ maxException = err\r
+ maxExcLoc = err.loc\r
+\r
+ if maxException is not None:\r
+ maxException.msg = self.errmsg\r
+ raise maxException\r
+ else:\r
+ raise ParseException(instring, loc, "no defined alternatives to match", self)\r
+\r
+\r
+ def __ixor__(self, other ):\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ return self.append( other ) #Or( [ self, other ] )\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "{" + " ^ ".join(_ustr(e) for e in self.exprs) + "}"\r
+\r
+ return self.strRepr\r
+\r
+ def checkRecursion( self, parseElementList ):\r
+ subRecCheckList = parseElementList[:] + [ self ]\r
+ for e in self.exprs:\r
+ e.checkRecursion( subRecCheckList )\r
+\r
+\r
+class MatchFirst(ParseExpression):\r
+ """\r
+ Requires that at least one C{ParseExpression} is found.\r
+ If two expressions match, the first one listed is the one that will match.\r
+ May be constructed using the C{'|'} operator.\r
+\r
+ Example::\r
+ # construct MatchFirst using '|' operator\r
+ \r
+ # watch the order of expressions to match\r
+ number = Word(nums) | Combine(Word(nums) + '.' + Word(nums))\r
+ print(number.searchString("123 3.1416 789")) # Fail! -> [['123'], ['3'], ['1416'], ['789']]\r
+\r
+ # put more selective expression first\r
+ number = Combine(Word(nums) + '.' + Word(nums)) | Word(nums)\r
+ print(number.searchString("123 3.1416 789")) # Better -> [['123'], ['3.1416'], ['789']]\r
+ """\r
+ def __init__( self, exprs, savelist = False ):\r
+ super(MatchFirst,self).__init__(exprs, savelist)\r
+ if self.exprs:\r
+ self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)\r
+ else:\r
+ self.mayReturnEmpty = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ maxExcLoc = -1\r
+ maxException = None\r
+ for e in self.exprs:\r
+ try:\r
+ ret = e._parse( instring, loc, doActions )\r
+ return ret\r
+ except ParseException as err:\r
+ if err.loc > maxExcLoc:\r
+ maxException = err\r
+ maxExcLoc = err.loc\r
+ except IndexError:\r
+ if len(instring) > maxExcLoc:\r
+ maxException = ParseException(instring,len(instring),e.errmsg,self)\r
+ maxExcLoc = len(instring)\r
+\r
+ # only got here if no expression matched, raise exception for match that made it the furthest\r
+ else:\r
+ if maxException is not None:\r
+ maxException.msg = self.errmsg\r
+ raise maxException\r
+ else:\r
+ raise ParseException(instring, loc, "no defined alternatives to match", self)\r
+\r
+ def __ior__(self, other ):\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass( other )\r
+ return self.append( other ) #MatchFirst( [ self, other ] )\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "{" + " | ".join(_ustr(e) for e in self.exprs) + "}"\r
+\r
+ return self.strRepr\r
+\r
+ def checkRecursion( self, parseElementList ):\r
+ subRecCheckList = parseElementList[:] + [ self ]\r
+ for e in self.exprs:\r
+ e.checkRecursion( subRecCheckList )\r
+\r
+\r
+class Each(ParseExpression):\r
+ """\r
+ Requires all given C{ParseExpression}s to be found, but in any order.\r
+ Expressions may be separated by whitespace.\r
+ May be constructed using the C{'&'} operator.\r
+\r
+ Example::\r
+ color = oneOf("RED ORANGE YELLOW GREEN BLUE PURPLE BLACK WHITE BROWN")\r
+ shape_type = oneOf("SQUARE CIRCLE TRIANGLE STAR HEXAGON OCTAGON")\r
+ integer = Word(nums)\r
+ shape_attr = "shape:" + shape_type("shape")\r
+ posn_attr = "posn:" + Group(integer("x") + ',' + integer("y"))("posn")\r
+ color_attr = "color:" + color("color")\r
+ size_attr = "size:" + integer("size")\r
+\r
+ # use Each (using operator '&') to accept attributes in any order \r
+ # (shape and posn are required, color and size are optional)\r
+ shape_spec = shape_attr & posn_attr & Optional(color_attr) & Optional(size_attr)\r
+\r
+ shape_spec.runTests('''\r
+ shape: SQUARE color: BLACK posn: 100, 120\r
+ shape: CIRCLE size: 50 color: BLUE posn: 50,80\r
+ color:GREEN size:20 shape:TRIANGLE posn:20,40\r
+ '''\r
+ )\r
+ prints::\r
+ shape: SQUARE color: BLACK posn: 100, 120\r
+ ['shape:', 'SQUARE', 'color:', 'BLACK', 'posn:', ['100', ',', '120']]\r
+ - color: BLACK\r
+ - posn: ['100', ',', '120']\r
+ - x: 100\r
+ - y: 120\r
+ - shape: SQUARE\r
+\r
+\r
+ shape: CIRCLE size: 50 color: BLUE posn: 50,80\r
+ ['shape:', 'CIRCLE', 'size:', '50', 'color:', 'BLUE', 'posn:', ['50', ',', '80']]\r
+ - color: BLUE\r
+ - posn: ['50', ',', '80']\r
+ - x: 50\r
+ - y: 80\r
+ - shape: CIRCLE\r
+ - size: 50\r
+\r
+\r
+ color: GREEN size: 20 shape: TRIANGLE posn: 20,40\r
+ ['color:', 'GREEN', 'size:', '20', 'shape:', 'TRIANGLE', 'posn:', ['20', ',', '40']]\r
+ - color: GREEN\r
+ - posn: ['20', ',', '40']\r
+ - x: 20\r
+ - y: 40\r
+ - shape: TRIANGLE\r
+ - size: 20\r
+ """\r
+ def __init__( self, exprs, savelist = True ):\r
+ super(Each,self).__init__(exprs, savelist)\r
+ self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)\r
+ self.skipWhitespace = True\r
+ self.initExprGroups = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if self.initExprGroups:\r
+ self.opt1map = dict((id(e.expr),e) for e in self.exprs if isinstance(e,Optional))\r
+ opt1 = [ e.expr for e in self.exprs if isinstance(e,Optional) ]\r
+ opt2 = [ e for e in self.exprs if e.mayReturnEmpty and not isinstance(e,Optional)]\r
+ self.optionals = opt1 + opt2\r
+ self.multioptionals = [ e.expr for e in self.exprs if isinstance(e,ZeroOrMore) ]\r
+ self.multirequired = [ e.expr for e in self.exprs if isinstance(e,OneOrMore) ]\r
+ self.required = [ e for e in self.exprs if not isinstance(e,(Optional,ZeroOrMore,OneOrMore)) ]\r
+ self.required += self.multirequired\r
+ self.initExprGroups = False\r
+ tmpLoc = loc\r
+ tmpReqd = self.required[:]\r
+ tmpOpt = self.optionals[:]\r
+ matchOrder = []\r
+\r
+ keepMatching = True\r
+ while keepMatching:\r
+ tmpExprs = tmpReqd + tmpOpt + self.multioptionals + self.multirequired\r
+ failed = []\r
+ for e in tmpExprs:\r
+ try:\r
+ tmpLoc = e.tryParse( instring, tmpLoc )\r
+ except ParseException:\r
+ failed.append(e)\r
+ else:\r
+ matchOrder.append(self.opt1map.get(id(e),e))\r
+ if e in tmpReqd:\r
+ tmpReqd.remove(e)\r
+ elif e in tmpOpt:\r
+ tmpOpt.remove(e)\r
+ if len(failed) == len(tmpExprs):\r
+ keepMatching = False\r
+\r
+ if tmpReqd:\r
+ missing = ", ".join(_ustr(e) for e in tmpReqd)\r
+ raise ParseException(instring,loc,"Missing one or more required elements (%s)" % missing )\r
+\r
+ # add any unmatched Optionals, in case they have default values defined\r
+ matchOrder += [e for e in self.exprs if isinstance(e,Optional) and e.expr in tmpOpt]\r
+\r
+ resultlist = []\r
+ for e in matchOrder:\r
+ loc,results = e._parse(instring,loc,doActions)\r
+ resultlist.append(results)\r
+\r
+ finalResults = sum(resultlist, ParseResults([]))\r
+ return loc, finalResults\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "{" + " & ".join(_ustr(e) for e in self.exprs) + "}"\r
+\r
+ return self.strRepr\r
+\r
+ def checkRecursion( self, parseElementList ):\r
+ subRecCheckList = parseElementList[:] + [ self ]\r
+ for e in self.exprs:\r
+ e.checkRecursion( subRecCheckList )\r
+\r
+\r
+class ParseElementEnhance(ParserElement):\r
+ """\r
+ Abstract subclass of C{ParserElement}, for combining and post-processing parsed tokens.\r
+ """\r
+ def __init__( self, expr, savelist=False ):\r
+ super(ParseElementEnhance,self).__init__(savelist)\r
+ if isinstance( expr, basestring ):\r
+ if issubclass(ParserElement._literalStringClass, Token):\r
+ expr = ParserElement._literalStringClass(expr)\r
+ else:\r
+ expr = ParserElement._literalStringClass(Literal(expr))\r
+ self.expr = expr\r
+ self.strRepr = None\r
+ if expr is not None:\r
+ self.mayIndexError = expr.mayIndexError\r
+ self.mayReturnEmpty = expr.mayReturnEmpty\r
+ self.setWhitespaceChars( expr.whiteChars )\r
+ self.skipWhitespace = expr.skipWhitespace\r
+ self.saveAsList = expr.saveAsList\r
+ self.callPreparse = expr.callPreparse\r
+ self.ignoreExprs.extend(expr.ignoreExprs)\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if self.expr is not None:\r
+ return self.expr._parse( instring, loc, doActions, callPreParse=False )\r
+ else:\r
+ raise ParseException("",loc,self.errmsg,self)\r
+\r
+ def leaveWhitespace( self ):\r
+ self.skipWhitespace = False\r
+ self.expr = self.expr.copy()\r
+ if self.expr is not None:\r
+ self.expr.leaveWhitespace()\r
+ return self\r
+\r
+ def ignore( self, other ):\r
+ if isinstance( other, Suppress ):\r
+ if other not in self.ignoreExprs:\r
+ super( ParseElementEnhance, self).ignore( other )\r
+ if self.expr is not None:\r
+ self.expr.ignore( self.ignoreExprs[-1] )\r
+ else:\r
+ super( ParseElementEnhance, self).ignore( other )\r
+ if self.expr is not None:\r
+ self.expr.ignore( self.ignoreExprs[-1] )\r
+ return self\r
+\r
+ def streamline( self ):\r
+ super(ParseElementEnhance,self).streamline()\r
+ if self.expr is not None:\r
+ self.expr.streamline()\r
+ return self\r
+\r
+ def checkRecursion( self, parseElementList ):\r
+ if self in parseElementList:\r
+ raise RecursiveGrammarException( parseElementList+[self] )\r
+ subRecCheckList = parseElementList[:] + [ self ]\r
+ if self.expr is not None:\r
+ self.expr.checkRecursion( subRecCheckList )\r
+\r
+ def validate( self, validateTrace=[] ):\r
+ tmp = validateTrace[:]+[self]\r
+ if self.expr is not None:\r
+ self.expr.validate(tmp)\r
+ self.checkRecursion( [] )\r
+\r
+ def __str__( self ):\r
+ try:\r
+ return super(ParseElementEnhance,self).__str__()\r
+ except Exception:\r
+ pass\r
+\r
+ if self.strRepr is None and self.expr is not None:\r
+ self.strRepr = "%s:(%s)" % ( self.__class__.__name__, _ustr(self.expr) )\r
+ return self.strRepr\r
+\r
+\r
+class FollowedBy(ParseElementEnhance):\r
+ """\r
+ Lookahead matching of the given parse expression. C{FollowedBy}\r
+ does I{not} advance the parsing position within the input string, it only\r
+ verifies that the specified parse expression matches at the current\r
+ position. C{FollowedBy} always returns a null token list.\r
+\r
+ Example::\r
+ # use FollowedBy to match a label only if it is followed by a ':'\r
+ data_word = Word(alphas)\r
+ label = data_word + FollowedBy(':')\r
+ attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join))\r
+ \r
+ OneOrMore(attr_expr).parseString("shape: SQUARE color: BLACK posn: upper left").pprint()\r
+ prints::\r
+ [['shape', 'SQUARE'], ['color', 'BLACK'], ['posn', 'upper left']]\r
+ """\r
+ def __init__( self, expr ):\r
+ super(FollowedBy,self).__init__(expr)\r
+ self.mayReturnEmpty = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ self.expr.tryParse( instring, loc )\r
+ return loc, []\r
+\r
+\r
+class NotAny(ParseElementEnhance):\r
+ """\r
+ Lookahead to disallow matching with the given parse expression. C{NotAny}\r
+ does I{not} advance the parsing position within the input string, it only\r
+ verifies that the specified parse expression does I{not} match at the current\r
+ position. Also, C{NotAny} does I{not} skip over leading whitespace. C{NotAny}\r
+ always returns a null token list. May be constructed using the '~' operator.\r
+\r
+ Example::\r
+ \r
+ """\r
+ def __init__( self, expr ):\r
+ super(NotAny,self).__init__(expr)\r
+ #~ self.leaveWhitespace()\r
+ self.skipWhitespace = False # do NOT use self.leaveWhitespace(), don't want to propagate to exprs\r
+ self.mayReturnEmpty = True\r
+ self.errmsg = "Found unwanted token, "+_ustr(self.expr)\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ if self.expr.canParseNext(instring, loc):\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+ return loc, []\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "~{" + _ustr(self.expr) + "}"\r
+\r
+ return self.strRepr\r
+\r
+class _MultipleMatch(ParseElementEnhance):\r
+ def __init__( self, expr, stopOn=None):\r
+ super(_MultipleMatch, self).__init__(expr)\r
+ self.saveAsList = True\r
+ ender = stopOn\r
+ if isinstance(ender, basestring):\r
+ ender = ParserElement._literalStringClass(ender)\r
+ self.not_ender = ~ender if ender is not None else None\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ self_expr_parse = self.expr._parse\r
+ self_skip_ignorables = self._skipIgnorables\r
+ check_ender = self.not_ender is not None\r
+ if check_ender:\r
+ try_not_ender = self.not_ender.tryParse\r
+ \r
+ # must be at least one (but first see if we are the stopOn sentinel;\r
+ # if so, fail)\r
+ if check_ender:\r
+ try_not_ender(instring, loc)\r
+ loc, tokens = self_expr_parse( instring, loc, doActions, callPreParse=False )\r
+ try:\r
+ hasIgnoreExprs = (not not self.ignoreExprs)\r
+ while 1:\r
+ if check_ender:\r
+ try_not_ender(instring, loc)\r
+ if hasIgnoreExprs:\r
+ preloc = self_skip_ignorables( instring, loc )\r
+ else:\r
+ preloc = loc\r
+ loc, tmptokens = self_expr_parse( instring, preloc, doActions )\r
+ if tmptokens or tmptokens.haskeys():\r
+ tokens += tmptokens\r
+ except (ParseException,IndexError):\r
+ pass\r
+\r
+ return loc, tokens\r
+ \r
+class OneOrMore(_MultipleMatch):\r
+ """\r
+ Repetition of one or more of the given expression.\r
+ \r
+ Parameters:\r
+ - expr - expression that must match one or more times\r
+ - stopOn - (default=C{None}) - expression for a terminating sentinel\r
+ (only required if the sentinel would ordinarily match the repetition \r
+ expression) \r
+\r
+ Example::\r
+ data_word = Word(alphas)\r
+ label = data_word + FollowedBy(':')\r
+ attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).setParseAction(' '.join))\r
+\r
+ text = "shape: SQUARE posn: upper left color: BLACK"\r
+ OneOrMore(attr_expr).parseString(text).pprint() # Fail! read 'color' as data instead of next label -> [['shape', 'SQUARE color']]\r
+\r
+ # use stopOn attribute for OneOrMore to avoid reading label string as part of the data\r
+ attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join))\r
+ OneOrMore(attr_expr).parseString(text).pprint() # Better -> [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'BLACK']]\r
+ \r
+ # could also be written as\r
+ (attr_expr * (1,)).parseString(text).pprint()\r
+ """\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "{" + _ustr(self.expr) + "}..."\r
+\r
+ return self.strRepr\r
+\r
+class ZeroOrMore(_MultipleMatch):\r
+ """\r
+ Optional repetition of zero or more of the given expression.\r
+ \r
+ Parameters:\r
+ - expr - expression that must match zero or more times\r
+ - stopOn - (default=C{None}) - expression for a terminating sentinel\r
+ (only required if the sentinel would ordinarily match the repetition \r
+ expression) \r
+\r
+ Example: similar to L{OneOrMore}\r
+ """\r
+ def __init__( self, expr, stopOn=None):\r
+ super(ZeroOrMore,self).__init__(expr, stopOn=stopOn)\r
+ self.mayReturnEmpty = True\r
+ \r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ try:\r
+ return super(ZeroOrMore, self).parseImpl(instring, loc, doActions)\r
+ except (ParseException,IndexError):\r
+ return loc, []\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "[" + _ustr(self.expr) + "]..."\r
+\r
+ return self.strRepr\r
+\r
+class _NullToken(object):\r
+ def __bool__(self):\r
+ return False\r
+ __nonzero__ = __bool__\r
+ def __str__(self):\r
+ return ""\r
+\r
+_optionalNotMatched = _NullToken()\r
+class Optional(ParseElementEnhance):\r
+ """\r
+ Optional matching of the given expression.\r
+\r
+ Parameters:\r
+ - expr - expression that must match zero or more times\r
+ - default (optional) - value to be returned if the optional expression is not found.\r
+\r
+ Example::\r
+ # US postal code can be a 5-digit zip, plus optional 4-digit qualifier\r
+ zip = Combine(Word(nums, exact=5) + Optional('-' + Word(nums, exact=4)))\r
+ zip.runTests('''\r
+ # traditional ZIP code\r
+ 12345\r
+ \r
+ # ZIP+4 form\r
+ 12101-0001\r
+ \r
+ # invalid ZIP\r
+ 98765-\r
+ ''')\r
+ prints::\r
+ # traditional ZIP code\r
+ 12345\r
+ ['12345']\r
+\r
+ # ZIP+4 form\r
+ 12101-0001\r
+ ['12101-0001']\r
+\r
+ # invalid ZIP\r
+ 98765-\r
+ ^\r
+ FAIL: Expected end of text (at char 5), (line:1, col:6)\r
+ """\r
+ def __init__( self, expr, default=_optionalNotMatched ):\r
+ super(Optional,self).__init__( expr, savelist=False )\r
+ self.saveAsList = self.expr.saveAsList\r
+ self.defaultValue = default\r
+ self.mayReturnEmpty = True\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ try:\r
+ loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False )\r
+ except (ParseException,IndexError):\r
+ if self.defaultValue is not _optionalNotMatched:\r
+ if self.expr.resultsName:\r
+ tokens = ParseResults([ self.defaultValue ])\r
+ tokens[self.expr.resultsName] = self.defaultValue\r
+ else:\r
+ tokens = [ self.defaultValue ]\r
+ else:\r
+ tokens = []\r
+ return loc, tokens\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+\r
+ if self.strRepr is None:\r
+ self.strRepr = "[" + _ustr(self.expr) + "]"\r
+\r
+ return self.strRepr\r
+\r
+class SkipTo(ParseElementEnhance):\r
+ """\r
+ Token for skipping over all undefined text until the matched expression is found.\r
+\r
+ Parameters:\r
+ - expr - target expression marking the end of the data to be skipped\r
+ - include - (default=C{False}) if True, the target expression is also parsed \r
+ (the skipped text and target expression are returned as a 2-element list).\r
+ - ignore - (default=C{None}) used to define grammars (typically quoted strings and \r
+ comments) that might contain false matches to the target expression\r
+ - failOn - (default=C{None}) define expressions that are not allowed to be \r
+ included in the skipped test; if found before the target expression is found, \r
+ the SkipTo is not a match\r
+\r
+ Example::\r
+ report = '''\r
+ Outstanding Issues Report - 1 Jan 2000\r
+\r
+ # | Severity | Description | Days Open\r
+ -----+----------+-------------------------------------------+-----------\r
+ 101 | Critical | Intermittent system crash | 6\r
+ 94 | Cosmetic | Spelling error on Login ('log|n') | 14\r
+ 79 | Minor | System slow when running too many reports | 47\r
+ '''\r
+ integer = Word(nums)\r
+ SEP = Suppress('|')\r
+ # use SkipTo to simply match everything up until the next SEP\r
+ # - ignore quoted strings, so that a '|' character inside a quoted string does not match\r
+ # - parse action will call token.strip() for each matched token, i.e., the description body\r
+ string_data = SkipTo(SEP, ignore=quotedString)\r
+ string_data.setParseAction(tokenMap(str.strip))\r
+ ticket_expr = (integer("issue_num") + SEP \r
+ + string_data("sev") + SEP \r
+ + string_data("desc") + SEP \r
+ + integer("days_open"))\r
+ \r
+ for tkt in ticket_expr.searchString(report):\r
+ print tkt.dump()\r
+ prints::\r
+ ['101', 'Critical', 'Intermittent system crash', '6']\r
+ - days_open: 6\r
+ - desc: Intermittent system crash\r
+ - issue_num: 101\r
+ - sev: Critical\r
+ ['94', 'Cosmetic', "Spelling error on Login ('log|n')", '14']\r
+ - days_open: 14\r
+ - desc: Spelling error on Login ('log|n')\r
+ - issue_num: 94\r
+ - sev: Cosmetic\r
+ ['79', 'Minor', 'System slow when running too many reports', '47']\r
+ - days_open: 47\r
+ - desc: System slow when running too many reports\r
+ - issue_num: 79\r
+ - sev: Minor\r
+ """\r
+ def __init__( self, other, include=False, ignore=None, failOn=None ):\r
+ super( SkipTo, self ).__init__( other )\r
+ self.ignoreExpr = ignore\r
+ self.mayReturnEmpty = True\r
+ self.mayIndexError = False\r
+ self.includeMatch = include\r
+ self.asList = False\r
+ if isinstance(failOn, basestring):\r
+ self.failOn = ParserElement._literalStringClass(failOn)\r
+ else:\r
+ self.failOn = failOn\r
+ self.errmsg = "No match found for "+_ustr(self.expr)\r
+\r
+ def parseImpl( self, instring, loc, doActions=True ):\r
+ startloc = loc\r
+ instrlen = len(instring)\r
+ expr = self.expr\r
+ expr_parse = self.expr._parse\r
+ self_failOn_canParseNext = self.failOn.canParseNext if self.failOn is not None else None\r
+ self_ignoreExpr_tryParse = self.ignoreExpr.tryParse if self.ignoreExpr is not None else None\r
+ \r
+ tmploc = loc\r
+ while tmploc <= instrlen:\r
+ if self_failOn_canParseNext is not None:\r
+ # break if failOn expression matches\r
+ if self_failOn_canParseNext(instring, tmploc):\r
+ break\r
+ \r
+ if self_ignoreExpr_tryParse is not None:\r
