# -*- coding: utf-8 -*- # module pyparsing.py # # Copyright (c) 2003-2019 Paul T. McGuire # # Permission is hereby granted, free of charge, to any person obtaining # a copy of this software and associated documentation files (the # "Software"), to deal in the Software without restriction, including # without limitation the rights to use, copy, modify, merge, publish, # distribute, sublicense, and/or sell copies of the Software, and to # permit persons to whom the Software is furnished to do so, subject to # the following conditions: # # The above copyright notice and this permission notice shall be # included in all copies or substantial portions of the Software. # # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, # EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF # MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. # IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY # CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. # __doc__ = """ pyparsing module - Classes and methods to define and execute parsing grammars ============================================================================= The pyparsing module is an alternative approach to creating and executing simple grammars, vs. the traditional lex/yacc approach, or the use of regular expressions. With pyparsing, you don't need to learn a new syntax for defining grammars or matching expressions - the parsing module provides a library of classes that you use to construct the grammar directly in Python. Here is a program to parse "Hello, World!" (or any greeting of the form ``", !"``), built up using :class:`Word`, :class:`Literal`, and :class:`And` elements (the :class:`'+'` operators create :class:`And` expressions, and the strings are auto-converted to :class:`Literal` expressions):: from pyparsing import Word, alphas # define grammar of a greeting greet = Word(alphas) + "," + Word(alphas) + "!" hello = "Hello, World!" print(hello, "->", greet.parseString(hello)) The program outputs the following:: Hello, World! -> ['Hello', ',', 'World', '!'] The Python representation of the grammar is quite readable, owing to the self-explanatory class names, and the use of '+', '|' and '^' operators. The :class:`ParseResults` object returned from :class:`ParserElement.parseString` can be accessed as a nested list, a dictionary, or an object with named attributes. The pyparsing module handles some of the problems that are typically vexing when writing text parsers: - extra or missing whitespace (the above program will also handle "Hello,World!", "Hello , World !", etc.) - quoted strings - embedded comments Getting Started - ----------------- Visit the classes :class:`ParserElement` and :class:`ParseResults` to see the base classes that most other pyparsing classes inherit from. Use the docstrings for examples of how to: - construct literal match expressions from :class:`Literal` and :class:`CaselessLiteral` classes - construct character word-group expressions using the :class:`Word` class - see how to create repetitive expressions using :class:`ZeroOrMore` and :class:`OneOrMore` classes - use :class:`'+'`, :class:`'|'`, :class:`'^'`, and :class:`'&'` operators to combine simple expressions into more complex ones - associate names with your parsed results using :class:`ParserElement.setResultsName` - access the parsed data, which is returned as a :class:`ParseResults` object - find some helpful expression short-cuts like :class:`delimitedList` and :class:`oneOf` - find more useful common expressions in the :class:`pyparsing_common` namespace class """ __version__ = "3.0.0a1" __versionTime__ = "13 Oct 2019 05:49 UTC" __author__ = "Paul McGuire " import string from weakref import ref as wkref import copy import sys import warnings import re import sre_constants import collections from collections.abc import Iterable, MutableMapping, Mapping import pprint import traceback import types from datetime import datetime from operator import itemgetter import itertools from functools import wraps from itertools import filterfalse from threading import RLock from contextlib import contextmanager import unittest class __config_flags: """Internal class for defining compatibility and debugging flags""" _all_names = [] _fixed_names = [] _type_desc = "configuration" @classmethod def _set(cls, dname, value): if dname in cls._fixed_names: warnings.warn( "{}.{} {} is {} and cannot be overridden".format( cls.__name__, dname, cls._type_desc, str(getattr(cls, dname)).upper(), ) ) return if dname in cls._all_names: setattr(cls, dname, value) else: raise ValueError("no such {} {!r}".format(cls._type_desc, dname)) enable = classmethod(lambda cls, name: cls._set(name, True)) disable = classmethod(lambda cls, name: cls._set(name, False)) class __compat__(__config_flags): """ A cross-version compatibility configuration for pyparsing features that will be released in a future version. By setting values in this configuration to True, those features can be enabled in prior versions for compatibility development and testing. - collect_all_And_tokens - flag to enable fix for Issue #63 that fixes erroneous grouping of results names when an And expression is nested within an Or or MatchFirst; maintained for compatibility, but setting to False no longer restores pre-2.3.1 behavior """ _type_desc = "compatibility" collect_all_And_tokens = True _all_names = [__ for __ in locals() if not __.startswith("_")] _fixed_names = """ collect_all_And_tokens """.split() class __diag__(__config_flags): """ Diagnostic configuration (all default to False) - warn_multiple_tokens_in_named_alternation - flag to enable warnings when a results name is defined on a MatchFirst or Or expression with one or more And subexpressions - warn_ungrouped_named_tokens_in_collection - flag to enable warnings when a results name is defined on a containing expression with ungrouped subexpressions that also have results names - warn_name_set_on_empty_Forward - flag to enable warnings when a Forward is defined with a results name, but has no contents defined - warn_on_multiple_string_args_to_oneof - flag to enable warnings when oneOf is incorrectly called with multiple str arguments - enable_debug_on_named_expressions - flag to auto-enable debug on all subsequent calls to ParserElement.setName() """ _type_desc = "diagnostic" warn_multiple_tokens_in_named_alternation = False warn_ungrouped_named_tokens_in_collection = False warn_name_set_on_empty_Forward = False warn_on_multiple_string_args_to_oneof = False enable_debug_on_named_expressions = False _all_names = [__ for __ in locals() if not __.startswith("_")] _warning_names = [name for name in _all_names if name.startswith("warn")] _debug_names = [name for name in _all_names if name.startswith("enable_debug")] @classmethod def enable_all_warnings(cls): for name in cls._warning_names: cls.enable(name) # hide abstract class del __config_flags # ~ sys.stderr.write("testing pyparsing module, version %s, %s\n" % (__version__, __versionTime__)) __all__ = [ "__version__", "__versionTime__", "__author__", "__compat__", "__diag__", "And", "CaselessKeyword", "CaselessLiteral", "CharsNotIn", "Combine", "Dict", "Each", "Empty", "FollowedBy", "Forward", "GoToColumn", "Group", "Keyword", "LineEnd", "LineStart", "Literal", "PrecededBy", "MatchFirst", "NoMatch", "NotAny", "OneOrMore", "OnlyOnce", "Optional", "Or", "ParseBaseException", "ParseElementEnhance", "ParseException", "ParseExpression", "ParseFatalException", "ParseResults", "ParseSyntaxException", "ParserElement", "QuotedString", "RecursiveGrammarException", "Regex", "SkipTo", "StringEnd", "StringStart", "Suppress", "Token", "TokenConverter", "White", "Word", "WordEnd", "WordStart", "ZeroOrMore", "Char", "alphanums", "alphas", "alphas8bit", "anyCloseTag", "anyOpenTag", "cStyleComment", "col", "commonHTMLEntity", "countedArray", "cppStyleComment", "dblQuotedString", "dblSlashComment", "delimitedList", "dictOf", "empty", "hexnums", "htmlComment", "javaStyleComment", "line", "lineEnd", "lineStart", "lineno", "makeHTMLTags", "makeXMLTags", "matchOnlyAtCol", "matchPreviousExpr", "matchPreviousLiteral", "nestedExpr", "nullDebugAction", "nums", "oneOf", "opAssoc", "printables", "punc8bit", "pythonStyleComment", "quotedString", "removeQuotes", "replaceHTMLEntity", "replaceWith", "restOfLine", "sglQuotedString", "srange", "stringEnd", "stringStart", "traceParseAction", "unicodeString", "withAttribute", "indentedBlock", "originalTextFor", "ungroup", "infixNotation", "locatedExpr", "withClass", "CloseMatch", "tokenMap", "pyparsing_common", "pyparsing_unicode", "unicode_set", "conditionAsParseAction", "pyparsing_test", "re", ] system_version = tuple(sys.version_info)[:3] _MAX_INT = sys.maxsize str_type = (str, bytes) # build list of single arg builtins, that can be used as parse actions singleArgBuiltins = [sum, len, sorted, reversed, list, tuple, set, any, all, min, max] _generatorType = types.GeneratorType alphas = string.ascii_uppercase + string.ascii_lowercase nums = "0123456789" hexnums = nums + "ABCDEFabcdef" alphanums = alphas + nums _bslash = chr(92) printables = "".join(c for c in string.printable if c not in string.whitespace) def conditionAsParseAction(fn, message=None, fatal=False): """ Function to convert a simple predicate function that returns True or False into a parse action. Can be used in places when a parse action is required and ParserElement.addCondition cannot be used (such as when adding a condition to an operator level in infixNotation). Optional keyword arguments: - message = define a custom message to be used in the raised exception - fatal = if True, will raise ParseFatalException to stop parsing immediately; otherwise will raise ParseException """ msg = message if message is not None else "failed user-defined condition" exc_type = ParseFatalException if fatal else ParseException fn = _trim_arity(fn) @wraps(fn) def pa(s, l, t): if not bool(fn(s, l, t)): raise exc_type(s, l, msg) return pa class ParseBaseException(Exception): """base exception class for all parsing runtime exceptions""" # Performance tuning: we construct a *lot* of these, so keep this # constructor as small and fast as possible def __init__(self, pstr, loc=0, msg=None, elem=None): self.loc = loc if msg is None: self.msg = pstr self.pstr = "" else: self.msg = msg self.pstr = pstr self.parserElement = elem self.args = (pstr, loc, msg) @classmethod def _from_exception(cls, pe): """ internal factory method to simplify creating one type of ParseException from another - avoids having __init__ signature conflicts among subclasses """ return cls(pe.pstr, pe.loc, pe.msg, pe.parserElement) def __getattr__(self, aname): """supported attributes by name are: - lineno - returns the line number of the exception text - col - returns the column number of the exception text - line - returns the line containing the exception text """ if aname == "lineno": return lineno(self.loc, self.pstr) elif aname in ("col", "column"): return col(self.loc, self.pstr) elif aname == "line": return line(self.loc, self.pstr) else: raise AttributeError(aname) def __str__(self): if self.pstr: if self.loc >= len(self.pstr): foundstr = ", found end of text" else: foundstr = (", found %r" % self.pstr[self.loc : self.loc + 1]).replace( r"\\", "\\" ) else: foundstr = "" return "%s%s (at char %d), (line:%d, col:%d)" % ( self.msg, foundstr, self.loc, self.lineno, self.column, ) def __repr__(self): return str(self) def markInputline(self, markerString=">!<"): """Extracts the exception line from the input string, and marks the location of the exception with a special symbol. """ line_str = self.line line_column = self.column - 1 if markerString: line_str = "".join( (line_str[:line_column], markerString, line_str[line_column:]) ) return line_str.strip() def __dir__(self): return "lineno col line".split() + dir(type(self)) class ParseException(ParseBaseException): """ Exception thrown when parse expressions don't match class; supported attributes by name are: - lineno - returns the line number of the exception text - col - returns the column number of the exception text - line - returns the line containing the exception text Example:: try: Word(nums).setName("integer").parseString("ABC") except ParseException as pe: print(pe) print("column: {}".format(pe.col)) prints:: Expected integer (at char 0), (line:1, col:1) column: 1 """ @staticmethod def explain(exc, depth=16): """ Method to take an exception and translate the Python internal traceback into a list of the pyparsing expressions that caused the exception to be raised. Parameters: - exc - exception raised during parsing (need not be a ParseException, in support of Python exceptions that might be raised in a parse action) - depth (default=16) - number of levels back in the stack trace to list expression and function names; if None, the full stack trace names will be listed; if 0, only the failing input line, marker, and exception string will be shown Returns a multi-line string listing the ParserElements and/or function names in the exception's stack trace. Note: the diagnostic output will include string representations of the expressions that failed to parse. These representations will be more helpful if you use `setName` to give identifiable names to your expressions. Otherwise they will use the default string forms, which may be cryptic to read. explain() is only supported under Python 3. """ import inspect if depth is None: depth = sys.getrecursionlimit() ret = [] if isinstance(exc, ParseBaseException): ret.append(exc.line) ret.append(" " * (exc.col - 1) + "^") ret.append("{}: {}".format(type(exc).__name__, exc)) if depth > 0: callers = inspect.getinnerframes(exc.__traceback__, context=depth) seen = set() for i, ff in enumerate(callers[-depth:]): frm = ff[0] f_self = frm.f_locals.get("self", None) if isinstance(f_self, ParserElement): if frm.f_code.co_name not in ("parseImpl", "_parseNoCache"): continue if f_self in seen: continue seen.add(f_self) self_type = type(f_self) ret.append( "{}.{} - {}".format( self_type.__module__, self_type.__name__, f_self ) ) elif f_self is not None: self_type = type(f_self) ret.append("{}.{}".format(self_type.__module__, self_type.__name__)) else: code = frm.f_code if code.co_name in ("wrapper", ""): continue ret.append("{}".format(code.co_name)) depth -= 1 if not depth: break return "\n".join(ret) class ParseFatalException(ParseBaseException): """user-throwable exception thrown when inconsistent parse content is found; stops all parsing immediately""" pass class ParseSyntaxException(ParseFatalException): """just like :class:`ParseFatalException`, but thrown internally when an :class:`ErrorStop` ('-' operator) indicates that parsing is to stop immediately because an unbacktrackable syntax error has been found. """ pass # ~ class ReparseException(ParseBaseException): # ~ """Experimental class - parse actions can raise this exception to cause # ~ pyparsing to reparse the input string: # ~ - with a modified input string, and/or # ~ - with a modified start location # ~ Set the values of the ReparseException in the constructor, and raise the # ~ exception in a parse action to cause pyparsing to use the new string/location. # ~ Setting the values as None causes no change to be made. # ~ """ # ~ def __init_( self, newstring, restartLoc ): # ~ self.newParseText = newstring # ~ self.reparseLoc = restartLoc class RecursiveGrammarException(Exception): """exception thrown by :class:`ParserElement.validate` if the grammar could be improperly recursive """ def __init__(self, parseElementList): self.parseElementTrace = parseElementList def __str__(self): return "RecursiveGrammarException: %s" % self.parseElementTrace class _ParseResultsWithOffset(object): def __init__(self, p1, p2): self.tup = (p1, p2) def __getitem__(self, i): return self.tup[i] def __repr__(self): return repr(self.tup[0]) def setOffset(self, i): self.tup = (self.tup[0], i) class ParseResults(object): """Structured parse results, to provide multiple means of access to the parsed data: - as a list (``len(results)``) - by list index (``results[0], results[1]``, etc.) - by attribute (``results.`` - see :class:`ParserElement.setResultsName`) Example:: integer = Word(nums) date_str = (integer.setResultsName("year") + '/' + integer.setResultsName("month") + '/' + integer.setResultsName("day")) # equivalent form: # date_str = integer("year") + '/' + integer("month") + '/' + integer("day") # parseString returns a ParseResults object result = date_str.parseString("1999/12/31") def test(s, fn=repr): print("%s -> %s" % (s, fn(eval(s)))) test("list(result)") test("result[0]") test("result['month']") test("result.day") test("'month' in result") test("'minutes' in result") test("result.dump()", str) prints:: list(result) -> ['1999', '/', '12', '/', '31'] result[0] -> '1999' result['month'] -> '12' result.day -> '31' 'month' in result -> True 'minutes' in result -> False result.dump() -> ['1999', '/', '12', '/', '31'] - day: 31 - month: 12 - year: 1999 """ def __new__(cls, toklist=None, name=None, asList=True, modal=True): if isinstance(toklist, ParseResults): return toklist retobj = object.__new__(cls) retobj.__doinit = True return retobj # Performance tuning: we construct a *lot* of these, so keep this # constructor as small and fast as possible def __init__( self, toklist=None, name=None, asList=True, modal=True, isinstance=isinstance ): if self.__doinit: self.__doinit = False self.__name = None self.__parent = None self.__accumNames = {} self.__asList = asList self.__modal = modal if toklist is None: toklist = [] if isinstance(toklist, list): self.__toklist = toklist[:] elif isinstance(toklist, _generatorType): self.__toklist = list(toklist) else: self.__toklist = [toklist] self.__tokdict = dict() if name is not None and name: if not modal: self.__accumNames[name] = 0 if isinstance(name, int): name = str(name) self.__name = name if not ( isinstance(toklist, (type(None), *str_type, list)) and toklist in (None, "", []) ): if isinstance(toklist, str_type): toklist = [toklist] if asList: if isinstance(toklist, ParseResults): self[name] = _ParseResultsWithOffset( ParseResults(toklist.__toklist), 0 ) else: self[name] = _ParseResultsWithOffset( ParseResults(toklist[0]), 0 ) self[name].__name = name else: try: self[name] = toklist[0] except (KeyError, TypeError, IndexError): self[name] = toklist def __getitem__(self, i): if isinstance(i, (int, slice)): return self.__toklist[i] else: if i not in self.__accumNames: return self.__tokdict[i][-1][0] else: return ParseResults([v[0] for v in self.__tokdict[i]]) def __setitem__(self, k, v, isinstance=isinstance): if isinstance(v, _ParseResultsWithOffset): self.__tokdict[k] = self.__tokdict.get(k, list()) + [v] sub = v[0] elif isinstance(k, (int, slice)): self.__toklist[k] = v sub = v else: self.__tokdict[k] = self.__tokdict.get(k, list()) + [ _ParseResultsWithOffset(v, 0) ] sub = v if isinstance(sub, ParseResults): sub.__parent = wkref(self) def __delitem__(self, i): if isinstance(i, (int, slice)): mylen = len(self.__toklist) del self.__toklist[i] # convert int to slice if isinstance(i, int): if i < 0: i += mylen i = slice(i, i + 1) # get removed indices removed = list(range(*i.indices(mylen))) removed.reverse() # fixup indices in token dictionary for name, occurrences in self.__tokdict.items(): for j in removed: for k, (value, position) in enumerate(occurrences): occurrences[k] = _ParseResultsWithOffset( value, position - (position > j) ) else: del self.__tokdict[i] def __contains__(self, k): return k in self.__tokdict def __len__(self): return len(self.__toklist) def __bool__(self): return not not self.__toklist def __iter__(self): return iter(self.__toklist) def __reversed__(self): return iter(self.__toklist[::-1]) def keys(self): return iter(self.__tokdict) def values(self): return (self[k] for k in self.keys()) def items(self): return ((k, self[k]) for k in self.keys()) def haskeys(self): """Since keys() returns an iterator, this method is helpful in bypassing code that looks for the existence of any defined results names.""" return bool(self.__tokdict) def pop(self, *args, **kwargs): """ Removes and returns item at specified index (default= ``last``). Supports both ``list`` and ``dict`` semantics for ``pop()``. If passed no argument or an integer argument, it will use ``list`` semantics and pop tokens from the list of parsed tokens. If passed a non-integer argument (most likely a string), it will use ``dict`` semantics and pop the corresponding value from any defined results names. A second default return value argument is supported, just as in ``dict.pop()``. Example:: def remove_first(tokens): tokens.pop(0) print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321'] print(OneOrMore(Word(nums)).addParseAction(remove_first).parseString("0 123 321")) # -> ['123', '321'] label = Word(alphas) patt = label("LABEL") + OneOrMore(Word(nums)) print(patt.parseString("AAB 123 321").dump()) # Use pop() in a parse action to remove named result (note that corresponding value is not # removed from list form of results) def remove_LABEL(tokens): tokens.pop("LABEL") return tokens patt.addParseAction(remove_LABEL) print(patt.parseString("AAB 123 321").dump()) prints:: ['AAB', '123', '321'] - LABEL: AAB ['AAB', '123', '321'] """ if not args: args = [-1] for k, v in kwargs.items(): if k == "default": args = (args[0], v) else: raise TypeError("pop() got an unexpected keyword argument '%s'" % k) if isinstance(args[0], int) or len(args) == 1 or args[0] in self: index = args[0] ret = self[index] del self[index] return ret else: defaultvalue = args[1] return defaultvalue def get(self, key, defaultValue=None): """ Returns named result matching the given key, or if there is no such name, then returns the given ``defaultValue`` or ``None`` if no ``defaultValue`` is specified. Similar to ``dict.get()``. Example:: integer = Word(nums) date_str = integer("year") + '/' + integer("month") + '/' + integer("day") result = date_str.parseString("1999/12/31") print(result.get("year")) # -> '1999' print(result.get("hour", "not specified")) # -> 'not specified' print(result.get("hour")) # -> None """ if key in self: return self[key] else: return defaultValue def insert(self, index, insStr): """ Inserts new element at location index in the list of parsed tokens. Similar to ``list.insert()``. Example:: print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321'] # use a parse action to insert the parse location in the front of the parsed results def insert_locn(locn, tokens): tokens.insert(0, locn) print(OneOrMore(Word(nums)).addParseAction(insert_locn).parseString("0 123 321")) # -> [0, '0', '123', '321'] """ self.__toklist.insert(index, insStr) # fixup indices in token dictionary for name, occurrences in self.__tokdict.items(): for k, (value, position) in enumerate(occurrences): occurrences[k] = _ParseResultsWithOffset( value, position + (position > index) ) def append(self, item): """ Add single element to end of ParseResults list of elements. Example:: print(OneOrMore(Word(nums)).parseString("0 123 321")) # -> ['0', '123', '321'] # use a parse action to compute the sum of the parsed integers, and add it to the end def append_sum(tokens): tokens.append(sum(map(int, tokens))) print(OneOrMore(Word(nums)).addParseAction(append_sum).parseString("0 123 321")) # -> ['0', '123', '321', 444] """ self.__toklist.append(item) def extend(self, itemseq): """ Add sequence of elements to end of ParseResults list of elements. Example:: patt = OneOrMore(Word(alphas)) # use a parse action to append the reverse of the matched strings, to make a palindrome def make_palindrome(tokens): tokens.extend(reversed([t[::-1] for t in tokens])) return ''.join(tokens) print(patt.addParseAction(make_palindrome).parseString("lskdj sdlkjf lksd")) # -> 'lskdjsdlkjflksddsklfjkldsjdksl' """ if isinstance(itemseq, ParseResults): self.__iadd__(itemseq) else: self.__toklist.extend(itemseq) def clear(self): """ Clear all elements and results names. """ del self.__toklist[:] self.__tokdict.clear() def __getattr__(self, name): try: return self[name] except KeyError: return "" def __add__(self, other): ret = self.copy() ret += other return ret def __iadd__(self, other): if other.__tokdict: offset = len(self.__toklist) addoffset = lambda a: offset if a < 0 else a + offset otheritems = other.__tokdict.items() otherdictitems = [ (k, _ParseResultsWithOffset(v[0], addoffset(v[1]))) for k, vlist in otheritems for v in vlist ] for k, v in otherdictitems: self[k] = v if isinstance(v[0], ParseResults): v[0].__parent = wkref(self) self.__toklist += other.__toklist self.__accumNames.update(other.__accumNames) return self def __radd__(self, other): if isinstance(other, int) and other == 0: # useful for merging many ParseResults using sum() builtin return self.copy() else: # this may raise a TypeError - so be it return other + self def __repr__(self): return "(%s, %s)" % (repr(self.__toklist), repr(self.__tokdict)) def __str__(self): return ( "[" + ", ".join( str(i) if isinstance(i, ParseResults) else repr(i) for i in self.__toklist ) + "]" ) def _asStringList(self, sep=""): out = [] for item in self.__toklist: if out and sep: out.append(sep) if isinstance(item, ParseResults): out += item._asStringList() else: out.append(str(item)) return out def asList(self): """ Returns the parse results as a nested list of matching tokens, all converted to strings. Example:: patt = OneOrMore(Word(alphas)) result = patt.parseString("sldkj lsdkj sldkj") # even though the result prints in string-like form, it is actually a pyparsing ParseResults print(type(result), result) # -> ['sldkj', 'lsdkj', 'sldkj'] # Use asList() to create an actual list result_list = result.asList() print(type(result_list), result_list) # -> ['sldkj', 'lsdkj', 'sldkj'] """ return [ res.asList() if isinstance(res, ParseResults) else res for res in self.