Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/pip/_vendor/pyparsing/core.py: 49%
2584 statements
« prev ^ index » next coverage.py v7.4.3, created at 2024-02-26 06:33 +0000
1#
2# core.py
3#
5from collections import deque
6import os
7import typing
8from typing import (
9 Any,
10 Callable,
11 Generator,
12 List,
13 NamedTuple,
14 Sequence,
15 Set,
16 TextIO,
17 Tuple,
18 Union,
19 cast,
20)
21from abc import ABC, abstractmethod
22from enum import Enum
23import string
24import copy
25import warnings
26import re
27import sys
28from collections.abc import Iterable
29import traceback
30import types
31from operator import itemgetter
32from functools import wraps
33from threading import RLock
34from pathlib import Path
36from .util import (
37 _FifoCache,
38 _UnboundedCache,
39 __config_flags,
40 _collapse_string_to_ranges,
41 _escape_regex_range_chars,
42 _bslash,
43 _flatten,
44 LRUMemo as _LRUMemo,
45 UnboundedMemo as _UnboundedMemo,
46 replaced_by_pep8,
47)
48from .exceptions import *
49from .actions import *
50from .results import ParseResults, _ParseResultsWithOffset
51from .unicode import pyparsing_unicode
53_MAX_INT = sys.maxsize
54str_type: Tuple[type, ...] = (str, bytes)
56#
57# Copyright (c) 2003-2022 Paul T. McGuire
58#
59# Permission is hereby granted, free of charge, to any person obtaining
60# a copy of this software and associated documentation files (the
61# "Software"), to deal in the Software without restriction, including
62# without limitation the rights to use, copy, modify, merge, publish,
63# distribute, sublicense, and/or sell copies of the Software, and to
64# permit persons to whom the Software is furnished to do so, subject to
65# the following conditions:
66#
67# The above copyright notice and this permission notice shall be
68# included in all copies or substantial portions of the Software.
69#
70# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
71# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
72# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
73# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
74# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
75# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
76# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
77#
80if sys.version_info >= (3, 8):
81 from functools import cached_property
82else:
84 class cached_property:
85 def __init__(self, func):
86 self._func = func
88 def __get__(self, instance, owner=None):
89 ret = instance.__dict__[self._func.__name__] = self._func(instance)
90 return ret
93class __compat__(__config_flags):
94 """
95 A cross-version compatibility configuration for pyparsing features that will be
96 released in a future version. By setting values in this configuration to True,
97 those features can be enabled in prior versions for compatibility development
98 and testing.
100 - ``collect_all_And_tokens`` - flag to enable fix for Issue #63 that fixes erroneous grouping
101 of results names when an :class:`And` expression is nested within an :class:`Or` or :class:`MatchFirst`;
102 maintained for compatibility, but setting to ``False`` no longer restores pre-2.3.1
103 behavior
104 """
106 _type_desc = "compatibility"
108 collect_all_And_tokens = True
110 _all_names = [__ for __ in locals() if not __.startswith("_")]
111 _fixed_names = """
112 collect_all_And_tokens
113 """.split()
116class __diag__(__config_flags):
117 _type_desc = "diagnostic"
119 warn_multiple_tokens_in_named_alternation = False
120 warn_ungrouped_named_tokens_in_collection = False
121 warn_name_set_on_empty_Forward = False
122 warn_on_parse_using_empty_Forward = False
123 warn_on_assignment_to_Forward = False
124 warn_on_multiple_string_args_to_oneof = False
125 warn_on_match_first_with_lshift_operator = False
126 enable_debug_on_named_expressions = False
128 _all_names = [__ for __ in locals() if not __.startswith("_")]
129 _warning_names = [name for name in _all_names if name.startswith("warn")]
130 _debug_names = [name for name in _all_names if name.startswith("enable_debug")]
132 @classmethod
133 def enable_all_warnings(cls) -> None:
134 for name in cls._warning_names:
135 cls.enable(name)
138class Diagnostics(Enum):
139 """
140 Diagnostic configuration (all default to disabled)
142 - ``warn_multiple_tokens_in_named_alternation`` - flag to enable warnings when a results
143 name is defined on a :class:`MatchFirst` or :class:`Or` expression with one or more :class:`And` subexpressions
144 - ``warn_ungrouped_named_tokens_in_collection`` - flag to enable warnings when a results
145 name is defined on a containing expression with ungrouped subexpressions that also
146 have results names
147 - ``warn_name_set_on_empty_Forward`` - flag to enable warnings when a :class:`Forward` is defined
148 with a results name, but has no contents defined
149 - ``warn_on_parse_using_empty_Forward`` - flag to enable warnings when a :class:`Forward` is
150 defined in a grammar but has never had an expression attached to it
151 - ``warn_on_assignment_to_Forward`` - flag to enable warnings when a :class:`Forward` is defined
152 but is overwritten by assigning using ``'='`` instead of ``'<<='`` or ``'<<'``
153 - ``warn_on_multiple_string_args_to_oneof`` - flag to enable warnings when :class:`one_of` is
154 incorrectly called with multiple str arguments
155 - ``enable_debug_on_named_expressions`` - flag to auto-enable debug on all subsequent
156 calls to :class:`ParserElement.set_name`
158 Diagnostics are enabled/disabled by calling :class:`enable_diag` and :class:`disable_diag`.
159 All warnings can be enabled by calling :class:`enable_all_warnings`.
160 """
162 warn_multiple_tokens_in_named_alternation = 0
163 warn_ungrouped_named_tokens_in_collection = 1
164 warn_name_set_on_empty_Forward = 2
165 warn_on_parse_using_empty_Forward = 3
166 warn_on_assignment_to_Forward = 4
167 warn_on_multiple_string_args_to_oneof = 5
168 warn_on_match_first_with_lshift_operator = 6
169 enable_debug_on_named_expressions = 7
172def enable_diag(diag_enum: Diagnostics) -> None:
173 """
174 Enable a global pyparsing diagnostic flag (see :class:`Diagnostics`).
175 """
176 __diag__.enable(diag_enum.name)
179def disable_diag(diag_enum: Diagnostics) -> None:
180 """
181 Disable a global pyparsing diagnostic flag (see :class:`Diagnostics`).
182 """
183 __diag__.disable(diag_enum.name)
186def enable_all_warnings() -> None:
187 """