+ # advance past ignore expressions\r
+ while 1:\r
+ try:\r
+ tmploc = self_ignoreExpr_tryParse(instring, tmploc)\r
+ except ParseBaseException:\r
+ break\r
+ \r
+ try:\r
+ expr_parse(instring, tmploc, doActions=False, callPreParse=False)\r
+ except (ParseException, IndexError):\r
+ # no match, advance loc in string\r
+ tmploc += 1\r
+ else:\r
+ # matched skipto expr, done\r
+ break\r
+\r
+ else:\r
+ # ran off the end of the input string without matching skipto expr, fail\r
+ raise ParseException(instring, loc, self.errmsg, self)\r
+\r
+ # build up return values\r
+ loc = tmploc\r
+ skiptext = instring[startloc:loc]\r
+ skipresult = ParseResults(skiptext)\r
+ \r
+ if self.includeMatch:\r
+ loc, mat = expr_parse(instring,loc,doActions,callPreParse=False)\r
+ skipresult += mat\r
+\r
+ return loc, skipresult\r
+\r
+class Forward(ParseElementEnhance):\r
+ """\r
+ Forward declaration of an expression to be defined later -\r
+ used for recursive grammars, such as algebraic infix notation.\r
+ When the expression is known, it is assigned to the C{Forward} variable using the '<<' operator.\r
+\r
+ Note: take care when assigning to C{Forward} not to overlook precedence of operators.\r
+ Specifically, '|' has a lower precedence than '<<', so that::\r
+ fwdExpr << a | b | c\r
+ will actually be evaluated as::\r
+ (fwdExpr << a) | b | c\r
+ thereby leaving b and c out as parseable alternatives. It is recommended that you\r
+ explicitly group the values inserted into the C{Forward}::\r
+ fwdExpr << (a | b | c)\r
+ Converting to use the '<<=' operator instead will avoid this problem.\r
+\r
+ See L{ParseResults.pprint} for an example of a recursive parser created using\r
+ C{Forward}.\r
+ """\r
+ def __init__( self, other=None ):\r
+ super(Forward,self).__init__( other, savelist=False )\r
+\r
+ def __lshift__( self, other ):\r
+ if isinstance( other, basestring ):\r
+ other = ParserElement._literalStringClass(other)\r
+ self.expr = other\r
+ self.strRepr = None\r
+ self.mayIndexError = self.expr.mayIndexError\r
+ self.mayReturnEmpty = self.expr.mayReturnEmpty\r
+ self.setWhitespaceChars( self.expr.whiteChars )\r
+ self.skipWhitespace = self.expr.skipWhitespace\r
+ self.saveAsList = self.expr.saveAsList\r
+ self.ignoreExprs.extend(self.expr.ignoreExprs)\r
+ return self\r
+ \r
+ def __ilshift__(self, other):\r
+ return self << other\r
+ \r
+ def leaveWhitespace( self ):\r
+ self.skipWhitespace = False\r
+ return self\r
+\r
+ def streamline( self ):\r
+ if not self.streamlined:\r
+ self.streamlined = True\r
+ if self.expr is not None:\r
+ self.expr.streamline()\r
+ return self\r
+\r
+ def validate( self, validateTrace=[] ):\r
+ if self not in validateTrace:\r
+ tmp = validateTrace[:]+[self]\r
+ if self.expr is not None:\r
+ self.expr.validate(tmp)\r
+ self.checkRecursion([])\r
+\r
+ def __str__( self ):\r
+ if hasattr(self,"name"):\r
+ return self.name\r
+ return self.__class__.__name__ + ": ..."\r
+\r
+ # stubbed out for now - creates awful memory and perf issues\r
+ self._revertClass = self.__class__\r
+ self.__class__ = _ForwardNoRecurse\r
+ try:\r
+ if self.expr is not None:\r
+ retString = _ustr(self.expr)\r
+ else:\r
+ retString = "None"\r
+ finally:\r
+ self.__class__ = self._revertClass\r
+ return self.__class__.__name__ + ": " + retString\r
+\r
+ def copy(self):\r
+ if self.expr is not None:\r
+ return super(Forward,self).copy()\r
+ else:\r
+ ret = Forward()\r
+ ret <<= self\r
+ return ret\r
+\r
+class _ForwardNoRecurse(Forward):\r
+ def __str__( self ):\r
+ return "..."\r
+\r
+class TokenConverter(ParseElementEnhance):\r
+ """\r
+ Abstract subclass of C{ParseExpression}, for converting parsed results.\r
+ """\r
+ def __init__( self, expr, savelist=False ):\r
+ super(TokenConverter,self).__init__( expr )#, savelist )\r
+ self.saveAsList = False\r
+\r
+class Combine(TokenConverter):\r
+ """\r
+ Converter to concatenate all matching tokens to a single string.\r
+ By default, the matching patterns must also be contiguous in the input string;\r
+ this can be disabled by specifying C{'adjacent=False'} in the constructor.\r
+\r
+ Example::\r
+ real = Word(nums) + '.' + Word(nums)\r
+ print(real.parseString('3.1416')) # -> ['3', '.', '1416']\r
+ # will also erroneously match the following\r
+ print(real.parseString('3. 1416')) # -> ['3', '.', '1416']\r
+\r
+ real = Combine(Word(nums) + '.' + Word(nums))\r
+ print(real.parseString('3.1416')) # -> ['3.1416']\r
+ # no match when there are internal spaces\r
+ print(real.parseString('3. 1416')) # -> Exception: Expected W:(0123...)\r
+ """\r
+ def __init__( self, expr, joinString="", adjacent=True ):\r
+ super(Combine,self).__init__( expr )\r
+ # suppress whitespace-stripping in contained parse expressions, but re-enable it on the Combine itself\r
+ if adjacent:\r
+ self.leaveWhitespace()\r
+ self.adjacent = adjacent\r
+ self.skipWhitespace = True\r
+ self.joinString = joinString\r
+ self.callPreparse = True\r
+\r
+ def ignore( self, other ):\r
+ if self.adjacent:\r
+ ParserElement.ignore(self, other)\r
+ else:\r
+ super( Combine, self).ignore( other )\r
+ return self\r
+\r
+ def postParse( self, instring, loc, tokenlist ):\r
+ retToks = tokenlist.copy()\r
+ del retToks[:]\r
+ retToks += ParseResults([ "".join(tokenlist._asStringList(self.joinString)) ], modal=self.modalResults)\r
+\r
+ if self.resultsName and retToks.haskeys():\r
+ return [ retToks ]\r
+ else:\r
+ return retToks\r
+\r
+class Group(TokenConverter):\r
+ """\r
+ Converter to return the matched tokens as a list - useful for returning tokens of C{L{ZeroOrMore}} and C{L{OneOrMore}} expressions.\r
+\r
+ Example::\r
+ ident = Word(alphas)\r
+ num = Word(nums)\r
+ term = ident | num\r
+ func = ident + Optional(delimitedList(term))\r
+ print(func.parseString("fn a,b,100")) # -> ['fn', 'a', 'b', '100']\r
+\r
+ func = ident + Group(Optional(delimitedList(term)))\r
+ print(func.parseString("fn a,b,100")) # -> ['fn', ['a', 'b', '100']]\r
+ """\r
+ def __init__( self, expr ):\r
+ super(Group,self).__init__( expr )\r
+ self.saveAsList = True\r
+\r
+ def postParse( self, instring, loc, tokenlist ):\r
+ return [ tokenlist ]\r
+\r
+class Dict(TokenConverter):\r
+ """\r
+ Converter to return a repetitive expression as a list, but also as a dictionary.\r
+ Each element can also be referenced using the first token in the expression as its key.\r
+ Useful for tabular report scraping when the first column can be used as a item key.\r
+\r
+ Example::\r
+ data_word = Word(alphas)\r
+ label = data_word + FollowedBy(':')\r
+ attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).setParseAction(' '.join))\r
+\r
+ text = "shape: SQUARE posn: upper left color: light blue texture: burlap"\r
+ attr_expr = (label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join))\r
+ \r
+ # print attributes as plain groups\r
+ print(OneOrMore(attr_expr).parseString(text).dump())\r
+ \r
+ # instead of OneOrMore(expr), parse using Dict(OneOrMore(Group(expr))) - Dict will auto-assign names\r
+ result = Dict(OneOrMore(Group(attr_expr))).parseString(text)\r
+ print(result.dump())\r
+ \r
+ # access named fields as dict entries, or output as dict\r
+ print(result['shape']) \r
+ print(result.asDict())\r
+ prints::\r
+ ['shape', 'SQUARE', 'posn', 'upper left', 'color', 'light blue', 'texture', 'burlap']\r
+\r
+ [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']]\r
+ - color: light blue\r
+ - posn: upper left\r
+ - shape: SQUARE\r
+ - texture: burlap\r
+ SQUARE\r
+ {'color': 'light blue', 'posn': 'upper left', 'texture': 'burlap', 'shape': 'SQUARE'}\r
+ See more examples at L{ParseResults} of accessing fields by results name.\r
+ """\r
+ def __init__( self, expr ):\r
+ super(Dict,self).__init__( expr )\r
+ self.saveAsList = True\r
+\r
+ def postParse( self, instring, loc, tokenlist ):\r
+ for i,tok in enumerate(tokenlist):\r
+ if len(tok) == 0:\r
+ continue\r
+ ikey = tok[0]\r
+ if isinstance(ikey,int):\r
+ ikey = _ustr(tok[0]).strip()\r
+ if len(tok)==1:\r
+ tokenlist[ikey] = _ParseResultsWithOffset("",i)\r
+ elif len(tok)==2 and not isinstance(tok[1],ParseResults):\r
+ tokenlist[ikey] = _ParseResultsWithOffset(tok[1],i)\r
+ else:\r
+ dictvalue = tok.copy() #ParseResults(i)\r
+ del dictvalue[0]\r
+ if len(dictvalue)!= 1 or (isinstance(dictvalue,ParseResults) and dictvalue.haskeys()):\r
+ tokenlist[ikey] = _ParseResultsWithOffset(dictvalue,i)\r
+ else:\r
+ tokenlist[ikey] = _ParseResultsWithOffset(dictvalue[0],i)\r
+\r
+ if self.resultsName:\r
+ return [ tokenlist ]\r
+ else:\r
+ return tokenlist\r
+\r
+\r
+class Suppress(TokenConverter):\r
+ """\r
+ Converter for ignoring the results of a parsed expression.\r
+\r
+ Example::\r
+ source = "a, b, c,d"\r
+ wd = Word(alphas)\r
+ wd_list1 = wd + ZeroOrMore(',' + wd)\r
+ print(wd_list1.parseString(source))\r
+\r
+ # often, delimiters that are useful during parsing are just in the\r
+ # way afterward - use Suppress to keep them out of the parsed output\r
+ wd_list2 = wd + ZeroOrMore(Suppress(',') + wd)\r
+ print(wd_list2.parseString(source))\r
+ prints::\r
+ ['a', ',', 'b', ',', 'c', ',', 'd']\r
+ ['a', 'b', 'c', 'd']\r
+ (See also L{delimitedList}.)\r
+ """\r
+ def postParse( self, instring, loc, tokenlist ):\r
+ return []\r
+\r
+ def suppress( self ):\r
+ return self\r
+\r
+\r
+class OnlyOnce(object):\r
+ """\r
+ Wrapper for parse actions, to ensure they are only called once.\r
+ """\r
+ def __init__(self, methodCall):\r
+ self.callable = _trim_arity(methodCall)\r
+ self.called = False\r
+ def __call__(self,s,l,t):\r
+ if not self.called:\r
+ results = self.callable(s,l,t)\r
+ self.called = True\r
+ return results\r
+ raise ParseException(s,l,"")\r
+ def reset(self):\r
+ self.called = False\r
+\r
+def traceParseAction(f):\r
+ """\r
+ Decorator for debugging parse actions. \r
+ \r
+ When the parse action is called, this decorator will print C{">> entering I{method-name}(line:I{current_source_line}, I{parse_location}, I{matched_tokens})".}\r