__toklist ] def asDict(self): """ Returns the named parse results as a nested dictionary. Example:: integer = Word(nums) date_str = integer("year") + '/' + integer("month") + '/' + integer("day") result = date_str.parseString('12/31/1999') print(type(result), repr(result)) # -> (['12', '/', '31', '/', '1999'], {'day': [('1999', 4)], 'year': [('12', 0)], 'month': [('31', 2)]}) result_dict = result.asDict() print(type(result_dict), repr(result_dict)) # -> {'day': '1999', 'year': '12', 'month': '31'} # even though a ParseResults supports dict-like access, sometime you just need to have a dict import json print(json.dumps(result)) # -> Exception: TypeError: ... is not JSON serializable print(json.dumps(result.asDict())) # -> {"month": "31", "day": "1999", "year": "12"} """ def to_item(obj): if isinstance(obj, ParseResults): return obj.asDict() if obj.haskeys() else [to_item(v) for v in obj] else: return obj return dict((k, to_item(v)) for k, v in self.items()) def copy(self): """ Returns a new copy of a :class:`ParseResults` object. """ ret = ParseResults(self.__toklist) ret.__tokdict = dict(self.__tokdict.items()) ret.__parent = self.__parent ret.__accumNames.update(self.__accumNames) ret.__name = self.__name return ret def getName(self): r""" Returns the results name for this token expression. Useful when several different expressions might match at a particular location. Example:: integer = Word(nums) ssn_expr = Regex(r"\d\d\d-\d\d-\d\d\d\d") house_number_expr = Suppress('#') + Word(nums, alphanums) user_data = (Group(house_number_expr)("house_number") | Group(ssn_expr)("ssn") | Group(integer)("age")) user_info = OneOrMore(user_data) result = user_info.parseString("22 111-22-3333 #221B") for item in result: print(item.getName(), ':', item[0]) prints:: age : 22 ssn : 111-22-3333 house_number : 221B """ if self.__name: return self.__name elif self.__parent: par = self.__parent() def lookup(self, sub): return next( ( k for k, vlist in par.__tokdict.items() for v, loc in vlist if sub is v ), None, ) return lookup(self) if par else None elif ( len(self) == 1 and len(self.__tokdict) == 1 and next(iter(self.__tokdict.values()))[0][1] in (0, -1) ): return next(iter(self.__tokdict.keys())) else: return None def dump(self, indent="", full=True, include_list=True, _depth=0): """ Diagnostic method for listing out the contents of a :class:`ParseResults`. Accepts an optional ``indent`` argument so that this string can be embedded in a nested display of other data. Example:: integer = Word(nums) date_str = integer("year") + '/' + integer("month") + '/' + integer("day") result = date_str.parseString('12/31/1999') print(result.dump()) prints:: ['12', '/', '31', '/', '1999'] - day: 1999 - month: 31 - year: 12 """ out = [] NL = "\n" out.append(indent + str(self.asList()) if include_list else "") if full: if self.haskeys(): items = sorted((str(k), v) for k, v in self.items()) for k, v in items: if out: out.append(NL) out.append("%s%s- %s: " % (indent, (" " * _depth), k)) if isinstance(v, ParseResults): if v: out.append( v.dump( indent=indent, full=full, include_list=include_list, _depth=_depth + 1, ) ) else: out.append(str(v)) else: out.append(repr(v)) elif any(isinstance(vv, ParseResults) for vv in self): v = self for i, vv in enumerate(v): if isinstance(vv, ParseResults): out.append( "\n%s%s[%d]:\n%s%s%s" % ( indent, (" " * (_depth)), i, indent, (" " * (_depth + 1)), vv.dump( indent=indent, full=full, include_list=include_list, _depth=_depth + 1, ), ) ) else: out.append( "\n%s%s[%d]:\n%s%s%s" % ( indent, (" " * (_depth)), i, indent, (" " * (_depth + 1)), str(vv), ) ) return "".join(out) def pprint(self, *args, **kwargs): """ Pretty-printer for parsed results as a list, using the `pprint `_ module. Accepts additional positional or keyword args as defined for `pprint.pprint `_ . Example:: ident = Word(alphas, alphanums) num = Word(nums) func = Forward() term = ident | num | Group('(' + func + ')') func <<= ident + Group(Optional(delimitedList(term))) result = func.parseString("fna a,b,(fnb c,d,200),100") result.pprint(width=40) prints:: ['fna', ['a', 'b', ['(', 'fnb', ['c', 'd', '200'], ')'], '100']] """ pprint.pprint(self.asList(), *args, **kwargs) # add support for pickle protocol def __getstate__(self): return ( self.__toklist, ( self.__tokdict.copy(), self.__parent is not None and self.__parent() or None, self.__accumNames, self.__name, ), ) def __setstate__(self, state): self.__toklist = state[0] self.__tokdict, par, inAccumNames, self.__name = state[1] self.__accumNames = {} self.__accumNames.update(inAccumNames) if par is not None: self.__parent = wkref(par) else: self.__parent = None def __getnewargs__(self): return self.__toklist, self.__name, self.__asList, self.__modal def __dir__(self): return dir(type(self)) + list(self.keys()) @classmethod def from_dict(cls, other, name=None): """ Helper classmethod to construct a ParseResults from a dict, preserving the name-value relations as results names. If an optional 'name' argument is given, a nested ParseResults will be returned """ def is_iterable(obj): try: iter(obj) except Exception: return False else: return not isinstance(obj, str_type) ret = cls([]) for k, v in other.items(): if isinstance(v, Mapping): ret += cls.from_dict(v, name=k) else: ret += cls([v], name=k, asList=is_iterable(v)) if name is not None: ret = cls([ret], name=name) return ret MutableMapping.register(ParseResults) def col(loc, strg): """Returns current column within a string, counting newlines as line separators. The first column is number 1. Note: the default parsing behavior is to expand tabs in the input string before starting the parsing process. See :class:`ParserElement.parseString` for more information on parsing strings containing ```` s, and suggested methods to maintain a consistent view of the parsed string, the parse location, and line and column positions within the parsed string. """ s = strg return 1 if 0 < loc < len(s) and s[loc - 1] == "\n" else loc - s.rfind("\n", 0, loc) def lineno(loc, strg): """Returns current line number within a string, counting newlines as line separators. The first line is number 1. Note - the default parsing behavior is to expand tabs in the input string before starting the parsing process. See :class:`ParserElement.parseString` for more information on parsing strings containing ```` s, and suggested methods to maintain a consistent view of the parsed string, the parse location, and line and column positions within the parsed string. """ return strg.count("\n", 0, loc) + 1 def line(loc, strg): """Returns the line of text containing loc within a string, counting newlines as line separators. """ lastCR = strg.rfind("\n", 0, loc) nextCR = strg.find("\n", loc) return strg[lastCR + 1 : nextCR] if nextCR >= 0 else strg[lastCR + 1 :] def _defaultStartDebugAction(instring, loc, expr): print( ( "Match " + str(expr) + " at loc " + str(loc) + "(%d,%d)" % (lineno(loc, instring), col(loc, instring)) ) ) def _defaultSuccessDebugAction(instring, startloc, endloc, expr, toks): print("Matched " + str(expr) + " -> " + str(toks.asList())) def _defaultExceptionDebugAction(instring, loc, expr, exc): print("Exception raised:" + str(exc)) def nullDebugAction(*args): """'Do-nothing' debug action, to suppress debugging output during parsing.""" pass def _trim_arity(func, maxargs=2): "decorator to trim function calls to match the arity of the target" if func in singleArgBuiltins: return lambda s, l, t: func(t) limit = 0 found_arity = False # traceback return data structure changed in Py3.5 - normalize back to plain tuples def extract_stack(limit=0): # special handling for Python 3.5.0 - extra deep call stack by 1 offset = -3 if system_version == (3, 5, 0) else -2 frame_summary = traceback.extract_stack(limit=-offset + limit - 1)[offset] return [frame_summary[:2]] def extract_tb(tb, limit=0): frames = traceback.extract_tb(tb, limit=limit) frame_summary = frames[-1] return [frame_summary[:2]] # synthesize what would be returned by traceback.extract_stack at the call to # user's parse action 'func', so that we don't incur call penalty at parse time LINE_DIFF = 7 # IF ANY CODE CHANGES, EVEN JUST COMMENTS OR BLANK LINES, BETWEEN THE NEXT LINE AND # THE CALL TO FUNC INSIDE WRAPPER, LINE_DIFF MUST BE MODIFIED!!!! this_line = extract_stack(limit=2)[-1] pa_call_line_synth = (this_line[0], this_line[1] + LINE_DIFF) def wrapper(*args): nonlocal found_arity, limit while 1: try: ret = func(*args[limit:]) found_arity = True return ret except TypeError: # re-raise TypeErrors if they did not come from our arity testing if found_arity: raise else: try: tb = sys.exc_info()[-1] if not extract_tb(tb, limit=2)[-1][:2] == pa_call_line_synth: raise finally: try: del tb except NameError: pass if limit <= maxargs: limit += 1 continue raise # copy func name to wrapper for sensible debug output func_name = "" try: func_name = getattr(func, "__name__", getattr(func, "__class__").__name__) except Exception: func_name = str(func) wrapper.__name__ = func_name return wrapper class ParserElement(object): """Abstract base level parser element class.""" DEFAULT_WHITE_CHARS = " \n\t\r" verbose_stacktrace = False @staticmethod def setDefaultWhitespaceChars(chars): r""" Overrides the default whitespace chars Example:: # default whitespace chars are space, and newline OneOrMore(Word(alphas)).parseString("abc def\nghi jkl") # -> ['abc', 'def', 'ghi', 'jkl'] # change to just treat newline as significant ParserElement.setDefaultWhitespaceChars(" \t") OneOrMore(Word(alphas)).parseString("abc def\nghi jkl") # -> ['abc', 'def'] """ ParserElement.DEFAULT_WHITE_CHARS = chars # update whitespace all parse expressions defined in this module for expr in _builtin_exprs: if expr.copyDefaultWhiteChars: expr.whiteChars = chars @staticmethod def inlineLiteralsUsing(cls): """ Set class to be used for inclusion of string literals into a parser. Example:: # default literal class used is Literal integer = Word(nums) date_str = integer("year") + '/' + integer("month") + '/' + integer("day") date_str.parseString("1999/12/31") # -> ['1999', '/', '12', '/', '31'] # change to Suppress ParserElement.inlineLiteralsUsing(Suppress) date_str = integer("year") + '/' + integer("month") + '/' + integer("day") date_str.parseString("1999/12/31") # -> ['1999', '12', '31'] """ ParserElement._literalStringClass = cls def __init__(self, savelist=False): self.parseAction = list() self.failAction = None # ~ self.name = "" # don't define self.name, let subclasses try/except upcall self.strRepr = None self.resultsName = None self.saveAsList = savelist self.skipWhitespace = True self.whiteChars = set(ParserElement.DEFAULT_WHITE_CHARS) self.copyDefaultWhiteChars = True self.mayReturnEmpty = False # used when checking for left-recursion self.keepTabs = False self.ignoreExprs = list() self.debug = False self.streamlined = False self.mayIndexError = True # used to optimize exception handling for subclasses that don't advance parse index self.errmsg = "" self.modalResults = True # used to mark results names as modal (report only last) or cumulative (list all) self.debugActions = (None, None, None) # custom debug actions self.re = None self.callPreparse = True # used to avoid redundant calls to preParse self.callDuringTry = False def copy(self): """ Make a copy of this :class:`ParserElement`. Useful for defining different parse actions for the same parsing pattern, using copies of the original parse element. Example:: integer = Word(nums).setParseAction(lambda toks: int(toks[0])) integerK = integer.copy().addParseAction(lambda toks: toks[0] * 1024) + Suppress("K") integerM = integer.copy().addParseAction(lambda toks: toks[0] * 1024 * 1024) + Suppress("M") print(OneOrMore(integerK | integerM | integer).parseString("5K 100 640K 256M")) prints:: [5120, 100, 655360, 268435456] Equivalent form of ``expr.copy()`` is just ``expr()``:: integerM = integer().addParseAction(lambda toks: toks[0] * 1024 * 1024) + Suppress("M") """ cpy = copy.copy(self) cpy.parseAction = self.parseAction[:] cpy.ignoreExprs = self.ignoreExprs[:] if self.copyDefaultWhiteChars: cpy.whiteChars = ParserElement.DEFAULT_WHITE_CHARS return cpy def setName(self, name): """ Define name for this expression, makes debugging and exception messages clearer. Example:: Word(nums).parseString("ABC") # -> Exception: Expected W:(0123...) (at char 0), (line:1, col:1) Word(nums).setName("integer").parseString("ABC") # -> Exception: Expected integer (at char 0), (line:1, col:1) """ self.name = name self.errmsg = "Expected " + self.name if __diag__.enable_debug_on_named_expressions: self.setDebug() return self def setResultsName(self, name, listAllMatches=False): """ Define name for referencing matching tokens as a nested attribute of the returned parse results. NOTE: this returns a *copy* of the original :class:`ParserElement` object; this is so that the client can define a basic element, such as an integer, and reference it in multiple places with different names. You can also set results names using the abbreviated syntax, ``expr("name")`` in place of ``expr.setResultsName("name")`` - see :class:`__call__`. Example:: date_str = (integer.setResultsName("year") + '/' + integer.setResultsName("month") + '/' + integer.setResultsName("day")) # equivalent form: date_str = integer("year") + '/' + integer("month") + '/' + integer("day") """ return self._setResultsName(name, listAllMatches) def _setResultsName(self, name, listAllMatches=False): newself = self.copy() if name.endswith("*"): name = name[:-1] listAllMatches = True newself.resultsName = name newself.modalResults = not listAllMatches return newself def setBreak(self, breakFlag=True): """Method to invoke the Python pdb debugger when this element is about to be parsed. Set ``breakFlag`` to True to enable, False to disable. """ if breakFlag: _parseMethod = self._parse def breaker(instring, loc, doActions=True, callPreParse=True): import pdb # this call to pdb.set_trace() is intentional, not a checkin error pdb.set_trace() return _parseMethod(instring, loc, doActions, callPreParse) breaker._originalParseMethod = _parseMethod self._parse = breaker else: if hasattr(self._parse, "_originalParseMethod"): self._parse = self._parse._originalParseMethod return self def setParseAction(self, *fns, **kwargs): """ Define one or more actions to perform when successfully matching parse element definition. Parse action fn is a callable method with 0-3 arguments, called as ``fn(s, loc, toks)`` , ``fn(loc, toks)`` , ``fn(toks)`` , or just ``fn()`` , where: - s = the original string being parsed (see note below) - loc = the location of the matching substring - toks = a list of the matched tokens, packaged as a :class:`ParseResults` object If the functions in fns modify the tokens, they can return them as the return value from fn, and the modified list of tokens will replace the original. Otherwise, fn does not need to return any value. If None is passed as the parse action, all previously added parse actions for this expression are cleared. Optional keyword arguments: - callDuringTry = (default= ``False``) indicate if parse action should be run during lookaheads and alternate testing Note: the default parsing behavior is to expand tabs in the input string before starting the parsing process. See :class:`parseString for more information on parsing strings containing ```` s, and suggested methods to maintain a consistent view of the parsed string, the parse location, and line and column positions within the parsed string. Example:: integer = Word(nums) date_str = integer + '/' + integer + '/' + integer date_str.parseString("1999/12/31") # -> ['1999', '/', '12', '/', '31'] # use parse action to convert to ints at parse time integer = Word(nums).setParseAction(lambda toks: int(toks[0])) date_str = integer + '/' + integer + '/' + integer # note that integer fields are now ints, not strings date_str.parseString("1999/12/31") # -> [1999, '/', 12, '/', 31] """ if list(fns) == [ None, ]: self.parseAction = [] else: if not all(callable(fn) for fn in fns): raise TypeError("parse actions must be callable") self.parseAction = list(map(_trim_arity, list(fns))) self.callDuringTry = kwargs.get("callDuringTry", False) return self def addParseAction(self, *fns, **kwargs): """ Add one or more parse actions to expression's list of parse actions. See :class:`setParseAction`. See examples in :class:`copy`. """ self.parseAction += list(map(_trim_arity, list(fns))) self.callDuringTry = self.callDuringTry or kwargs.get("callDuringTry", False) return self def addCondition(self, *fns, **kwargs): """Add a boolean predicate function to expression's list of parse actions. See :class:`setParseAction` for function call signatures. Unlike ``setParseAction``, functions passed to ``addCondition`` need to return boolean success/fail of the condition. Optional keyword arguments: - message = define a custom message to be used in the raised exception - fatal = if True, will raise ParseFatalException to stop parsing immediately; otherwise will raise ParseException - callDuringTry = boolean to indicate if this method should be called during internal tryParse calls, default=False Example:: integer = Word(nums).setParseAction(lambda toks: int(toks[0])) year_int = integer.copy() year_int.addCondition(lambda toks: toks[0] >= 2000, message="Only support years 2000 and later") date_str = year_int + '/' + integer + '/' + integer result = date_str.parseString("1999/12/31") # -> Exception: Only support years 2000 and later (at char 0), (line:1, col:1) """ for fn in fns: self.parseAction.append( conditionAsParseAction( fn, message=kwargs.get("message"), fatal=kwargs.get("fatal", False) ) ) self.callDuringTry = self.callDuringTry or kwargs.get("callDuringTry", False) return self def setFailAction(self, fn): """Define action to perform if parsing fails at this expression. Fail acton fn is a callable function that takes the arguments ``fn(s, loc, expr, err)`` where: - s = string being parsed - loc = location where expression match was attempted and failed - expr = the parse expression that failed - err = the exception thrown The function returns no value. It may throw :class:`ParseFatalException` if it is desired to stop parsing immediately.""" self.failAction = fn return self def _skipIgnorables(self, instring, loc): exprsFound = True while exprsFound: exprsFound = False for e in self.ignoreExprs: try: while 1: loc, dummy = e._parse(instring, loc) exprsFound = True except ParseException: pass return loc def preParse(self, instring, loc): if self.ignoreExprs: loc = self._skipIgnorables(instring, loc) if self.skipWhitespace: wt = self.whiteChars instrlen = len(instring) while loc < instrlen and instring[loc] in wt: loc += 1 return loc def parseImpl(self, instring, loc, doActions=True): return loc, [] def postParse(self, instring, loc, tokenlist): return tokenlist # ~ @profile def _parseNoCache(self, instring, loc, doActions=True, callPreParse=True): TRY, MATCH, FAIL = 0, 1, 2 debugging = self.debug # and doActions) if debugging or self.failAction: # ~ print("Match", self, "at loc", loc, "(%d, %d)" % (lineno(loc, instring), col(loc, instring))) if self.debugActions[TRY]: self.debugActions[TRY](instring, loc, self) try: if callPreParse and self.callPreparse: preloc = self.preParse(instring, loc) else: preloc = loc tokensStart = preloc if self.mayIndexError or preloc >= len(instring): try: loc, tokens = self.parseImpl(instring, preloc, doActions) except IndexError: raise ParseException(instring, len(instring), self.errmsg, self) else: loc, tokens = self.parseImpl(instring, preloc, doActions) except Exception as err: # ~ print("Exception raised:", err) if self.debugActions[FAIL]: self.debugActions[FAIL](instring, tokensStart, self, err) if self.failAction: self.failAction(instring, tokensStart, self, err) raise else: if callPreParse and self.callPreparse: preloc = self.preParse(instring, loc) else: preloc = loc tokensStart = preloc if self.mayIndexError or preloc >= len(instring): try: loc, tokens = self.parseImpl(instring, preloc, doActions) except IndexError: raise ParseException(instring, len(instring), self.errmsg, self) else: loc, tokens = self.parseImpl(instring, preloc, doActions) tokens = self.postParse(instring, loc, tokens) retTokens = ParseResults( tokens, self.resultsName, asList=self.saveAsList, modal=self.modalResults ) if self.parseAction and (doActions or self.callDuringTry): if debugging: try: for fn in self.parseAction: try: tokens = fn(instring, tokensStart, retTokens) except IndexError as parse_action_exc: exc = ParseException("exception raised in parse action") exc.__cause__ = parse_action_exc raise exc if tokens is not None and tokens is not retTokens: retTokens = ParseResults( tokens, self.resultsName, asList=self.saveAsList and isinstance(tokens, (ParseResults, list)), modal=self.modalResults, ) except Exception as err: # ~ print "Exception raised in user parse action:", err if self.debugActions[FAIL]: self.debugActions[FAIL](instring, tokensStart, self, err) raise else: for fn in self.parseAction: try: tokens = fn(instring, tokensStart, retTokens) except IndexError as parse_action_exc: exc = ParseException("exception raised in parse action") exc.__cause__ = parse_action_exc raise exc if tokens is not None and tokens is not retTokens: retTokens = ParseResults( tokens, self.resultsName, asList=self.saveAsList and isinstance(tokens, (ParseResults, list)), modal=self.modalResults, ) if debugging: # ~ print("Matched", self, "->", retTokens.asList()) if self.debugActions[MATCH]: self.debugActions[MATCH](instring, tokensStart, loc, self, retTokens) return loc, retTokens def tryParse(self, instring, loc, raise_fatal=False): try: return self._parse(instring, loc, doActions=False)[0] except ParseFatalException: if raise_fatal: raise raise ParseException(instring, loc, self.errmsg, self) def canParseNext(self, instring, loc): try: self.tryParse(instring, loc) except (ParseException, IndexError): return False else: return True class _UnboundedCache(object): def __init__(self): cache = {} self.not_in_cache = not_in_cache = object() def get(self, key): return cache.get(key, not_in_cache) def set(self, key, value): cache[key] = value def clear(self): cache.clear() def cache_len(self): return len(cache) self.get = types.MethodType(get, self) self.set = types.MethodType(set, self) self.clear = types.MethodType(clear, self) self.__len__ = types.MethodType(cache_len, self) class _FifoCache(object): def __init__(self, size): self.not_in_cache = not_in_cache = object() cache = collections.OrderedDict() def get(self, key): return cache.get(key, not_in_cache) def set(self, key, value): cache[key] = value while len(cache) > size: try: cache.popitem(False) except KeyError: pass def clear(self): cache.clear() def cache_len(self): return len(cache) self.get = types.MethodType(get, self) self.set = types.MethodType(set, self) self.clear = types.MethodType(clear, self) self.__len__ = types.MethodType(cache_len, self) # argument cache for optimizing repeated calls when backtracking through recursive expressions packrat_cache = ( {} ) # this is set later by enabledPackrat(); this is here so that resetCache() doesn't fail packrat_cache_lock = RLock() packrat_cache_stats = [0, 0] # this method gets repeatedly called during backtracking with the same arguments - # we can cache these arguments and save ourselves the trouble of re-parsing the contained expression def _parseCache(self, instring, loc, doActions=True, callPreParse=True): HIT, MISS = 0, 1 lookup = (self, instring, loc, callPreParse, doActions) with ParserElement.packrat_cache_lock: cache = ParserElement.packrat_cache value = cache.get(lookup) if value is cache.not_in_cache: ParserElement.packrat_cache_stats[MISS] += 1 try: value = self._parseNoCache(instring, loc, doActions, callPreParse) except ParseBaseException as pe: # cache a copy of the exception, without the traceback cache.set(lookup, pe.