188 Enable all global pyparsing diagnostic warnings (see :class:`Diagnostics`).
189 """
190 __diag__.enable_all_warnings()
193# hide abstract class
194del __config_flags
197def _should_enable_warnings(
198 cmd_line_warn_options: typing.Iterable[str], warn_env_var: typing.Optional[str]
199) -> bool:
200 enable = bool(warn_env_var)
201 for warn_opt in cmd_line_warn_options:
202 w_action, w_message, w_category, w_module, w_line = (warn_opt + "::::").split(
203 ":"
204 )[:5]
205 if not w_action.lower().startswith("i") and (
206 not (w_message or w_category or w_module) or w_module == "pyparsing"
207 ):
208 enable = True
209 elif w_action.lower().startswith("i") and w_module in ("pyparsing", ""):
210 enable = False
211 return enable
214if _should_enable_warnings(
215 sys.warnoptions, os.environ.get("PYPARSINGENABLEALLWARNINGS")
216):
217 enable_all_warnings()
220# build list of single arg builtins, that can be used as parse actions
221_single_arg_builtins = {
222 sum,
223 len,
224 sorted,
225 reversed,
226 list,
227 tuple,
228 set,
229 any,
230 all,
231 min,
232 max,
233}
235_generatorType = types.GeneratorType
236ParseImplReturnType = Tuple[int, Any]
237PostParseReturnType = Union[ParseResults, Sequence[ParseResults]]
238ParseAction = Union[
239 Callable[[], Any],
240 Callable[[ParseResults], Any],
241 Callable[[int, ParseResults], Any],
242 Callable[[str, int, ParseResults], Any],
243]
244ParseCondition = Union[
245 Callable[[], bool],
246 Callable[[ParseResults], bool],
247 Callable[[int, ParseResults], bool],
248 Callable[[str, int, ParseResults], bool],
249]
250ParseFailAction = Callable[[str, int, "ParserElement", Exception], None]
251DebugStartAction = Callable[[str, int, "ParserElement", bool], None]
252DebugSuccessAction = Callable[
253 [str, int, int, "ParserElement", ParseResults, bool], None
254]
255DebugExceptionAction = Callable[[str, int, "ParserElement", Exception, bool], None]
258alphas = string.ascii_uppercase + string.ascii_lowercase
259identchars = pyparsing_unicode.Latin1.identchars
260identbodychars = pyparsing_unicode.Latin1.identbodychars
261nums = "0123456789"
262hexnums = nums + "ABCDEFabcdef"
263alphanums = alphas + nums
264printables = "".join([c for c in string.printable if c not in string.whitespace])
266_trim_arity_call_line: traceback.StackSummary = None # type: ignore[assignment]
269def _trim_arity(func, max_limit=3):
270 """decorator to trim function calls to match the arity of the target"""
271 global _trim_arity_call_line
273 if func in _single_arg_builtins:
274 return lambda s, l, t: func(t)
276 limit = 0
277 found_arity = False
279 # synthesize what would be returned by traceback.extract_stack at the call to
280 # user's parse action 'func', so that we don't incur call penalty at parse time
282 # fmt: off
283 LINE_DIFF = 7
284 # IF ANY CODE CHANGES, EVEN JUST COMMENTS OR BLANK LINES, BETWEEN THE NEXT LINE AND
285 # THE CALL TO FUNC INSIDE WRAPPER, LINE_DIFF MUST BE MODIFIED!!!!
286 _trim_arity_call_line = (_trim_arity_call_line or traceback.extract_stack(limit=2)[-1])
287 pa_call_line_synth = (_trim_arity_call_line[0], _trim_arity_call_line[1] + LINE_DIFF)
289 def wrapper(*args):
290 nonlocal found_arity, limit
291 while 1:
292 try:
293 ret = func(*args[limit:])
294 found_arity = True
295 return ret
296 except TypeError as te:
297 # re-raise TypeErrors if they did not come from our arity testing
298 if found_arity:
299 raise
300 else:
301 tb = te.__traceback__
302 frames = traceback.extract_tb(tb, limit=2)
303 frame_summary = frames[-1]
304 trim_arity_type_error = (
305 [frame_summary[:2]][-1][:2] == pa_call_line_synth
306 )
307 del tb
309 if trim_arity_type_error:
310 if limit < max_limit:
311 limit += 1
312 continue
314 raise
315 # fmt: on
317 # copy func name to wrapper for sensible debug output
318 # (can't use functools.wraps, since that messes with function signature)
319 func_name = getattr(func, "__name__", getattr(func, "__class__").__name__)
320 wrapper.__name__ = func_name
321 wrapper.__doc__ = func.__doc__
323 return wrapper
326def condition_as_parse_action(
327 fn: ParseCondition, message: typing.Optional[str] = None, fatal: bool = False
328) -> ParseAction:
329 """
330 Function to convert a simple predicate function that returns ``True`` or ``False``
331 into a parse action. Can be used in places when a parse action is required
332 and :class:`ParserElement.add_condition` cannot be used (such as when adding a condition
333 to an operator level in :class:`infix_notation`).
335 Optional keyword arguments:
337 - ``message`` - define a custom message to be used in the raised exception
338 - ``fatal`` - if True, will raise :class:`ParseFatalException` to stop parsing immediately;
339 otherwise will raise :class:`ParseException`
341 """
342 msg = message if message is not None else "failed user-defined condition"
343 exc_type = ParseFatalException if fatal else ParseException
344 fn = _trim_arity(fn)
346 @wraps(fn)
347 def pa(s, l, t):
348 if not bool(fn(s, l, t)):
349 raise exc_type(s, l, msg)
351 return pa
354def _default_start_debug_action(
355 instring: str, loc: int, expr: "ParserElement", cache_hit: bool = False
356):
357 cache_hit_str = "*" if cache_hit else ""
358 print(
359 (
360 f"{cache_hit_str}Match {expr} at loc {loc}({lineno(loc, instring)},{col(loc, instring)})\n"
361 f" {line(loc, instring)}\n"
362 f" {' ' * (col(loc, instring) - 1)}^"
363 )
364 )
367def _default_success_debug_action(
368 instring: str,
369 startloc: int,
370 endloc: int,
371 expr: "ParserElement",
372 toks: ParseResults,
373 cache_hit: bool = False,
374):
375 cache_hit_str = "*" if cache_hit else ""
376 print(f"{cache_hit_str}Matched {expr} -> {toks.as_list()}")
379def _default_exception_debug_action(
380 instring: str,
381 loc: int,
382 expr: "ParserElement",
383 exc: Exception,
384 cache_hit: bool = False,
385):
386 cache_hit_str = "*" if cache_hit else ""
387 print(f"{cache_hit_str}Match {expr} failed, {type(exc).__name__} raised: {exc}")
390def null_debug_action(*args):
391 """'Do-nothing' debug action, to suppress debugging output during parsing."""
394class ParserElement(ABC):
395 """Abstract base level parser element class."""
397 DEFAULT_WHITE_CHARS: str = " \n\t\r"
398 verbose_stacktrace: bool = False
399 _literalStringClass: type = None # type: ignore[assignment]
401 @staticmethod
402 def set_default_whitespace_chars(chars: str) -> None:
403 r"""
404 Overrides the default whitespace chars
406 Example::
408 # default whitespace chars are space, <TAB> and newline
409 Word(alphas)[1, ...].parse_string("abc def\nghi jkl") # -> ['abc', 'def', 'ghi', 'jkl']
411 # change to just treat newline as significant
412 ParserElement.set_default_whitespace_chars(" \t")
413 Word(alphas)[1, ...].parse_string("abc def\nghi jkl") # -> ['abc', 'def']
414 """
415 ParserElement.DEFAULT_WHITE_CHARS = chars
417 # update whitespace all parse expressions defined in this module
418 for expr in _builtin_exprs:
419 if expr.copyDefaultWhiteChars:
420 expr.whiteChars = set(chars)
422 @staticmethod
423 def inline_literals_using(cls: type) -> None:
424 """
425 Set class to be used for inclusion of string literals into a parser.
427 Example::
429 # default literal class used is Literal
430 integer = Word(nums)
431 date_str = integer("year") + '/' + integer("month") + '/' + integer("day")
433 date_str.parse_string("1999/12/31") # -> ['1999', '/', '12', '/', '31']
436 # change to Suppress
437 ParserElement.inline_literals_using(Suppress)
438 date_str = integer("year") + '/' + integer("month") + '/' + integer("day")
440 date_str.parse_string("1999/12/31") # -> ['1999', '12', '31']
441 """
442 ParserElement._literalStringClass = cls
444 @classmethod
445 def using_each(cls, seq, **class_kwargs):
446 """