+ When the parse action completes, the decorator will print C{"<<"} followed by the returned value, or any exception that the parse action raised.\r
+\r
+ Example::\r
+ wd = Word(alphas)\r
+\r
+ @traceParseAction\r
+ def remove_duplicate_chars(tokens):\r
+ return ''.join(sorted(set(''.join(tokens)))\r
+\r
+ wds = OneOrMore(wd).setParseAction(remove_duplicate_chars)\r
+ print(wds.parseString("slkdjs sld sldd sdlf sdljf"))\r
+ prints::\r
+ >>entering remove_duplicate_chars(line: 'slkdjs sld sldd sdlf sdljf', 0, (['slkdjs', 'sld', 'sldd', 'sdlf', 'sdljf'], {}))\r
+ <<leaving remove_duplicate_chars (ret: 'dfjkls')\r
+ ['dfjkls']\r
+ """\r
+ f = _trim_arity(f)\r
+ def z(*paArgs):\r
+ thisFunc = f.__name__\r
+ s,l,t = paArgs[-3:]\r
+ if len(paArgs)>3:\r
+ thisFunc = paArgs[0].__class__.__name__ + '.' + thisFunc\r
+ sys.stderr.write( ">>entering %s(line: '%s', %d, %r)\n" % (thisFunc,line(l,s),l,t) )\r
+ try:\r
+ ret = f(*paArgs)\r
+ except Exception as exc:\r
+ sys.stderr.write( "<<leaving %s (exception: %s)\n" % (thisFunc,exc) )\r
+ raise\r
+ sys.stderr.write( "<<leaving %s (ret: %r)\n" % (thisFunc,ret) )\r
+ return ret\r
+ try:\r
+ z.__name__ = f.__name__\r
+ except AttributeError:\r
+ pass\r
+ return z\r
+\r
+#\r
+# global helpers\r
+#\r
+def delimitedList( expr, delim=",", combine=False ):\r
+ """\r
+ Helper to define a delimited list of expressions - the delimiter defaults to ','.\r
+ By default, the list elements and delimiters can have intervening whitespace, and\r
+ comments, but this can be overridden by passing C{combine=True} in the constructor.\r
+ If C{combine} is set to C{True}, the matching tokens are returned as a single token\r
+ string, with the delimiters included; otherwise, the matching tokens are returned\r
+ as a list of tokens, with the delimiters suppressed.\r
+\r
+ Example::\r
+ delimitedList(Word(alphas)).parseString("aa,bb,cc") # -> ['aa', 'bb', 'cc']\r
+ delimitedList(Word(hexnums), delim=':', combine=True).parseString("AA:BB:CC:DD:EE") # -> ['AA:BB:CC:DD:EE']\r
+ """\r
+ dlName = _ustr(expr)+" ["+_ustr(delim)+" "+_ustr(expr)+"]..."\r
+ if combine:\r
+ return Combine( expr + ZeroOrMore( delim + expr ) ).setName(dlName)\r
+ else:\r
+ return ( expr + ZeroOrMore( Suppress( delim ) + expr ) ).setName(dlName)\r
+\r
+def countedArray( expr, intExpr=None ):\r
+ """\r
+ Helper to define a counted list of expressions.\r
+ This helper defines a pattern of the form::\r
+ integer expr expr expr...\r
+ where the leading integer tells how many expr expressions follow.\r
+ The matched tokens returns the array of expr tokens as a list - the leading count token is suppressed.\r
+ \r
+ If C{intExpr} is specified, it should be a pyparsing expression that produces an integer value.\r
+\r
+ Example::\r
+ countedArray(Word(alphas)).parseString('2 ab cd ef') # -> ['ab', 'cd']\r
+\r
+ # in this parser, the leading integer value is given in binary,\r
+ # '10' indicating that 2 values are in the array\r
+ binaryConstant = Word('01').setParseAction(lambda t: int(t[0], 2))\r
+ countedArray(Word(alphas), intExpr=binaryConstant).parseString('10 ab cd ef') # -> ['ab', 'cd']\r
+ """\r
+ arrayExpr = Forward()\r
+ def countFieldParseAction(s,l,t):\r
+ n = t[0]\r
+ arrayExpr << (n and Group(And([expr]*n)) or Group(empty))\r
+ return []\r
+ if intExpr is None:\r
+ intExpr = Word(nums).setParseAction(lambda t:int(t[0]))\r
+ else:\r
+ intExpr = intExpr.copy()\r
+ intExpr.setName("arrayLen")\r
+ intExpr.addParseAction(countFieldParseAction, callDuringTry=True)\r
+ return ( intExpr + arrayExpr ).setName('(len) ' + _ustr(expr) + '...')\r
+\r
+def _flatten(L):\r
+ ret = []\r
+ for i in L:\r
+ if isinstance(i,list):\r
+ ret.extend(_flatten(i))\r
+ else:\r
+ ret.append(i)\r
+ return ret\r
+\r
+def matchPreviousLiteral(expr):\r
+ """\r
+ Helper to define an expression that is indirectly defined from\r
+ the tokens matched in a previous expression, that is, it looks\r
+ for a 'repeat' of a previous expression. For example::\r
+ first = Word(nums)\r
+ second = matchPreviousLiteral(first)\r
+ matchExpr = first + ":" + second\r
+ will match C{"1:1"}, but not C{"1:2"}. Because this matches a\r
+ previous literal, will also match the leading C{"1:1"} in C{"1:10"}.\r
+ If this is not desired, use C{matchPreviousExpr}.\r
+ Do I{not} use with packrat parsing enabled.\r
+ """\r
+ rep = Forward()\r
+ def copyTokenToRepeater(s,l,t):\r
+ if t:\r
+ if len(t) == 1:\r
+ rep << t[0]\r
+ else:\r
+ # flatten t tokens\r
+ tflat = _flatten(t.asList())\r
+ rep << And(Literal(tt) for tt in tflat)\r
+ else:\r
+ rep << Empty()\r
+ expr.addParseAction(copyTokenToRepeater, callDuringTry=True)\r
+ rep.setName('(prev) ' + _ustr(expr))\r
+ return rep\r
+\r
+def matchPreviousExpr(expr):\r
+ """\r
+ Helper to define an expression that is indirectly defined from\r
+ the tokens matched in a previous expression, that is, it looks\r
+ for a 'repeat' of a previous expression. For example::\r
+ first = Word(nums)\r
+ second = matchPreviousExpr(first)\r
+ matchExpr = first + ":" + second\r
+ will match C{"1:1"}, but not C{"1:2"}. Because this matches by\r
+ expressions, will I{not} match the leading C{"1:1"} in C{"1:10"};\r
+ the expressions are evaluated first, and then compared, so\r
+ C{"1"} is compared with C{"10"}.\r
+ Do I{not} use with packrat parsing enabled.\r
+ """\r
+ rep = Forward()\r
+ e2 = expr.copy()\r
+ rep <<= e2\r
+ def copyTokenToRepeater(s,l,t):\r
+ matchTokens = _flatten(t.asList())\r
+ def mustMatchTheseTokens(s,l,t):\r
+ theseTokens = _flatten(t.asList())\r
+ if theseTokens != matchTokens:\r
+ raise ParseException("",0,"")\r
+ rep.setParseAction( mustMatchTheseTokens, callDuringTry=True )\r
+ expr.addParseAction(copyTokenToRepeater, callDuringTry=True)\r
+ rep.setName('(prev) ' + _ustr(expr))\r
+ return rep\r
+\r
+def _escapeRegexRangeChars(s):\r
+ #~ escape these chars: ^-]\r
+ for c in r"\^-]":\r
+ s = s.replace(c,_bslash+c)\r
+ s = s.replace("\n",r"\n")\r
+ s = s.replace("\t",r"\t")\r
+ return _ustr(s)\r
+\r
+def oneOf( strs, caseless=False, useRegex=True ):\r
+ """\r
+ Helper to quickly define a set of alternative Literals, and makes sure to do\r
+ longest-first testing when there is a conflict, regardless of the input order,\r
+ but returns a C{L{MatchFirst}} for best performance.\r
+\r
+ Parameters:\r
+ - strs - a string of space-delimited literals, or a collection of string literals\r
+ - caseless - (default=C{False}) - treat all literals as caseless\r
+ - useRegex - (default=C{True}) - as an optimization, will generate a Regex\r
+ object; otherwise, will generate a C{MatchFirst} object (if C{caseless=True}, or\r
+ if creating a C{Regex} raises an exception)\r
+\r
+ Example::\r
+ comp_oper = oneOf("< = > <= >= !=")\r
+ var = Word(alphas)\r
+ number = Word(nums)\r
+ term = var | number\r
+ comparison_expr = term + comp_oper + term\r
+ print(comparison_expr.searchString("B = 12 AA=23 B<=AA AA>12"))\r
+ prints::\r
+ [['B', '=', '12'], ['AA', '=', '23'], ['B', '<=', 'AA'], ['AA', '>', '12']]\r
+ """\r
+ if caseless:\r
+ isequal = ( lambda a,b: a.upper() == b.upper() )\r
+ masks = ( lambda a,b: b.upper().startswith(a.upper()) )\r
+ parseElementClass = CaselessLiteral\r
+ else:\r
+ isequal = ( lambda a,b: a == b )\r
+ masks = ( lambda a,b: b.startswith(a) )\r
+ parseElementClass = Literal\r
+\r
+ symbols = []\r
+ if isinstance(strs,basestring):\r
+ symbols = strs.split()\r
+ elif isinstance(strs, collections.Iterable):\r
+ symbols = list(strs)\r
+ else:\r
+ warnings.warn("Invalid argument to oneOf, expected string or iterable",\r
+ SyntaxWarning, stacklevel=2)\r
+ if not symbols:\r
+ return NoMatch()\r
+\r
+ i = 0\r
+ while i < len(symbols)-1:\r
+ cur = symbols[i]\r
+ for j,other in enumerate(symbols[i+1:]):\r
+ if ( isequal(other, cur) ):\r
+ del symbols[i+j+1]\r
+ break\r
+ elif ( masks(cur, other) ):\r
+ del symbols[i+j+1]\r
+ symbols.insert(i,other)\r
+ cur = other\r
+ break\r
+ else:\r
+ i += 1\r
+\r
+ if not caseless and useRegex:\r
+ #~ print (strs,"->", "|".join( [ _escapeRegexChars(sym) for sym in symbols] ))\r
+ try:\r
+ if len(symbols)==len("".join(symbols)):\r
+ return Regex( "[%s]" % "".join(_escapeRegexRangeChars(sym) for sym in symbols) ).setName(' | '.join(symbols))\r
+ else:\r
+ return Regex( "|".join(re.escape(sym) for sym in symbols) ).setName(' | '.join(symbols))\r
+ except Exception:\r
+ warnings.warn("Exception creating Regex for oneOf, building MatchFirst",\r
+ SyntaxWarning, stacklevel=2)\r
+\r
+\r
+ # last resort, just use MatchFirst\r
+ return MatchFirst(parseElementClass(sym) for sym in symbols).setName(' | '.join(symbols))\r
+\r
+def dictOf( key, value ):\r
+ """\r
+ Helper to easily and clearly define a dictionary by specifying the respective patterns\r
+ for the key and value. Takes care of defining the C{L{Dict}}, C{L{ZeroOrMore}}, and C{L{Group}} tokens\r
+ in the proper order. The key pattern can include delimiting markers or punctuation,\r
+ as long as they are suppressed, thereby leaving the significant key text. The value\r
+ pattern can include named results, so that the C{Dict} results can include named token\r
+ fields.\r
+\r
+ Example::\r
+ text = "shape: SQUARE posn: upper left color: light blue texture: burlap"\r
+ attr_expr = (label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join))\r
+ print(OneOrMore(attr_expr).parseString(text).dump())\r
+ \r
+ attr_label = label\r
+ attr_value = Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)\r
+\r
+ # similar to Dict, but simpler call format\r
+ result = dictOf(attr_label, attr_value).parseString(text)\r
+ print(result.dump())\r
+ print(result['shape'])\r
+ print(result.shape) # object attribute access works too\r
+ print(result.asDict())\r
+ prints::\r
+ [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']]\r
+ - color: light blue\r
+ - posn: upper left\r
+ - shape: SQUARE\r
+ - texture: burlap\r
+ SQUARE\r
+ SQUARE\r
+ {'color': 'light blue', 'shape': 'SQUARE', 'posn': 'upper left', 'texture': 'burlap'}\r
+ """\r
+ return Dict( ZeroOrMore( Group ( key + value ) ) )\r
+\r
+def originalTextFor(expr, asString=True):\r