__class__(*pe.args)) raise else: cache.set(lookup, (value[0], value[1].copy())) return value else: ParserElement.packrat_cache_stats[HIT] += 1 if isinstance(value, Exception): raise value return value[0], value[1].copy() _parse = _parseNoCache @staticmethod def resetCache(): ParserElement.packrat_cache.clear() ParserElement.packrat_cache_stats[:] = [0] * len( ParserElement.packrat_cache_stats ) _packratEnabled = False @staticmethod def enablePackrat(cache_size_limit=128): """Enables "packrat" parsing, which adds memoizing to the parsing logic. Repeated parse attempts at the same string location (which happens often in many complex grammars) can immediately return a cached value, instead of re-executing parsing/validating code. Memoizing is done of both valid results and parsing exceptions. Parameters: - cache_size_limit - (default= ``128``) - if an integer value is provided will limit the size of the packrat cache; if None is passed, then the cache size will be unbounded; if 0 is passed, the cache will be effectively disabled. This speedup may break existing programs that use parse actions that have side-effects. For this reason, packrat parsing is disabled when you first import pyparsing. To activate the packrat feature, your program must call the class method :class:`ParserElement.enablePackrat`. For best results, call ``enablePackrat()`` immediately after importing pyparsing. Example:: import pyparsing pyparsing.ParserElement.enablePackrat() """ if not ParserElement._packratEnabled: ParserElement._packratEnabled = True if cache_size_limit is None: ParserElement.packrat_cache = ParserElement._UnboundedCache() else: ParserElement.packrat_cache = ParserElement._FifoCache(cache_size_limit) ParserElement._parse = ParserElement._parseCache def parseString(self, instring, parseAll=False): """ Parse a string with respect to the parser definition. This function is intended as the primary interface to the client code. :param instring: The input string to be parsed. :param parseAll: If set, the entire input string must match the grammar. :raises ParseException: Raised if ``parseAll`` is set and the input string does not match the whole grammar. :returns: the parsed data as a :class:`ParseResults` object, which may be accessed as a `list`, a `dict`, or an object with attributes if the given parser includes results names. If the input string is required to match the entire grammar, ``parseAll`` flag must be set to True. This is also equivalent to ending the grammar with ``StringEnd()``. To report proper column numbers, ``parseString`` operates on a copy of the input string where all tabs are converted to spaces (8 spaces per tab, as per the default in ``string.expandtabs``). If the input string contains tabs and the grammar uses parse actions that use the ``loc`` argument to index into the string being parsed, one can ensure a consistent view of the input string by doing one of the following: - calling ``parseWithTabs`` on your grammar before calling ``parseString`` (see :class:`parseWithTabs`), - define your parse action using the full ``(s,loc,toks)`` signature, and reference the input string using the parse action's ``s`` argument, or - explicitly expand the tabs in your input string before calling ``parseString``. Examples: By default, partial matches are OK. >>> res = Word('a').parseString('aaaaabaaa') >>> print(res) ['aaaaa'] The parsing behavior varies by the inheriting class of this abstract class. Please refer to the children directly to see more examples. It raises an exception if parseAll flag is set and instring does not match the whole grammar. >>> res = Word('a').parseString('aaaaabaaa', parseAll=True) Traceback (most recent call last): ... pyparsing.ParseException: Expected end of text, found 'b' (at char 5), (line:1, col:6) """ ParserElement.resetCache() if not self.streamlined: self.streamline() # ~ self.saveAsList = True for e in self.ignoreExprs: e.streamline() if not self.keepTabs: instring = instring.expandtabs() try: loc, tokens = self._parse(instring, 0) if parseAll: loc = self.preParse(instring, loc) se = Empty() + StringEnd() se._parse(instring, loc) except ParseBaseException as exc: if ParserElement.verbose_stacktrace: raise else: # catch and re-raise exception from here, clears out pyparsing internal stack trace raise exc else: return tokens def scanString(self, instring, maxMatches=_MAX_INT, overlap=False): """ Scan the input string for expression matches. Each match will return the matching tokens, start location, and end location. May be called with optional ``maxMatches`` argument, to clip scanning after 'n' matches are found. If ``overlap`` is specified, then overlapping matches will be reported. Note that the start and end locations are reported relative to the string being parsed. See :class:`parseString` for more information on parsing strings with embedded tabs. Example:: source = "sldjf123lsdjjkf345sldkjf879lkjsfd987" print(source) for tokens, start, end in Word(alphas).scanString(source): print(' '*start + '^'*(end-start)) print(' '*start + tokens[0]) prints:: sldjf123lsdjjkf345sldkjf879lkjsfd987 ^^^^^ sldjf ^^^^^^^ lsdjjkf ^^^^^^ sldkjf ^^^^^^ lkjsfd """ if not self.streamlined: self.streamline() for e in self.ignoreExprs: e.streamline() if not self.keepTabs: instring = str(instring).expandtabs() instrlen = len(instring) loc = 0 preparseFn = self.preParse parseFn = self._parse ParserElement.resetCache() matches = 0 try: while loc <= instrlen and matches < maxMatches: try: preloc = preparseFn(instring, loc) nextLoc, tokens = parseFn(instring, preloc, callPreParse=False) except ParseException: loc = preloc + 1 else: if nextLoc > loc: matches += 1 yield tokens, preloc, nextLoc if overlap: nextloc = preparseFn(instring, loc) if nextloc > loc: loc = nextLoc else: loc += 1 else: loc = nextLoc else: loc = preloc + 1 except ParseBaseException as exc: if ParserElement.verbose_stacktrace: raise else: # catch and re-raise exception from here, clears out pyparsing internal stack trace raise exc def transformString(self, instring): """ Extension to :class:`scanString`, to modify matching text with modified tokens that may be returned from a parse action. To use ``transformString``, define a grammar and attach a parse action to it that modifies the returned token list. Invoking ``transformString()`` on a target string will then scan for matches, and replace the matched text patterns according to the logic in the parse action. ``transformString()`` returns the resulting transformed string. Example:: wd = Word(alphas) wd.setParseAction(lambda toks: toks[0].title()) print(wd.transformString("now is the winter of our discontent made glorious summer by this sun of york.")) prints:: Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York. """ out = [] lastE = 0 # force preservation of s, to minimize unwanted transformation of string, and to # keep string locs straight between transformString and scanString self.keepTabs = True try: for t, s, e in self.scanString(instring): out.append(instring[lastE:s]) if t: if isinstance(t, ParseResults): out += t.asList() elif isinstance(t, list): out += t else: out.append(t) lastE = e out.append(instring[lastE:]) out = [o for o in out if o] return "".join(map(str, _flatten(out))) except ParseBaseException as exc: if ParserElement.verbose_stacktrace: raise else: # catch and re-raise exception from here, clears out pyparsing internal stack trace raise exc def searchString(self, instring, maxMatches=_MAX_INT): """ Another extension to :class:`scanString`, simplifying the access to the tokens found to match the given parse expression. May be called with optional ``maxMatches`` argument, to clip searching after 'n' matches are found. Example:: # a capitalized word starts with an uppercase letter, followed by zero or more lowercase letters cap_word = Word(alphas.upper(), alphas.lower()) print(cap_word.searchString("More than Iron, more than Lead, more than Gold I need Electricity")) # the sum() builtin can be used to merge results into a single ParseResults object print(sum(cap_word.searchString("More than Iron, more than Lead, more than Gold I need Electricity"))) prints:: [['More'], ['Iron'], ['Lead'], ['Gold'], ['I'], ['Electricity']] ['More', 'Iron', 'Lead', 'Gold', 'I', 'Electricity'] """ try: return ParseResults( [t for t, s, e in self.scanString(instring, maxMatches)] ) except ParseBaseException as exc: if ParserElement.verbose_stacktrace: raise else: # catch and re-raise exception from here, clears out pyparsing internal stack trace raise exc def split(self, instring, maxsplit=_MAX_INT, includeSeparators=False): """ Generator method to split a string using the given expression as a separator. May be called with optional ``maxsplit`` argument, to limit the number of splits; and the optional ``includeSeparators`` argument (default= ``False``), if the separating matching text should be included in the split results. Example:: punc = oneOf(list(".,;:/-!?")) print(list(punc.split("This, this?, this sentence, is badly punctuated!"))) prints:: ['This', ' this', '', ' this sentence', ' is badly punctuated', ''] """ splits = 0 last = 0 for t, s, e in self.scanString(instring, maxMatches=maxsplit): yield instring[last:s] if includeSeparators: yield t[0] last = e yield instring[last:] def __add__(self, other): """ Implementation of + operator - returns :class:`And`. Adding strings to a ParserElement converts them to :class:`Literal`s by default. Example:: greet = Word(alphas) + "," + Word(alphas) + "!" hello = "Hello, World!" print(hello, "->", greet.parseString(hello)) prints:: Hello, World! -> ['Hello', ',', 'World', '!'] ``...`` may be used as a parse expression as a short form of :class:`SkipTo`. Literal('start') + ... + Literal('end') is equivalent to: Literal('start') + SkipTo('end')("_skipped*") + Literal('end') Note that the skipped text is returned with '_skipped' as a results name, and to support having multiple skips in the same parser, the value returned is a list of all skipped text. """ if other is Ellipsis: return _PendingSkip(self) if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return And([self, other]) def __radd__(self, other): """ Implementation of + operator when left operand is not a :class:`ParserElement` """ if other is Ellipsis: return SkipTo(self)("_skipped*") + self if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return other + self def __sub__(self, other): """ Implementation of - operator, returns :class:`And` with error stop """ if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return self + And._ErrorStop() + other def __rsub__(self, other): """ Implementation of - operator when left operand is not a :class:`ParserElement` """ if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return other - self def __mul__(self, other): """ Implementation of * operator, allows use of ``expr * 3`` in place of ``expr + expr + expr``. Expressions may also me multiplied by a 2-integer tuple, similar to ``{min, max}`` multipliers in regular expressions. Tuples may also include ``None`` as in: - ``expr*(n, None)`` or ``expr*(n, )`` is equivalent to ``expr*n + ZeroOrMore(expr)`` (read as "at least n instances of ``expr``") - ``expr*(None, n)`` is equivalent to ``expr*(0, n)`` (read as "0 to n instances of ``expr``") - ``expr*(None, None)`` is equivalent to ``ZeroOrMore(expr)`` - ``expr*(1, None)`` is equivalent to ``OneOrMore(expr)`` Note that ``expr*(None, n)`` does not raise an exception if more than n exprs exist in the input stream; that is, ``expr*(None, n)`` does not enforce a maximum number of expr occurrences. If this behavior is desired, then write ``expr*(None, n) + ~expr`` """ if other is Ellipsis: other = (0, None) elif isinstance(other, tuple) and other[:1] == (Ellipsis,): other = ((0,) + other[1:] + (None,))[:2] if isinstance(other, int): minElements, optElements = other, 0 elif isinstance(other, tuple): other = tuple(o if o is not Ellipsis else None for o in other) other = (other + (None, None))[:2] if other[0] is None: other = (0, other[1]) if isinstance(other[0], int) and other[1] is None: if other[0] == 0: return ZeroOrMore(self) if other[0] == 1: return OneOrMore(self) else: return self * other[0] + ZeroOrMore(self) elif isinstance(other[0], int) and isinstance(other[1], int): minElements, optElements = other optElements -= minElements else: raise TypeError( "cannot multiply 'ParserElement' and ('%s', '%s') objects", type(other[0]), type(other[1]), ) else: raise TypeError( "cannot multiply 'ParserElement' and '%s' objects", type(other) ) if minElements < 0: raise ValueError("cannot multiply ParserElement by negative value") if optElements < 0: raise ValueError( "second tuple value must be greater or equal to first tuple value" ) if minElements == optElements == 0: raise ValueError("cannot multiply ParserElement by 0 or (0, 0)") if optElements: def makeOptionalList(n): if n > 1: return Optional(self + makeOptionalList(n - 1)) else: return Optional(self) if minElements: if minElements == 1: ret = self + makeOptionalList(optElements) else: ret = And([self] * minElements) + makeOptionalList(optElements) else: ret = makeOptionalList(optElements) else: if minElements == 1: ret = self else: ret = And([self] * minElements) return ret def __rmul__(self, other): return self.__mul__(other) def __or__(self, other): """ Implementation of | operator - returns :class:`MatchFirst` """ if other is Ellipsis: return _PendingSkip(self, must_skip=True) if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return MatchFirst([self, other]) def __ror__(self, other): """ Implementation of | operator when left operand is not a :class:`ParserElement` """ if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return other | self def __xor__(self, other): """ Implementation of ^ operator - returns :class:`Or` """ if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return Or([self, other]) def __rxor__(self, other): """ Implementation of ^ operator when left operand is not a :class:`ParserElement` """ if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return other ^ self def __and__(self, other): """ Implementation of & operator - returns :class:`Each` """ if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return Each([self, other]) def __rand__(self, other): """ Implementation of & operator when left operand is not a :class:`ParserElement` """ if isinstance(other, str_type): other = self._literalStringClass(other) if not isinstance(other, ParserElement): warnings.warn( "Cannot combine element of type %s with ParserElement" % type(other), SyntaxWarning, stacklevel=2, ) return None return other & self def __invert__(self): """ Implementation of ~ operator - returns :class:`NotAny` """ return NotAny(self) def __iter__(self): # must implement __iter__ to override legacy use of sequential access to __getitem__ to # iterate over a sequence raise TypeError("%r object is not iterable" % self.__class__.__name__) def __getitem__(self, key): """ use ``[]`` indexing notation as a short form for expression repetition: - ``expr[n]`` is equivalent to ``expr*n`` - ``expr[m, n]`` is equivalent to ``expr*(m, n)`` - ``expr[n, ...]`` or ``expr[n,]`` is equivalent to ``expr*n + ZeroOrMore(expr)`` (read as "at least n instances of ``expr``") - ``expr[..., n]`` is equivalent to ``expr*(0, n)`` (read as "0 to n instances of ``expr``") - ``expr[...]`` and ``expr[0, ...]`` are equivalent to ``ZeroOrMore(expr)`` - ``expr[1, ...]`` is equivalent to ``OneOrMore(expr)`` ``None`` may be used in place of ``...``. Note that ``expr[..., n]`` and ``expr[m, n]``do not raise an exception if more than ``n`` ``expr``s exist in the input stream. If this behavior is desired, then write ``expr[..., n] + ~expr``. """ # convert single arg keys to tuples try: if isinstance(key, str_type): key = (key,) iter(key) except TypeError: key = (key, key) if len(key) > 2: warnings.warn( "only 1 or 2 index arguments supported ({}{})".format( key[:5], "... [{}]".format(len(key)) if len(key) > 5 else "" ) ) # clip to 2 elements ret = self * tuple(key[:2]) return ret def __call__(self, name=None): """ Shortcut for :class:`setResultsName`, with ``listAllMatches=False``. If ``name`` is given with a trailing ``'*'`` character, then ``listAllMatches`` will be passed as ``True``. If ``name` is omitted, same as calling :class:`copy`. Example:: # these are equivalent userdata = Word(alphas).setResultsName("name") + Word(nums + "-").setResultsName("socsecno") userdata = Word(alphas)("name") + Word(nums + "-")("socsecno") """ if name is not None: return self._setResultsName(name) else: return self.copy() def suppress(self): """ Suppresses the output of this :class:`ParserElement`; useful to keep punctuation from cluttering up returned output. """ return Suppress(self) def leaveWhitespace(self): """ Disables the skipping of whitespace before matching the characters in the :class:`ParserElement`'s defined pattern. This is normally only used internally by the pyparsing module, but may be needed in some whitespace-sensitive grammars. """ self.skipWhitespace = False return self def setWhitespaceChars(self, chars, copy_defaults=False): """ Overrides the default whitespace chars """ self.skipWhitespace = True self.whiteChars = chars self.copyDefaultWhiteChars = copy_defaults return self def parseWithTabs(self): """ Overrides default behavior to expand ````s to spaces before parsing the input string. Must be called before ``parseString`` when the input grammar contains elements that match ```` characters. """ self.keepTabs = True return self def ignore(self, other): """ Define expression to be ignored (e.g., comments) while doing pattern matching; may be called repeatedly, to define multiple comment or other ignorable patterns. Example:: patt = OneOrMore(Word(alphas)) patt.parseString('ablaj /* comment */ lskjd') # -> ['ablaj'] patt.ignore(cStyleComment) patt.parseString('ablaj /* comment */ lskjd') # -> ['ablaj', 'lskjd'] """ if isinstance(other, str_type): other = Suppress(other) if isinstance(other, Suppress): if other not in self.ignoreExprs: self.ignoreExprs.append(other) else: self.ignoreExprs.append(Suppress(other.copy())) return self def setDebugActions(self, startAction, successAction, exceptionAction): """ Enable display of debugging messages while doing pattern matching. """ self.debugActions = ( startAction or _defaultStartDebugAction, successAction or _defaultSuccessDebugAction, exceptionAction or _defaultExceptionDebugAction, ) self.debug = True return self def setDebug(self, flag=True): """ Enable display of debugging messages while doing pattern matching. Set ``flag`` to True to enable, False to disable. Example:: wd = Word(alphas).setName("alphaword") integer = Word(nums).setName("numword") term = wd | integer # turn on debugging for wd wd.setDebug() OneOrMore(term).parseString("abc 123 xyz 890") prints:: Match alphaword at loc 0(1,1) Matched alphaword -> ['abc'] Match alphaword at loc 3(1,4) Exception raised:Expected alphaword (at char 4), (line:1, col:5) Match alphaword at loc 7(1,8) Matched alphaword -> ['xyz'] Match alphaword at loc 11(1,12) Exception raised:Expected alphaword (at char 12), (line:1, col:13) Match alphaword at loc 15(1,16) Exception raised:Expected alphaword (at char 15), (line:1, col:16) The output shown is that produced by the default debug actions - custom debug actions can be specified using :class:`setDebugActions`. Prior to attempting to match the ``wd`` expression, the debugging message ``"Match at loc (,)"`` is shown. Then if the parse succeeds, a ``"Matched"`` message is shown, or an ``"Exception raised"`` message is shown. Also note the use of :class:`setName` to assign a human-readable name to the expression, which makes debugging and exception messages easier to understand - for instance, the default name created for the :class:`Word` expression without calling ``setName`` is ``"W:(ABCD...)"``. """ if flag: self.setDebugActions( _defaultStartDebugAction, _defaultSuccessDebugAction, _defaultExceptionDebugAction, ) else: self.debug = False return self def __str__(self): return self.name def __repr__(self): return str(self) def streamline(self): self.streamlined = True self.strRepr = None return self def checkRecursion(self, parseElementList): pass def validate(self, validateTrace=None): """ Check defined expressions for valid structure, check for infinite recursive definitions. """ self.checkRecursion([]) def parseFile(self, file_or_filename, parseAll=False): """ Execute the parse expression on the given file or filename. If a filename is specified (instead of a file object), the entire file is opened, read, and closed before parsing. """ try: file_contents = file_or_filename.read() except AttributeError: with open(file_or_filename, "r") as f: file_contents = f.read() try: return self.parseString(file_contents, parseAll) except ParseBaseException as exc: if ParserElement.verbose_stacktrace: raise else: # catch and re-raise exception from here, clears out pyparsing internal stack trace raise exc def __eq__(self, other): if self is other: return True elif isinstance(other, str_type): return self.matches(other) elif isinstance(other, ParserElement): return vars(self) == vars(other) return False def __hash__(self): return id(self) def __req__(self, other): return self == other def __rne__(self, other): return not (self == other) def matches(self, testString, parseAll=True): """ Method for quick testing of a parser against a test string. Good for simple inline microtests of sub expressions while building up larger parser. Parameters: - testString - to test against this expression for a match - parseAll - (default= ``True``) - flag to pass to :class:`parseString` when running tests Example:: expr = Word(nums) assert expr.matches("100") """ try: self.parseString(str(testString), parseAll=parseAll) return True except ParseBaseException: return False def runTests( self, tests, parseAll=True, comment="#", fullDump=True, printResults=True, failureTests=False, postParse=None, file=None, ): """ Execute the parse expression on a series of test strings, showing each test, the parsed results or where the parse failed. Quick and easy way to run a parse expression against a list of sample strings. Parameters: - tests - a list of separate test strings, or a multiline string of test strings - parseAll - (default= ``True``) - flag to pass to :class:`parseString` when running tests - comment - (default= ``'#'``) - expression for indicating embedded comments in the test string; pass None to disable comment filtering - fullDump - (default= ``True``) - dump results as list followed by results names in nested outline; if False, only dump nested list - printResults - (default= ``True``) prints test output to stdout - failureTests - (default= ``False``) indicates if these tests are expected to fail parsing - postParse - (default= ``None``) optional callback for successful parse results; called as `fn(test_string, parse_results)` and returns a string to be added to the test output - file - (default=``None``) optional file-like object to which test output will be written; if None, will default to ``sys.stdout`` Returns: a (success, results) tuple, where success indicates that all tests succeeded (or failed if ``failureTests`` is True), and the results contain a list of lines of each test's output Example:: number_expr = pyparsing_common.number.copy() result = number_expr.runTests(''' # unsigned integer 100 # negative integer -100 # float with scientific notation 6.02e23 # integer with scientific notation 1e-12 ''') print("Success" if result[0] else "Failed!") result = number_expr.runTests(''' # stray character 100Z # missing leading digit before '.' -.100 # too many '.' 3.14.159 ''', failureTests=True) print("Success" if result[0] else "Failed!") prints:: # unsigned integer 100 [100] # negative integer -100 [-100] # float with scientific notation 6.02e23 [6.02e+23] # integer with scientific notation 1e-12 [1e-12] Success # stray character 100Z ^ FAIL: Expected end of text (at char 3), (line:1, col:4) # missing leading digit before '.' -.100 ^ FAIL: Expected {real number with scientific notation | real number | signed integer} (at char 0), (line:1, col:1) # too many '.' 3.14.159 ^ FAIL: Expected end of text (at char 4), (line:1, col:5) Success Each test string must be on a single line. If you want to test a string that spans multiple lines, create a test like this:: expr.runTest(r"this is a test\\n of strings that spans \\n 3 lines") (Note that this is a raw string literal, you must include the leading 'r'.) """ if isinstance(tests, str_type): tests = list(map(type(tests).strip, tests.rstrip().splitlines())) if isinstance(comment, str_type): comment = Literal(comment) if file is None: file = sys.stdout print_ = file.write allResults = [] comments = [] success = True NL = Literal(r"\n").addParseAction(replaceWith("\n")).ignore(quotedString) BOM = "\ufeff" for t in tests: if comment is not None and comment.matches(t, False) or comments and not t: comments.append(t) continue if not t: continue out = ["\n".join(comments), t] comments = [] try: # convert newline marks to actual newlines, and strip leading BOM if present t = NL.transformString(t.lstrip(BOM)) result = self.parseString(t, parseAll=parseAll) except ParseBaseException as pe: fatal = "(FATAL)" if isinstance(pe, ParseFatalException) else "" if "\n" in t: out.append(line(pe.loc, t)) out.append(" " * (col(pe.loc, t) - 1) + "^" + fatal) else: out.append(" " * pe.loc + "^" + fatal) out.append("FAIL: " + str(pe)) success = success and failureTests result = pe except Exception as exc: out.append("FAIL-EXCEPTION: " + str(exc)) success = success and failureTests result = exc else: success = success and not failureTests if postParse is not None: try: pp_value = postParse(t, result) if pp_value is not None: if isinstance(pp_value, ParseResults): out.append(pp_value.dump()) else: out.append(str(pp_value)) else: out.append(result.dump()) except Exception as e: out.append(result.dump(full=fullDump)) out.append( "{} failed: {}: {}".format( postParse.__name__, type(e).__name__, e ) ) else: out.append(result.dump(full=fullDump)) out.append("") if printResults: print_("\n".join(out)) allResults.append((t, result)) return success, allResults class _PendingSkip(ParserElement): # internal placeholder class to hold a place were '...' is added to a parser element, # once another ParserElement is added, this placeholder will be replaced with a SkipTo def __init__(self, expr, must_skip=False): super().__init__() self.strRepr = str(expr + Empty()).replace("Empty", "...") self.name = self.strRepr self.anchor = expr self.must_skip = must_skip def __add__(self, other): skipper = SkipTo(other).setName("...")