447 Yields a sequence of class(obj, **class_kwargs) for obj in seq.
449 Example::
451 LPAR, RPAR, LBRACE, RBRACE, SEMI = Suppress.using_each("(){};")
453 """
454 yield from (cls(obj, **class_kwargs) for obj in seq)
456 class DebugActions(NamedTuple):
457 debug_try: typing.Optional[DebugStartAction]
458 debug_match: typing.Optional[DebugSuccessAction]
459 debug_fail: typing.Optional[DebugExceptionAction]
461 def __init__(self, savelist: bool = False):
462 self.parseAction: List[ParseAction] = list()
463 self.failAction: typing.Optional[ParseFailAction] = None
464 self.customName: str = None # type: ignore[assignment]
465 self._defaultName: typing.Optional[str] = None
466 self.resultsName: str = None # type: ignore[assignment]
467 self.saveAsList = savelist
468 self.skipWhitespace = True
469 self.whiteChars = set(ParserElement.DEFAULT_WHITE_CHARS)
470 self.copyDefaultWhiteChars = True
471 # used when checking for left-recursion
472 self.mayReturnEmpty = False
473 self.keepTabs = False
474 self.ignoreExprs: List["ParserElement"] = list()
475 self.debug = False
476 self.streamlined = False
477 # optimize exception handling for subclasses that don't advance parse index
478 self.mayIndexError = True
479 self.errmsg = ""
480 # mark results names as modal (report only last) or cumulative (list all)
481 self.modalResults = True
482 # custom debug actions
483 self.debugActions = self.DebugActions(None, None, None)
484 # avoid redundant calls to preParse
485 self.callPreparse = True
486 self.callDuringTry = False
487 self.suppress_warnings_: List[Diagnostics] = []
489 def suppress_warning(self, warning_type: Diagnostics) -> "ParserElement":
490 """
491 Suppress warnings emitted for a particular diagnostic on this expression.
493 Example::
495 base = pp.Forward()
496 base.suppress_warning(Diagnostics.warn_on_parse_using_empty_Forward)
498 # statement would normally raise a warning, but is now suppressed
499 print(base.parse_string("x"))
501 """
502 self.suppress_warnings_.append(warning_type)
503 return self
505 def visit_all(self):
506 """General-purpose method to yield all expressions and sub-expressions
507 in a grammar. Typically just for internal use.
508 """
509 to_visit = deque([self])
510 seen = set()
511 while to_visit:
512 cur = to_visit.popleft()
514 # guard against looping forever through recursive grammars
515 if cur in seen:
516 continue
517 seen.add(cur)
519 to_visit.extend(cur.recurse())
520 yield cur
522 def copy(self) -> "ParserElement":
523 """
524 Make a copy of this :class:`ParserElement`. Useful for defining
525 different parse actions for the same parsing pattern, using copies of
526 the original parse element.
528 Example::
530 integer = Word(nums).set_parse_action(lambda toks: int(toks[0]))
531 integerK = integer.copy().add_parse_action(lambda toks: toks[0] * 1024) + Suppress("K")
532 integerM = integer.copy().add_parse_action(lambda toks: toks[0] * 1024 * 1024) + Suppress("M")
534 print((integerK | integerM | integer)[1, ...].parse_string("5K 100 640K 256M"))
536 prints::
538 [5120, 100, 655360, 268435456]
540 Equivalent form of ``expr.copy()`` is just ``expr()``::
542 integerM = integer().add_parse_action(lambda toks: toks[0] * 1024 * 1024) + Suppress("M")
543 """
544 cpy = copy.copy(self)
545 cpy.parseAction = self.parseAction[:]
546 cpy.ignoreExprs = self.ignoreExprs[:]
547 if self.copyDefaultWhiteChars:
548 cpy.whiteChars = set(ParserElement.DEFAULT_WHITE_CHARS)
549 return cpy
551 def set_results_name(
552 self, name: str, list_all_matches: bool = False, *, listAllMatches: bool = False
553 ) -> "ParserElement":
554 """
555 Define name for referencing matching tokens as a nested attribute
556 of the returned parse results.
558 Normally, results names are assigned as you would assign keys in a dict:
559 any existing value is overwritten by later values. If it is necessary to
560 keep all values captured for a particular results name, call ``set_results_name``
561 with ``list_all_matches`` = True.
563 NOTE: ``set_results_name`` returns a *copy* of the original :class:`ParserElement` object;
564 this is so that the client can define a basic element, such as an
565 integer, and reference it in multiple places with different names.
567 You can also set results names using the abbreviated syntax,
568 ``expr("name")`` in place of ``expr.set_results_name("name")``
569 - see :class:`__call__`. If ``list_all_matches`` is required, use
570 ``expr("name*")``.
572 Example::
574 date_str = (integer.set_results_name("year") + '/'
575 + integer.set_results_name("month") + '/'
576 + integer.set_results_name("day"))
578 # equivalent form:
579 date_str = integer("year") + '/' + integer("month") + '/' + integer("day")
580 """
581 listAllMatches = listAllMatches or list_all_matches
582 return self._setResultsName(name, listAllMatches)
584 def _setResultsName(self, name, listAllMatches=False):
585 if name is None:
586 return self
587 newself = self.copy()
588 if name.endswith("*"):
589 name = name[:-1]
590 listAllMatches = True
591 newself.resultsName = name
592 newself.modalResults = not listAllMatches
593 return newself
595 def set_break(self, break_flag: bool = True) -> "ParserElement":
596 """
597 Method to invoke the Python pdb debugger when this element is
598 about to be parsed. Set ``break_flag`` to ``True`` to enable, ``False`` to
599 disable.
600 """
601 if break_flag:
602 _parseMethod = self._parse
604 def breaker(instring, loc, doActions=True, callPreParse=True):
605 import pdb
607 # this call to pdb.set_trace() is intentional, not a checkin error
608 pdb.set_trace()
609 return _parseMethod(instring, loc, doActions, callPreParse)
611 breaker._originalParseMethod = _parseMethod # type: ignore [attr-defined]
612 self._parse = breaker # type: ignore [assignment]
613 else:
614 if hasattr(self._parse, "_originalParseMethod"):
615 self._parse = self._parse._originalParseMethod # type: ignore [attr-defined, assignment]
616 return self
618 def set_parse_action(self, *fns: ParseAction, **kwargs) -> "ParserElement":
619 """
620 Define one or more actions to perform when successfully matching parse element definition.
622 Parse actions can be called to perform data conversions, do extra validation,
623 update external data structures, or enhance or replace the parsed tokens.
624 Each parse action ``fn`` is a callable method with 0-3 arguments, called as
625 ``fn(s, loc, toks)`` , ``fn(loc, toks)`` , ``fn(toks)`` , or just ``fn()`` , where:
627 - ``s`` = the original string being parsed (see note below)
628 - ``loc`` = the location of the matching substring
629 - ``toks`` = a list of the matched tokens, packaged as a :class:`ParseResults` object
631 The parsed tokens are passed to the parse action as ParseResults. They can be
632 modified in place using list-style append, extend, and pop operations to update
633 the parsed list elements; and with dictionary-style item set and del operations
634 to add, update, or remove any named results. If the tokens are modified in place,
635 it is not necessary to return them with a return statement.
637 Parse actions can also completely replace the given tokens, with another ``ParseResults``
638 object, or with some entirely different object (common for parse actions that perform data
639 conversions). A convenient way to build a new parse result is to define the values
640 using a dict, and then create the return value using :class:`ParseResults.from_dict`.
642 If None is passed as the ``fn`` parse action, all previously added parse actions for this
643 expression are cleared.
645 Optional keyword arguments:
647 - ``call_during_try`` = (default= ``False``) indicate if parse action should be run during
648 lookaheads and alternate testing. For parse actions that have side effects, it is
649 important to only call the parse action once it is determined that it is being
650 called as part of a successful parse. For parse actions that perform additional
651 validation, then call_during_try should be passed as True, so that the validation
652 code is included in the preliminary "try" parses.
654 Note: the default parsing behavior is to expand tabs in the input string
655 before starting the parsing process. See :class:`parse_string` for more
656 information on parsing strings containing ``<TAB>`` s, and suggested
657 methods to maintain a consistent view of the parsed string, the parse
658 location, and line and column positions within the parsed string.