+ """\r
+ Helper to return the original, untokenized text for a given expression. Useful to\r
+ restore the parsed fields of an HTML start tag into the raw tag text itself, or to\r
+ revert separate tokens with intervening whitespace back to the original matching\r
+ input text. By default, returns astring containing the original parsed text. \r
+ \r
+ If the optional C{asString} argument is passed as C{False}, then the return value is a \r
+ C{L{ParseResults}} containing any results names that were originally matched, and a \r
+ single token containing the original matched text from the input string. So if \r
+ the expression passed to C{L{originalTextFor}} contains expressions with defined\r
+ results names, you must set C{asString} to C{False} if you want to preserve those\r
+ results name values.\r
+\r
+ Example::\r
+ src = "this is test <b> bold <i>text</i> </b> normal text "\r
+ for tag in ("b","i"):\r
+ opener,closer = makeHTMLTags(tag)\r
+ patt = originalTextFor(opener + SkipTo(closer) + closer)\r
+ print(patt.searchString(src)[0])\r
+ prints::\r
+ ['<b> bold <i>text</i> </b>']\r
+ ['<i>text</i>']\r
+ """\r
+ locMarker = Empty().setParseAction(lambda s,loc,t: loc)\r
+ endlocMarker = locMarker.copy()\r
+ endlocMarker.callPreparse = False\r
+ matchExpr = locMarker("_original_start") + expr + endlocMarker("_original_end")\r
+ if asString:\r
+ extractText = lambda s,l,t: s[t._original_start:t._original_end]\r
+ else:\r
+ def extractText(s,l,t):\r
+ t[:] = [s[t.pop('_original_start'):t.pop('_original_end')]]\r
+ matchExpr.setParseAction(extractText)\r
+ matchExpr.ignoreExprs = expr.ignoreExprs\r
+ return matchExpr\r
+\r
+def ungroup(expr): \r
+ """\r
+ Helper to undo pyparsing's default grouping of And expressions, even\r
+ if all but one are non-empty.\r
+ """\r
+ return TokenConverter(expr).setParseAction(lambda t:t[0])\r
+\r
+def locatedExpr(expr):\r
+ """\r
+ Helper to decorate a returned token with its starting and ending locations in the input string.\r
+ This helper adds the following results names:\r
+ - locn_start = location where matched expression begins\r
+ - locn_end = location where matched expression ends\r
+ - value = the actual parsed results\r
+\r
+ Be careful if the input text contains C{<TAB>} characters, you may want to call\r
+ C{L{ParserElement.parseWithTabs}}\r
+\r
+ Example::\r
+ wd = Word(alphas)\r
+ for match in locatedExpr(wd).searchString("ljsdf123lksdjjf123lkkjj1222"):\r
+ print(match)\r
+ prints::\r
+ [[0, 'ljsdf', 5]]\r
+ [[8, 'lksdjjf', 15]]\r
+ [[18, 'lkkjj', 23]]\r
+ """\r
+ locator = Empty().setParseAction(lambda s,l,t: l)\r
+ return Group(locator("locn_start") + expr("value") + locator.copy().leaveWhitespace()("locn_end"))\r
+\r
+\r
+# convenience constants for positional expressions\r
+empty = Empty().setName("empty")\r
+lineStart = LineStart().setName("lineStart")\r
+lineEnd = LineEnd().setName("lineEnd")\r
+stringStart = StringStart().setName("stringStart")\r
+stringEnd = StringEnd().setName("stringEnd")\r
+\r
+_escapedPunc = Word( _bslash, r"\[]-*.$+^?()~ ", exact=2 ).setParseAction(lambda s,l,t:t[0][1])\r
+_escapedHexChar = Regex(r"\\0?[xX][0-9a-fA-F]+").setParseAction(lambda s,l,t:unichr(int(t[0].lstrip(r'\0x'),16)))\r
+_escapedOctChar = Regex(r"\\0[0-7]+").setParseAction(lambda s,l,t:unichr(int(t[0][1:],8)))\r
+_singleChar = _escapedPunc | _escapedHexChar | _escapedOctChar | Word(printables, excludeChars=r'\]', exact=1) | Regex(r"\w", re.UNICODE)\r
+_charRange = Group(_singleChar + Suppress("-") + _singleChar)\r
+_reBracketExpr = Literal("[") + Optional("^").setResultsName("negate") + Group( OneOrMore( _charRange | _singleChar ) ).setResultsName("body") + "]"\r
+\r
+def srange(s):\r
+ r"""\r
+ Helper to easily define string ranges for use in Word construction. Borrows\r
+ syntax from regexp '[]' string range definitions::\r
+ srange("[0-9]") -> "0123456789"\r
+ srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz"\r
+ srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_"\r
+ The input string must be enclosed in []'s, and the returned string is the expanded\r
+ character set joined into a single string.\r
+ The values enclosed in the []'s may be:\r
+ - a single character\r
+ - an escaped character with a leading backslash (such as C{\-} or C{\]})\r
+ - an escaped hex character with a leading C{'\x'} (C{\x21}, which is a C{'!'} character) \r
+ (C{\0x##} is also supported for backwards compatibility) \r
+ - an escaped octal character with a leading C{'\0'} (C{\041}, which is a C{'!'} character)\r
+ - a range of any of the above, separated by a dash (C{'a-z'}, etc.)\r
+ - any combination of the above (C{'aeiouy'}, C{'a-zA-Z0-9_$'}, etc.)\r
+ """\r
+ _expanded = lambda p: p if not isinstance(p,ParseResults) else ''.join(unichr(c) for c in range(ord(p[0]),ord(p[1])+1))\r
+ try:\r
+ return "".join(_expanded(part) for part in _reBracketExpr.parseString(s).body)\r
+ except Exception:\r
+ return ""\r
+\r
+def matchOnlyAtCol(n):\r
+ """\r
+ Helper method for defining parse actions that require matching at a specific\r
+ column in the input text.\r
+ """\r
+ def verifyCol(strg,locn,toks):\r
+ if col(locn,strg) != n:\r
+ raise ParseException(strg,locn,"matched token not at column %d" % n)\r
+ return verifyCol\r
+\r
+def replaceWith(replStr):\r
+ """\r
+ Helper method for common parse actions that simply return a literal value. Especially\r
+ useful when used with C{L{transformString<ParserElement.transformString>}()}.\r
+\r
+ Example::\r
+ num = Word(nums).setParseAction(lambda toks: int(toks[0]))\r
+ na = oneOf("N/A NA").setParseAction(replaceWith(math.nan))\r
+ term = na | num\r
+ \r
+ OneOrMore(term).parseString("324 234 N/A 234") # -> [324, 234, nan, 234]\r
+ """\r
+ return lambda s,l,t: [replStr]\r
+\r
+def removeQuotes(s,l,t):\r
+ """\r
+ Helper parse action for removing quotation marks from parsed quoted strings.\r
+\r
+ Example::\r
+ # by default, quotation marks are included in parsed results\r
+ quotedString.parseString("'Now is the Winter of our Discontent'") # -> ["'Now is the Winter of our Discontent'"]\r
+\r
+ # use removeQuotes to strip quotation marks from parsed results\r
+ quotedString.setParseAction(removeQuotes)\r
+ quotedString.parseString("'Now is the Winter of our Discontent'") # -> ["Now is the Winter of our Discontent"]\r
+ """\r
+ return t[0][1:-1]\r
+\r
+def tokenMap(func, *args):\r
+ """\r
+ Helper to define a parse action by mapping a function to all elements of a ParseResults list.If any additional \r
+ args are passed, they are forwarded to the given function as additional arguments after\r
+ the token, as in C{hex_integer = Word(hexnums).setParseAction(tokenMap(int, 16))}, which will convert the\r
+ parsed data to an integer using base 16.\r
+\r
+ Example (compare the last to example in L{ParserElement.transformString}::\r
+ hex_ints = OneOrMore(Word(hexnums)).setParseAction(tokenMap(int, 16))\r
+ hex_ints.runTests('''\r
+ 00 11 22 aa FF 0a 0d 1a\r
+ ''')\r
+ \r
+ upperword = Word(alphas).setParseAction(tokenMap(str.upper))\r
+ OneOrMore(upperword).runTests('''\r
+ my kingdom for a horse\r
+ ''')\r
+\r
+ wd = Word(alphas).setParseAction(tokenMap(str.title))\r
+ OneOrMore(wd).setParseAction(' '.join).runTests('''\r
+ now is the winter of our discontent made glorious summer by this sun of york\r
+ ''')\r
+ prints::\r
+ 00 11 22 aa FF 0a 0d 1a\r
+ [0, 17, 34, 170, 255, 10, 13, 26]\r
+\r
+ my kingdom for a horse\r
+ ['MY', 'KINGDOM', 'FOR', 'A', 'HORSE']\r
+\r
+ now is the winter of our discontent made glorious summer by this sun of york\r
+ ['Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York']\r
+ """\r
+ def pa(s,l,t):\r
+ return [func(tokn, *args) for tokn in t]\r
+\r
+ try:\r
+ func_name = getattr(func, '__name__', \r
+ getattr(func, '__class__').__name__)\r
+ except Exception:\r
+ func_name = str(func)\r
+ pa.__name__ = func_name\r
+\r
+ return pa\r
+\r
+upcaseTokens = tokenMap(lambda t: _ustr(t).upper())\r
+"""(Deprecated) Helper parse action to convert tokens to upper case. Deprecated in favor of L{pyparsing_common.upcaseTokens}"""\r
+\r
+downcaseTokens = tokenMap(lambda t: _ustr(t).lower())\r
+"""(Deprecated) Helper parse action to convert tokens to lower case. Deprecated in favor of L{pyparsing_common.downcaseTokens}"""\r
+ \r
+def _makeTags(tagStr, xml):\r
+ """Internal helper to construct opening and closing tag expressions, given a tag name"""\r
+ if isinstance(tagStr,basestring):\r
+ resname = tagStr\r
+ tagStr = Keyword(tagStr, caseless=not xml)\r
+ else:\r
+ resname = tagStr.name\r
+\r
+ tagAttrName = Word(alphas,alphanums+"_-:")\r
+ if (xml):\r
+ tagAttrValue = dblQuotedString.copy().setParseAction( removeQuotes )\r
+ openTag = Suppress("<") + tagStr("tag") + \\r
+ Dict(ZeroOrMore(Group( tagAttrName + Suppress("=") + tagAttrValue ))) + \\r
+ Optional("/",default=[False]).setResultsName("empty").setParseAction(lambda s,l,t:t[0]=='/') + Suppress(">")\r
+ else:\r
+ printablesLessRAbrack = "".join(c for c in printables if c not in ">")\r
+ tagAttrValue = quotedString.copy().setParseAction( removeQuotes ) | Word(printablesLessRAbrack)\r
+ openTag = Suppress("<") + tagStr("tag") + \\r
+ Dict(ZeroOrMore(Group( tagAttrName.setParseAction(downcaseTokens) + \\r
+ Optional( Suppress("=") + tagAttrValue ) ))) + \\r
+ Optional("/",default=[False]).setResultsName("empty").setParseAction(lambda s,l,t:t[0]=='/') + Suppress(">")\r
+ closeTag = Combine(_L("</") + tagStr + ">")\r
+\r
+ openTag = openTag.setResultsName("start"+"".join(resname.replace(":"," ").title().split())).setName("<%s>" % resname)\r
+ closeTag = closeTag.setResultsName("end"+"".join(resname.replace(":"," ").title().split())).setName("</%s>" % resname)\r
+ openTag.tag = resname\r
+ closeTag.tag = resname\r
+ return openTag, closeTag\r
+\r
+def makeHTMLTags(tagStr):\r
+ """\r
+ Helper to construct opening and closing tag expressions for HTML, given a tag name. Matches\r
+ tags in either upper or lower case, attributes with namespaces and with quoted or unquoted values.\r