("_skipped*") if self.must_skip: def must_skip(t): if not t._skipped or t._skipped.asList() == [""]: del t[0] t.pop("_skipped", None) def show_skip(t): if t._skipped.asList()[-1:] == [""]: skipped = t.pop("_skipped") t["_skipped"] = "missing <" + repr(self.anchor) + ">" return ( self.anchor + skipper().addParseAction(must_skip) | skipper().addParseAction(show_skip) ) + other return self.anchor + skipper + other def __repr__(self): return self.strRepr def parseImpl(self, *args): raise Exception( "use of `...` expression without following SkipTo target expression" ) class Token(ParserElement): """Abstract :class:`ParserElement` subclass, for defining atomic matching patterns. """ def __init__(self): super().__init__(savelist=False) class Empty(Token): """An empty token, will always match. """ def __init__(self): super().__init__() self.name = "Empty" self.mayReturnEmpty = True self.mayIndexError = False class NoMatch(Token): """A token that will never match. """ def __init__(self): super().__init__() self.name = "NoMatch" self.mayReturnEmpty = True self.mayIndexError = False self.errmsg = "Unmatchable token" def parseImpl(self, instring, loc, doActions=True): raise ParseException(instring, loc, self.errmsg, self) class Literal(Token): """Token to exactly match a specified string. Example:: Literal('blah').parseString('blah') # -> ['blah'] Literal('blah').parseString('blahfooblah') # -> ['blah'] Literal('blah').parseString('bla') # -> Exception: Expected "blah" For case-insensitive matching, use :class:`CaselessLiteral`. For keyword matching (force word break before and after the matched string), use :class:`Keyword` or :class:`CaselessKeyword`. """ def __init__(self, matchString): super().__init__() self.match = matchString self.matchLen = len(matchString) try: self.firstMatchChar = matchString[0] except IndexError: warnings.warn( "null string passed to Literal; use Empty() instead", SyntaxWarning, stacklevel=2, ) self.__class__ = Empty self.name = '"%s"' % str(self.match) self.errmsg = "Expected " + self.name self.mayReturnEmpty = False self.mayIndexError = False # Performance tuning: modify __class__ to select # a parseImpl optimized for single-character check if self.matchLen == 1 and type(self) is Literal: self.__class__ = _SingleCharLiteral def parseImpl(self, instring, loc, doActions=True): if instring[loc] == self.firstMatchChar and instring.startswith( self.match, loc ): return loc + self.matchLen, self.match raise ParseException(instring, loc, self.errmsg, self) class _SingleCharLiteral(Literal): def parseImpl(self, instring, loc, doActions=True): if instring[loc] == self.firstMatchChar: return loc + 1, self.match raise ParseException(instring, loc, self.errmsg, self) _L = Literal ParserElement._literalStringClass = Literal class Keyword(Token): """Token to exactly match a specified string as a keyword, that is, it must be immediately followed by a non-keyword character. Compare with :class:`Literal`: - ``Literal("if")`` will match the leading ``'if'`` in ``'ifAndOnlyIf'``. - ``Keyword("if")`` will not; it will only match the leading ``'if'`` in ``'if x=1'``, or ``'if(y==2)'`` Accepts two optional constructor arguments in addition to the keyword string: - ``identChars`` is a string of characters that would be valid identifier characters, defaulting to all alphanumerics + "_" and "$" - ``caseless`` allows case-insensitive matching, default is ``False``. Example:: Keyword("start").parseString("start") # -> ['start'] Keyword("start").parseString("starting") # -> Exception For case-insensitive matching, use :class:`CaselessKeyword`. """ DEFAULT_KEYWORD_CHARS = alphanums + "_$" def __init__(self, matchString, identChars=None, caseless=False): super().__init__() if identChars is None: identChars = Keyword.DEFAULT_KEYWORD_CHARS self.match = matchString self.matchLen = len(matchString) try: self.firstMatchChar = matchString[0] except IndexError: warnings.warn( "null string passed to Keyword; use Empty() instead", SyntaxWarning, stacklevel=2, ) self.name = '"%s"' % self.match self.errmsg = "Expected " + self.name self.mayReturnEmpty = False self.mayIndexError = False self.caseless = caseless if caseless: self.caselessmatch = matchString.upper() identChars = identChars.upper() self.identChars = set(identChars) def parseImpl(self, instring, loc, doActions=True): if self.caseless: if ( (instring[loc : loc + self.matchLen].upper() == self.caselessmatch) and ( loc >= len(instring) - self.matchLen or instring[loc + self.matchLen].upper() not in self.identChars ) and (loc == 0 or instring[loc - 1].upper() not in self.identChars) ): return loc + self.matchLen, self.match else: if instring[loc] == self.firstMatchChar: if ( (self.matchLen == 1 or instring.startswith(self.match, loc)) and ( loc >= len(instring) - self.matchLen or instring[loc + self.matchLen] not in self.identChars ) and (loc == 0 or instring[loc - 1] not in self.identChars) ): return loc + self.matchLen, self.match raise ParseException(instring, loc, self.errmsg, self) def copy(self): c = super().copy() c.identChars = Keyword.DEFAULT_KEYWORD_CHARS return c @staticmethod def setDefaultKeywordChars(chars): """Overrides the default Keyword chars """ Keyword.DEFAULT_KEYWORD_CHARS = chars class CaselessLiteral(Literal): """Token to match a specified string, ignoring case of letters. Note: the matched results will always be in the case of the given match string, NOT the case of the input text. Example:: OneOrMore(CaselessLiteral("CMD")).parseString("cmd CMD Cmd10") # -> ['CMD', 'CMD', 'CMD'] (Contrast with example for :class:`CaselessKeyword`.) """ def __init__(self, matchString): super().__init__(matchString.upper()) # Preserve the defining literal. self.returnString = matchString self.name = "'%s'" % self.returnString self.errmsg = "Expected " + self.name def parseImpl(self, instring, loc, doActions=True): if instring[loc : loc + self.matchLen].upper() == self.match: return loc + self.matchLen, self.returnString raise ParseException(instring, loc, self.errmsg, self) class CaselessKeyword(Keyword): """ Caseless version of :class:`Keyword`. Example:: OneOrMore(CaselessKeyword("CMD")).parseString("cmd CMD Cmd10") # -> ['CMD', 'CMD'] (Contrast with example for :class:`CaselessLiteral`.) """ def __init__(self, matchString, identChars=None): super().__init__(matchString, identChars, caseless=True) class CloseMatch(Token): """A variation on :class:`Literal` which matches "close" matches, that is, strings with at most 'n' mismatching characters. :class:`CloseMatch` takes parameters: - ``match_string`` - string to be matched - ``maxMismatches`` - (``default=1``) maximum number of mismatches allowed to count as a match The results from a successful parse will contain the matched text from the input string and the following named results: - ``mismatches`` - a list of the positions within the match_string where mismatches were found - ``original`` - the original match_string used to compare against the input string If ``mismatches`` is an empty list, then the match was an exact match. Example:: patt = CloseMatch("ATCATCGAATGGA") patt.parseString("ATCATCGAAXGGA") # -> (['ATCATCGAAXGGA'], {'mismatches': [[9]], 'original': ['ATCATCGAATGGA']}) patt.parseString("ATCAXCGAAXGGA") # -> Exception: Expected 'ATCATCGAATGGA' (with up to 1 mismatches) (at char 0), (line:1, col:1) # exact match patt.parseString("ATCATCGAATGGA") # -> (['ATCATCGAATGGA'], {'mismatches': [[]], 'original': ['ATCATCGAATGGA']}) # close match allowing up to 2 mismatches patt = CloseMatch("ATCATCGAATGGA", maxMismatches=2) patt.parseString("ATCAXCGAAXGGA") # -> (['ATCAXCGAAXGGA'], {'mismatches': [[4, 9]], 'original': ['ATCATCGAATGGA']}) """ def __init__(self, match_string, maxMismatches=1): super().__init__() self.name = match_string self.match_string = match_string self.maxMismatches = maxMismatches self.errmsg = "Expected %r (with up to %d mismatches)" % ( self.match_string, self.maxMismatches, ) self.mayIndexError = False self.mayReturnEmpty = False def parseImpl(self, instring, loc, doActions=True): start = loc instrlen = len(instring) maxloc = start + len(self.match_string) if maxloc <= instrlen: match_string = self.match_string match_stringloc = 0 mismatches = [] maxMismatches = self.maxMismatches for match_stringloc, s_m in enumerate( zip(instring[loc:maxloc], match_string) ): src, mat = s_m if src != mat: mismatches.append(match_stringloc) if len(mismatches) > maxMismatches: break else: loc = start + match_stringloc + 1 results = ParseResults([instring[start:loc]]) results["original"] = match_string results["mismatches"] = mismatches return loc, results raise ParseException(instring, loc, self.errmsg, self) class Word(Token): """Token for matching words composed of allowed character sets. Defined with string containing all allowed initial characters, an optional string containing allowed body characters (if omitted, defaults to the initial character set), and an optional minimum, maximum, and/or exact length. The default value for ``min`` is 1 (a minimum value < 1 is not valid); the default values for ``max`` and ``exact`` are 0, meaning no maximum or exact length restriction. An optional ``excludeChars`` parameter can list characters that might be found in the input ``bodyChars`` string; useful to define a word of all printables except for one or two characters, for instance. :class:`srange` is useful for defining custom character set strings for defining ``Word`` expressions, using range notation from regular expression character sets. A common mistake is to use :class:`Word` to match a specific literal string, as in ``Word("Address")``. Remember that :class:`Word` uses the string argument to define *sets* of matchable characters. This expression would match "Add", "AAA", "dAred", or any other word made up of the characters 'A', 'd', 'r', 'e', and 's'. To match an exact literal string, use :class:`Literal` or :class:`Keyword`. pyparsing includes helper strings for building Words: - :class:`alphas` - :class:`nums` - :class:`alphanums` - :class:`hexnums` - :class:`alphas8bit` (alphabetic characters in ASCII range 128-255 - accented, tilded, umlauted, etc.) - :class:`punc8bit` (non-alphabetic characters in ASCII range 128-255 - currency, symbols, superscripts, diacriticals, etc.) - :class:`printables` (any non-whitespace character) Example:: # a word composed of digits integer = Word(nums) # equivalent to Word("0123456789") or Word(srange("0-9")) # a word with a leading capital, and zero or more lowercase capital_word = Word(alphas.upper(), alphas.lower()) # hostnames are alphanumeric, with leading alpha, and '-' hostname = Word(alphas, alphanums + '-') # roman numeral (not a strict parser, accepts invalid mix of characters) roman = Word("IVXLCDM") # any string of non-whitespace characters, except for ',' csv_value = Word(printables, excludeChars=",") """ def __init__( self, initChars, bodyChars=None, min=1, max=0, exact=0, asKeyword=False, excludeChars=None, ): super().__init__() if excludeChars: excludeChars = set(excludeChars) initChars = "".join(c for c in initChars if c not in excludeChars) if bodyChars: bodyChars = "".join(c for c in bodyChars if c not in excludeChars) self.initCharsOrig = initChars self.initChars = set(initChars) if bodyChars: self.bodyCharsOrig = bodyChars self.bodyChars = set(bodyChars) else: self.bodyCharsOrig = initChars self.bodyChars = set(initChars) self.maxSpecified = max > 0 if min < 1: raise ValueError( "cannot specify a minimum length < 1; use Optional(Word()) if zero-length word is permitted" ) self.minLen = min if max > 0: self.maxLen = max else: self.maxLen = _MAX_INT if exact > 0: self.maxLen = exact self.minLen = exact self.name = str(self) self.errmsg = "Expected " + self.name self.mayIndexError = False self.asKeyword = asKeyword if " " not in self.initCharsOrig + self.bodyCharsOrig and ( min == 1 and max == 0 and exact == 0 ): if self.bodyCharsOrig == self.initCharsOrig: self.reString = "[%s]+" % _collapseAndEscapeRegexRangeChars( self.initCharsOrig ) elif len(self.initCharsOrig) == 1: self.reString = "%s[%s]*" % ( re.escape(self.initCharsOrig), _collapseAndEscapeRegexRangeChars(self.bodyCharsOrig), ) else: self.reString = "[%s][%s]*" % ( _collapseAndEscapeRegexRangeChars(self.initCharsOrig), _collapseAndEscapeRegexRangeChars(self.bodyCharsOrig), ) if self.asKeyword: self.reString = r"\b" + self.reString + r"\b" try: self.re = re.compile(self.reString) except Exception: self.re = None else: self.re_match = self.re.match self.__class__ = _WordRegex def parseImpl(self, instring, loc, doActions=True): if instring[loc] not in self.initChars: raise ParseException(instring, loc, self.errmsg, self) start = loc loc += 1 instrlen = len(instring) bodychars = self.bodyChars maxloc = start + self.maxLen maxloc = min(maxloc, instrlen) while loc < maxloc and instring[loc] in bodychars: loc += 1 throwException = False if loc - start < self.minLen: throwException = True elif self.maxSpecified and loc < instrlen and instring[loc] in bodychars: throwException = True elif self.asKeyword: if ( start > 0 and instring[start - 1] in bodychars or loc < instrlen and instring[loc] in bodychars ): throwException = True if throwException: raise ParseException(instring, loc, self.errmsg, self) return loc, instring[start:loc] def __str__(self): try: return super().__str__() except Exception: pass if self.strRepr is None: def charsAsStr(s): if len(s) > 4: return s[:4] + "..." else: return s if self.initCharsOrig != self.bodyCharsOrig: self.strRepr = "W:(%s, %s)" % ( charsAsStr(self.initCharsOrig), charsAsStr(self.bodyCharsOrig), ) else: self.strRepr = "W:(%s)" % charsAsStr(self.initCharsOrig) return self.strRepr class _WordRegex(Word): def parseImpl(self, instring, loc, doActions=True): result = self.re_match(instring, loc) if not result: raise ParseException(instring, loc, self.errmsg, self) loc = result.end() return loc, result.group() class Char(_WordRegex): """A short-cut class for defining ``Word(characters, exact=1)``, when defining a match of any single character in a string of characters. """ def __init__(self, charset, asKeyword=False, excludeChars=None): super().__init__( charset, exact=1, asKeyword=asKeyword, excludeChars=excludeChars ) self.reString = "[%s]" % _collapseAndEscapeRegexRangeChars(self.initChars) if asKeyword: self.reString = r"\b%s\b" % self.reString self.re = re.compile(self.reString) self.re_match = self.re.match class Regex(Token): r"""Token for matching strings that match a given regular expression. Defined with string specifying the regular expression in a form recognized by the stdlib Python `re module `_. If the given regex contains named groups (defined using ``(?P...)``), these will be preserved as named parse results. If instead of the Python stdlib re module you wish to use a different RE module (such as the `regex` module), you can replace it by either building your Regex object with a compiled RE that was compiled using regex, or by replacing the imported `re` module in pyparsing with the `regex` module: Example:: realnum = Regex(r"[+-]?\d+\.\d*") date = Regex(r'(?P\d{4})-(?P\d\d?)-(?P\d\d?)') # ref: https://stackoverflow.com/questions/267399/how-do-you-match-only-valid-roman-numerals-with-a-regular-expression roman = Regex(r"M{0,4}(CM|CD|D?{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})") import regex parser = pp.Regex(regex.compile(r'[0-9]')) # or import pyparsing pyparsing.re = regex # both of these will use the regex module to compile their internal re's parser = pp.Regex(r'[0-9]') parser = pp.Word(pp.nums) """ def __init__(self, pattern, flags=0, asGroupList=False, asMatch=False): """The parameters ``pattern`` and ``flags`` are passed to the ``re.compile()`` function as-is. See the Python `re module `_ module for an explanation of the acceptable patterns and flags. """ super().__init__() if isinstance(pattern, str_type): if not pattern: warnings.warn( "null string passed to Regex; use Empty() instead", SyntaxWarning, stacklevel=2, ) self.pattern = pattern self.flags = flags try: self.re = re.compile(self.pattern, self.flags) self.reString = self.pattern except sre_constants.error: warnings.warn( "invalid pattern (%s) passed to Regex" % pattern, SyntaxWarning, stacklevel=2, ) raise elif hasattr(pattern, "pattern") and hasattr(pattern, "match"): self.re = pattern self.pattern = self.reString = pattern.pattern self.flags = flags else: raise TypeError( "Regex may only be constructed with a string or a compiled RE object" ) self.re_match = self.re.match self.name = str(self) self.errmsg = "Expected " + self.name self.mayIndexError = False self.mayReturnEmpty = True self.asGroupList = asGroupList self.asMatch = asMatch if self.asGroupList: self.parseImpl = self.parseImplAsGroupList if self.asMatch: self.parseImpl = self.parseImplAsMatch def parseImpl(self, instring, loc, doActions=True): result = self.re_match(instring, loc) if not result: raise ParseException(instring, loc, self.errmsg, self) loc = result.end() ret = ParseResults(result.group()) d = result.groupdict() if d: for k, v in d.items(): ret[k] = v return loc, ret def parseImplAsGroupList(self, instring, loc, doActions=True): result = self.re_match(instring, loc) if not result: raise ParseException(instring, loc, self.errmsg, self) loc = result.end() ret = result.groups() return loc, ret def parseImplAsMatch(self, instring, loc, doActions=True): result = self.re_match(instring, loc) if not result: raise ParseException(instring, loc, self.errmsg, self) loc = result.end() ret = result return loc, ret def __str__(self): try: return super().__str__() except Exception: pass if self.strRepr is None: self.strRepr = "Re:(%s)" % repr(self.pattern) return self.strRepr def sub(self, repl): r""" Return Regex with an attached parse action to transform the parsed result as if called using `re.sub(expr, repl, string) `_. Example:: make_html = Regex(r"(\w+):(.*?):").sub(r"<\1>\2") print(make_html.transformString("h1:main title:")) # prints "

main title

" """ if self.asGroupList: warnings.warn( "cannot use sub() with Regex(asGroupList=True)", SyntaxWarning, stacklevel=2, ) raise SyntaxError() if self.asMatch and callable(repl): warnings.warn( "cannot use sub() with a callable with Regex(asMatch=True)", SyntaxWarning, stacklevel=2, ) raise SyntaxError() if self.asMatch: def pa(tokens): return tokens[0].expand(repl) else: def pa(tokens): return self.re.sub(repl, tokens[0]) return self.addParseAction(pa) class QuotedString(Token): r""" Token for matching strings that are delimited by quoting characters. Defined with the following parameters: - quoteChar - string of one or more characters defining the quote delimiting string - escChar - character to escape quotes, typically backslash (default= ``None``) - escQuote - special quote sequence to escape an embedded quote string (such as SQL's ``""`` to escape an embedded ``"``) (default= ``None``) - multiline - boolean indicating whether quotes can span multiple lines (default= ``False``) - unquoteResults - boolean indicating whether the matched text should be unquoted (default= ``True``) - endQuoteChar - string of one or more characters defining the end of the quote delimited string (default= ``None`` => same as quoteChar) - convertWhitespaceEscapes - convert escaped whitespace (``'\t'``, ``'\n'``, etc.) to actual whitespace (default= ``True``) Example:: qs = QuotedString('"') print(qs.searchString('lsjdf "This is the quote" sldjf')) complex_qs = QuotedString('{{', endQuoteChar='}}') print(complex_qs.searchString('lsjdf {{This is the "quote"}} sldjf')) sql_qs = QuotedString('"', escQuote='""') print(sql_qs.searchString('lsjdf "This is the quote with ""embedded"" quotes" sldjf')) prints:: [['This is the quote']] [['This is the "quote"']] [['This is the quote with "embedded" quotes']] """ def __init__( self, quoteChar, escChar=None, escQuote=None, multiline=False, unquoteResults=True, endQuoteChar=None, convertWhitespaceEscapes=True, ): super().__init__() # remove white space from quote chars - wont work anyway quoteChar = quoteChar.strip() if not quoteChar: warnings.warn( "quoteChar cannot be the empty string", SyntaxWarning, stacklevel=2 ) raise SyntaxError() if endQuoteChar is None: endQuoteChar = quoteChar else: endQuoteChar = endQuoteChar.strip() if not endQuoteChar: warnings.warn( "endQuoteChar cannot be the empty string", SyntaxWarning, stacklevel=2, ) raise SyntaxError() self.quoteChar = quoteChar self.quoteCharLen = len(quoteChar) self.firstQuoteChar = quoteChar[0] self.endQuoteChar = endQuoteChar self.endQuoteCharLen = len(endQuoteChar) self.escChar = escChar self.escQuote = escQuote self.unquoteResults = unquoteResults self.convertWhitespaceEscapes = convertWhitespaceEscapes if multiline: self.flags = re.MULTILINE | re.DOTALL self.pattern = r"%s(?:[^%s%s]" % ( re.escape(self.quoteChar), _escapeRegexRangeChars(self.endQuoteChar[0]), (escChar is not None and _escapeRegexRangeChars(escChar) or ""), ) else: self.flags = 0 self.pattern = r"%s(?:[^%s\n\r%s]" % ( re.escape(self.quoteChar), _escapeRegexRangeChars(self.endQuoteChar[0]), (escChar is not None and _escapeRegexRangeChars(escChar) or ""), ) if len(self.endQuoteChar) > 1: self.pattern += ( "|(?:" + ")|(?:".join( "%s[^%s]" % ( re.escape(self.endQuoteChar[:i]), _escapeRegexRangeChars(self.endQuoteChar[i]), ) for i in range(len(self.endQuoteChar) - 1, 0, -1) ) + ")" ) if escQuote: self.pattern += r"|(?:%s)" % re.escape(escQuote) if escChar: self.pattern += r"|(?:%s.)" % re.escape(escChar) self.escCharReplacePattern = re.escape(self.escChar) + "(.)" self.pattern += r")*%s" % re.escape(self.endQuoteChar) try: self.re = re.compile(self.pattern, self.flags) self.reString = self.pattern self.re_match = self.re.match except sre_constants.error: warnings.warn( "invalid pattern (%s) passed to Regex" % self.pattern, SyntaxWarning, stacklevel=2, ) raise self.name = str(self) self.errmsg = "Expected " + self.name self.mayIndexError = False self.mayReturnEmpty = True def parseImpl(self, instring, loc, doActions=True): result = ( instring[loc] == self.firstQuoteChar and self.re_match(instring, loc) or None ) if not result: raise ParseException(instring, loc, self.errmsg, self) loc = result.end() ret = result.group() if self.unquoteResults: # strip off quotes ret = ret[self.quoteCharLen : -self.endQuoteCharLen] if isinstance(ret, str_type): # replace escaped whitespace if "\\" in ret and self.convertWhitespaceEscapes: ws_map = { r"\t": "\t", r"\n": "\n", r"\f": "\f", r"\r": "\r", } for wslit, wschar in ws_map.items(): ret = ret.replace(wslit, wschar) # replace escaped characters if self.escChar: ret = re.sub(self.escCharReplacePattern, r"\g<1>", ret) # replace escaped quotes if self.escQuote: ret = ret.replace(self.escQuote, self.endQuoteChar) return loc, ret def __str__(self): try: return super().__str__() except Exception: pass if self.strRepr is None: self.strRepr = "quoted string, starting with %s ending with %s" % ( self.quoteChar, self.endQuoteChar, ) return self.strRepr class CharsNotIn(Token): """Token for matching words composed of characters *not* in a given set (will include whitespace in matched characters if not listed in the provided exclusion set - see example). Defined with string containing all disallowed characters, and an optional minimum, maximum, and/or exact length. The default value for ``min`` is 1 (a minimum value < 1 is not valid); the default values for ``max`` and ``exact`` are 0, meaning no maximum or exact length restriction. Example:: # define a comma-separated-value as anything that is not a ',' csv_value = CharsNotIn(',') print(delimitedList(csv_value).parseString("dkls,lsdkjf,s12 34,@!#,213")) prints:: ['dkls', 'lsdkjf', 's12 34', '@!#', '213'] """ def __init__(self, notChars, min=1, max=0, exact=0): super().__init__() self.skipWhitespace = False self.notChars = notChars if min < 1: raise ValueError( "cannot specify a minimum length < 1; use " "Optional(CharsNotIn()) if zero-length char group is permitted" ) self.minLen = min if max > 0: self.maxLen = max else: self.maxLen = _MAX_INT if exact > 0: self.maxLen = exact self.minLen = exact self.name = str(self) self.errmsg = "Expected " + self.name self.mayReturnEmpty = self.minLen == 0 self.mayIndexError = False def parseImpl(self, instring, loc, doActions=True): if instring[loc] in self.notChars: raise ParseException(instring, loc, self.errmsg, self) start = loc loc += 1 notchars = self.notChars maxlen = min(start + self.maxLen, len(instring)) while loc < maxlen and instring[loc] not in notchars: loc += 1 if loc - start < self.minLen: raise ParseException(instring, loc, self.errmsg, self) return loc, instring[start:loc] def __str__(self): try: return super().__str__() except Exception: pass if self.strRepr is None: if len(self.notChars) > 4: self.strRepr = "!W:(%s...)" % self.notChars[:4] else: self.strRepr = "!W:(%s)" % self.notChars return self.strRepr class White(Token): """Special matching class for matching whitespace. Normally, whitespace is ignored by pyparsing grammars. This class is included when some whitespace structures are significant. Define with a string containing the whitespace characters to be matched; default is ``" \\t\\r\\n"``. Also takes optional ``min``, ``max``, and ``exact`` arguments, as defined for the :class:`Word` class. """ whiteStrs = { " ": "", "\t": "", "\n": "", "\r": "", "\f": "", "u\00A0": "", "u\1680": "", "u\180E": "", "u\2000": "", "u\2001": "", "u\2002": "", "u\2003": "", "u\2004": "", "u\2005": "", "u\2006": "", "u\2007": "", "u\2008": "", "u\2009": "", "u\200A": "", "u\200B": "", "u\202F": "", "u\205F": "", "u\3000": "", } def __init__(self, ws=" \t\r\n", min=1, max=0, exact=0): super().