660 Example::
662 # parse dates in the form YYYY/MM/DD
664 # use parse action to convert toks from str to int at parse time
665 def convert_to_int(toks):
666 return int(toks[0])
668 # use a parse action to verify that the date is a valid date
669 def is_valid_date(instring, loc, toks):
670 from datetime import date
671 year, month, day = toks[::2]
672 try:
673 date(year, month, day)
674 except ValueError:
675 raise ParseException(instring, loc, "invalid date given")
677 integer = Word(nums)
678 date_str = integer + '/' + integer + '/' + integer
680 # add parse actions
681 integer.set_parse_action(convert_to_int)
682 date_str.set_parse_action(is_valid_date)
684 # note that integer fields are now ints, not strings
685 date_str.run_tests('''
686 # successful parse - note that integer fields were converted to ints
687 1999/12/31
689 # fail - invalid date
690 1999/13/31
691 ''')
692 """
693 if list(fns) == [None]:
694 self.parseAction = []
695 else:
696 if not all(callable(fn) for fn in fns):
697 raise TypeError("parse actions must be callable")
698 self.parseAction = [_trim_arity(fn) for fn in fns]
699 self.callDuringTry = kwargs.get(
700 "call_during_try", kwargs.get("callDuringTry", False)
701 )
702 return self
704 def add_parse_action(self, *fns: ParseAction, **kwargs) -> "ParserElement":
705 """
706 Add one or more parse actions to expression's list of parse actions. See :class:`set_parse_action`.
708 See examples in :class:`copy`.
709 """
710 self.parseAction += [_trim_arity(fn) for fn in fns]
711 self.callDuringTry = self.callDuringTry or kwargs.get(
712 "call_during_try", kwargs.get("callDuringTry", False)
713 )
714 return self
716 def add_condition(self, *fns: ParseCondition, **kwargs) -> "ParserElement":
717 """Add a boolean predicate function to expression's list of parse actions. See
718 :class:`set_parse_action` for function call signatures. Unlike ``set_parse_action``,
719 functions passed to ``add_condition`` need to return boolean success/fail of the condition.
721 Optional keyword arguments:
723 - ``message`` = define a custom message to be used in the raised exception
724 - ``fatal`` = if True, will raise ParseFatalException to stop parsing immediately; otherwise will raise
725 ParseException
726 - ``call_during_try`` = boolean to indicate if this method should be called during internal tryParse calls,
727 default=False
729 Example::
731 integer = Word(nums).set_parse_action(lambda toks: int(toks[0]))
732 year_int = integer.copy()
733 year_int.add_condition(lambda toks: toks[0] >= 2000, message="Only support years 2000 and later")
734 date_str = year_int + '/' + integer + '/' + integer
736 result = date_str.parse_string("1999/12/31") # -> Exception: Only support years 2000 and later (at char 0),
737 (line:1, col:1)
738 """
739 for fn in fns:
740 self.parseAction.append(
741 condition_as_parse_action(
742 fn,
743 message=str(kwargs.get("message")),
744 fatal=bool(kwargs.get("fatal", False)),
745 )
746 )
748 self.callDuringTry = self.callDuringTry or kwargs.get(
749 "call_during_try", kwargs.get("callDuringTry", False)
750 )
751 return self
753 def set_fail_action(self, fn: ParseFailAction) -> "ParserElement":
754 """
755 Define action to perform if parsing fails at this expression.
756 Fail acton fn is a callable function that takes the arguments
757 ``fn(s, loc, expr, err)`` where:
759 - ``s`` = string being parsed
760 - ``loc`` = location where expression match was attempted and failed
761 - ``expr`` = the parse expression that failed
762 - ``err`` = the exception thrown
764 The function returns no value. It may throw :class:`ParseFatalException`
765 if it is desired to stop parsing immediately."""
766 self.failAction = fn
767 return self
769 def _skipIgnorables(self, instring: str, loc: int) -> int:
770 if not self.ignoreExprs:
771 return loc
772 exprsFound = True
773 ignore_expr_fns = [e._parse for e in self.ignoreExprs]
774 while exprsFound:
775 exprsFound = False
776 for ignore_fn in ignore_expr_fns:
777 try:
778 while 1:
779 loc, dummy = ignore_fn(instring, loc)
780 exprsFound = True
781 except ParseException:
782 pass
783 return loc
785 def preParse(self, instring: str, loc: int) -> int:
786 if self.ignoreExprs:
787 loc = self._skipIgnorables(instring, loc)
789 if self.skipWhitespace:
790 instrlen = len(instring)
791 white_chars = self.whiteChars
792 while loc < instrlen and instring[loc] in white_chars:
793 loc += 1
795 return loc
797 def parseImpl(self, instring, loc, doActions=True):
798 return loc, []
800 def postParse(self, instring, loc, tokenlist):
801 return tokenlist
803 # @profile
804 def _parseNoCache(
805 self, instring, loc, doActions=True, callPreParse=True
806 ) -> Tuple[int, ParseResults]:
807 TRY, MATCH, FAIL = 0, 1, 2
808 debugging = self.debug # and doActions)
809 len_instring = len(instring)
811 if debugging or self.failAction:
812 # print("Match {} at loc {}({}, {})".format(self, loc, lineno(loc, instring), col(loc, instring)))
813 try:
814 if callPreParse and self.callPreparse:
815 pre_loc = self.preParse(instring, loc)
816 else:
817 pre_loc = loc
818 tokens_start = pre_loc
819 if self.debugActions.debug_try:
820 self.debugActions.debug_try(instring, tokens_start, self, False)
821 if self.mayIndexError or pre_loc >= len_instring:
822 try:
823 loc, tokens = self.parseImpl(instring, pre_loc, doActions)
824 except IndexError:
825 raise ParseException(instring, len_instring, self.errmsg, self)
826 else:
827 loc, tokens = self.parseImpl(instring, pre_loc, doActions)
828 except Exception as err:
829 # print("Exception raised:", err)
830 if self.debugActions.debug_fail:
831 self.debugActions.debug_fail(
832 instring, tokens_start, self, err, False
833 )
834 if self.failAction:
835 self.failAction(instring, tokens_start, self, err)
836 raise
837 else:
838 if callPreParse and self.callPreparse:
839 pre_loc = self.preParse(instring, loc)
840 else:
841 pre_loc = loc
842 tokens_start = pre_loc
843 if self.mayIndexError or pre_loc >= len_instring:
844 try:
845 loc, tokens = self.parseImpl(instring, pre_loc, doActions)
846 except IndexError:
847 raise ParseException(instring, len_instring, self.errmsg, self)
848 else:
849 loc, tokens = self.parseImpl(instring, pre_loc, doActions)
851 tokens = self.postParse(instring, loc, tokens)
853 ret_tokens = ParseResults(
854 tokens, self.resultsName, asList=self.saveAsList, modal=self.modalResults
855 )
856 if self.parseAction and (doActions or self.callDuringTry):
857 if debugging:
858 try:
859 for fn in self.parseAction:
860 try:
861 tokens = fn(instring, tokens_start, ret_tokens) # type: ignore [call-arg, arg-type]
862 except IndexError as parse_action_exc:
863 exc = ParseException("exception raised in parse action")
864 raise exc from parse_action_exc
866 if tokens is not None and tokens is not ret_tokens:
867 ret_tokens = ParseResults(
868 tokens,
869 self.resultsName,
870 asList=self.saveAsList
871 and isinstance(tokens, (ParseResults, list)),
872 modal=self.modalResults,
873 )
874 except Exception as err:
875 # print "Exception raised in user parse action:", err
876 if self.debugActions.debug_fail:
877 self.debugActions.debug_fail(
878 instring, tokens_start, self, err, False
879 )
880 raise
881 else:
882 for fn in self.parseAction:
883 try:
884 tokens = fn(instring, tokens_start, ret_tokens) # type: ignore [call-arg, arg-type]
885 except IndexError as parse_action_exc:
886 exc = ParseException("exception raised in parse action")
887 raise exc from parse_action_exc
889 if tokens is not None and tokens is not ret_tokens:
890 ret_tokens = ParseResults(
891 tokens,
892 self.resultsName,
893 asList=self.saveAsList
894 and isinstance(tokens, (ParseResults, list)),
895 modal=self.modalResults,
896 )
897 if debugging:
898 # print("Matched", self, "->", ret_tokens.as_list())
899 if self.debugActions.debug_match:
900 self.debugActions.debug_match(
901 instring, tokens_start, loc, self, ret_tokens, False
902 )
904 return loc, ret_tokens
906 def try_parse(
907 self,
908 instring: str,
909 loc: int,
910 *,
911 raise_fatal: bool = False,
912 do_actions: bool = False,
913 ) -> int:
914 try:
915 return self._parse(instring, loc, doActions=do_actions)[0]
916 except ParseFatalException:
917 if raise_fatal:
918 raise
919 raise ParseException(instring, loc, self.errmsg, self)
921 def can_parse_next(self, instring: str, loc: int, do_actions: bool = False) -> bool:
922 try:
923 self.try_parse(instring, loc, do_actions=do_actions)
924 except (ParseException, IndexError):
925 return False
926 else:
927 return True
929 # cache for left-recursion in Forward references
930 recursion_lock = RLock()
931 recursion_memos: typing.Dict[
932 Tuple[int, "Forward", bool], Tuple[int, Union[ParseResults, Exception]]
933 ] = {}
935 class _CacheType(dict):
936 """
937 class to help type checking
938 """
940 not_in_cache: bool
942 def get(self, *args):
943 ...