+\r
+ Example::\r
+ text = '<td>More info at the <a href="http://pyparsing.wikispaces.com">pyparsing</a> wiki page</td>'\r
+ # makeHTMLTags returns pyparsing expressions for the opening and closing tags as a 2-tuple\r
+ a,a_end = makeHTMLTags("A")\r
+ link_expr = a + SkipTo(a_end)("link_text") + a_end\r
+ \r
+ for link in link_expr.searchString(text):\r
+ # attributes in the <A> tag (like "href" shown here) are also accessible as named results\r
+ print(link.link_text, '->', link.href)\r
+ prints::\r
+ pyparsing -> http://pyparsing.wikispaces.com\r
+ """\r
+ return _makeTags( tagStr, False )\r
+\r
+def makeXMLTags(tagStr):\r
+ """\r
+ Helper to construct opening and closing tag expressions for XML, given a tag name. Matches\r
+ tags only in the given upper/lower case.\r
+\r
+ Example: similar to L{makeHTMLTags}\r
+ """\r
+ return _makeTags( tagStr, True )\r
+\r
+def withAttribute(*args,**attrDict):\r
+ """\r
+ Helper to create a validating parse action to be used with start tags created\r
+ with C{L{makeXMLTags}} or C{L{makeHTMLTags}}. Use C{withAttribute} to qualify a starting tag\r
+ with a required attribute value, to avoid false matches on common tags such as\r
+ C{<TD>} or C{<DIV>}.\r
+\r
+ Call C{withAttribute} with a series of attribute names and values. Specify the list\r
+ of filter attributes names and values as:\r
+ - keyword arguments, as in C{(align="right")}, or\r
+ - as an explicit dict with C{**} operator, when an attribute name is also a Python\r
+ reserved word, as in C{**{"class":"Customer", "align":"right"}}\r
+ - a list of name-value tuples, as in ( ("ns1:class", "Customer"), ("ns2:align","right") )\r
+ For attribute names with a namespace prefix, you must use the second form. Attribute\r
+ names are matched insensitive to upper/lower case.\r
+ \r
+ If just testing for C{class} (with or without a namespace), use C{L{withClass}}.\r
+\r
+ To verify that the attribute exists, but without specifying a value, pass\r
+ C{withAttribute.ANY_VALUE} as the value.\r
+\r
+ Example::\r
+ html = '''\r
+ <div>\r
+ Some text\r
+ <div type="grid">1 4 0 1 0</div>\r
+ <div type="graph">1,3 2,3 1,1</div>\r
+ <div>this has no type</div>\r
+ </div>\r
+ \r
+ '''\r
+ div,div_end = makeHTMLTags("div")\r
+\r
+ # only match div tag having a type attribute with value "grid"\r
+ div_grid = div().setParseAction(withAttribute(type="grid"))\r
+ grid_expr = div_grid + SkipTo(div | div_end)("body")\r
+ for grid_header in grid_expr.searchString(html):\r
+ print(grid_header.body)\r
+ \r
+ # construct a match with any div tag having a type attribute, regardless of the value\r
+ div_any_type = div().setParseAction(withAttribute(type=withAttribute.ANY_VALUE))\r
+ div_expr = div_any_type + SkipTo(div | div_end)("body")\r
+ for div_header in div_expr.searchString(html):\r
+ print(div_header.body)\r
+ prints::\r
+ 1 4 0 1 0\r
+\r
+ 1 4 0 1 0\r
+ 1,3 2,3 1,1\r
+ """\r
+ if args:\r
+ attrs = args[:]\r
+ else:\r
+ attrs = attrDict.items()\r
+ attrs = [(k,v) for k,v in attrs]\r
+ def pa(s,l,tokens):\r
+ for attrName,attrValue in attrs:\r
+ if attrName not in tokens:\r
+ raise ParseException(s,l,"no matching attribute " + attrName)\r
+ if attrValue != withAttribute.ANY_VALUE and tokens[attrName] != attrValue:\r
+ raise ParseException(s,l,"attribute '%s' has value '%s', must be '%s'" %\r
+ (attrName, tokens[attrName], attrValue))\r
+ return pa\r
+withAttribute.ANY_VALUE = object()\r
+\r
+def withClass(classname, namespace=''):\r
+ """\r
+ Simplified version of C{L{withAttribute}} when matching on a div class - made\r
+ difficult because C{class} is a reserved word in Python.\r
+\r
+ Example::\r
+ html = '''\r
+ <div>\r
+ Some text\r
+ <div class="grid">1 4 0 1 0</div>\r
+ <div class="graph">1,3 2,3 1,1</div>\r
+ <div>this <div> has no class</div>\r
+ </div>\r
+ \r
+ '''\r
+ div,div_end = makeHTMLTags("div")\r
+ div_grid = div().setParseAction(withClass("grid"))\r
+ \r
+ grid_expr = div_grid + SkipTo(div | div_end)("body")\r
+ for grid_header in grid_expr.searchString(html):\r
+ print(grid_header.body)\r
+ \r
+ div_any_type = div().setParseAction(withClass(withAttribute.ANY_VALUE))\r
+ div_expr = div_any_type + SkipTo(div | div_end)("body")\r
+ for div_header in div_expr.searchString(html):\r
+ print(div_header.body)\r
+ prints::\r
+ 1 4 0 1 0\r
+\r
+ 1 4 0 1 0\r
+ 1,3 2,3 1,1\r
+ """\r
+ classattr = "%s:class" % namespace if namespace else "class"\r
+ return withAttribute(**{classattr : classname}) \r
+\r
+opAssoc = _Constants()\r
+opAssoc.LEFT = object()\r
+opAssoc.RIGHT = object()\r
+\r
+def infixNotation( baseExpr, opList, lpar=Suppress('('), rpar=Suppress(')') ):\r
+ """\r
+ Helper method for constructing grammars of expressions made up of\r
+ operators working in a precedence hierarchy. Operators may be unary or\r
+ binary, left- or right-associative. Parse actions can also be attached\r
+ to operator expressions. The generated parser will also recognize the use \r
+ of parentheses to override operator precedences (see example below).\r
+ \r
+ Note: if you define a deep operator list, you may see performance issues\r
+ when using infixNotation. See L{ParserElement.enablePackrat} for a\r
+ mechanism to potentially improve your parser performance.\r
+\r
+ Parameters:\r
+ - baseExpr - expression representing the most basic element for the nested\r
+ - opList - list of tuples, one for each operator precedence level in the\r
+ expression grammar; each tuple is of the form\r
+ (opExpr, numTerms, rightLeftAssoc, parseAction), where:\r
+ - opExpr is the pyparsing expression for the operator;\r
+ may also be a string, which will be converted to a Literal;\r
+ if numTerms is 3, opExpr is a tuple of two expressions, for the\r
+ two operators separating the 3 terms\r
+ - numTerms is the number of terms for this operator (must\r
+ be 1, 2, or 3)\r
+ - rightLeftAssoc is the indicator whether the operator is\r
+ right or left associative, using the pyparsing-defined\r
+ constants C{opAssoc.RIGHT} and C{opAssoc.LEFT}.\r
+ - parseAction is the parse action to be associated with\r
+ expressions matching this operator expression (the\r
+ parse action tuple member may be omitted)\r
+ - lpar - expression for matching left-parentheses (default=C{Suppress('(')})\r
+ - rpar - expression for matching right-parentheses (default=C{Suppress(')')})\r
+\r
+ Example::\r
+ # simple example of four-function arithmetic with ints and variable names\r
+ integer = pyparsing_common.signed_integer\r
+ varname = pyparsing_common.identifier \r
+ \r
+ arith_expr = infixNotation(integer | varname,\r
+ [\r
+ ('-', 1, opAssoc.RIGHT),\r
+ (oneOf('* /'), 2, opAssoc.LEFT),\r
+ (oneOf('+ -'), 2, opAssoc.LEFT),\r
+ ])\r
+ \r
+ arith_expr.runTests('''\r
+ 5+3*6\r
+ (5+3)*6\r
+ -2--11\r
+ ''', fullDump=False)\r
+ prints::\r
+ 5+3*6\r
+ [[5, '+', [3, '*', 6]]]\r
+\r
+ (5+3)*6\r
+ [[[5, '+', 3], '*', 6]]\r
+\r
+ -2--11\r
+ [[['-', 2], '-', ['-', 11]]]\r
+ """\r
+ ret = Forward()\r
+ lastExpr = baseExpr | ( lpar + ret + rpar )\r
+ for i,operDef in enumerate(opList):\r
+ opExpr,arity,rightLeftAssoc,pa = (operDef + (None,))[:4]\r
+ termName = "%s term" % opExpr if arity < 3 else "%s%s term" % opExpr\r
+ if arity == 3:\r
+ if opExpr is None or len(opExpr) != 2:\r
+ raise ValueError("if numterms=3, opExpr must be a tuple or list of two expressions")\r
+ opExpr1, opExpr2 = opExpr\r
+ thisExpr = Forward().setName(termName)\r
+ if rightLeftAssoc == opAssoc.LEFT:\r
+ if arity == 1:\r
+ matchExpr = FollowedBy(lastExpr + opExpr) + Group( lastExpr + OneOrMore( opExpr ) )\r
+ elif arity == 2:\r
+ if opExpr is not None:\r
+ matchExpr = FollowedBy(lastExpr + opExpr + lastExpr) + Group( lastExpr + OneOrMore( opExpr + lastExpr ) )\r
+ else:\r
+ matchExpr = FollowedBy(lastExpr+lastExpr) + Group( lastExpr + OneOrMore(lastExpr) )\r
+ elif arity == 3:\r
+ matchExpr = FollowedBy(lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr) + \\r
+ Group( lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr )\r
+ else:\r
+ raise ValueError("operator must be unary (1), binary (2), or ternary (3)")\r
+ elif rightLeftAssoc == opAssoc.RIGHT:\r
+ if arity == 1:\r
+ # try to avoid LR with this extra test\r
+ if not isinstance(opExpr, Optional):\r
+ opExpr = Optional(opExpr)\r
+ matchExpr = FollowedBy(opExpr.expr + thisExpr) + Group( opExpr + thisExpr )\r
+ elif arity == 2:\r
+ if opExpr is not None:\r
+ matchExpr = FollowedBy(lastExpr + opExpr + thisExpr) + Group( lastExpr + OneOrMore( opExpr + thisExpr ) )\r
+ else:\r
+ matchExpr = FollowedBy(lastExpr + thisExpr) + Group( lastExpr + OneOrMore( thisExpr ) )\r
+ elif arity == 3:\r
+ matchExpr = FollowedBy(lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr) + \\r
+ Group( lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr )\r
+ else:\r
+ raise ValueError("operator must be unary (1), binary (2), or ternary (3)")\r
+ else:\r
+ raise ValueError("operator must indicate right or left associativity")\r
+ if pa:\r
+ matchExpr.setParseAction( pa )\r
+ thisExpr <<= ( matchExpr.setName(termName) | lastExpr )\r
+ lastExpr = thisExpr\r
+ ret <<= lastExpr\r
+ return ret\r
+\r
+operatorPrecedence = infixNotation\r
+"""(Deprecated) Former name of C{L{infixNotation}}, will be dropped in a future release."""\r
+\r
+dblQuotedString = Combine(Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*')+'"').setName("string enclosed in double quotes")\r
+sglQuotedString = Combine(Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*")+"'").setName("string enclosed in single quotes")\r
+quotedString = Combine(Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*')+'"'|\r
+ Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*")+"'").setName("quotedString using single or double quotes")\r
+unicodeString = Combine(_L('u') + quotedString.copy()).setName("unicode string literal")\r
+\r
+def nestedExpr(opener="(", closer=")", content=None, ignoreExpr=quotedString.copy()):\r
+ """\r
+ Helper method for defining nested lists enclosed in opening and closing\r