__init__() self.matchWhite = ws self.setWhitespaceChars( "".join(c for c in self.whiteChars if c not in self.matchWhite), copy_defaults=True, ) # ~ self.leaveWhitespace() self.name = "".join(White.whiteStrs[c] for c in self.matchWhite) self.mayReturnEmpty = True self.errmsg = "Expected " + self.name self.minLen = min if max > 0: self.maxLen = max else: self.maxLen = _MAX_INT if exact > 0: self.maxLen = exact self.minLen = exact def parseImpl(self, instring, loc, doActions=True): if instring[loc] not in self.matchWhite: raise ParseException(instring, loc, self.errmsg, self) start = loc loc += 1 maxloc = start + self.maxLen maxloc = min(maxloc, len(instring)) while loc < maxloc and instring[loc] in self.matchWhite: loc += 1 if loc - start < self.minLen: raise ParseException(instring, loc, self.errmsg, self) return loc, instring[start:loc] class _PositionToken(Token): def __init__(self): super().__init__() self.name = self.__class__.__name__ self.mayReturnEmpty = True self.mayIndexError = False class GoToColumn(_PositionToken): """Token to advance to a specific column of input text; useful for tabular report scraping. """ def __init__(self, colno): super().__init__() self.col = colno def preParse(self, instring, loc): if col(loc, instring) != self.col: instrlen = len(instring) if self.ignoreExprs: loc = self._skipIgnorables(instring, loc) while ( loc < instrlen and instring[loc].isspace() and col(loc, instring) != self.col ): loc += 1 return loc def parseImpl(self, instring, loc, doActions=True): thiscol = col(loc, instring) if thiscol > self.col: raise ParseException(instring, loc, "Text not in expected column", self) newloc = loc + self.col - thiscol ret = instring[loc:newloc] return newloc, ret class LineStart(_PositionToken): r"""Matches if current position is at the beginning of a line within the parse string Example:: test = '''\ AAA this line AAA and this line AAA but not this one B AAA and definitely not this one ''' for t in (LineStart() + 'AAA' + restOfLine).searchString(test): print(t) prints:: ['AAA', ' this line'] ['AAA', ' and this line'] """ def __init__(self): super().__init__() self.errmsg = "Expected start of line" def parseImpl(self, instring, loc, doActions=True): if col(loc, instring) == 1: return loc, [] raise ParseException(instring, loc, self.errmsg, self) class LineEnd(_PositionToken): """Matches if current position is at the end of a line within the parse string """ def __init__(self): super().__init__() self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n", ""), copy_defaults=False ) self.errmsg = "Expected end of line" def parseImpl(self, instring, loc, doActions=True): if loc < len(instring): if instring[loc] == "\n": return loc + 1, "\n" else: raise ParseException(instring, loc, self.errmsg, self) elif loc == len(instring): return loc + 1, [] else: raise ParseException(instring, loc, self.errmsg, self) class StringStart(_PositionToken): """Matches if current position is at the beginning of the parse string """ def __init__(self): super().__init__() self.errmsg = "Expected start of text" def parseImpl(self, instring, loc, doActions=True): if loc != 0: # see if entire string up to here is just whitespace and ignoreables if loc != self.preParse(instring, 0): raise ParseException(instring, loc, self.errmsg, self) return loc, [] class StringEnd(_PositionToken): """Matches if current position is at the end of the parse string """ def __init__(self): super().__init__() self.errmsg = "Expected end of text" def parseImpl(self, instring, loc, doActions=True): if loc < len(instring): raise ParseException(instring, loc, self.errmsg, self) elif loc == len(instring): return loc + 1, [] elif loc > len(instring): return loc, [] else: raise ParseException(instring, loc, self.errmsg, self) class WordStart(_PositionToken): """Matches if the current position is at the beginning of a Word, and is not preceded by any character in a given set of ``wordChars`` (default= ``printables``). To emulate the ``\b`` behavior of regular expressions, use ``WordStart(alphanums)``. ``WordStart`` will also match at the beginning of the string being parsed, or at the beginning of a line. """ def __init__(self, wordChars=printables): super().__init__() self.wordChars = set(wordChars) self.errmsg = "Not at the start of a word" def parseImpl(self, instring, loc, doActions=True): if loc != 0: if ( instring[loc - 1] in self.wordChars or instring[loc] not in self.wordChars ): raise ParseException(instring, loc, self.errmsg, self) return loc, [] class WordEnd(_PositionToken): """Matches if the current position is at the end of a Word, and is not followed by any character in a given set of ``wordChars`` (default= ``printables``). To emulate the ``\b`` behavior of regular expressions, use ``WordEnd(alphanums)``. ``WordEnd`` will also match at the end of the string being parsed, or at the end of a line. """ def __init__(self, wordChars=printables): super().__init__() self.wordChars = set(wordChars) self.skipWhitespace = False self.errmsg = "Not at the end of a word" def parseImpl(self, instring, loc, doActions=True): instrlen = len(instring) if instrlen > 0 and loc < instrlen: if ( instring[loc] in self.wordChars or instring[loc - 1] not in self.wordChars ): raise ParseException(instring, loc, self.errmsg, self) return loc, [] class ParseExpression(ParserElement): """Abstract subclass of ParserElement, for combining and post-processing parsed tokens. """ def __init__(self, exprs, savelist=False): super().__init__(savelist) if isinstance(exprs, _generatorType): exprs = list(exprs) if isinstance(exprs, str_type): self.exprs = [self._literalStringClass(exprs)] elif isinstance(exprs, ParserElement): self.exprs = [exprs] elif isinstance(exprs, Iterable): exprs = list(exprs) # if sequence of strings provided, wrap with Literal if any(isinstance(expr, str_type) for expr in exprs): exprs = ( self._literalStringClass(e) if isinstance(e, str_type) else e for e in exprs ) self.exprs = list(exprs) else: try: self.exprs = list(exprs) except TypeError: self.exprs = [exprs] self.callPreparse = False def append(self, other): self.exprs.append(other) self.strRepr = None return self def leaveWhitespace(self): """Extends ``leaveWhitespace`` defined in base class, and also invokes ``leaveWhitespace`` on all contained expressions.""" self.skipWhitespace = False self.exprs = [e.copy() for e in self.exprs] for e in self.exprs: e.leaveWhitespace() return self def ignore(self, other): if isinstance(other, Suppress): if other not in self.ignoreExprs: super().ignore(other) for e in self.exprs: e.ignore(self.ignoreExprs[-1]) else: super().ignore(other) for e in self.exprs: e.ignore(self.ignoreExprs[-1]) return self def __str__(self): try: return super().__str__() except Exception: pass if self.strRepr is None: self.strRepr = "%s:(%s)" % (self.__class__.__name__, str(self.exprs)) return self.strRepr def streamline(self): super().streamline() for e in self.exprs: e.streamline() # collapse nested And's of the form And(And(And(a, b), c), d) to And(a, b, c, d) # but only if there are no parse actions or resultsNames on the nested And's # (likewise for Or's and MatchFirst's) if len(self.exprs) == 2: other = self.exprs[0] if ( isinstance(other, self.__class__) and not other.parseAction and other.resultsName is None and not other.debug ): self.exprs = other.exprs[:] + [self.exprs[1]] self.strRepr = None self.mayReturnEmpty |= other.mayReturnEmpty self.mayIndexError |= other.mayIndexError other = self.exprs[-1] if ( isinstance(other, self.__class__) and not other.parseAction and other.resultsName is None and not other.debug ): self.exprs = self.exprs[:-1] + other.exprs[:] self.strRepr = None self.mayReturnEmpty |= other.mayReturnEmpty self.mayIndexError |= other.mayIndexError self.errmsg = "Expected " + str(self) return self def validate(self, validateTrace=None): tmp = (validateTrace if validateTrace is not None else [])[:] + [self] for e in self.exprs: e.validate(tmp) self.checkRecursion([]) def copy(self): ret = super().copy() ret.exprs = [e.copy() for e in self.exprs] return ret def _setResultsName(self, name, listAllMatches=False): if __diag__.warn_ungrouped_named_tokens_in_collection: for e in self.exprs: if isinstance(e, ParserElement) and e.resultsName: warnings.warn( "{}: setting results name {!r} on {} expression " "collides with {!r} on contained expression".format( "warn_ungrouped_named_tokens_in_collection", name, type(self).__name__, e.resultsName, ), stacklevel=3, ) return super()._setResultsName(name, listAllMatches) class And(ParseExpression): """ Requires all given :class:`ParseExpression` s to be found in the given order. Expressions may be separated by whitespace. May be constructed using the ``'+'`` operator. May also be constructed using the ``'-'`` operator, which will suppress backtracking. Example:: integer = Word(nums) name_expr = OneOrMore(Word(alphas)) expr = And([integer("id"), name_expr("name"), integer("age")]) # more easily written as: expr = integer("id") + name_expr("name") + integer("age") """ class _ErrorStop(Empty): def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) self.name = "-" self.leaveWhitespace() def __init__(self, exprs, savelist=True): if exprs and Ellipsis in exprs: tmp = [] for i, expr in enumerate(exprs): if expr is Ellipsis: if i < len(exprs) - 1: skipto_arg = (Empty() + exprs[i + 1]).exprs[-1] tmp.append(SkipTo(skipto_arg)("_skipped*")) else: raise Exception( "cannot construct And with sequence ending in ..." ) else: tmp.append(expr) exprs[:] = tmp super().__init__(exprs, savelist) self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs) self.setWhitespaceChars( self.exprs[0].whiteChars, copy_defaults=self.exprs[0].copyDefaultWhiteChars ) self.skipWhitespace = self.exprs[0].skipWhitespace self.callPreparse = True def streamline(self): # collapse any _PendingSkip's if self.exprs: if any( isinstance(e, ParseExpression) and e.exprs and isinstance(e.exprs[-1], _PendingSkip) for e in self.exprs[:-1] ): for i, e in enumerate(self.exprs[:-1]): if e is None: continue if ( isinstance(e, ParseExpression) and e.exprs and isinstance(e.exprs[-1], _PendingSkip) ): e.exprs[-1] = e.exprs[-1] + self.exprs[i + 1] self.exprs[i + 1] = None self.exprs = [e for e in self.exprs if e is not None] super().streamline() self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs) return self def parseImpl(self, instring, loc, doActions=True): # pass False as last arg to _parse for first element, since we already # pre-parsed the string as part of our And pre-parsing loc, resultlist = self.exprs[0]._parse( instring, loc, doActions, callPreParse=False ) errorStop = False for e in self.exprs[1:]: if isinstance(e, And._ErrorStop): errorStop = True continue if errorStop: try: loc, exprtokens = e._parse(instring, loc, doActions) except ParseSyntaxException: raise except ParseBaseException as pe: pe.__traceback__ = None raise ParseSyntaxException._from_exception(pe) except IndexError: raise ParseSyntaxException( instring, len(instring), self.errmsg, self ) else: loc, exprtokens = e._parse(instring, loc, doActions) if exprtokens or exprtokens.haskeys(): resultlist += exprtokens return loc, resultlist def __iadd__(self, other): if isinstance(other, str_type): other = self._literalStringClass(other) return self.append(other) # And([self, other]) def checkRecursion(self, parseElementList): subRecCheckList = parseElementList[:] + [self] for e in self.exprs: e.checkRecursion(subRecCheckList) if not e.mayReturnEmpty: break def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "{" + " ".join(str(e) for e in self.exprs) + "}" return self.strRepr class Or(ParseExpression): """Requires that at least one :class:`ParseExpression` is found. If two expressions match, the expression that matches the longest string will be used. May be constructed using the ``'^'`` operator. Example:: # construct Or using '^' operator number = Word(nums) ^ Combine(Word(nums) + '.' + Word(nums)) print(number.searchString("123 3.1416 789")) prints:: [['123'], ['3.1416'], ['789']] """ def __init__(self, exprs, savelist=False): super().__init__(exprs, savelist) if self.exprs: self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs) else: self.mayReturnEmpty = True def streamline(self): super().streamline() self.saveAsList = any(e.saveAsList for e in self.exprs) return self def parseImpl(self, instring, loc, doActions=True): maxExcLoc = -1 maxException = None matches = [] fatals = [] for e in self.exprs: try: loc2 = e.tryParse(instring, loc, raise_fatal=True) except ParseFatalException as pfe: pfe.__traceback__ = None pfe.parserElement = e fatals.append(pfe) maxException = None maxExcLoc = -1 except ParseException as err: if not fatals: err.__traceback__ = None if err.loc > maxExcLoc: maxException = err maxExcLoc = err.loc except IndexError: if len(instring) > maxExcLoc: maxException = ParseException( instring, len(instring), e.errmsg, self ) maxExcLoc = len(instring) else: # save match among all matches, to retry longest to shortest matches.append((loc2, e)) if matches: # re-evaluate all matches in descending order of length of match, in case attached actions # might change whether or how much they match of the input. matches.sort(key=itemgetter(0), reverse=True) if not doActions: # no further conditions or parse actions to change the selection of # alternative, so the first match will be the best match best_expr = matches[0][1] return best_expr._parse(instring, loc, doActions) longest = -1, None for loc1, expr1 in matches: if loc1 <= longest[0]: # already have a longer match than this one will deliver, we are done return longest try: loc2, toks = expr1._parse(instring, loc, doActions) except ParseException as err: err.__traceback__ = None if err.loc > maxExcLoc: maxException = err maxExcLoc = err.loc else: if loc2 >= loc1: return loc2, toks # didn't match as much as before elif loc2 > longest[0]: longest = loc2, toks if longest != (-1, None): return longest if fatals: if len(fatals) > 1: fatals.sort(key=lambda e: -e.loc) if fatals[0].loc == fatals[1].loc: fatals.sort(key=lambda e: (-e.loc, -len(str(e.parserElement)))) max_fatal = fatals[0] raise max_fatal if maxException is not None: maxException.msg = self.errmsg raise maxException else: raise ParseException( instring, loc, "no defined alternatives to match", self ) def __ixor__(self, other): if isinstance(other, str_type): other = self._literalStringClass(other) return self.append(other) # Or([self, other]) def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "{" + " ^ ".join(str(e) for e in self.exprs) + "}" return self.strRepr def checkRecursion(self, parseElementList): subRecCheckList = parseElementList[:] + [self] for e in self.exprs: e.checkRecursion(subRecCheckList) def _setResultsName(self, name, listAllMatches=False): if __diag__.warn_multiple_tokens_in_named_alternation: if any(isinstance(e, And) for e in self.exprs): warnings.warn( "{}: setting results name {!r} on {} expression " "will return a list of all parsed tokens in an And alternative, " "in prior versions only the first token was returned".format( "warn_multiple_tokens_in_named_alternation", name, type(self).__name__, ), stacklevel=3, ) return super()._setResultsName(name, listAllMatches) class MatchFirst(ParseExpression): """Requires that at least one :class:`ParseExpression` is found. If two expressions match, the first one listed is the one that will match. May be constructed using the ``'|'`` operator. Example:: # construct MatchFirst using '|' operator # watch the order of expressions to match number = Word(nums) | Combine(Word(nums) + '.' + Word(nums)) print(number.searchString("123 3.1416 789")) # Fail! -> [['123'], ['3'], ['1416'], ['789']] # put more selective expression first number = Combine(Word(nums) + '.' + Word(nums)) | Word(nums) print(number.searchString("123 3.1416 789")) # Better -> [['123'], ['3.1416'], ['789']] """ def __init__(self, exprs, savelist=False): super().__init__(exprs, savelist) if self.exprs: self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs) else: self.mayReturnEmpty = True def streamline(self): super().streamline() self.saveAsList = any(e.saveAsList for e in self.exprs) return self def parseImpl(self, instring, loc, doActions=True): maxExcLoc = -1 maxException = None fatals = [] for e in self.exprs: try: ret = e._parse(instring, loc, doActions) return ret except ParseFatalException as pfe: pfe.__traceback__ = None pfe.parserElement = e fatals.append(pfe) maxException = None except ParseException as err: if not fatals and err.loc > maxExcLoc: maxException = err maxExcLoc = err.loc except IndexError: if len(instring) > maxExcLoc: maxException = ParseException( instring, len(instring), e.errmsg, self ) maxExcLoc = len(instring) # only got here if no expression matched, raise exception for match that made it the furthest if fatals: if len(fatals) > 1: fatals.sort(key=lambda e: -e.loc) if fatals[0].loc == fatals[1].loc: fatals.sort(key=lambda e: (-e.loc, -len(str(e.parserElement)))) max_fatal = fatals[0] raise max_fatal if maxException is not None: maxException.msg = self.errmsg raise maxException else: raise ParseException( instring, loc, "no defined alternatives to match", self ) def __ior__(self, other): if isinstance(other, str_type): other = self._literalStringClass(other) return self.append(other) # MatchFirst([self, other]) def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "{" + " | ".join(str(e) for e in self.exprs) + "}" return self.strRepr def checkRecursion(self, parseElementList): subRecCheckList = parseElementList[:] + [self] for e in self.exprs: e.checkRecursion(subRecCheckList) def _setResultsName(self, name, listAllMatches=False): if __diag__.warn_multiple_tokens_in_named_alternation: if any(isinstance(e, And) for e in self.exprs): warnings.warn( "{}: setting results name {!r} on {} expression " "may only return a single token for an And alternative, " "in future will return the full list of tokens".format( "warn_multiple_tokens_in_named_alternation", name, type(self).__name__, ), stacklevel=3, ) return super()._setResultsName(name, listAllMatches) class Each(ParseExpression): """Requires all given :class:`ParseExpression` s to be found, but in any order. Expressions may be separated by whitespace. May be constructed using the ``'&'`` operator. Example:: color = oneOf("RED ORANGE YELLOW GREEN BLUE PURPLE BLACK WHITE BROWN") shape_type = oneOf("SQUARE CIRCLE TRIANGLE STAR HEXAGON OCTAGON") integer = Word(nums) shape_attr = "shape:" + shape_type("shape") posn_attr = "posn:" + Group(integer("x") + ',' + integer("y"))("posn") color_attr = "color:" + color("color") size_attr = "size:" + integer("size") # use Each (using operator '&') to accept attributes in any order # (shape and posn are required, color and size are optional) shape_spec = shape_attr & posn_attr & Optional(color_attr) & Optional(size_attr) shape_spec.runTests(''' shape: SQUARE color: BLACK posn: 100, 120 shape: CIRCLE size: 50 color: BLUE posn: 50,80 color:GREEN size:20 shape:TRIANGLE posn:20,40 ''' ) prints:: shape: SQUARE color: BLACK posn: 100, 120 ['shape:', 'SQUARE', 'color:', 'BLACK', 'posn:', ['100', ',', '120']] - color: BLACK - posn: ['100', ',', '120'] - x: 100 - y: 120 - shape: SQUARE shape: CIRCLE size: 50 color: BLUE posn: 50,80 ['shape:', 'CIRCLE', 'size:', '50', 'color:', 'BLUE', 'posn:', ['50', ',', '80']] - color: BLUE - posn: ['50', ',', '80'] - x: 50 - y: 80 - shape: CIRCLE - size: 50 color: GREEN size: 20 shape: TRIANGLE posn: 20,40 ['color:', 'GREEN', 'size:', '20', 'shape:', 'TRIANGLE', 'posn:', ['20', ',', '40']] - color: GREEN - posn: ['20', ',', '40'] - x: 20 - y: 40 - shape: TRIANGLE - size: 20 """ def __init__(self, exprs, savelist=True): super().__init__(exprs, savelist) self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs) self.skipWhitespace = True self.initExprGroups = True self.saveAsList = True def streamline(self): super().streamline() self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs) return self def parseImpl(self, instring, loc, doActions=True): if self.initExprGroups: self.opt1map = dict( (id(e.expr), e) for e in self.exprs if isinstance(e, Optional) ) opt1 = [e.expr for e in self.exprs if isinstance(e, Optional)] opt2 = [ e for e in self.exprs if e.mayReturnEmpty and not isinstance(e, Optional) ] self.optionals = opt1 + opt2 self.multioptionals = [ e.expr for e in self.exprs if isinstance(e, ZeroOrMore) ] self.multirequired = [ e.expr for e in self.exprs if isinstance(e, OneOrMore) ] self.required = [ e for e in self.exprs if not isinstance(e, (Optional, ZeroOrMore, OneOrMore)) ] self.required += self.multirequired self.initExprGroups = False tmpLoc = loc tmpReqd = self.required[:] tmpOpt = self.optionals[:] matchOrder = [] keepMatching = True failed = [] fatals = [] while keepMatching: tmpExprs = tmpReqd + tmpOpt + self.multioptionals + self.multirequired failed.clear() fatals.clear() for e in tmpExprs: try: tmpLoc = e.tryParse(instring, tmpLoc, raise_fatal=True) except ParseFatalException as pfe: pfe.__traceback__ = None pfe.parserElement = e fatals.append(pfe) failed.append(e) except ParseException: failed.append(e) else: matchOrder.append(self.opt1map.get(id(e), e)) if e in tmpReqd: tmpReqd.remove(e) elif e in tmpOpt: tmpOpt.remove(e) if len(failed) == len(tmpExprs): keepMatching = False # look for any ParseFatalExceptions if fatals: if len(fatals) > 1: fatals.sort(key=lambda e: -e.loc) if fatals[0].loc == fatals[1].loc: fatals.sort(key=lambda e: (-e.loc, -len(str(e.parserElement)))) max_fatal = fatals[0] raise max_fatal if tmpReqd: missing = ", ".join(str(e) for e in tmpReqd) raise ParseException( instring, loc, "Missing one or more required elements (%s)" % missing ) # add any unmatched Optionals, in case they have default values defined matchOrder += [ e for e in self.exprs if isinstance(e, Optional) and e.expr in tmpOpt ] resultlist = [] for e in matchOrder: loc, results = e._parse(instring, loc, doActions) resultlist.append(results) finalResults = sum(resultlist, ParseResults([])) return loc, finalResults def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "{" + " & ".join(str(e) for e in self.exprs) + "}" return self.strRepr def checkRecursion(self, parseElementList): subRecCheckList = parseElementList[:] + [self] for e in self.exprs: e.checkRecursion(subRecCheckList) class ParseElementEnhance(ParserElement): """Abstract subclass of :class:`ParserElement`, for combining and post-processing parsed tokens. """ def __init__(self, expr, savelist=False): super().__init__(savelist) if isinstance(expr, str_type): if issubclass(self._literalStringClass, Token): expr = self._literalStringClass(expr) elif issubclass(type(self), self._literalStringClass): expr = Literal(expr) else: expr = self._literalStringClass(Literal(expr)) self.expr = expr self.strRepr = None if expr is not None: self.mayIndexError = expr.mayIndexError self.mayReturnEmpty = expr.mayReturnEmpty self.setWhitespaceChars( expr.whiteChars, copy_defaults=expr.copyDefaultWhiteChars ) self.skipWhitespace = expr.skipWhitespace self.saveAsList = expr.saveAsList self.callPreparse = expr.callPreparse self.ignoreExprs.extend(expr.ignoreExprs) def parseImpl(self, instring, loc, doActions=True): if self.expr is not None: return self.expr._parse(instring, loc, doActions, callPreParse=False) else: raise ParseException("", loc, self.errmsg, self) def leaveWhitespace(self): self.skipWhitespace = False self.expr = self.expr.copy() if self.expr is not None: self.expr.leaveWhitespace() return self def ignore(self, other): if isinstance(other, Suppress): if other not in self.ignoreExprs: super().ignore(other) if self.expr is not None: self.expr.ignore(self.ignoreExprs[-1]) else: super().ignore(other) if self.expr is not None: self.expr.ignore(self.ignoreExprs[-1]) return self def streamline(self): super().streamline() if self.expr is not None: self.expr.streamline() return self def checkRecursion(self, parseElementList): if self in parseElementList: raise RecursiveGrammarException(parseElementList + [self]) subRecCheckList = parseElementList[:] + [self] if self.expr is not None: self.expr.checkRecursion(subRecCheckList) def validate(self, validateTrace=None): if validateTrace is None: validateTrace = [] tmp = validateTrace[:] + [self] if self.expr is not None: self.expr.validate(tmp) self.checkRecursion([]) def __str__(self): try: return super().__str__() except Exception: pass if self.strRepr is None and self.expr is not None: self.strRepr = "%s:(%s)" % (self.__class__.__name__, str(self.expr)) return self.strRepr class FollowedBy(ParseElementEnhance): """Lookahead matching of the given parse expression. ``FollowedBy`` does *not* advance the parsing position within the input string, it only verifies that the specified parse expression matches at the current position. ``FollowedBy`` always returns a null token list. If any results names are defined in the lookahead expression, those *will* be returned for access by name. Example:: # use FollowedBy to match a label only if it is followed by a ':' data_word = Word(alphas) label = data_word + FollowedBy(':') attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) OneOrMore(attr_expr).parseString("shape: SQUARE color: BLACK posn: upper left").pprint() prints:: [['shape', 'SQUARE'], ['color', 'BLACK'], ['posn', 'upper left']] """ def __init__(self, expr): super().