945 def set(self, *args):
946 ...
948 # argument cache for optimizing repeated calls when backtracking through recursive expressions
949 packrat_cache = (
950 _CacheType()
951 ) # set later by enable_packrat(); this is here so that reset_cache() doesn't fail
952 packrat_cache_lock = RLock()
953 packrat_cache_stats = [0, 0]
955 # this method gets repeatedly called during backtracking with the same arguments -
956 # we can cache these arguments and save ourselves the trouble of re-parsing the contained expression
957 def _parseCache(
958 self, instring, loc, doActions=True, callPreParse=True
959 ) -> Tuple[int, ParseResults]:
960 HIT, MISS = 0, 1
961 TRY, MATCH, FAIL = 0, 1, 2
962 lookup = (self, instring, loc, callPreParse, doActions)
963 with ParserElement.packrat_cache_lock:
964 cache = ParserElement.packrat_cache
965 value = cache.get(lookup)
966 if value is cache.not_in_cache:
967 ParserElement.packrat_cache_stats[MISS] += 1
968 try:
969 value = self._parseNoCache(instring, loc, doActions, callPreParse)
970 except ParseBaseException as pe:
971 # cache a copy of the exception, without the traceback
972 cache.set(lookup, pe.__class__(*pe.args))
973 raise
974 else:
975 cache.set(lookup, (value[0], value[1].copy(), loc))
976 return value
977 else:
978 ParserElement.packrat_cache_stats[HIT] += 1
979 if self.debug and self.debugActions.debug_try:
980 try:
981 self.debugActions.debug_try(instring, loc, self, cache_hit=True) # type: ignore [call-arg]
982 except TypeError:
983 pass
984 if isinstance(value, Exception):
985 if self.debug and self.debugActions.debug_fail:
986 try:
987 self.debugActions.debug_fail(
988 instring, loc, self, value, cache_hit=True # type: ignore [call-arg]
989 )
990 except TypeError:
991 pass
992 raise value
994 value = cast(Tuple[int, ParseResults, int], value)
995 loc_, result, endloc = value[0], value[1].copy(), value[2]
996 if self.debug and self.debugActions.debug_match:
997 try:
998 self.debugActions.debug_match(
999 instring, loc_, endloc, self, result, cache_hit=True # type: ignore [call-arg]
1000 )
1001 except TypeError:
1002 pass
1004 return loc_, result
1006 _parse = _parseNoCache
1008 @staticmethod
1009 def reset_cache() -> None:
1010 ParserElement.packrat_cache.clear()
1011 ParserElement.packrat_cache_stats[:] = [0] * len(
1012 ParserElement.packrat_cache_stats
1013 )
1014 ParserElement.recursion_memos.clear()
1016 _packratEnabled = False
1017 _left_recursion_enabled = False
1019 @staticmethod
1020 def disable_memoization() -> None:
1021 """
1022 Disables active Packrat or Left Recursion parsing and their memoization
1024 This method also works if neither Packrat nor Left Recursion are enabled.
1025 This makes it safe to call before activating Packrat nor Left Recursion
1026 to clear any previous settings.
1027 """
1028 ParserElement.reset_cache()
1029 ParserElement._left_recursion_enabled = False
1030 ParserElement._packratEnabled = False
1031 ParserElement._parse = ParserElement._parseNoCache
1033 @staticmethod
1034 def enable_left_recursion(
1035 cache_size_limit: typing.Optional[int] = None, *, force=False
1036 ) -> None:
1037 """
1038 Enables "bounded recursion" parsing, which allows for both direct and indirect
1039 left-recursion. During parsing, left-recursive :class:`Forward` elements are
1040 repeatedly matched with a fixed recursion depth that is gradually increased
1041 until finding the longest match.
1043 Example::
1045 from pip._vendor import pyparsing as pp
1046 pp.ParserElement.enable_left_recursion()
1048 E = pp.Forward("E")
1049 num = pp.Word(pp.nums)
1050 # match `num`, or `num '+' num`, or `num '+' num '+' num`, ...
1051 E <<= E + '+' - num | num
1053 print(E.parse_string("1+2+3"))
1055 Recursion search naturally memoizes matches of ``Forward`` elements and may
1056 thus skip reevaluation of parse actions during backtracking. This may break
1057 programs with parse actions which rely on strict ordering of side-effects.
1059 Parameters:
1061 - ``cache_size_limit`` - (default=``None``) - memoize at most this many
1062 ``Forward`` elements during matching; if ``None`` (the default),
1063 memoize all ``Forward`` elements.
1065 Bounded Recursion parsing works similar but not identical to Packrat parsing,
1066 thus the two cannot be used together. Use ``force=True`` to disable any
1067 previous, conflicting settings.
1068 """
1069 if force:
1070 ParserElement.disable_memoization()
1071 elif ParserElement._packratEnabled:
1072 raise RuntimeError("Packrat and Bounded Recursion are not compatible")
1073 if cache_size_limit is None:
1074 ParserElement.recursion_memos = _UnboundedMemo() # type: ignore[assignment]
1075 elif cache_size_limit > 0:
1076 ParserElement.recursion_memos = _LRUMemo(capacity=cache_size_limit) # type: ignore[assignment]
1077 else:
1078 raise NotImplementedError("Memo size of %s" % cache_size_limit)
1079 ParserElement._left_recursion_enabled = True
1081 @staticmethod
1082 def enable_packrat(cache_size_limit: int = 128, *, force: bool = False) -> None:
1083 """
1084 Enables "packrat" parsing, which adds memoizing to the parsing logic.
1085 Repeated parse attempts at the same string location (which happens
1086 often in many complex grammars) can immediately return a cached value,
1087 instead of re-executing parsing/validating code. Memoizing is done of
1088 both valid results and parsing exceptions.
1090 Parameters:
1092 - ``cache_size_limit`` - (default= ``128``) - if an integer value is provided
1093 will limit the size of the packrat cache; if None is passed, then
1094 the cache size will be unbounded; if 0 is passed, the cache will
1095 be effectively disabled.
1097 This speedup may break existing programs that use parse actions that
1098 have side-effects. For this reason, packrat parsing is disabled when
1099 you first import pyparsing. To activate the packrat feature, your
1100 program must call the class method :class:`ParserElement.enable_packrat`.
1101 For best results, call ``enable_packrat()`` immediately after
1102 importing pyparsing.
1104 Example::
1106 from pip._vendor import pyparsing
1107 pyparsing.ParserElement.enable_packrat()
1109 Packrat parsing works similar but not identical to Bounded Recursion parsing,
1110 thus the two cannot be used together. Use ``force=True`` to disable any
1111 previous, conflicting settings.