+ delimiters ("(" and ")" are the default).\r
+\r
+ Parameters:\r
+ - opener - opening character for a nested list (default=C{"("}); can also be a pyparsing expression\r
+ - closer - closing character for a nested list (default=C{")"}); can also be a pyparsing expression\r
+ - content - expression for items within the nested lists (default=C{None})\r
+ - ignoreExpr - expression for ignoring opening and closing delimiters (default=C{quotedString})\r
+\r
+ If an expression is not provided for the content argument, the nested\r
+ expression will capture all whitespace-delimited content between delimiters\r
+ as a list of separate values.\r
+\r
+ Use the C{ignoreExpr} argument to define expressions that may contain\r
+ opening or closing characters that should not be treated as opening\r
+ or closing characters for nesting, such as quotedString or a comment\r
+ expression. Specify multiple expressions using an C{L{Or}} or C{L{MatchFirst}}.\r
+ The default is L{quotedString}, but if no expressions are to be ignored,\r
+ then pass C{None} for this argument.\r
+\r
+ Example::\r
+ data_type = oneOf("void int short long char float double")\r
+ decl_data_type = Combine(data_type + Optional(Word('*')))\r
+ ident = Word(alphas+'_', alphanums+'_')\r
+ number = pyparsing_common.number\r
+ arg = Group(decl_data_type + ident)\r
+ LPAR,RPAR = map(Suppress, "()")\r
+\r
+ code_body = nestedExpr('{', '}', ignoreExpr=(quotedString | cStyleComment))\r
+\r
+ c_function = (decl_data_type("type") \r
+ + ident("name")\r
+ + LPAR + Optional(delimitedList(arg), [])("args") + RPAR \r
+ + code_body("body"))\r
+ c_function.ignore(cStyleComment)\r
+ \r
+ source_code = '''\r
+ int is_odd(int x) { \r
+ return (x%2); \r
+ }\r
+ \r
+ int dec_to_hex(char hchar) { \r
+ if (hchar >= '0' && hchar <= '9') { \r
+ return (ord(hchar)-ord('0')); \r
+ } else { \r
+ return (10+ord(hchar)-ord('A'));\r
+ } \r
+ }\r
+ '''\r
+ for func in c_function.searchString(source_code):\r
+ print("%(name)s (%(type)s) args: %(args)s" % func)\r
+\r
+ prints::\r
+ is_odd (int) args: [['int', 'x']]\r
+ dec_to_hex (int) args: [['char', 'hchar']]\r
+ """\r
+ if opener == closer:\r
+ raise ValueError("opening and closing strings cannot be the same")\r
+ if content is None:\r
+ if isinstance(opener,basestring) and isinstance(closer,basestring):\r
+ if len(opener) == 1 and len(closer)==1:\r
+ if ignoreExpr is not None:\r
+ content = (Combine(OneOrMore(~ignoreExpr +\r
+ CharsNotIn(opener+closer+ParserElement.DEFAULT_WHITE_CHARS,exact=1))\r
+ ).setParseAction(lambda t:t[0].strip()))\r
+ else:\r
+ content = (empty.copy()+CharsNotIn(opener+closer+ParserElement.DEFAULT_WHITE_CHARS\r
+ ).setParseAction(lambda t:t[0].strip()))\r
+ else:\r
+ if ignoreExpr is not None:\r
+ content = (Combine(OneOrMore(~ignoreExpr + \r
+ ~Literal(opener) + ~Literal(closer) +\r
+ CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS,exact=1))\r
+ ).setParseAction(lambda t:t[0].strip()))\r
+ else:\r
+ content = (Combine(OneOrMore(~Literal(opener) + ~Literal(closer) +\r
+ CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS,exact=1))\r
+ ).setParseAction(lambda t:t[0].strip()))\r
+ else:\r
+ raise ValueError("opening and closing arguments must be strings if no content expression is given")\r
+ ret = Forward()\r
+ if ignoreExpr is not None:\r
+ ret <<= Group( Suppress(opener) + ZeroOrMore( ignoreExpr | ret | content ) + Suppress(closer) )\r
+ else:\r
+ ret <<= Group( Suppress(opener) + ZeroOrMore( ret | content ) + Suppress(closer) )\r
+ ret.setName('nested %s%s expression' % (opener,closer))\r
+ return ret\r
+\r
+def indentedBlock(blockStatementExpr, indentStack, indent=True):\r
+ """\r
+ Helper method for defining space-delimited indentation blocks, such as\r
+ those used to define block statements in Python source code.\r
+\r
+ Parameters:\r
+ - blockStatementExpr - expression defining syntax of statement that\r
+ is repeated within the indented block\r
+ - indentStack - list created by caller to manage indentation stack\r
+ (multiple statementWithIndentedBlock expressions within a single grammar\r
+ should share a common indentStack)\r
+ - indent - boolean indicating whether block must be indented beyond the\r
+ the current level; set to False for block of left-most statements\r
+ (default=C{True})\r
+\r
+ A valid block must contain at least one C{blockStatement}.\r
+\r
+ Example::\r
+ data = '''\r
+ def A(z):\r
+ A1\r
+ B = 100\r
+ G = A2\r
+ A2\r
+ A3\r
+ B\r
+ def BB(a,b,c):\r
+ BB1\r
+ def BBA():\r
+ bba1\r
+ bba2\r
+ bba3\r
+ C\r
+ D\r
+ def spam(x,y):\r
+ def eggs(z):\r
+ pass\r
+ '''\r
+\r
+\r
+ indentStack = [1]\r
+ stmt = Forward()\r
+\r
+ identifier = Word(alphas, alphanums)\r
+ funcDecl = ("def" + identifier + Group( "(" + Optional( delimitedList(identifier) ) + ")" ) + ":")\r
+ func_body = indentedBlock(stmt, indentStack)\r
+ funcDef = Group( funcDecl + func_body )\r
+\r
+ rvalue = Forward()\r
+ funcCall = Group(identifier + "(" + Optional(delimitedList(rvalue)) + ")")\r
+ rvalue << (funcCall | identifier | Word(nums))\r
+ assignment = Group(identifier + "=" + rvalue)\r
+ stmt << ( funcDef | assignment | identifier )\r
+\r
+ module_body = OneOrMore(stmt)\r
+\r
+ parseTree = module_body.parseString(data)\r
+ parseTree.pprint()\r
+ prints::\r
+ [['def',\r
+ 'A',\r
+ ['(', 'z', ')'],\r
+ ':',\r
+ [['A1'], [['B', '=', '100']], [['G', '=', 'A2']], ['A2'], ['A3']]],\r
+ 'B',\r
+ ['def',\r
+ 'BB',\r
+ ['(', 'a', 'b', 'c', ')'],\r
+ ':',\r
+ [['BB1'], [['def', 'BBA', ['(', ')'], ':', [['bba1'], ['bba2'], ['bba3']]]]]],\r
+ 'C',\r
+ 'D',\r
+ ['def',\r
+ 'spam',\r
+ ['(', 'x', 'y', ')'],\r
+ ':',\r
+ [[['def', 'eggs', ['(', 'z', ')'], ':', [['pass']]]]]]] \r
+ """\r
+ def checkPeerIndent(s,l,t):\r
+ if l >= len(s): return\r
+ curCol = col(l,s)\r
+ if curCol != indentStack[-1]:\r
+ if curCol > indentStack[-1]:\r
+ raise ParseFatalException(s,l,"illegal nesting")\r
+ raise ParseException(s,l,"not a peer entry")\r
+\r
+ def checkSubIndent(s,l,t):\r
+ curCol = col(l,s)\r
+ if curCol > indentStack[-1]:\r
+ indentStack.append( curCol )\r
+ else:\r
+ raise ParseException(s,l,"not a subentry")\r
+\r
+ def checkUnindent(s,l,t):\r
+ if l >= len(s): return\r
+ curCol = col(l,s)\r
+ if not(indentStack and curCol < indentStack[-1] and curCol <= indentStack[-2]):\r
+ raise ParseException(s,l,"not an unindent")\r
+ indentStack.pop()\r
+\r
+ NL = OneOrMore(LineEnd().setWhitespaceChars("\t ").suppress())\r
+ INDENT = (Empty() + Empty().setParseAction(checkSubIndent)).setName('INDENT')\r
+ PEER = Empty().setParseAction(checkPeerIndent).setName('')\r
+ UNDENT = Empty().setParseAction(checkUnindent).setName('UNINDENT')\r
+ if indent:\r
+ smExpr = Group( Optional(NL) +\r
+ #~ FollowedBy(blockStatementExpr) +\r
+ INDENT + (OneOrMore( PEER + Group(blockStatementExpr) + Optional(NL) )) + UNDENT)\r
+ else:\r
+ smExpr = Group( Optional(NL) +\r
+ (OneOrMore( PEER + Group(blockStatementExpr) + Optional(NL) )) )\r
+ blockStatementExpr.ignore(_bslash + LineEnd())\r
+ return smExpr.setName('indented block')\r
+\r
+alphas8bit = srange(r"[\0xc0-\0xd6\0xd8-\0xf6\0xf8-\0xff]")\r
+punc8bit = srange(r"[\0xa1-\0xbf\0xd7\0xf7]")\r
+\r
+anyOpenTag,anyCloseTag = makeHTMLTags(Word(alphas,alphanums+"_:").setName('any tag'))\r
+_htmlEntityMap = dict(zip("gt lt amp nbsp quot apos".split(),'><& "\''))\r
+commonHTMLEntity = Regex('&(?P<entity>' + '|'.join(_htmlEntityMap.keys()) +");").setName("common HTML entity")\r
+def replaceHTMLEntity(t):\r
+ """Helper parser action to replace common HTML entities with their special characters"""\r
+ return _htmlEntityMap.get(t.entity)\r
+\r
+# it's easy to get these comment structures wrong - they're very common, so may as well make them available\r
+cStyleComment = Combine(Regex(r"/\*(?:[^*]|\*(?!/))*") + '*/').setName("C style comment")\r
+"Comment of the form C{/* ... */}"\r
+\r
+htmlComment = Regex(r"<!--[\s\S]*?-->").setName("HTML comment")\r
+"Comment of the form C{<!-- ... -->}"\r
+\r
+restOfLine = Regex(r".*").leaveWhitespace().setName("rest of line")\r
+dblSlashComment = Regex(r"//(?:\\\n|[^\n])*").setName("// comment")\r
+"Comment of the form C{// ... (to end of line)}"\r
+\r
+cppStyleComment = Combine(Regex(r"/\*(?:[^*]|\*(?!/))*") + '*/'| dblSlashComment).setName("C++ style comment")\r
+"Comment of either form C{L{cStyleComment}} or C{L{dblSlashComment}}"\r
+\r
+javaStyleComment = cppStyleComment\r
+"Same as C{L{cppStyleComment}}"\r
+\r
+pythonStyleComment = Regex(r"#.*").setName("Python style comment")\r
+"Comment of the form C{# ... (to end of line)}"\r
+\r
+_commasepitem = Combine(OneOrMore(Word(printables, excludeChars=',') +\r
+ Optional( Word(" \t") +\r
+ ~Literal(",") + ~LineEnd() ) ) ).streamline().setName("commaItem")\r
+commaSeparatedList = delimitedList( Optional( quotedString.copy() | _commasepitem, default="") ).setName("commaSeparatedList")\r
+"""(Deprecated) Predefined expression of 1 or more printable words or quoted strings, separated by commas.\r
+ This expression is deprecated in favor of L{pyparsing_common.comma_separated_list}."""\r
+\r
+# some other useful expressions - using lower-case class name since we are really using this as a namespace\r
+class pyparsing_common:\r
+ """\r
+ Here are some common low-level expressions that may be useful in jump-starting parser development:\r
+ - numeric forms (L{integers<integer>}, L{reals<real>}, L{scientific notation<sci_real>})\r
+ - common L{programming identifiers<identifier>}\r
+ - network addresses (L{MAC<mac_address>}, L{IPv4<ipv4_address>}, L{IPv6<ipv6_address>})\r
+ - ISO8601 L{dates<iso8601_date>} and L{datetime<iso8601_datetime>}\r
+ - L{UUID<uuid>}\r
+ - L{comma-separated list<comma_separated_list>}\r
+ Parse actions:\r
+ - C{L{convertToInteger}}\r
+ - C{L{convertToFloat}}\r
+ - C{L{convertToDate}}\r
+ - C{L{convertToDatetime}}\r
+ - C{L{stripHTMLTags}}\r
+ - C{L{upcaseTokens}}\r
+ - C{L{downcaseTokens}}\r
+\r
+ Example::\r
+ pyparsing_common.number.runTests('''\r
+ # any int or real number, returned as the appropriate type\r
+ 100\r