__init__(expr) self.mayReturnEmpty = True def parseImpl(self, instring, loc, doActions=True): # by using self._expr.parse and deleting the contents of the returned ParseResults list # we keep any named results that were defined in the FollowedBy expression _, ret = self.expr._parse(instring, loc, doActions=doActions) del ret[:] return loc, ret class PrecededBy(ParseElementEnhance): """Lookbehind matching of the given parse expression. ``PrecededBy`` does not advance the parsing position within the input string, it only verifies that the specified parse expression matches prior to the current position. ``PrecededBy`` always returns a null token list, but if a results name is defined on the given expression, it is returned. Parameters: - expr - expression that must match prior to the current parse location - retreat - (default= ``None``) - (int) maximum number of characters to lookbehind prior to the current parse location If the lookbehind expression is a string, Literal, Keyword, or a Word or CharsNotIn with a specified exact or maximum length, then the retreat parameter is not required. Otherwise, retreat must be specified to give a maximum number of characters to look back from the current parse position for a lookbehind match. Example:: # VB-style variable names with type prefixes int_var = PrecededBy("#") + pyparsing_common.identifier str_var = PrecededBy("$") + pyparsing_common.identifier """ def __init__(self, expr, retreat=None): super().__init__(expr) self.expr = self.expr().leaveWhitespace() self.mayReturnEmpty = True self.mayIndexError = False self.exact = False if isinstance(expr, str_type): retreat = len(expr) self.exact = True elif isinstance(expr, (Literal, Keyword)): retreat = expr.matchLen self.exact = True elif isinstance(expr, (Word, CharsNotIn)) and expr.maxLen != _MAX_INT: retreat = expr.maxLen self.exact = True elif isinstance(expr, _PositionToken): retreat = 0 self.exact = True self.retreat = retreat self.errmsg = "not preceded by " + str(expr) self.skipWhitespace = False self.parseAction.append(lambda s, l, t: t.__delitem__(slice(None, None))) def parseImpl(self, instring, loc=0, doActions=True): if self.exact: if loc < self.retreat: raise ParseException(instring, loc, self.errmsg) start = loc - self.retreat _, ret = self.expr._parse(instring, start) else: # retreat specified a maximum lookbehind window, iterate test_expr = self.expr + StringEnd() instring_slice = instring[max(0, loc - self.retreat) : loc] last_expr = ParseException(instring, loc, self.errmsg) for offset in range(1, min(loc, self.retreat + 1) + 1): try: # print('trying', offset, instring_slice, repr(instring_slice[loc - offset:])) _, ret = test_expr._parse( instring_slice, len(instring_slice) - offset ) except ParseBaseException as pbe: last_expr = pbe else: break else: raise last_expr return loc, ret class NotAny(ParseElementEnhance): """Lookahead to disallow matching with the given parse expression. ``NotAny`` does *not* advance the parsing position within the input string, it only verifies that the specified parse expression does *not* match at the current position. Also, ``NotAny`` does *not* skip over leading whitespace. ``NotAny`` always returns a null token list. May be constructed using the '~' operator. Example:: AND, OR, NOT = map(CaselessKeyword, "AND OR NOT".split()) # take care not to mistake keywords for identifiers ident = ~(AND | OR | NOT) + Word(alphas) boolean_term = Optional(NOT) + ident # very crude boolean expression - to support parenthesis groups and # operation hierarchy, use infixNotation boolean_expr = boolean_term + ZeroOrMore((AND | OR) + boolean_term) # integers that are followed by "." are actually floats integer = Word(nums) + ~Char(".") """ def __init__(self, expr): super().__init__(expr) # ~ self.leaveWhitespace() self.skipWhitespace = ( False # do NOT use self.leaveWhitespace(), don't want to propagate to exprs ) self.mayReturnEmpty = True self.errmsg = "Found unwanted token, " + str(self.expr) def parseImpl(self, instring, loc, doActions=True): if self.expr.canParseNext(instring, loc): raise ParseException(instring, loc, self.errmsg, self) return loc, [] def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "~{" + str(self.expr) + "}" return self.strRepr class _MultipleMatch(ParseElementEnhance): def __init__(self, expr, stopOn=None): super().__init__(expr) self.saveAsList = True ender = stopOn if isinstance(ender, str_type): ender = self._literalStringClass(ender) self.stopOn(ender) def stopOn(self, ender): if isinstance(ender, str_type): ender = self._literalStringClass(ender) self.not_ender = ~ender if ender is not None else None return self def parseImpl(self, instring, loc, doActions=True): self_expr_parse = self.expr._parse self_skip_ignorables = self._skipIgnorables check_ender = self.not_ender is not None if check_ender: try_not_ender = self.not_ender.tryParse # must be at least one (but first see if we are the stopOn sentinel; # if so, fail) if check_ender: try_not_ender(instring, loc) loc, tokens = self_expr_parse(instring, loc, doActions, callPreParse=False) try: hasIgnoreExprs = not not self.ignoreExprs while 1: if check_ender: try_not_ender(instring, loc) if hasIgnoreExprs: preloc = self_skip_ignorables(instring, loc) else: preloc = loc loc, tmptokens = self_expr_parse(instring, preloc, doActions) if tmptokens or tmptokens.haskeys(): tokens += tmptokens except (ParseException, IndexError): pass return loc, tokens def _setResultsName(self, name, listAllMatches=False): if __diag__.warn_ungrouped_named_tokens_in_collection: for e in [self.expr] + getattr(self.expr, "exprs", []): if isinstance(e, ParserElement) and e.resultsName: warnings.warn( "{}: setting results name {!r} on {} expression " "collides with {!r} on contained expression".format( "warn_ungrouped_named_tokens_in_collection", name, type(self).__name__, e.resultsName, ), stacklevel=3, ) return super()._setResultsName(name, listAllMatches) class OneOrMore(_MultipleMatch): """Repetition of one or more of the given expression. Parameters: - expr - expression that must match one or more times - stopOn - (default= ``None``) - expression for a terminating sentinel (only required if the sentinel would ordinarily match the repetition expression) Example:: data_word = Word(alphas) label = data_word + FollowedBy(':') attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).setParseAction(' '.join)) text = "shape: SQUARE posn: upper left color: BLACK" OneOrMore(attr_expr).parseString(text).pprint() # Fail! read 'color' as data instead of next label -> [['shape', 'SQUARE color']] # use stopOn attribute for OneOrMore to avoid reading label string as part of the data attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) OneOrMore(attr_expr).parseString(text).pprint() # Better -> [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'BLACK']] # could also be written as (attr_expr * (1,)).parseString(text).pprint() """ def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "{" + str(self.expr) + "}..." return self.strRepr class ZeroOrMore(_MultipleMatch): """Optional repetition of zero or more of the given expression. Parameters: - expr - expression that must match zero or more times - stopOn - (default= ``None``) - expression for a terminating sentinel (only required if the sentinel would ordinarily match the repetition expression) Example: similar to :class:`OneOrMore` """ def __init__(self, expr, stopOn=None): super().__init__(expr, stopOn=stopOn) self.mayReturnEmpty = True def parseImpl(self, instring, loc, doActions=True): try: return super().parseImpl(instring, loc, doActions) except (ParseException, IndexError): return loc, ParseResults([], name=self.resultsName) def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "[" + str(self.expr) + "]..." return self.strRepr class _NullToken(object): def __bool__(self): return False def __str__(self): return "" class Optional(ParseElementEnhance): """Optional matching of the given expression. Parameters: - expr - expression that must match zero or more times - default (optional) - value to be returned if the optional expression is not found. Example:: # US postal code can be a 5-digit zip, plus optional 4-digit qualifier zip = Combine(Word(nums, exact=5) + Optional('-' + Word(nums, exact=4))) zip.runTests(''' # traditional ZIP code 12345 # ZIP+4 form 12101-0001 # invalid ZIP 98765- ''') prints:: # traditional ZIP code 12345 ['12345'] # ZIP+4 form 12101-0001 ['12101-0001'] # invalid ZIP 98765- ^ FAIL: Expected end of text (at char 5), (line:1, col:6) """ __optionalNotMatched = _NullToken() def __init__(self, expr, default=__optionalNotMatched): super().__init__(expr, savelist=False) self.saveAsList = self.expr.saveAsList self.defaultValue = default self.mayReturnEmpty = True def parseImpl(self, instring, loc, doActions=True): try: loc, tokens = self.expr._parse(instring, loc, doActions, callPreParse=False) except (ParseException, IndexError): if self.defaultValue is not self.__optionalNotMatched: if self.expr.resultsName: tokens = ParseResults([self.defaultValue]) tokens[self.expr.resultsName] = self.defaultValue else: tokens = [self.defaultValue] else: tokens = [] return loc, tokens def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is None: self.strRepr = "[" + str(self.expr) + "]" return self.strRepr class SkipTo(ParseElementEnhance): """Token for skipping over all undefined text until the matched expression is found. Parameters: - expr - target expression marking the end of the data to be skipped - include - (default= ``False``) if True, the target expression is also parsed (the skipped text and target expression are returned as a 2-element list). - ignore - (default= ``None``) used to define grammars (typically quoted strings and comments) that might contain false matches to the target expression - failOn - (default= ``None``) define expressions that are not allowed to be included in the skipped test; if found before the target expression is found, the SkipTo is not a match Example:: report = ''' Outstanding Issues Report - 1 Jan 2000 # | Severity | Description | Days Open -----+----------+-------------------------------------------+----------- 101 | Critical | Intermittent system crash | 6 94 | Cosmetic | Spelling error on Login ('log|n') | 14 79 | Minor | System slow when running too many reports | 47 ''' integer = Word(nums) SEP = Suppress('|') # use SkipTo to simply match everything up until the next SEP # - ignore quoted strings, so that a '|' character inside a quoted string does not match # - parse action will call token.strip() for each matched token, i.e., the description body string_data = SkipTo(SEP, ignore=quotedString) string_data.setParseAction(tokenMap(str.strip)) ticket_expr = (integer("issue_num") + SEP + string_data("sev") + SEP + string_data("desc") + SEP + integer("days_open")) for tkt in ticket_expr.searchString(report): print tkt.dump() prints:: ['101', 'Critical', 'Intermittent system crash', '6'] - days_open: 6 - desc: Intermittent system crash - issue_num: 101 - sev: Critical ['94', 'Cosmetic', "Spelling error on Login ('log|n')", '14'] - days_open: 14 - desc: Spelling error on Login ('log|n') - issue_num: 94 - sev: Cosmetic ['79', 'Minor', 'System slow when running too many reports', '47'] - days_open: 47 - desc: System slow when running too many reports - issue_num: 79 - sev: Minor """ def __init__(self, other, include=False, ignore=None, failOn=None): super().__init__(other) self.ignoreExpr = ignore self.mayReturnEmpty = True self.mayIndexError = False self.includeMatch = include self.saveAsList = False if isinstance(failOn, str_type): self.failOn = self._literalStringClass(failOn) else: self.failOn = failOn self.errmsg = "No match found for " + str(self.expr) def parseImpl(self, instring, loc, doActions=True): startloc = loc instrlen = len(instring) expr = self.expr expr_parse = self.expr._parse self_failOn_canParseNext = ( self.failOn.canParseNext if self.failOn is not None else None ) self_ignoreExpr_tryParse = ( self.ignoreExpr.tryParse if self.ignoreExpr is not None else None ) tmploc = loc while tmploc <= instrlen: if self_failOn_canParseNext is not None: # break if failOn expression matches if self_failOn_canParseNext(instring, tmploc): break if self_ignoreExpr_tryParse is not None: # advance past ignore expressions while 1: try: tmploc = self_ignoreExpr_tryParse(instring, tmploc) except ParseBaseException: break try: expr_parse(instring, tmploc, doActions=False, callPreParse=False) except (ParseException, IndexError): # no match, advance loc in string tmploc += 1 else: # matched skipto expr, done break else: # ran off the end of the input string without matching skipto expr, fail raise ParseException(instring, loc, self.errmsg, self) # build up return values loc = tmploc skiptext = instring[startloc:loc] skipresult = ParseResults(skiptext) if self.includeMatch: loc, mat = expr_parse(instring, loc, doActions, callPreParse=False) skipresult += mat return loc, skipresult class Forward(ParseElementEnhance): """Forward declaration of an expression to be defined later - used for recursive grammars, such as algebraic infix notation. When the expression is known, it is assigned to the ``Forward`` variable using the '<<' operator. Note: take care when assigning to ``Forward`` not to overlook precedence of operators. Specifically, '|' has a lower precedence than '<<', so that:: fwdExpr << a | b | c will actually be evaluated as:: (fwdExpr << a) | b | c thereby leaving b and c out as parseable alternatives. It is recommended that you explicitly group the values inserted into the ``Forward``:: fwdExpr << (a | b | c) Converting to use the '<<=' operator instead will avoid this problem. See :class:`ParseResults.pprint` for an example of a recursive parser created using ``Forward``. """ def __init__(self, other=None): super().__init__(other, savelist=False) def __lshift__(self, other): if isinstance(other, str_type): other = self._literalStringClass(other) self.expr = other self.strRepr = None self.mayIndexError = self.expr.mayIndexError self.mayReturnEmpty = self.expr.mayReturnEmpty self.setWhitespaceChars( self.expr.whiteChars, copy_defaults=self.expr.copyDefaultWhiteChars ) self.skipWhitespace = self.expr.skipWhitespace self.saveAsList = self.expr.saveAsList self.ignoreExprs.extend(self.expr.ignoreExprs) return self def __ilshift__(self, other): return self << other def leaveWhitespace(self): self.skipWhitespace = False return self def streamline(self): if not self.streamlined: self.streamlined = True if self.expr is not None: self.expr.streamline() return self def validate(self, validateTrace=None): if validateTrace is None: validateTrace = [] if self not in validateTrace: tmp = validateTrace[:] + [self] if self.expr is not None: self.expr.validate(tmp) self.checkRecursion([]) def __str__(self): if hasattr(self, "name"): return self.name if self.strRepr is not None: return self.strRepr # Avoid infinite recursion by setting a temporary strRepr self.strRepr = ": ..." # Use the string representation of main expression. retString = "..." try: if self.expr is not None: retString = str(self.expr)[:1000] else: retString = "None" finally: self.strRepr = self.__class__.__name__ + ": " + retString return self.strRepr def copy(self): if self.expr is not None: return super().copy() else: ret = Forward() ret <<= self return ret def _setResultsName(self, name, listAllMatches=False): if __diag__.warn_name_set_on_empty_Forward: if self.expr is None: warnings.warn( "{}: setting results name {!r} on {} expression " "that has no contained expression".format( "warn_name_set_on_empty_Forward", name, type(self).__name__ ), stacklevel=3, ) return super()._setResultsName(name, listAllMatches) class TokenConverter(ParseElementEnhance): """ Abstract subclass of :class:`ParseExpression`, for converting parsed results. """ def __init__(self, expr, savelist=False): super().__init__(expr) # , savelist) self.saveAsList = False class Combine(TokenConverter): """Converter to concatenate all matching tokens to a single string. By default, the matching patterns must also be contiguous in the input string; this can be disabled by specifying ``'adjacent=False'`` in the constructor. Example:: real = Word(nums) + '.' + Word(nums) print(real.parseString('3.1416')) # -> ['3', '.', '1416'] # will also erroneously match the following print(real.parseString('3. 1416')) # -> ['3', '.', '1416'] real = Combine(Word(nums) + '.' + Word(nums)) print(real.parseString('3.1416')) # -> ['3.1416'] # no match when there are internal spaces print(real.parseString('3. 1416')) # -> Exception: Expected W:(0123...) """ def __init__(self, expr, joinString="", adjacent=True): super().__init__(expr) # suppress whitespace-stripping in contained parse expressions, but re-enable it on the Combine itself if adjacent: self.leaveWhitespace() self.adjacent = adjacent self.skipWhitespace = True self.joinString = joinString self.callPreparse = True def ignore(self, other): if self.adjacent: ParserElement.ignore(self, other) else: super().ignore(other) return self def postParse(self, instring, loc, tokenlist): retToks = tokenlist.copy() del retToks[:] retToks += ParseResults( ["".join(tokenlist._asStringList(self.joinString))], modal=self.modalResults ) if self.resultsName and retToks.haskeys(): return [retToks] else: return retToks class Group(TokenConverter): """Converter to return the matched tokens as a list - useful for returning tokens of :class:`ZeroOrMore` and :class:`OneOrMore` expressions. Example:: ident = Word(alphas) num = Word(nums) term = ident | num func = ident + Optional(delimitedList(term)) print(func.parseString("fn a, b, 100")) # -> ['fn', 'a', 'b', '100'] func = ident + Group(Optional(delimitedList(term))) print(func.parseString("fn a, b, 100")) # -> ['fn', ['a', 'b', '100']] """ def __init__(self, expr): super().__init__(expr) self.saveAsList = True def postParse(self, instring, loc, tokenlist): return [tokenlist] class Dict(TokenConverter): """Converter to return a repetitive expression as a list, but also as a dictionary. Each element can also be referenced using the first token in the expression as its key. Useful for tabular report scraping when the first column can be used as a item key. Example:: data_word = Word(alphas) label = data_word + FollowedBy(':') attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).setParseAction(' '.join)) text = "shape: SQUARE posn: upper left color: light blue texture: burlap" attr_expr = (label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) # print attributes as plain groups print(OneOrMore(attr_expr).parseString(text).dump()) # instead of OneOrMore(expr), parse using Dict(OneOrMore(Group(expr))) - Dict will auto-assign names result = Dict(OneOrMore(Group(attr_expr))).parseString(text) print(result.dump()) # access named fields as dict entries, or output as dict print(result['shape']) print(result.asDict()) prints:: ['shape', 'SQUARE', 'posn', 'upper left', 'color', 'light blue', 'texture', 'burlap'] [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']] - color: light blue - posn: upper left - shape: SQUARE - texture: burlap SQUARE {'color': 'light blue', 'posn': 'upper left', 'texture': 'burlap', 'shape': 'SQUARE'} See more examples at :class:`ParseResults` of accessing fields by results name. """ def __init__(self, expr): super().__init__(expr) self.saveAsList = True def postParse(self, instring, loc, tokenlist): for i, tok in enumerate(tokenlist): if len(tok) == 0: continue ikey = tok[0] if isinstance(ikey, int): ikey = str(tok[0]).strip() if len(tok) == 1: tokenlist[ikey] = _ParseResultsWithOffset("", i) elif len(tok) == 2 and not isinstance(tok[1], ParseResults): tokenlist[ikey] = _ParseResultsWithOffset(tok[1], i) else: dictvalue = tok.copy() # ParseResults(i) del dictvalue[0] if len(dictvalue) != 1 or ( isinstance(dictvalue, ParseResults) and dictvalue.haskeys() ): tokenlist[ikey] = _ParseResultsWithOffset(dictvalue, i) else: tokenlist[ikey] = _ParseResultsWithOffset(dictvalue[0], i) if self.resultsName: return [tokenlist] else: return tokenlist class Suppress(TokenConverter): """Converter for ignoring the results of a parsed expression. Example:: source = "a, b, c,d" wd = Word(alphas) wd_list1 = wd + ZeroOrMore(',' + wd) print(wd_list1.parseString(source)) # often, delimiters that are useful during parsing are just in the # way afterward - use Suppress to keep them out of the parsed output wd_list2 = wd + ZeroOrMore(Suppress(',') + wd) print(wd_list2.parseString(source)) prints:: ['a', ',', 'b', ',', 'c', ',', 'd'] ['a', 'b', 'c', 'd'] (See also :class:`delimitedList`.) """ def postParse(self, instring, loc, tokenlist): return [] def suppress(self): return self class OnlyOnce(object): """Wrapper for parse actions, to ensure they are only called once. """ def __init__(self, methodCall): self.callable = _trim_arity(methodCall) self.called = False def __call__(self, s, l, t): if not self.called: results = self.callable(s, l, t) self.called = True return results raise ParseException(s, l, "") def reset(self): self.called = False def traceParseAction(f): """Decorator for debugging parse actions. When the parse action is called, this decorator will print ``">> entering method-name(line:, , )"``. When the parse action completes, the decorator will print ``"<<"`` followed by the returned value, or any exception that the parse action raised. Example:: wd = Word(alphas) @traceParseAction def remove_duplicate_chars(tokens): return ''.join(sorted(set(''.join(tokens)))) wds = OneOrMore(wd).setParseAction(remove_duplicate_chars) print(wds.parseString("slkdjs sld sldd sdlf sdljf")) prints:: >>entering remove_duplicate_chars(line: 'slkdjs sld sldd sdlf sdljf', 0, (['slkdjs', 'sld', 'sldd', 'sdlf', 'sdljf'], {})) < 3: thisFunc = paArgs[0].__class__.__name__ + "." + thisFunc sys.stderr.write( ">>entering %s(line: '%s', %d, %r)\n" % (thisFunc, line(l, s), l, t) ) try: ret = f(*paArgs) except Exception as exc: sys.stderr.write("< ['aa', 'bb', 'cc'] delimitedList(Word(hexnums), delim=':', combine=True).parseString("AA:BB:CC:DD:EE") # -> ['AA:BB:CC:DD:EE'] """ dlName = str(expr) + " [" + str(delim) + " " + str(expr) + "]..." if combine: return Combine(expr + ZeroOrMore(delim + expr)).setName(dlName) else: return (expr + ZeroOrMore(Suppress(delim) + expr)).setName(dlName) def countedArray(expr, intExpr=None): """Helper to define a counted list of expressions. This helper defines a pattern of the form:: integer expr expr expr... where the leading integer tells how many expr expressions follow. The matched tokens returns the array of expr tokens as a list - the leading count token is suppressed. If ``intExpr`` is specified, it should be a pyparsing expression that produces an integer value. Example:: countedArray(Word(alphas)).parseString('2 ab cd ef') # -> ['ab', 'cd'] # in this parser, the leading integer value is given in binary, # '10' indicating that 2 values are in the array binaryConstant = Word('01').setParseAction(lambda t: int(t[0], 2)) countedArray(Word(alphas), intExpr=binaryConstant).parseString('10 ab cd ef') # -> ['ab', 'cd'] """ arrayExpr = Forward() def countFieldParseAction(s, l, t): n = t[0] arrayExpr << (n and Group(And([expr] * n)) or Group(empty)) return [] if intExpr is None: intExpr = Word(nums).setParseAction(lambda t: int(t[0])) else: intExpr = intExpr.copy() intExpr.setName("arrayLen") intExpr.addParseAction(countFieldParseAction, callDuringTry=True) return (intExpr + arrayExpr).setName("(len) " + str(expr) + "...") def _flatten(L): ret = [] for i in L: if isinstance(i, list): ret.extend(_flatten(i)) else: ret.append(i) return ret def matchPreviousLiteral(expr): """Helper to define an expression that is indirectly defined from the tokens matched in a previous expression, that is, it looks for a 'repeat' of a previous expression. For example:: first = Word(nums) second = matchPreviousLiteral(first) matchExpr = first + ":" + second will match ``"1:1"``, but not ``"1:2"``. Because this matches a previous literal, will also match the leading ``"1:1"`` in ``"1:10"``. If this is not desired, use :class:`matchPreviousExpr`. Do *not* use with packrat parsing enabled. """ rep = Forward() def copyTokenToRepeater(s, l, t): if t: if len(t) == 1: rep << t[0] else: # flatten t tokens tflat = _flatten(t.asList()) rep << And(Literal(tt) for tt in tflat) else: rep << Empty() expr.addParseAction(copyTokenToRepeater, callDuringTry=True) rep.setName("(prev) " + str(expr)) return rep def matchPreviousExpr(expr): """Helper to define an expression that is indirectly defined from the tokens matched in a previous expression, that is, it looks for a 'repeat' of a previous expression. For example:: first = Word(nums) second = matchPreviousExpr(first) matchExpr = first + ":" + second will match ``"1:1"``, but not ``"1:2"``. Because this matches by expressions, will *not* match the leading ``"1:1"`` in ``"1:10"``; the expressions are evaluated first, and then compared, so ``"1"`` is compared with ``"10"``. Do *not* use with packrat parsing enabled. """ rep = Forward() e2 = expr.copy() rep <<= e2 def copyTokenToRepeater(s, l, t): matchTokens = _flatten(t.asList()) def mustMatchTheseTokens(s, l, t): theseTokens = _flatten(t.asList()) if theseTokens != matchTokens: raise ParseException("", 0, "") rep.setParseAction(mustMatchTheseTokens, callDuringTry=True) expr.addParseAction(copyTokenToRepeater, callDuringTry=True) rep.setName("(prev) " + str(expr)) return rep def _escapeRegexRangeChars(s): # ~ escape these chars: ^-] for c in r"\^-]": s = s.replace(c, _bslash + c) s = s.replace("\n", r"\n") s = s.replace("\t", r"\t") return str(s) def _collapseAndEscapeRegexRangeChars(s): def is_consecutive(c): c_int = ord(c) is_consecutive.prev, prev = c_int, is_consecutive.prev if c_int - prev > 1: is_consecutive.value = next(is_consecutive.counter) return is_consecutive.value is_consecutive.prev = 0 is_consecutive.counter = itertools.count() is_consecutive.value = -1 def escape_re_range_char(c): return "\\" + c if c in r"\^-]" else c ret = [] for _, chars in itertools.groupby(sorted(s), key=is_consecutive): first = last = next(chars) for c in chars: last = c if first == last: ret.append(escape_re_range_char(first)) else: ret.append( "{}-{}".format(escape_re_range_char(first), escape_re_range_char(last)) ) return "".join(ret) def oneOf(strs, caseless=False, useRegex=True, asKeyword=False): """Helper to quickly define a set of alternative Literals, and makes sure to do longest-first testing when there is a conflict, regardless of the input order, but returns a :class:`MatchFirst` for best performance. Parameters: - strs - a string of space-delimited literals, or a collection of string literals - caseless - (default= ``False``) - treat all literals as caseless - useRegex - (default= ``True``) - as an optimization, will generate a Regex object; otherwise, will generate a :class:`MatchFirst` object (if ``caseless=True`` or ``asKeyword=True``, or if creating a :class:`Regex` raises an exception) - asKeyword - (default=``False``) - enforce Keyword-style matching on the generated expressions Example:: comp_oper = oneOf("< = > <= >= !