1112 """
1113 if force:
1114 ParserElement.disable_memoization()
1115 elif ParserElement._left_recursion_enabled:
1116 raise RuntimeError("Packrat and Bounded Recursion are not compatible")
1117 if not ParserElement._packratEnabled:
1118 ParserElement._packratEnabled = True
1119 if cache_size_limit is None:
1120 ParserElement.packrat_cache = _UnboundedCache()
1121 else:
1122 ParserElement.packrat_cache = _FifoCache(cache_size_limit) # type: ignore[assignment]
1123 ParserElement._parse = ParserElement._parseCache
1125 def parse_string(
1126 self, instring: str, parse_all: bool = False, *, parseAll: bool = False
1127 ) -> ParseResults:
1128 """
1129 Parse a string with respect to the parser definition. This function is intended as the primary interface to the
1130 client code.
1132 :param instring: The input string to be parsed.
1133 :param parse_all: If set, the entire input string must match the grammar.
1134 :param parseAll: retained for pre-PEP8 compatibility, will be removed in a future release.
1135 :raises ParseException: Raised if ``parse_all`` is set and the input string does not match the whole grammar.
1136 :returns: the parsed data as a :class:`ParseResults` object, which may be accessed as a `list`, a `dict`, or
1137 an object with attributes if the given parser includes results names.
1139 If the input string is required to match the entire grammar, ``parse_all`` flag must be set to ``True``. This
1140 is also equivalent to ending the grammar with :class:`StringEnd`\\ ().
1142 To report proper column numbers, ``parse_string`` operates on a copy of the input string where all tabs are
1143 converted to spaces (8 spaces per tab, as per the default in ``string.expandtabs``). If the input string
1144 contains tabs and the grammar uses parse actions that use the ``loc`` argument to index into the string
1145 being parsed, one can ensure a consistent view of the input string by doing one of the following:
1147 - calling ``parse_with_tabs`` on your grammar before calling ``parse_string`` (see :class:`parse_with_tabs`),
1148 - define your parse action using the full ``(s,loc,toks)`` signature, and reference the input string using the
1149 parse action's ``s`` argument, or
1150 - explicitly expand the tabs in your input string before calling ``parse_string``.
1152 Examples:
1154 By default, partial matches are OK.
1156 >>> res = Word('a').parse_string('aaaaabaaa')
1157 >>> print(res)
1158 ['aaaaa']
1160 The parsing behavior varies by the inheriting class of this abstract class. Please refer to the children
1161 directly to see more examples.
1163 It raises an exception if parse_all flag is set and instring does not match the whole grammar.
1165 >>> res = Word('a').parse_string('aaaaabaaa', parse_all=True)
1166 Traceback (most recent call last):
1167 ...
1168 pyparsing.ParseException: Expected end of text, found 'b' (at char 5), (line:1, col:6)
1169 """
1170 parseAll = parse_all or parseAll
1172 ParserElement.reset_cache()
1173 if not self.streamlined:
1174 self.streamline()
1175 for e in self.ignoreExprs:
1176 e.streamline()
1177 if not self.keepTabs:
1178 instring = instring.expandtabs()
1179 try:
1180 loc, tokens = self._parse(instring, 0)
1181 if parseAll:
1182 loc = self.preParse(instring, loc)
1183 se = Empty() + StringEnd()
1184 se._parse(instring, loc)
1185 except ParseBaseException as exc:
1186 if ParserElement.verbose_stacktrace:
1187 raise
1188 else:
1189 # catch and re-raise exception from here, clearing out pyparsing internal stack trace
1190 raise exc.with_traceback(None)
1191 else:
1192 return tokens
1194 def scan_string(
1195 self,
1196 instring: str,
1197 max_matches: int = _MAX_INT,
1198 overlap: bool = False,
1199 *,
1200 debug: bool = False,
1201 maxMatches: int = _MAX_INT,
1202 ) -> Generator[Tuple[ParseResults, int, int], None, None]:
1203 """
1204 Scan the input string for expression matches. Each match will return the
1205 matching tokens, start location, and end location. May be called with optional
1206 ``max_matches`` argument, to clip scanning after 'n' matches are found. If
1207 ``overlap`` is specified, then overlapping matches will be reported.
1209 Note that the start and end locations are reported relative to the string
1210 being parsed. See :class:`parse_string` for more information on parsing
1211 strings with embedded tabs.
1213 Example::
1215 source = "sldjf123lsdjjkf345sldkjf879lkjsfd987"
1216 print(source)
1217 for tokens, start, end in Word(alphas).scan_string(source):
1218 print(' '*start + '^'*(end-start))
1219 print(' '*start + tokens[0])
1221 prints::
1223 sldjf123lsdjjkf345sldkjf879lkjsfd987
1224 ^^^^^
1225 sldjf
1226 ^^^^^^^
1227 lsdjjkf
1228 ^^^^^^
1229 sldkjf
1230 ^^^^^^
1231 lkjsfd
1232 """
1233 maxMatches = min(maxMatches, max_matches)
1234 if not self.streamlined:
1235 self.streamline()
1236 for e in self.ignoreExprs:
1237 e.streamline()
1239 if not self.keepTabs:
1240 instring = str(instring).expandtabs()
1241 instrlen = len(instring)
1242 loc = 0
1243 preparseFn = self.preParse
1244 parseFn = self._parse
1245 ParserElement.resetCache()
1246 matches = 0
1247 try:
1248 while loc <= instrlen and matches < maxMatches:
1249 try:
1250 preloc: int = preparseFn(instring, loc)
1251 nextLoc: int
1252 tokens: ParseResults
1253 nextLoc, tokens = parseFn(instring, preloc, callPreParse=False)
1254 except ParseException:
1255 loc = preloc + 1
1256 else:
1257 if nextLoc > loc:
1258 matches += 1
1259 if debug:
1260 print(
1261 {
1262 "tokens": tokens.asList(),
1263 "start": preloc,
1264 "end": nextLoc,
1265 }
1266 )
1267 yield tokens, preloc, nextLoc
1268 if overlap:
1269 nextloc = preparseFn(instring, loc)
1270 if nextloc > loc:
1271 loc = nextLoc
1272 else:
1273 loc += 1
1274 else:
1275 loc = nextLoc
1276 else:
1277 loc = preloc + 1
1278 except ParseBaseException as exc:
1279 if ParserElement.verbose_stacktrace:
1280 raise
1281 else:
1282 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1283 raise exc.with_traceback(None)
1285 def transform_string(self, instring: str, *, debug: bool = False) -> str:
1286 """
1287 Extension to :class:`scan_string`, to modify matching text with modified tokens that may
1288 be returned from a parse action. To use ``transform_string``, define a grammar and
1289 attach a parse action to it that modifies the returned token list.
1290 Invoking ``transform_string()`` on a target string will then scan for matches,
1291 and replace the matched text patterns according to the logic in the parse
1292 action. ``transform_string()`` returns the resulting transformed string.
1294 Example::
1296 wd = Word(alphas)
1297 wd.set_parse_action(lambda toks: toks[0].title())
1299 print(wd.transform_string("now is the winter of our discontent made glorious summer by this sun of york."))
1301 prints::
1303 Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York.
1304 """
1305 out: List[str] = []
1306 lastE = 0
1307 # force preservation of <TAB>s, to minimize unwanted transformation of string, and to
1308 # keep string locs straight between transform_string and scan_string
1309 self.keepTabs = True
1310 try:
1311 for t, s, e in self.scan_string(instring, debug=debug):
1312 out.append(instring[lastE:s])
1313 if t:
1314 if isinstance(t, ParseResults):
1315 out += t.as_list()
1316 elif isinstance(t, Iterable) and not isinstance(t, str_type):
1317 out.extend(t)
1318 else:
1319 out.append(t)
1320 lastE = e
1321 out.append(instring[lastE:])
1322 out = [o for o in out if o]
1323 return "".join([str(s) for s in _flatten(out)])
1324 except ParseBaseException as exc:
1325 if ParserElement.verbose_stacktrace:
1326 raise
1327 else:
1328 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1329 raise exc.with_traceback(None)
1331 def search_string(
1332 self,
1333 instring: str,
1334 max_matches: int = _MAX_INT,
1335 *,
1336 debug: bool = False,
1337 maxMatches: int = _MAX_INT,
1338 ) -> ParseResults:
1339 """
1340 Another extension to :class:`scan_string`, simplifying the access to the tokens found
1341 to match the given parse expression. May be called with optional
1342 ``max_matches`` argument, to clip searching after 'n' matches are found.