+ -100\r
+ +100\r
+ 3.14159\r
+ 6.02e23\r
+ 1e-12\r
+ ''')\r
+\r
+ pyparsing_common.fnumber.runTests('''\r
+ # any int or real number, returned as float\r
+ 100\r
+ -100\r
+ +100\r
+ 3.14159\r
+ 6.02e23\r
+ 1e-12\r
+ ''')\r
+\r
+ pyparsing_common.hex_integer.runTests('''\r
+ # hex numbers\r
+ 100\r
+ FF\r
+ ''')\r
+\r
+ pyparsing_common.fraction.runTests('''\r
+ # fractions\r
+ 1/2\r
+ -3/4\r
+ ''')\r
+\r
+ pyparsing_common.mixed_integer.runTests('''\r
+ # mixed fractions\r
+ 1\r
+ 1/2\r
+ -3/4\r
+ 1-3/4\r
+ ''')\r
+\r
+ import uuid\r
+ pyparsing_common.uuid.setParseAction(tokenMap(uuid.UUID))\r
+ pyparsing_common.uuid.runTests('''\r
+ # uuid\r
+ 12345678-1234-5678-1234-567812345678\r
+ ''')\r
+ prints::\r
+ # any int or real number, returned as the appropriate type\r
+ 100\r
+ [100]\r
+\r
+ -100\r
+ [-100]\r
+\r
+ +100\r
+ [100]\r
+\r
+ 3.14159\r
+ [3.14159]\r
+\r
+ 6.02e23\r
+ [6.02e+23]\r
+\r
+ 1e-12\r
+ [1e-12]\r
+\r
+ # any int or real number, returned as float\r
+ 100\r
+ [100.0]\r
+\r
+ -100\r
+ [-100.0]\r
+\r
+ +100\r
+ [100.0]\r
+\r
+ 3.14159\r
+ [3.14159]\r
+\r
+ 6.02e23\r
+ [6.02e+23]\r
+\r
+ 1e-12\r
+ [1e-12]\r
+\r
+ # hex numbers\r
+ 100\r
+ [256]\r
+\r
+ FF\r
+ [255]\r
+\r
+ # fractions\r
+ 1/2\r
+ [0.5]\r
+\r
+ -3/4\r
+ [-0.75]\r
+\r
+ # mixed fractions\r
+ 1\r
+ [1]\r
+\r
+ 1/2\r
+ [0.5]\r
+\r
+ -3/4\r
+ [-0.75]\r
+\r
+ 1-3/4\r
+ [1.75]\r
+\r
+ # uuid\r
+ 12345678-1234-5678-1234-567812345678\r
+ [UUID('12345678-1234-5678-1234-567812345678')]\r
+ """\r
+\r
+ convertToInteger = tokenMap(int)\r
+ """\r
+ Parse action for converting parsed integers to Python int\r
+ """\r
+\r
+ convertToFloat = tokenMap(float)\r
+ """\r
+ Parse action for converting parsed numbers to Python float\r
+ """\r
+\r
+ integer = Word(nums).setName("integer").setParseAction(convertToInteger)\r
+ """expression that parses an unsigned integer, returns an int"""\r
+\r
+ hex_integer = Word(hexnums).setName("hex integer").setParseAction(tokenMap(int,16))\r
+ """expression that parses a hexadecimal integer, returns an int"""\r
+\r
+ signed_integer = Regex(r'[+-]?\d+').setName("signed integer").setParseAction(convertToInteger)\r
+ """expression that parses an integer with optional leading sign, returns an int"""\r
+\r
+ fraction = (signed_integer().setParseAction(convertToFloat) + '/' + signed_integer().setParseAction(convertToFloat)).setName("fraction")\r
+ """fractional expression of an integer divided by an integer, returns a float"""\r
+ fraction.addParseAction(lambda t: t[0]/t[-1])\r
+\r
+ mixed_integer = (fraction | signed_integer + Optional(Optional('-').suppress() + fraction)).setName("fraction or mixed integer-fraction")\r
+ """mixed integer of the form 'integer - fraction', with optional leading integer, returns float"""\r
+ mixed_integer.addParseAction(sum)\r
+\r
+ real = Regex(r'[+-]?\d+\.\d*').setName("real number").setParseAction(convertToFloat)\r
+ """expression that parses a floating point number and returns a float"""\r
+\r
+ sci_real = Regex(r'[+-]?\d+([eE][+-]?\d+|\.\d*([eE][+-]?\d+)?)').setName("real number with scientific notation").setParseAction(convertToFloat)\r
+ """expression that parses a floating point number with optional scientific notation and returns a float"""\r
+\r
+ # streamlining this expression makes the docs nicer-looking\r
+ number = (sci_real | real | signed_integer).streamline()\r
+ """any numeric expression, returns the corresponding Python type"""\r
+\r
+ fnumber = Regex(r'[+-]?\d+\.?\d*([eE][+-]?\d+)?').setName("fnumber").setParseAction(convertToFloat)\r
+ """any int or real number, returned as float"""\r
+ \r
+ identifier = Word(alphas+'_', alphanums+'_').setName("identifier")\r
+ """typical code identifier (leading alpha or '_', followed by 0 or more alphas, nums, or '_')"""\r
+ \r
+ 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")\r
+ "IPv4 address (C{0.0.0.0 - 255.255.255.255})"\r
+\r
+ _ipv6_part = Regex(r'[0-9a-fA-F]{1,4}').setName("hex_integer")\r
+ _full_ipv6_address = (_ipv6_part + (':' + _ipv6_part)*7).setName("full IPv6 address")\r
+ _short_ipv6_address = (Optional(_ipv6_part + (':' + _ipv6_part)*(0,6)) + "::" + Optional(_ipv6_part + (':' + _ipv6_part)*(0,6))).setName("short IPv6 address")\r
+ _short_ipv6_address.addCondition(lambda t: sum(1 for tt in t if pyparsing_common._ipv6_part.matches(tt)) < 8)\r
+ _mixed_ipv6_address = ("::ffff:" + ipv4_address).setName("mixed IPv6 address")\r
+ ipv6_address = Combine((_full_ipv6_address | _mixed_ipv6_address | _short_ipv6_address).setName("IPv6 address")).setName("IPv6 address")\r
+ "IPv6 address (long, short, or mixed form)"\r
+ \r
+ mac_address = Regex(r'[0-9a-fA-F]{2}([:.-])[0-9a-fA-F]{2}(?:\1[0-9a-fA-F]{2}){4}').setName("MAC address")\r
+ "MAC address xx:xx:xx:xx:xx (may also have '-' or '.' delimiters)"\r
+\r
+ @staticmethod\r
+ def convertToDate(fmt="%Y-%m-%d"):\r
+ """\r
+ Helper to create a parse action for converting parsed date string to Python datetime.date\r
+\r
+ Params -\r
+ - fmt - format to be passed to datetime.strptime (default=C{"%Y-%m-%d"})\r
+\r
+ Example::\r
+ date_expr = pyparsing_common.iso8601_date.copy()\r
+ date_expr.setParseAction(pyparsing_common.convertToDate())\r
+ print(date_expr.parseString("1999-12-31"))\r
+ prints::\r
+ [datetime.date(1999, 12, 31)]\r
+ """\r
+ def cvt_fn(s,l,t):\r
+ try:\r
+ return datetime.strptime(t[0], fmt).date()\r
+ except ValueError as ve:\r
+ raise ParseException(s, l, str(ve))\r
+ return cvt_fn\r
+\r
+ @staticmethod\r
+ def convertToDatetime(fmt="%Y-%m-%dT%H:%M:%S.%f"):\r
+ """\r
+ Helper to create a parse action for converting parsed datetime string to Python datetime.datetime\r
+\r
+ Params -\r
+ - fmt - format to be passed to datetime.strptime (default=C{"%Y-%m-%dT%H:%M:%S.%f"})\r
+\r
+ Example::\r
+ dt_expr = pyparsing_common.iso8601_datetime.copy()\r
+ dt_expr.setParseAction(pyparsing_common.convertToDatetime())\r
+ print(dt_expr.parseString("1999-12-31T23:59:59.999"))\r
+ prints::\r
+ [datetime.datetime(1999, 12, 31, 23, 59, 59, 999000)]\r
+ """\r
+ def cvt_fn(s,l,t):\r
+ try:\r
+ return datetime.strptime(t[0], fmt)\r
+ except ValueError as ve:\r
+ raise ParseException(s, l, str(ve))\r
+ return cvt_fn\r
+\r
+ iso8601_date = Regex(r'(?P<year>\d{4})(?:-(?P<month>\d\d)(?:-(?P<day>\d\d))?)?').setName("ISO8601 date")\r
+ "ISO8601 date (C{yyyy-mm-dd})"\r
+\r
+ iso8601_datetime = Regex(r'(?P<year>\d{4})-(?P<month>\d\d)-(?P<day>\d\d)[T ](?P<hour>\d\d):(?P<minute>\d\d)(:(?P<second>\d\d(\.\d*)?)?)?(?P<tz>Z|[+-]\d\d:?\d\d)?').setName("ISO8601 datetime")\r
+ "ISO8601 datetime (C{yyyy-mm-ddThh:mm:ss.s(Z|+-00:00)}) - trailing seconds, milliseconds, and timezone optional; accepts separating C{'T'} or C{' '}"\r
+\r
+ uuid = Regex(r'[0-9a-fA-F]{8}(-[0-9a-fA-F]{4}){3}-[0-9a-fA-F]{12}').setName("UUID")\r
+ "UUID (C{xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx})"\r
+\r
+ _html_stripper = anyOpenTag.suppress() | anyCloseTag.suppress()\r
+ @staticmethod\r
+ def stripHTMLTags(s, l, tokens):\r
+ """\r
+ Parse action to remove HTML tags from web page HTML source\r
+\r
+ Example::\r
+ # strip HTML links from normal text \r
+ text = '<td>More info at the <a href="http://pyparsing.wikispaces.com">pyparsing</a> wiki page</td>'\r
+ td,td_end = makeHTMLTags("TD")\r
+ table_text = td + SkipTo(td_end).setParseAction(pyparsing_common.stripHTMLTags)("body") + td_end\r
+ \r
+ print(table_text.parseString(text).body) # -> 'More info at the pyparsing wiki page'\r
+ """\r
+ return pyparsing_common._html_stripper.transformString(tokens[0])\r
+\r
+ _commasepitem = Combine(OneOrMore(~Literal(",") + ~LineEnd() + Word(printables, excludeChars=',') \r
+ + Optional( White(" \t") ) ) ).streamline().setName("commaItem")\r
+ comma_separated_list = delimitedList( Optional( quotedString.copy() | _commasepitem, default="") ).setName("comma separated list")\r
+ """Predefined expression of 1 or more printable words or quoted strings, separated by commas."""\r
+\r
+ upcaseTokens = staticmethod(tokenMap(lambda t: _ustr(t).upper()))\r
+ """Parse action to convert tokens to upper case."""\r
+\r
+ downcaseTokens = staticmethod(tokenMap(lambda t: _ustr(t).lower()))\r
+ """Parse action to convert tokens to lower case."""\r
+\r
+\r
+if __name__ == "__main__":\r
+\r
+ selectToken = CaselessLiteral("select")\r
+ fromToken = CaselessLiteral("from")\r
+\r
+ ident = Word(alphas, alphanums + "_$")\r
+\r
+ columnName = delimitedList(ident, ".", combine=True).setParseAction(upcaseTokens)\r
+ columnNameList = Group(delimitedList(columnName)).setName("columns")\r
+ columnSpec = ('*' | columnNameList)\r
+\r
+ tableName = delimitedList(ident, ".", combine=True).setParseAction(upcaseTokens)\r
+ tableNameList = Group(delimitedList(tableName)).setName("tables")\r
+ \r
+ simpleSQL = selectToken("command") + columnSpec("columns") + fromToken + tableNameList("tables")\r
+\r
+ # demo runTests method, including embedded comments in test string\r
+ simpleSQL.runTests("""\r
+ # '*' as column list and dotted table name\r
+ select * from SYS.XYZZY\r
+\r
+ # caseless match on "SELECT", and casts back to "select"\r
+ SELECT * from XYZZY, ABC\r
+\r
+ # list of column names, and mixed case SELECT keyword\r
+ Select AA,BB,CC from Sys.dual\r
+\r
+ # multiple tables\r
+ Select A, B, C from Sys.dual, Table2\r
+\r
+ # invalid SELECT keyword - should fail\r
+ Xelect A, B, C from Sys.dual\r
+\r
+ # incomplete command - should fail\r
+ Select\r
+\r
+ # invalid column name - should fail\r
+ Select ^^^ frox Sys.dual\r
+\r
+ """)\r
+\r
+ pyparsing_common.number.runTests("""\r
+ 100\r
+ -100\r
+ +100\r
+ 3.14159\r
+ 6.02e23\r
+ 1e-12\r
+ """)\r
+\r
+ # any int or real number, returned as float\r
+ pyparsing_common.fnumber.runTests("""\r
+ 100\r
+ -100\r
+ +100\r
+ 3.14159\r
+ 6.02e23\r
+ 1e-12\r
+ """)\r
+\r
+ pyparsing_common.hex_integer.runTests("""\r
+ 100\r
+ FF\r
+ """)\r
+\r
+ import uuid\r
+ pyparsing_common.uuid.setParseAction(tokenMap(uuid.UUID))\r
+ pyparsing_common.uuid.runTests("""\r
+ 12345678-1234-5678-1234-567812345678\r
+ """)\r