=") var = Word(alphas) number = Word(nums) term = var | number comparison_expr = term + comp_oper + term print(comparison_expr.searchString("B = 12 AA=23 B<=AA AA>12")) prints:: [['B', '=', '12'], ['AA', '=', '23'], ['B', '<=', 'AA'], ['AA', '>', '12']] """ if isinstance(caseless, str_type): warnings.warn( "More than one string argument passed to oneOf, pass " "choices as a list or space-delimited string", stacklevel=2, ) if caseless: isequal = lambda a, b: a.upper() == b.upper() masks = lambda a, b: b.upper().startswith(a.upper()) parseElementClass = CaselessKeyword if asKeyword else CaselessLiteral else: isequal = lambda a, b: a == b masks = lambda a, b: b.startswith(a) parseElementClass = Keyword if asKeyword else Literal symbols = [] if isinstance(strs, str_type): symbols = strs.split() elif isinstance(strs, Iterable): symbols = list(strs) else: warnings.warn( "Invalid argument to oneOf, expected string or iterable", SyntaxWarning, stacklevel=2, ) if not symbols: return NoMatch() if not asKeyword: # if not producing keywords, need to reorder to take care to avoid masking # longer choices with shorter ones i = 0 while i < len(symbols) - 1: cur = symbols[i] for j, other in enumerate(symbols[i + 1 :]): if isequal(other, cur): del symbols[i + j + 1] break elif masks(cur, other): del symbols[i + j + 1] symbols.insert(i, other) break else: i += 1 if not (caseless or asKeyword) and useRegex: # ~ print(strs, "->", "|".join([_escapeRegexChars(sym) for sym in symbols])) try: if len(symbols) == len("".join(symbols)): return Regex( "[%s]" % "".join(_escapeRegexRangeChars(sym) for sym in symbols) ).setName(" | ".join(symbols)) else: return Regex("|".join(re.escape(sym) for sym in symbols)).setName( " | ".join(symbols) ) except Exception: warnings.warn( "Exception creating Regex for oneOf, building MatchFirst", SyntaxWarning, stacklevel=2, ) # last resort, just use MatchFirst return MatchFirst(parseElementClass(sym) for sym in symbols).setName( " | ".join(symbols) ) def dictOf(key, value): """Helper to easily and clearly define a dictionary by specifying the respective patterns for the key and value. Takes care of defining the :class:`Dict`, :class:`ZeroOrMore`, and :class:`Group` tokens in the proper order. The key pattern can include delimiting markers or punctuation, as long as they are suppressed, thereby leaving the significant key text. The value pattern can include named results, so that the :class:`Dict` results can include named token fields. Example:: text = "shape: SQUARE posn: upper left color: light blue texture: burlap" attr_expr = (label + Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join)) print(OneOrMore(attr_expr).parseString(text).dump()) attr_label = label attr_value = Suppress(':') + OneOrMore(data_word, stopOn=label).setParseAction(' '.join) # similar to Dict, but simpler call format result = dictOf(attr_label, attr_value).parseString(text) print(result.dump()) print(result['shape']) print(result.shape) # object attribute access works too print(result.asDict()) prints:: [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']] - color: light blue - posn: upper left - shape: SQUARE - texture: burlap SQUARE SQUARE {'color': 'light blue', 'shape': 'SQUARE', 'posn': 'upper left', 'texture': 'burlap'} """ return Dict(OneOrMore(Group(key + value))) def originalTextFor(expr, asString=True): """Helper to return the original, untokenized text for a given expression. Useful to restore the parsed fields of an HTML start tag into the raw tag text itself, or to revert separate tokens with intervening whitespace back to the original matching input text. By default, returns astring containing the original parsed text. If the optional ``asString`` argument is passed as ``False``, then the return value is a :class:`ParseResults` containing any results names that were originally matched, and a single token containing the original matched text from the input string. So if the expression passed to :class:`originalTextFor` contains expressions with defined results names, you must set ``asString`` to ``False`` if you want to preserve those results name values. Example:: src = "this is test bold text normal text " for tag in ("b", "i"): opener, closer = makeHTMLTags(tag) patt = originalTextFor(opener + SkipTo(closer) + closer) print(patt.searchString(src)[0]) prints:: [' bold text '] ['text'] """ locMarker = Empty().setParseAction(lambda s, loc, t: loc) endlocMarker = locMarker.copy() endlocMarker.callPreparse = False matchExpr = locMarker("_original_start") + expr + endlocMarker("_original_end") if asString: extractText = lambda s, l, t: s[t._original_start : t._original_end] else: def extractText(s, l, t): t[:] = [s[t.pop("_original_start") : t.pop("_original_end")]] matchExpr.setParseAction(extractText) matchExpr.ignoreExprs = expr.ignoreExprs return matchExpr def ungroup(expr): """Helper to undo pyparsing's default grouping of And expressions, even if all but one are non-empty. """ return TokenConverter(expr).addParseAction(lambda t: t[0]) def locatedExpr(expr): """Helper to decorate a returned token with its starting and ending locations in the input string. This helper adds the following results names: - locn_start = location where matched expression begins - locn_end = location where matched expression ends - value = the actual parsed results Be careful if the input text contains ```` characters, you may want to call :class:`ParserElement.parseWithTabs` Example:: wd = Word(alphas) for match in locatedExpr(wd).searchString("ljsdf123lksdjjf123lkkjj1222"): print(match) prints:: [[0, 'ljsdf', 5]] [[8, 'lksdjjf', 15]] [[18, 'lkkjj', 23]] """ locator = Empty().setParseAction(lambda s, l, t: l) return Group( locator("locn_start") + expr("value") + locator.copy().leaveWhitespace()("locn_end") ) # convenience constants for positional expressions empty = Empty().setName("empty") lineStart = LineStart().setName("lineStart") lineEnd = LineEnd().setName("lineEnd") stringStart = StringStart().setName("stringStart") stringEnd = StringEnd().setName("stringEnd") _escapedPunc = Word(_bslash, r"\[]-*.$+^?()~ ", exact=2).setParseAction( lambda s, l, t: t[0][1] ) _escapedHexChar = Regex(r"\\0?[xX][0-9a-fA-F]+").setParseAction( lambda s, l, t: chr(int(t[0].lstrip(r"\0x"), 16)) ) _escapedOctChar = Regex(r"\\0[0-7]+").setParseAction( lambda s, l, t: chr(int(t[0][1:], 8)) ) _singleChar = ( _escapedPunc | _escapedHexChar | _escapedOctChar | CharsNotIn(r"\]", exact=1) ) _charRange = Group(_singleChar + Suppress("-") + _singleChar) _reBracketExpr = ( Literal("[") + Optional("^").setResultsName("negate") + Group(OneOrMore(_charRange | _singleChar)).setResultsName("body") + "]" ) def srange(s): r"""Helper to easily define string ranges for use in Word construction. Borrows syntax from regexp '[]' string range definitions:: srange("[0-9]") -> "0123456789" srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz" srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_" The input string must be enclosed in []'s, and the returned string is the expanded character set joined into a single string. The values enclosed in the []'s may be: - a single character - an escaped character with a leading backslash (such as ``\-`` or ``\]``) - an escaped hex character with a leading ``'\x'`` (``\x21``, which is a ``'!'`` character) (``\0x##`` is also supported for backwards compatibility) - an escaped octal character with a leading ``'\0'`` (``\041``, which is a ``'!'`` character) - a range of any of the above, separated by a dash (``'a-z'``, etc.) - any combination of the above (``'aeiouy'``, ``'a-zA-Z0-9_$'``, etc.) """ _expanded = ( lambda p: p if not isinstance(p, ParseResults) else "".join(chr(c) for c in range(ord(p[0]), ord(p[1]) + 1)) ) try: return "".join(_expanded(part) for part in _reBracketExpr.parseString(s).body) except Exception: return "" def matchOnlyAtCol(n): """Helper method for defining parse actions that require matching at a specific column in the input text. """ def verifyCol(strg, locn, toks): if col(locn, strg) != n: raise ParseException(strg, locn, "matched token not at column %d" % n) return verifyCol def replaceWith(replStr): """Helper method for common parse actions that simply return a literal value. Especially useful when used with :class:`transformString` (). Example:: num = Word(nums).setParseAction(lambda toks: int(toks[0])) na = oneOf("N/A NA").setParseAction(replaceWith(math.nan)) term = na | num OneOrMore(term).parseString("324 234 N/A 234") # -> [324, 234, nan, 234] """ return lambda s, l, t: [replStr] def removeQuotes(s, l, t): """Helper parse action for removing quotation marks from parsed quoted strings. Example:: # by default, quotation marks are included in parsed results quotedString.parseString("'Now is the Winter of our Discontent'") # -> ["'Now is the Winter of our Discontent'"] # use removeQuotes to strip quotation marks from parsed results quotedString.setParseAction(removeQuotes) quotedString.parseString("'Now is the Winter of our Discontent'") # -> ["Now is the Winter of our Discontent"] """ return t[0][1:-1] def tokenMap(func, *args): """Helper to define a parse action by mapping a function to all elements of a ParseResults list. If any additional args are passed, they are forwarded to the given function as additional arguments after the token, as in ``hex_integer = Word(hexnums).setParseAction(tokenMap(int, 16))``, which will convert the parsed data to an integer using base 16. Example (compare the last to example in :class:`ParserElement.transformString`:: hex_ints = OneOrMore(Word(hexnums)).setParseAction(tokenMap(int, 16)) hex_ints.runTests(''' 00 11 22 aa FF 0a 0d 1a ''') upperword = Word(alphas).setParseAction(tokenMap(str.upper)) OneOrMore(upperword).runTests(''' my kingdom for a horse ''') wd = Word(alphas).setParseAction(tokenMap(str.title)) OneOrMore(wd).setParseAction(' '.join).runTests(''' now is the winter of our discontent made glorious summer by this sun of york ''') prints:: 00 11 22 aa FF 0a 0d 1a [0, 17, 34, 170, 255, 10, 13, 26] my kingdom for a horse ['MY', 'KINGDOM', 'FOR', 'A', 'HORSE'] now is the winter of our discontent made glorious summer by this sun of york ['Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York'] """ def pa(s, l, t): return [func(tokn, *args) for tokn in t] try: func_name = getattr(func, "__name__", getattr(func, "__class__").__name__) except Exception: func_name = str(func) pa.__name__ = func_name return pa def _makeTags(tagStr, xml, suppress_LT=Suppress("<"), suppress_GT=Suppress(">")): """Internal helper to construct opening and closing tag expressions, given a tag name""" if isinstance(tagStr, str_type): resname = tagStr tagStr = Keyword(tagStr, caseless=not xml) else: resname = tagStr.name tagAttrName = Word(alphas, alphanums + "_-:") if xml: tagAttrValue = dblQuotedString.copy().setParseAction(removeQuotes) openTag = ( suppress_LT + tagStr("tag") + Dict(ZeroOrMore(Group(tagAttrName + Suppress("=") + tagAttrValue))) + Optional("/", default=[False])("empty").setParseAction( lambda s, l, t: t[0] == "/" ) + suppress_GT ) else: tagAttrValue = quotedString.copy().setParseAction(removeQuotes) | Word( printables, excludeChars=">" ) openTag = ( suppress_LT + tagStr("tag") + Dict( ZeroOrMore( Group( tagAttrName.setParseAction(lambda t: t[0].lower()) + Optional(Suppress("=") + tagAttrValue) ) ) ) + Optional("/", default=[False])("empty").setParseAction( lambda s, l, t: t[0] == "/" ) + suppress_GT ) closeTag = Combine(_L("", adjacent=False) openTag.setName("<%s>" % resname) # add start results name in parse action now that ungrouped names are not reported at two levels openTag.addParseAction( lambda t: t.__setitem__( "start" + "".join(resname.replace(":", " ").title().split()), t.copy() ) ) closeTag = closeTag( "end" + "".join(resname.replace(":", " ").title().split()) ).setName("" % resname) openTag.tag = resname closeTag.tag = resname openTag.tag_body = SkipTo(closeTag()) return openTag, closeTag def makeHTMLTags(tagStr): """Helper to construct opening and closing tag expressions for HTML, given a tag name. Matches tags in either upper or lower case, attributes with namespaces and with quoted or unquoted values. Example:: text = 'More info at the pyparsing wiki page' # makeHTMLTags returns pyparsing expressions for the opening and # closing tags as a 2-tuple a, a_end = makeHTMLTags("A") link_expr = a + SkipTo(a_end)("link_text") + a_end for link in link_expr.searchString(text): # attributes in the tag (like "href" shown here) are # also accessible as named results print(link.link_text, '->', link.href) prints:: pyparsing -> https://github.com/pyparsing/pyparsing/wiki """ return _makeTags(tagStr, False) def makeXMLTags(tagStr): """Helper to construct opening and closing tag expressions for XML, given a tag name. Matches tags only in the given upper/lower case. Example: similar to :class:`makeHTMLTags` """ return _makeTags(tagStr, True) def withAttribute(*args, **attrDict): """Helper to create a validating parse action to be used with start tags created with :class:`makeXMLTags` or :class:`makeHTMLTags`. Use ``withAttribute`` to qualify a starting tag with a required attribute value, to avoid false matches on common tags such as ```` or ``
``. Call ``withAttribute`` with a series of attribute names and values. Specify the list of filter attributes names and values as: - keyword arguments, as in ``(align="right")``, or - as an explicit dict with ``**`` operator, when an attribute name is also a Python reserved word, as in ``**{"class":"Customer", "align":"right"}`` - a list of name-value tuples, as in ``(("ns1:class", "Customer"), ("ns2:align", "right"))`` For attribute names with a namespace prefix, you must use the second form. Attribute names are matched insensitive to upper/lower case. If just testing for ``class`` (with or without a namespace), use :class:`withClass`. To verify that the attribute exists, but without specifying a value, pass ``withAttribute.ANY_VALUE`` as the value. Example:: html = '''
Some text
1 4 0 1 0
1,3 2,3 1,1
this has no type
''' div,div_end = makeHTMLTags("div") # only match div tag having a type attribute with value "grid" div_grid = div().setParseAction(withAttribute(type="grid")) grid_expr = div_grid + SkipTo(div | div_end)("body") for grid_header in grid_expr.searchString(html): print(grid_header.body) # construct a match with any div tag having a type attribute, regardless of the value div_any_type = div().setParseAction(withAttribute(type=withAttribute.ANY_VALUE)) div_expr = div_any_type + SkipTo(div | div_end)("body") for div_header in div_expr.searchString(html): print(div_header.body) prints:: 1 4 0 1 0 1 4 0 1 0 1,3 2,3 1,1 """ if args: attrs = args[:] else: attrs = attrDict.items() attrs = [(k, v) for k, v in attrs] def pa(s, l, tokens): for attrName, attrValue in attrs: if attrName not in tokens: raise ParseException(s, l, "no matching attribute " + attrName) if attrValue != withAttribute.ANY_VALUE and tokens[attrName] != attrValue: raise ParseException( s, l, "attribute '%s' has value '%s', must be '%s'" % (attrName, tokens[attrName], attrValue), ) return pa withAttribute.ANY_VALUE = object() def withClass(classname, namespace=""): """Simplified version of :class:`withAttribute` when matching on a div class - made difficult because ``class`` is a reserved word in Python. Example:: html = '''
Some text
1 4 0 1 0
1,3 2,3 1,1
this <div> has no class
''' div,div_end = makeHTMLTags("div") div_grid = div().setParseAction(withClass("grid")) grid_expr = div_grid + SkipTo(div | div_end)("body") for grid_header in grid_expr.searchString(html): print(grid_header.body) div_any_type = div().setParseAction(withClass(withAttribute.ANY_VALUE)) div_expr = div_any_type + SkipTo(div | div_end)("body") for div_header in div_expr.searchString(html): print(div_header.body) prints:: 1 4 0 1 0 1 4 0 1 0 1,3 2,3 1,1 """ classattr = "%s:class" % namespace if namespace else "class" return withAttribute(**{classattr: classname}) opAssoc = types.SimpleNamespace() opAssoc.LEFT = object() opAssoc.RIGHT = object() def infixNotation(baseExpr, opList, lpar=Suppress("("), rpar=Suppress(")")): """Helper method for constructing grammars of expressions made up of operators working in a precedence hierarchy. Operators may be unary or binary, left- or right-associative. Parse actions can also be attached to operator expressions. The generated parser will also recognize the use of parentheses to override operator precedences (see example below). Note: if you define a deep operator list, you may see performance issues when using infixNotation. See :class:`ParserElement.enablePackrat` for a mechanism to potentially improve your parser performance. Parameters: - baseExpr - expression representing the most basic element for the nested - opList - list of tuples, one for each operator precedence level in the expression grammar; each tuple is of the form ``(opExpr, numTerms, rightLeftAssoc, parseAction)``, where: - opExpr is the pyparsing expression for the operator; may also be a string, which will be converted to a Literal; if numTerms is 3, opExpr is a tuple of two expressions, for the two operators separating the 3 terms - numTerms is the number of terms for this operator (must be 1, 2, or 3) - rightLeftAssoc is the indicator whether the operator is right or left associative, using the pyparsing-defined constants ``opAssoc.RIGHT`` and ``opAssoc.LEFT``. - parseAction is the parse action to be associated with expressions matching this operator expression (the parse action tuple member may be omitted); if the parse action is passed a tuple or list of functions, this is equivalent to calling ``setParseAction(*fn)`` (:class:`ParserElement.setParseAction`) - lpar - expression for matching left-parentheses (default= ``Suppress('(')``) - rpar - expression for matching right-parentheses (default= ``Suppress(')')``) Example:: # simple example of four-function arithmetic with ints and # variable names integer = pyparsing_common.signed_integer varname = pyparsing_common.identifier arith_expr = infixNotation(integer | varname, [ ('-', 1, opAssoc.RIGHT), (oneOf('* /'), 2, opAssoc.LEFT), (oneOf('+ -'), 2, opAssoc.LEFT), ]) arith_expr.runTests(''' 5+3*6 (5+3)*6 -2--11 ''', fullDump=False) prints:: 5+3*6 [[5, '+', [3, '*', 6]]] (5+3)*6 [[[5, '+', 3], '*', 6]] -2--11 [[['-', 2], '-', ['-', 11]]] """ # captive version of FollowedBy that does not do parse actions or capture results names class _FB(FollowedBy): def parseImpl(self, instring, loc, doActions=True): self.expr.tryParse(instring, loc) return loc, [] ret = Forward() lastExpr = baseExpr | (lpar + ret + rpar) for i, operDef in enumerate(opList): opExpr, arity, rightLeftAssoc, pa = (operDef + (None,))[:4] termName = "%s term" % opExpr if arity < 3 else "%s%s term" % opExpr if arity == 3: if opExpr is None or len(opExpr) != 2: raise ValueError( "if numterms=3, opExpr must be a tuple or list of two expressions" ) opExpr1, opExpr2 = opExpr thisExpr = Forward().setName(termName) if rightLeftAssoc == opAssoc.LEFT: if arity == 1: matchExpr = _FB(lastExpr + opExpr) + Group(lastExpr + OneOrMore(opExpr)) elif arity == 2: if opExpr is not None: matchExpr = _FB(lastExpr + opExpr + lastExpr) + Group( lastExpr + OneOrMore(opExpr + lastExpr) ) else: matchExpr = _FB(lastExpr + lastExpr) + Group( lastExpr + OneOrMore(lastExpr) ) elif arity == 3: matchExpr = _FB( lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr ) + Group(lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr) else: raise ValueError( "operator must be unary (1), binary (2), or ternary (3)" ) elif rightLeftAssoc == opAssoc.RIGHT: if arity == 1: # try to avoid LR with this extra test if not isinstance(opExpr, Optional): opExpr = Optional(opExpr) matchExpr = _FB(opExpr.expr + thisExpr) + Group(opExpr + thisExpr) elif arity == 2: if opExpr is not None: matchExpr = _FB(lastExpr + opExpr + thisExpr) + Group( lastExpr + OneOrMore(opExpr + thisExpr) ) else: matchExpr = _FB(lastExpr + thisExpr) + Group( lastExpr + OneOrMore(thisExpr) ) elif arity == 3: matchExpr = _FB( lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr ) + Group(lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr) else: raise ValueError( "operator must be unary (1), binary (2), or ternary (3)" ) else: raise ValueError("operator must indicate right or left associativity") if pa: if isinstance(pa, (tuple, list)): matchExpr.setParseAction(*pa) else: matchExpr.setParseAction(pa) thisExpr <<= matchExpr.setName(termName) | lastExpr lastExpr = thisExpr ret <<= lastExpr return ret dblQuotedString = Combine( Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"' ).setName("string enclosed in double quotes") sglQuotedString = Combine( Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'" ).setName("string enclosed in single quotes") quotedString = Combine( Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"' | Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'" ).setName("quotedString using single or double quotes") unicodeString = Combine(_L("u") + quotedString.copy()).setName("unicode string literal") def nestedExpr(opener="(", closer=")", content=None, ignoreExpr=quotedString.copy()): """Helper method for defining nested lists enclosed in opening and closing delimiters ("(" and ")" are the default). Parameters: - opener - opening character for a nested list (default= ``"("``); can also be a pyparsing expression - closer - closing character for a nested list (default= ``")"``); can also be a pyparsing expression - content - expression for items within the nested lists (default= ``None``) - ignoreExpr - expression for ignoring opening and closing delimiters (default= :class:`quotedString`) If an expression is not provided for the content argument, the nested expression will capture all whitespace-delimited content between delimiters as a list of separate values. Use the ``ignoreExpr`` argument to define expressions that may contain opening or closing characters that should not be treated as opening or closing characters for nesting, such as quotedString or a comment expression. Specify multiple expressions using an :class:`Or` or :class:`MatchFirst`. The default is :class:`quotedString`, but if no expressions are to be ignored, then pass ``None`` for this argument. Example:: data_type = oneOf("void int short long char float double") decl_data_type = Combine(data_type + Optional(Word('*'))) ident = Word(alphas+'_', alphanums+'_') number = pyparsing_common.number arg = Group(decl_data_type + ident) LPAR, RPAR = map(Suppress, "()") code_body = nestedExpr('{', '}', ignoreExpr=(quotedString | cStyleComment)) c_function = (decl_data_type("type") + ident("name") + LPAR + Optional(delimitedList(arg), [])("args") + RPAR + code_body("body")) c_function.ignore(cStyleComment) source_code = ''' int is_odd(int x) { return (x%2); } int dec_to_hex(char hchar) { if (hchar >= '0' && hchar <= '9') { return (ord(hchar)-ord('0')); } else { return (10+ord(hchar)-ord('A')); } } ''' for func in c_function.searchString(source_code): print("%(name)s (%(type)s) args: %(args)s" % func) prints:: is_odd (int) args: [['int', 'x']] dec_to_hex (int) args: [['char', 'hchar']] """ if opener == closer: raise ValueError("opening and closing strings cannot be the same") if content is None: if isinstance(opener, str_type) and isinstance(closer, str_type): if len(opener) == 1 and len(closer) == 1: if ignoreExpr is not None: content = Combine( OneOrMore( ~ignoreExpr + CharsNotIn( opener + closer + ParserElement.DEFAULT_WHITE_CHARS, exact=1, ) ) ).setParseAction(lambda t: t[0].strip()) else: content = empty.copy() + CharsNotIn( opener + closer + ParserElement.DEFAULT_WHITE_CHARS ).setParseAction(lambda t: t[0].strip()) else: if ignoreExpr is not None: content = Combine( OneOrMore( ~ignoreExpr + ~Literal(opener) + ~Literal(closer) + CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS, exact=1) ) ).setParseAction(lambda t: t[0].strip()) else: content = Combine( OneOrMore( ~Literal(opener) + ~Literal(closer) + CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS, exact=1) ) ).setParseAction(lambda