1344 Example::
1346 # a capitalized word starts with an uppercase letter, followed by zero or more lowercase letters
1347 cap_word = Word(alphas.upper(), alphas.lower())
1349 print(cap_word.search_string("More than Iron, more than Lead, more than Gold I need Electricity"))
1351 # the sum() builtin can be used to merge results into a single ParseResults object
1352 print(sum(cap_word.search_string("More than Iron, more than Lead, more than Gold I need Electricity")))
1354 prints::
1356 [['More'], ['Iron'], ['Lead'], ['Gold'], ['I'], ['Electricity']]
1357 ['More', 'Iron', 'Lead', 'Gold', 'I', 'Electricity']
1358 """
1359 maxMatches = min(maxMatches, max_matches)
1360 try:
1361 return ParseResults(
1362 [t for t, s, e in self.scan_string(instring, maxMatches, debug=debug)]
1363 )
1364 except ParseBaseException as exc:
1365 if ParserElement.verbose_stacktrace:
1366 raise
1367 else:
1368 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1369 raise exc.with_traceback(None)
1371 def split(
1372 self,
1373 instring: str,
1374 maxsplit: int = _MAX_INT,
1375 include_separators: bool = False,
1376 *,
1377 includeSeparators=False,
1378 ) -> Generator[str, None, None]:
1379 """
1380 Generator method to split a string using the given expression as a separator.
1381 May be called with optional ``maxsplit`` argument, to limit the number of splits;
1382 and the optional ``include_separators`` argument (default= ``False``), if the separating
1383 matching text should be included in the split results.
1385 Example::
1387 punc = one_of(list(".,;:/-!?"))
1388 print(list(punc.split("This, this?, this sentence, is badly punctuated!")))
1390 prints::
1392 ['This', ' this', '', ' this sentence', ' is badly punctuated', '']
1393 """
1394 includeSeparators = includeSeparators or include_separators
1395 last = 0
1396 for t, s, e in self.scan_string(instring, max_matches=maxsplit):
1397 yield instring[last:s]
1398 if includeSeparators:
1399 yield t[0]
1400 last = e
1401 yield instring[last:]
1403 def __add__(self, other) -> "ParserElement":
1404 """
1405 Implementation of ``+`` operator - returns :class:`And`. Adding strings to a :class:`ParserElement`
1406 converts them to :class:`Literal`\\ s by default.
1408 Example::
1410 greet = Word(alphas) + "," + Word(alphas) + "!"
1411 hello = "Hello, World!"
1412 print(hello, "->", greet.parse_string(hello))
1414 prints::
1416 Hello, World! -> ['Hello', ',', 'World', '!']
1418 ``...`` may be used as a parse expression as a short form of :class:`SkipTo`::
1420 Literal('start') + ... + Literal('end')
1422 is equivalent to::
1424 Literal('start') + SkipTo('end')("_skipped*") + Literal('end')
1426 Note that the skipped text is returned with '_skipped' as a results name,
1427 and to support having multiple skips in the same parser, the value returned is
1428 a list of all skipped text.
1429 """
1430 if other is Ellipsis:
1431 return _PendingSkip(self)
1433 if isinstance(other, str_type):
1434 other = self._literalStringClass(other)
1435 if not isinstance(other, ParserElement):
1436 return NotImplemented
1437 return And([self, other])
1439 def __radd__(self, other) -> "ParserElement":
1440 """
1441 Implementation of ``+`` operator when left operand is not a :class:`ParserElement`
1442 """
1443 if other is Ellipsis:
1444 return SkipTo(self)("_skipped*") + self
1446 if isinstance(other, str_type):
1447 other = self._literalStringClass(other)
1448 if not isinstance(other, ParserElement):
1449 return NotImplemented
1450 return other + self
1452 def __sub__(self, other) -> "ParserElement":
1453 """
1454 Implementation of ``-`` operator, returns :class:`And` with error stop
1455 """
1456 if isinstance(other, str_type):
1457 other = self._literalStringClass(other)
1458 if not isinstance(other, ParserElement):
1459 return NotImplemented
1460 return self + And._ErrorStop() + other
1462 def __rsub__(self, other) -> "ParserElement":
1463 """
1464 Implementation of ``-`` operator when left operand is not a :class:`ParserElement`
1465 """
1466 if isinstance(other, str_type):
1467 other = self._literalStringClass(other)
1468 if not isinstance(other, ParserElement):
1469 return NotImplemented
1470 return other - self
1472 def __mul__(self, other) -> "ParserElement":
1473 """
1474 Implementation of ``*`` operator, allows use of ``expr * 3`` in place of
1475 ``expr + expr + expr``. Expressions may also be multiplied by a 2-integer
1476 tuple, similar to ``{min, max}`` multipliers in regular expressions. Tuples
1477 may also include ``None`` as in:
1479 - ``expr*(n, None)`` or ``expr*(n, )`` is equivalent
1480 to ``expr*n + ZeroOrMore(expr)``
1481 (read as "at least n instances of ``expr``")
1482 - ``expr*(None, n)`` is equivalent to ``expr*(0, n)``
1483 (read as "0 to n instances of ``expr``")
1484 - ``expr*(None, None)`` is equivalent to ``ZeroOrMore(expr)``
1485 - ``expr*(1, None)`` is equivalent to ``OneOrMore(expr)``
1487 Note that ``expr*(None, n)`` does not raise an exception if
1488 more than n exprs exist in the input stream; that is,
1489 ``expr*(None, n)`` does not enforce a maximum number of expr
1490 occurrences. If this behavior is desired, then write
1491 ``expr*(None, n) + ~expr``
1492 """
1493 if other is Ellipsis:
1494 other = (0, None)
1495 elif isinstance(other, tuple) and other[:1] == (Ellipsis,):
1496 other = ((0,) + other[1:] + (None,))[:2]
1498 if isinstance(other, int):
1499 minElements, optElements = other, 0
1500 elif isinstance(other, tuple):
1501 other = tuple(o if o is not Ellipsis else None for o in other)
1502 other = (other + (None, None))[:2]
1503 if other[0] is None:
1504 other = (0, other[1])
1505 if isinstance(other[0], int) and other[1] is None:
1506 if other[0] == 0:
1507 return ZeroOrMore(self)
1508 if other[0] == 1:
1509 return OneOrMore(self)
1510 else:
1511 return self * other[0] + ZeroOrMore(self)
1512 elif isinstance(other[0], int) and isinstance(other[1], int):
1513 minElements, optElements = other
1514 optElements -= minElements
1515 else:
1516 return NotImplemented
1517 else:
1518 return NotImplemented
1520 if minElements < 0:
1521 raise ValueError("cannot multiply ParserElement by negative value")
1522 if optElements < 0:
1523 raise ValueError(
1524 "second tuple value must be greater or equal to first tuple value"
1525 )
1526 if minElements == optElements == 0:
1527 return And([])
1529 if optElements:
1531 def makeOptionalList(n):
1532 if n > 1:
1533 return Opt(self + makeOptionalList(n - 1))
1534 else:
1535 return Opt(self)
1537 if minElements:
1538 if minElements == 1:
1539 ret = self + makeOptionalList(optElements)
1540 else:
1541 ret = And([self] * minElements) + makeOptionalList(optElements)
1542 else:
1543 ret = makeOptionalList(optElements)
1544 else:
1545 if minElements == 1:
1546 ret = self
1547 else:
1548 ret = And([self] * minElements)
1549 return ret
1551 def __rmul__(self, other) -> "ParserElement":
1552 return self.__mul__(other)
1554 def __or__(self, other) -> "ParserElement":
1555 """