t: t[0].strip()) else: raise ValueError( "opening and closing arguments must be strings if no content expression is given" ) ret = Forward() if ignoreExpr is not None: ret <<= Group( Suppress(opener) + ZeroOrMore(ignoreExpr | ret | content) + Suppress(closer) ) else: ret <<= Group(Suppress(opener) + ZeroOrMore(ret | content) + Suppress(closer)) ret.setName("nested %s%s expression" % (opener, closer)) return ret def indentedBlock(blockStatementExpr, indentStack, indent=True, backup_stacks=[]): """Helper method for defining space-delimited indentation blocks, such as those used to define block statements in Python source code. Parameters: - blockStatementExpr - expression defining syntax of statement that is repeated within the indented block - indentStack - list created by caller to manage indentation stack (multiple statementWithIndentedBlock expressions within a single grammar should share a common indentStack) - indent - boolean indicating whether block must be indented beyond the current level; set to False for block of left-most statements (default= ``True``) A valid block must contain at least one ``blockStatement``. Example:: data = ''' def A(z): A1 B = 100 G = A2 A2 A3 B def BB(a,b,c): BB1 def BBA(): bba1 bba2 bba3 C D def spam(x,y): def eggs(z): pass ''' indentStack = [1] stmt = Forward() identifier = Word(alphas, alphanums) funcDecl = ("def" + identifier + Group("(" + Optional(delimitedList(identifier)) + ")") + ":") func_body = indentedBlock(stmt, indentStack) funcDef = Group(funcDecl + func_body) rvalue = Forward() funcCall = Group(identifier + "(" + Optional(delimitedList(rvalue)) + ")") rvalue << (funcCall | identifier | Word(nums)) assignment = Group(identifier + "=" + rvalue) stmt << (funcDef | assignment | identifier) module_body = OneOrMore(stmt) parseTree = module_body.parseString(data) parseTree.pprint() prints:: [['def', 'A', ['(', 'z', ')'], ':', [['A1'], [['B', '=', '100']], [['G', '=', 'A2']], ['A2'], ['A3']]], 'B', ['def', 'BB', ['(', 'a', 'b', 'c', ')'], ':', [['BB1'], [['def', 'BBA', ['(', ')'], ':', [['bba1'], ['bba2'], ['bba3']]]]]], 'C', 'D', ['def', 'spam', ['(', 'x', 'y', ')'], ':', [[['def', 'eggs', ['(', 'z', ')'], ':', [['pass']]]]]]] """ backup_stacks.append(indentStack[:]) def reset_stack(): indentStack[:] = backup_stacks[-1] def checkPeerIndent(s, l, t): if l >= len(s): return curCol = col(l, s) if curCol != indentStack[-1]: if curCol > indentStack[-1]: raise ParseException(s, l, "illegal nesting") raise ParseException(s, l, "not a peer entry") def checkSubIndent(s, l, t): curCol = col(l, s) if curCol > indentStack[-1]: indentStack.append(curCol) else: raise ParseException(s, l, "not a subentry") def checkUnindent(s, l, t): if l >= len(s): return curCol = col(l, s) if not (indentStack and curCol in indentStack): raise ParseException(s, l, "not an unindent") if curCol < indentStack[-1]: indentStack.pop() NL = OneOrMore(LineEnd().setWhitespaceChars("\t ").suppress()) INDENT = (Empty() + Empty().setParseAction(checkSubIndent)).setName("INDENT") PEER = Empty().setParseAction(checkPeerIndent).setName("") UNDENT = Empty().setParseAction(checkUnindent).setName("UNINDENT") if indent: smExpr = Group( Optional(NL) + INDENT + OneOrMore(PEER + Group(blockStatementExpr) + Optional(NL)) + UNDENT ) else: smExpr = Group( Optional(NL) + OneOrMore(PEER + Group(blockStatementExpr) + Optional(NL)) + Optional(UNDENT) ) # add a parse action to remove backup_stack from list of backups smExpr.addParseAction( lambda: backup_stacks.pop(-1) and None if backup_stacks else None ) smExpr.setFailAction(lambda a, b, c, d: reset_stack()) blockStatementExpr.ignore(_bslash + LineEnd()) return smExpr.setName("indented block") alphas8bit = srange(r"[\0xc0-\0xd6\0xd8-\0xf6\0xf8-\0xff]") punc8bit = srange(r"[\0xa1-\0xbf\0xd7\0xf7]") anyOpenTag, anyCloseTag = makeHTMLTags( Word(alphas, alphanums + "_:").setName("any tag") ) _htmlEntityMap = dict(zip("gt lt amp nbsp quot apos".split(), "><& \"'")) commonHTMLEntity = Regex( "&(?P" + "|".join(_htmlEntityMap.keys()) + ");" ).setName("common HTML entity") def replaceHTMLEntity(t): """Helper parser action to replace common HTML entities with their special characters""" return _htmlEntityMap.get(t.entity) # it's easy to get these comment structures wrong - they're very common, so may as well make them available cStyleComment = Combine(Regex(r"/\*(?:[^*]|\*(?!/))*") + "*/").setName( "C style comment" ) "Comment of the form ``/* ... */``" htmlComment = Regex(r"").setName("HTML comment") "Comment of the form ````" restOfLine = Regex(r".*").leaveWhitespace().setName("rest of line") dblSlashComment = Regex(r"//(?:\\\n|[^\n])*").setName("// comment") "Comment of the form ``// ... (to end of line)``" cppStyleComment = Combine( Regex(r"/\*(?:[^*]|\*(?!/))*") + "*/" | dblSlashComment ).setName("C++ style comment") "Comment of either form :class:`cStyleComment` or :class:`dblSlashComment`" javaStyleComment = cppStyleComment "Same as :class:`cppStyleComment`" pythonStyleComment = Regex(r"#.*").setName("Python style comment") "Comment of the form ``# ... (to end of line)``" # some other useful expressions - using lower-case class name since we are really using this as a namespace class pyparsing_common: """Here are some common low-level expressions that may be useful in jump-starting parser development: - numeric forms (:class:`integers`, :class:`reals`, :class:`scientific notation`) - common :class:`programming identifiers` - network addresses (:class:`MAC`, :class:`IPv4`, :class:`IPv6`) - ISO8601 :class:`dates` and :class:`datetime` - :class:`UUID` - :class:`comma-separated list` Parse actions: - :class:`convertToInteger` - :class:`convertToFloat` - :class:`convertToDate` - :class:`convertToDatetime` - :class:`stripHTMLTags` - :class:`upcaseTokens` - :class:`downcaseTokens` Example:: pyparsing_common.number.runTests(''' # any int or real number, returned as the appropriate type 100 -100 +100 3.14159 6.02e23 1e-12 ''') pyparsing_common.fnumber.runTests(''' # any int or real number, returned as float 100 -100 +100 3.14159 6.02e23 1e-12 ''') pyparsing_common.hex_integer.runTests(''' # hex numbers 100 FF ''') pyparsing_common.fraction.runTests(''' # fractions 1/2 -3/4 ''') pyparsing_common.mixed_integer.runTests(''' # mixed fractions 1 1/2 -3/4 1-3/4 ''') import uuid pyparsing_common.uuid.setParseAction(tokenMap(uuid.UUID)) pyparsing_common.uuid.runTests(''' # uuid 12345678-1234-5678-1234-567812345678 ''') prints:: # any int or real number, returned as the appropriate type 100 [100] -100 [-100] +100 [100] 3.14159 [3.14159] 6.02e23 [6.02e+23] 1e-12 [1e-12] # any int or real number, returned as float 100 [100.0] -100 [-100.0] +100 [100.0] 3.14159 [3.14159] 6.02e23 [6.02e+23] 1e-12 [1e-12] # hex numbers 100 [256] FF [255] # fractions 1/2 [0.5] -3/4 [-0.75] # mixed fractions 1 [1] 1/2 [0.5] -3/4 [-0.75] 1-3/4 [1.75] # uuid 12345678-1234-5678-1234-567812345678 [UUID('12345678-1234-5678-1234-567812345678')] """ convertToInteger = tokenMap(int) """ Parse action for converting parsed integers to Python int """ convertToFloat = tokenMap(float) """ Parse action for converting parsed numbers to Python float """ integer = Word(nums).setName("integer").setParseAction(convertToInteger) """expression that parses an unsigned integer, returns an int""" hex_integer = Word(hexnums).setName("hex integer").setParseAction(tokenMap(int, 16)) """expression that parses a hexadecimal integer, returns an int""" signed_integer = ( Regex(r"[+-]?\d+").setName("signed integer").setParseAction(convertToInteger) ) """expression that parses an integer with optional leading sign, returns an int""" fraction = ( signed_integer().setParseAction(convertToFloat) + "/" + signed_integer().setParseAction(convertToFloat) ).setName("fraction") """fractional expression of an integer divided by an integer, returns a float""" fraction.addParseAction(lambda t: t[0] / t[-1]) mixed_integer = ( fraction | signed_integer + Optional(Optional("-").suppress() + fraction) ).setName("fraction or mixed integer-fraction") """mixed integer of the form 'integer - fraction', with optional leading integer, returns float""" mixed_integer.addParseAction(sum) real = ( Regex(r"[+-]?(:?\d+\.\d*|\.\d+)") .setName("real number") .setParseAction(convertToFloat) ) """expression that parses a floating point number and returns a float""" sci_real = ( Regex(r"[+-]?(:?\d+(:?[eE][+-]?\d+)|(:?\d+\.\d*|\.\d+)(:?[eE][+-]?\d+)?)") .setName("real number with scientific notation") .setParseAction(convertToFloat) ) """expression that parses a floating point number with optional scientific notation and returns a float""" # streamlining this expression makes the docs nicer-looking number = (sci_real | real | signed_integer).streamline() """any numeric expression, returns the corresponding Python type""" fnumber = ( Regex(r"[+-]?\d+\.?\d*([eE][+-]?\d+)?") .setName("fnumber") .setParseAction(convertToFloat) ) """any int or real number, returned as float""" identifier = Word(alphas + "_", alphanums + "_").setName("identifier") """typical code identifier (leading alpha or '_', followed by 0 or more alphas, nums, or '_')""" ipv4_address = Regex( r"(25[0-5]|2[0-4][0-9]|1?[0-9]{1,2})(\.(25[0-5]|2[0-4][0-9]|1?[0-9]{1,2})){3}" ).setName("IPv4 address") "IPv4 address (``0.0.0.0 - 255.255.255.255``)" _ipv6_part = Regex(r"[0-9a-fA-F]{1,4}").setName("hex_integer") _full_ipv6_address = (_ipv6_part + (":" + _ipv6_part) * 7).setName( "full IPv6 address" ) _short_ipv6_address = ( Optional(_ipv6_part + (":" + _ipv6_part) * (0, 6)) + "::" + Optional(_ipv6_part + (":" + _ipv6_part) * (0, 6)) ).setName("short IPv6 address") _short_ipv6_address.addCondition( lambda t: sum(1 for tt in t if pyparsing_common._ipv6_part.matches(tt)) < 8 ) _mixed_ipv6_address = ("::ffff:" + ipv4_address).setName("mixed IPv6 address") ipv6_address = Combine( (_full_ipv6_address | _mixed_ipv6_address | _short_ipv6_address).setName( "IPv6 address" ) ).setName("IPv6 address") "IPv6 address (long, short, or mixed form)" mac_address = Regex( r"[0-9a-fA-F]{2}([:.-])[0-9a-fA-F]{2}(?:\1[0-9a-fA-F]{2}){4}" ).setName("MAC address") "MAC address xx:xx:xx:xx:xx (may also have '-' or '.' delimiters)" @staticmethod def convertToDate(fmt="%Y-%m-%d"): """ Helper to create a parse action for converting parsed date string to Python datetime.date Params - - fmt - format to be passed to datetime.strptime (default= ``"%Y-%m-%d"``) Example:: date_expr = pyparsing_common.iso8601_date.copy() date_expr.setParseAction(pyparsing_common.convertToDate()) print(date_expr.parseString("1999-12-31")) prints:: [datetime.date(1999, 12, 31)] """ def cvt_fn(s, l, t): try: return datetime.strptime(t[0], fmt).date() except ValueError as ve: raise ParseException(s, l, str(ve)) return cvt_fn @staticmethod def convertToDatetime(fmt="%Y-%m-%dT%H:%M:%S.%f"): """Helper to create a parse action for converting parsed datetime string to Python datetime.datetime Params - - fmt - format to be passed to datetime.strptime (default= ``"%Y-%m-%dT%H:%M:%S.%f"``) Example:: dt_expr = pyparsing_common.iso8601_datetime.copy() dt_expr.setParseAction(pyparsing_common.convertToDatetime()) print(dt_expr.parseString("1999-12-31T23:59:59.999")) prints:: [datetime.datetime(1999, 12, 31, 23, 59, 59, 999000)] """ def cvt_fn(s, l, t): try: return datetime.strptime(t[0], fmt) except ValueError as ve: raise ParseException(s, l, str(ve)) return cvt_fn iso8601_date = Regex( r"(?P\d{4})(?:-(?P\d\d)(?:-(?P\d\d))?)?" ).setName("ISO8601 date") "ISO8601 date (``yyyy-mm-dd``)" iso8601_datetime = Regex( r"(?P\d{4})-(?P\d\d)-(?P\d\d)[T ](?P\d\d):(?P\d\d)(:(?P\d\d(\.\d*)?)?)?(?PZ|[+-]\d\d:?\d\d)?" ).setName("ISO8601 datetime") "ISO8601 datetime (``yyyy-mm-ddThh:mm:ss.s(Z|+-00:00)``) - trailing seconds, milliseconds, and timezone optional; accepts separating ``'T'`` or ``' '``" uuid = Regex(r"[0-9a-fA-F]{8}(-[0-9a-fA-F]{4}){3}-[0-9a-fA-F]{12}").setName("UUID") "UUID (``xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx``)" _html_stripper = anyOpenTag.suppress() | anyCloseTag.suppress() @staticmethod def stripHTMLTags(s, l, tokens): """Parse action to remove HTML tags from web page HTML source Example:: # strip HTML links from normal text text = 'More info at the
pyparsing wiki page' td, td_end = makeHTMLTags("TD") table_text = td + SkipTo(td_end).setParseAction(pyparsing_common.stripHTMLTags)("body") + td_end print(table_text.parseString(text).body) Prints:: More info at the pyparsing wiki page """ return pyparsing_common._html_stripper.transformString(tokens[0]) _commasepitem = ( Combine( OneOrMore( ~Literal(",") + ~LineEnd() + Word(printables, excludeChars=",") + Optional(White(" \t") + ~FollowedBy(LineEnd() | ",")) ) ) .streamline() .setName("commaItem") ) comma_separated_list = delimitedList( Optional(quotedString.copy() | _commasepitem, default="") ).setName("comma separated list") """Predefined expression of 1 or more printable words or quoted strin gs, separated by commas.""" upcaseTokens = staticmethod(tokenMap(lambda t: t.upper())) """Parse action to convert tokens to upper case.""" downcaseTokens = staticmethod(tokenMap(lambda t: t.lower())) """Parse action to convert tokens to lower case.""" class _lazyclassproperty(object): def __init__(self, fn): self.fn = fn self.__doc__ = fn.__doc__ self.__name__ = fn.__name__ def __get__(self, obj, cls): if cls is None: cls = type(obj) if not hasattr(cls, "_intern") or any( cls._intern is getattr(superclass, "_intern", []) for superclass in cls.__mro__[1:] ): cls._intern = {} attrname = self.fn.__name__ if attrname not in cls._intern: cls._intern[attrname] = self.fn(cls) return cls._intern[attrname] class unicode_set(object): """ A set of Unicode characters, for language-specific strings for ``alphas``, ``nums``, ``alphanums``, and ``printables``. A unicode_set is defined by a list of ranges in the Unicode character set, in a class attribute ``_ranges``, such as:: _ranges = [(0x0020, 0x007e), (0x00a0, 0x00ff),] A unicode set can also be defined using multiple inheritance of other unicode sets:: class CJK(Chinese, Japanese, Korean): pass """ _ranges = [] @classmethod def _get_chars_for_ranges(cls): ret = [] for cc in cls.__mro__: if cc is unicode_set: break for rr in cc._ranges: ret.extend(range(rr[0], rr[-1] + 1)) return [chr(c) for c in sorted(set(ret))] @_lazyclassproperty def printables(cls): "all non-whitespace characters in this range" return "".join(filterfalse(str.isspace, cls._get_chars_for_ranges())) @_lazyclassproperty def alphas(cls): "all alphabetic characters in this range" return "".join(filter(str.isalpha, cls._get_chars_for_ranges())) @_lazyclassproperty def nums(cls): "all numeric digit characters in this range" return "".join(filter(str.isdigit, cls._get_chars_for_ranges())) @_lazyclassproperty def alphanums(cls): "all alphanumeric characters in this range" return cls.alphas + cls.nums class pyparsing_unicode(unicode_set): """ A namespace class for defining common language unicode_sets. """ _ranges = [(32, sys.maxunicode)] class Latin1(unicode_set): "Unicode set for Latin-1 Unicode Character Range" _ranges = [ (0x0020, 0x007E), (0x00A0, 0x00FF), ] class LatinA(unicode_set): "Unicode set for Latin-A Unicode Character Range" _ranges = [ (0x0100, 0x017F), ] class LatinB(unicode_set): "Unicode set for Latin-B Unicode Character Range" _ranges = [ (0x0180, 0x024F), ] class Greek(unicode_set): "Unicode set for Greek Unicode Character Ranges" _ranges = [ (0x0370, 0x03FF), (0x1F00, 0x1F15), (0x1F18, 0x1F1D), (0x1F20, 0x1F45), (0x1F48, 0x1F4D), (0x1F50, 0x1F57), (0x1F59,), (0x1F5B,), (0x1F5D,), (0x1F5F, 0x1F7D), (0x1F80, 0x1FB4), (0x1FB6, 0x1FC4), (0x1FC6, 0x1FD3), (0x1FD6, 0x1FDB), (0x1FDD, 0x1FEF), (0x1FF2, 0x1FF4), (0x1FF6, 0x1FFE), ] class Cyrillic(unicode_set): "Unicode set for Cyrillic Unicode Character Range" _ranges = [(0x0400, 0x04FF)] class Chinese(unicode_set): "Unicode set for Chinese Unicode Character Range" _ranges = [ (0x4E00, 0x9FFF), (0x3000, 0x303F), ] class Japanese(unicode_set): "Unicode set for Japanese Unicode Character Range, combining Kanji, Hiragana, and Katakana ranges" _ranges = [] class Kanji(unicode_set): "Unicode set for Kanji Unicode Character Range" _ranges = [ (0x4E00, 0x9FBF), (0x3000, 0x303F), ] class Hiragana(unicode_set): "Unicode set for Hiragana Unicode Character Range" _ranges = [ (0x3040, 0x309F), ] class Katakana(unicode_set): "Unicode set for Katakana Unicode Character Range" _ranges = [ (0x30A0, 0x30FF), ] class Korean(unicode_set): "Unicode set for Korean Unicode Character Range" _ranges = [ (0xAC00, 0xD7AF), (0x1100, 0x11FF), (0x3130, 0x318F), (0xA960, 0xA97F), (0xD7B0, 0xD7FF), (0x3000, 0x303F), ] class CJK(Chinese, Japanese, Korean): "Unicode set for combined Chinese, Japanese, and Korean (CJK) Unicode Character Range" pass class Thai(unicode_set): "Unicode set for Thai Unicode Character Range" _ranges = [ (0x0E01, 0x0E3A), (0x0E3F, 0x0E5B), ] class Arabic(unicode_set): "Unicode set for Arabic Unicode Character Range" _ranges = [ (0x0600, 0x061B), (0x061E, 0x06FF), (0x0700, 0x077F), ] class Hebrew(unicode_set): "Unicode set for Hebrew Unicode Character Range" _ranges = [ (0x0590, 0x05FF), ] class Devanagari(unicode_set): "Unicode set for Devanagari Unicode Character Range" _ranges = [(0x0900, 0x097F), (0xA8E0, 0xA8FF)] pyparsing_unicode.Japanese._ranges = ( pyparsing_unicode.Japanese.Kanji._ranges + pyparsing_unicode.Japanese.Hiragana._ranges + pyparsing_unicode.Japanese.Katakana._ranges ) # define ranges in language character sets pyparsing_unicode.العربية = pyparsing_unicode.Arabic pyparsing_unicode.中文 = pyparsing_unicode.Chinese pyparsing_unicode.кириллица = pyparsing_unicode.Cyrillic pyparsing_unicode.Ελληνικά = pyparsing_unicode.Greek pyparsing_unicode.עִברִית = pyparsing_unicode.Hebrew pyparsing_unicode.日本語 = pyparsing_unicode.Japanese pyparsing_unicode.Japanese.漢字 = pyparsing_unicode.Japanese.Kanji pyparsing_unicode.Japanese.カタカナ = pyparsing_unicode.Japanese.Katakana pyparsing_unicode.Japanese.ひらがな = pyparsing_unicode.Japanese.Hiragana pyparsing_unicode.한국어 = pyparsing_unicode.Korean pyparsing_unicode.ไทย = pyparsing_unicode.Thai pyparsing_unicode.देवनागरी = pyparsing_unicode.Devanagari class pyparsing_test: """ namespace class for classes useful in writing unit tests """ class reset_pyparsing_context: """ Context manager to be used when writing unit tests that modify pyparsing config values: - packrat parsing - default whitespace characters - default keyword characters - literal string auto-conversion class - __diag__ settings Example: with reset_pyparsing_context(): # test that literals used to construct a grammar are automatically suppressed ParserElement.inlineLiteralsUsing(Suppress) term = Word(alphas) | Word(nums) group = Group('(' + term[...] + ')') # assert that the '()' characters are not included in the parsed tokens self.assertParseAndCheckLisst(group, "(abc 123 def)", ['abc', '123', 'def']) # after exiting context manager, literals are converted to Literal expressions again """ def __init__(self): self._save_context = {} def __enter__(self): self.save() def __exit__(self, *args): self.restore() def save(self): self._save_context["default_whitespace"] = ParserElement.DEFAULT_WHITE_CHARS self._save_context["default_keyword_chars"] = Keyword.DEFAULT_KEYWORD_CHARS self._save_context[ "literal_string_class" ] = ParserElement._literalStringClass self._save_context["packrat_enabled"] = ParserElement._packratEnabled self._save_context["packrat_parse"] = ParserElement._parse self._save_context["__diag__"] = { name: getattr(__diag__, name) for name in __diag__._all_names } self._save_context["__compat__"] = { "collect_all_And_tokens": __compat__.collect_all_And_tokens } return self def restore(self): # restore pyparsing global state to what was saved ParserElement.setDefaultWhitespaceChars( self._save_context["default_whitespace"] ) Keyword.DEFAULT_KEYWORD_CHARS = self._save_context["default_keyword_chars"] ParserElement.inlineLiteralsUsing( self._save_context["literal_string_class"] ) ParserElement._packratEnabled = self._save_context["packrat_enabled"] ParserElement._parse = self._save_context["packrat_parse"] for name, value in self._save_context["__diag__"].items(): (__diag__.enable if value else __diag__.disable)(name) for name, value in self._save_context["__compat__"].items(): setattr(__compat__, name, value) class TestParseResultsAsserts(unittest.TestCase): def assertParseResultsEquals( self, result, expected_list=None, expected_dict=None, msg=None ): """ Unit test assertion to compare a ParseResults object with an optional expected_list, and compare any defined results names with an optional expected_dict. """ if expected_list is not None: self.assertEqual(expected_list, result.asList(), msg=msg) if expected_dict is not None: self.assertEqual(expected_dict, result.asDict(), msg=msg) def assertParseAndCheckList( self, expr, test_string, expected_list, msg=None, verbose=True ): """ Convenience wrapper assert to test a parser element and input string, and assert that the resulting ParseResults.asList() is equal to the expected_list. """ result = expr.parseString(test_string, parseAll=True) if verbose: print(result.dump()) self.assertParseResultsEquals(result, expected_list=expected_list, msg=msg) def assertParseAndCheckDict( self, expr, test_string, expected_dict, msg=None, verbose=True ): """ Convenience wrapper assert to test a parser element and input string, and assert that the resulting ParseResults.asDict() is equal to the expected_dict. """ result = expr.parseString(test_string, parseAll=True) if verbose: print(result.dump()) self.assertParseResultsEquals(result, expected_dict=expected_dict, msg=msg) def assertRunTestResults( self, run_tests_report, expected_parse_results=None, msg=None ): """ Unit test assertion to evaluate output of ParserElement.runTests(). If a list of list-dict tuples is given as the expected_parse_results argument, then these are zipped with the report tuples returned by runTests and evaluated using assertParseResultsEquals. Finally, asserts that the overall runTests() success value is True. :param run_tests_report: tuple(bool, [tuple(str, ParseResults or Exception)]) returned from runTests :param expected_parse_results (optional): [tuple(str, list, dict, Exception)] """ run_test_success, run_test_results = run_tests_report if expected_parse_results is not None: merged = [ (*rpt, expected) for rpt, expected in zip(run_test_results, expected_parse_results) ] for test_string, result, expected in merged: # expected should be a tuple containing a list and/or a dict or an exception, # and optional failure message string # an empty tuple will skip any result validation fail_msg = next( (exp for exp in expected if isinstance(exp, str)), None ) expected_exception = next( ( exp for exp in expected if isinstance(exp, type) and issubclass(exp, Exception) ), None, ) if expected_exception is not None: with self.assertRaises( expected_exception=expected_exception, msg=fail_msg or msg ): if isinstance(result, Exception): raise result else: expected_list = next( (exp for exp in expected if isinstance(exp, list)), None ) expected_dict = next( (exp for exp in expected if isinstance(exp, dict)), None ) if (expected_list, expected_dict) != (None, None): self.assertParseResultsEquals( result, expected_list=expected_list, expected_dict=expected_dict, msg=fail_msg or msg, ) else: # warning here maybe? print("no validation for {!r}".format(test_string)) # do this last, in case some specific test results can be reported instead self.assertTrue( run_test_success, msg=msg if msg is not None else "failed runTests" ) @contextmanager def assertRaisesParseException(self, exc_type=ParseException, msg=None): with self.assertRaises(exc_type, msg=msg): yield # build list of built-in expressions, for future reference if a global default value # gets updated _builtin_exprs = [ v for v in itertools.chain(vars().values(), vars(pyparsing_common).values()) if isinstance(v, ParserElement) ] if __name__ == "__main__": selectToken = CaselessLiteral("select") fromToken = CaselessLiteral("from") ident = Word(alphas, alphanums + "_$") columnName = delimitedList(ident, ".", combine=True).setParseAction( pyparsing_common.upcaseTokens ) columnNameList = Group(delimitedList(columnName)).setName("columns") columnSpec = "*" | columnNameList tableName = delimitedList(ident, ".", combine=True).setParseAction( pyparsing_common.upcaseTokens ) tableNameList = Group(delimitedList(tableName)).setName("tables") simpleSQL = ( selectToken("command") + columnSpec("columns") + fromToken + tableNameList("tables") ) # demo runTests method, including embedded comments in test string simpleSQL.runTests( """ # '*' as column list and dotted table name select * from SYS.XYZZY # caseless match on "SELECT", and casts back to "select" SELECT * from XYZZY, ABC # list of column names, and mixed case SELECT keyword Select AA,BB,CC from Sys.dual # multiple tables Select A, B, C from Sys.dual, Table2 # invalid SELECT keyword - should fail Xelect A, B, C from Sys.dual # incomplete command - should fail Select # invalid column name - should fail Select ^^^ frox Sys.dual """ ) pyparsing_common.number.runTests( """ 100 -100 +100 3.14159 6.02e23 1e-12 """ ) # any int or real number, returned as float pyparsing_common.fnumber.runTests( """ 100 -100 +100 3.14159 6.02e23 1e-12 """ ) pyparsing_common.hex_integer.runTests( """ 100 FF """ ) import uuid pyparsing_common.uuid.setParseAction(tokenMap(uuid.UUID)) pyparsing_common.uuid.runTests( """ 12345678-1234-5678-1234-567812345678 """ )