1556 Implementation of ``|`` operator - returns :class:`MatchFirst`
1557 """
1558 if other is Ellipsis:
1559 return _PendingSkip(self, must_skip=True)
1561 if isinstance(other, str_type):
1562 # `expr | ""` is equivalent to `Opt(expr)`
1563 if other == "":
1564 return Opt(self)
1565 other = self._literalStringClass(other)
1566 if not isinstance(other, ParserElement):
1567 return NotImplemented
1568 return MatchFirst([self, other])
1570 def __ror__(self, other) -> "ParserElement":
1571 """
1572 Implementation of ``|`` operator when left operand is not a :class:`ParserElement`
1573 """
1574 if isinstance(other, str_type):
1575 other = self._literalStringClass(other)
1576 if not isinstance(other, ParserElement):
1577 return NotImplemented
1578 return other | self
1580 def __xor__(self, other) -> "ParserElement":
1581 """
1582 Implementation of ``^`` operator - returns :class:`Or`
1583 """
1584 if isinstance(other, str_type):
1585 other = self._literalStringClass(other)
1586 if not isinstance(other, ParserElement):
1587 return NotImplemented
1588 return Or([self, other])
1590 def __rxor__(self, other) -> "ParserElement":
1591 """
1592 Implementation of ``^`` operator when left operand is not a :class:`ParserElement`
1593 """
1594 if isinstance(other, str_type):
1595 other = self._literalStringClass(other)
1596 if not isinstance(other, ParserElement):
1597 return NotImplemented
1598 return other ^ self
1600 def __and__(self, other) -> "ParserElement":
1601 """
1602 Implementation of ``&`` operator - returns :class:`Each`
1603 """
1604 if isinstance(other, str_type):
1605 other = self._literalStringClass(other)
1606 if not isinstance(other, ParserElement):
1607 return NotImplemented
1608 return Each([self, other])
1610 def __rand__(self, other) -> "ParserElement":
1611 """
1612 Implementation of ``&`` operator when left operand is not a :class:`ParserElement`
1613 """
1614 if isinstance(other, str_type):
1615 other = self._literalStringClass(other)
1616 if not isinstance(other, ParserElement):
1617 return NotImplemented
1618 return other & self
1620 def __invert__(self) -> "ParserElement":
1621 """
1622 Implementation of ``~`` operator - returns :class:`NotAny`
1623 """
1624 return NotAny(self)
1626 # disable __iter__ to override legacy use of sequential access to __getitem__ to
1627 # iterate over a sequence
1628 __iter__ = None
1630 def __getitem__(self, key):
1631 """
1632 use ``[]`` indexing notation as a short form for expression repetition:
1634 - ``expr[n]`` is equivalent to ``expr*n``
1635 - ``expr[m, n]`` is equivalent to ``expr*(m, n)``
1636 - ``expr[n, ...]`` or ``expr[n,]`` is equivalent
1637 to ``expr*n + ZeroOrMore(expr)``
1638 (read as "at least n instances of ``expr``")
1639 - ``expr[..., n]`` is equivalent to ``expr*(0, n)``
1640 (read as "0 to n instances of ``expr``")
1641 - ``expr[...]`` and ``expr[0, ...]`` are equivalent to ``ZeroOrMore(expr)``
1642 - ``expr[1, ...]`` is equivalent to ``OneOrMore(expr)``
1644 ``None`` may be used in place of ``...``.
1646 Note that ``expr[..., n]`` and ``expr[m, n]`` do not raise an exception
1647 if more than ``n`` ``expr``\\ s exist in the input stream. If this behavior is
1648 desired, then write ``expr[..., n] + ~expr``.
1650 For repetition with a stop_on expression, use slice notation:
1652 - ``expr[...: end_expr]`` and ``expr[0, ...: end_expr]`` are equivalent to ``ZeroOrMore(expr, stop_on=end_expr)``
1653 - ``expr[1, ...: end_expr]`` is equivalent to ``OneOrMore(expr, stop_on=end_expr)``
1655 """
1657 stop_on_defined = False
1658 stop_on = NoMatch()
1659 if isinstance(key, slice):
1660 key, stop_on = key.start, key.stop
1661 if key is None:
1662 key = ...
1663 stop_on_defined = True
1664 elif isinstance(key, tuple) and isinstance(key[-1], slice):
1665 key, stop_on = (key[0], key[1].start), key[1].stop
1666 stop_on_defined = True
1668 # convert single arg keys to tuples
1669 if isinstance(key, str_type):
1670 key = (key,)
1671 try:
1672 iter(key)
1673 except TypeError:
1674 key = (key, key)
1676 if len(key) > 2:
1677 raise TypeError(
1678 f"only 1 or 2 index arguments supported ({key[:5]}{f'... [{len(key)}]' if len(key) > 5 else ''})"
1679 )
1681 # clip to 2 elements
1682 ret = self * tuple(key[:2])
1683 ret = typing.cast(_MultipleMatch, ret)
1685 if stop_on_defined:
1686 ret.stopOn(stop_on)
1688 return ret
1690 def __call__(self, name: typing.Optional[str] = None) -> "ParserElement":
1691 """
1692 Shortcut for :class:`set_results_name`, with ``list_all_matches=False``.
1694 If ``name`` is given with a trailing ``'*'`` character, then ``list_all_matches`` will be
1695 passed as ``True``.
1697 If ``name`` is omitted, same as calling :class:`copy`.
1699 Example::
1701 # these are equivalent
1702 userdata = Word(alphas).set_results_name("name") + Word(nums + "-").set_results_name("socsecno")
1703 userdata = Word(alphas)("name") + Word(nums + "-")("socsecno")
1704 """
1705 if name is not None:
1706 return self._setResultsName(name)
1707 else:
1708 return self.copy()
1710 def suppress(self) -> "ParserElement":
1711 """
1712 Suppresses the output of this :class:`ParserElement`; useful to keep punctuation from
1713 cluttering up returned output.
1714 """
1715 return Suppress(self)
1717 def ignore_whitespace(self, recursive: bool = True) -> "ParserElement":
1718 """
1719 Enables the skipping of whitespace before matching the characters in the
1720 :class:`ParserElement`'s defined pattern.
1722 :param recursive: If ``True`` (the default), also enable whitespace skipping in child elements (if any)
1723 """
1724 self.skipWhitespace = True
1725 return self
1727 def leave_whitespace(self, recursive: bool = True) -> "ParserElement":
1728 """
1729 Disables the skipping of whitespace before matching the characters in the
1730 :class:`ParserElement`'s defined pattern. This is normally only used internally by
1731 the pyparsing module, but may be needed in some whitespace-sensitive grammars.
1733 :param recursive: If true (the default), also disable whitespace skipping in child elements (if any)
1734 """
1735 self.skipWhitespace = False
1736 return self
1738 def set_whitespace_chars(
1739 self, chars: Union[Set[str], str], copy_defaults: bool = False
1740 ) -> "ParserElement":
1741 """
1742 Overrides the default whitespace chars
1743 """
1744 self.skipWhitespace = True
1745 self.whiteChars = set(chars)
1746 self.copyDefaultWhiteChars = copy_defaults
1747 return self
1749 def parse_with_tabs(self) -> "ParserElement":
1750 """
1751 Overrides default behavior to expand ``<TAB>`` s to spaces before parsing the input string.
1752 Must be called before ``parse_string`` when the input grammar contains elements that
1753 match ``<TAB>`` characters.
1754 """
1755 self.keepTabs = True
1756 return self
1758 def ignore(self, other: "ParserElement") -> "ParserElement":
1759 """
1760 Define expression to be ignored (e.g., comments) while doing pattern
1761 matching; may be called repeatedly, to define multiple comment or other
1762 ignorable patterns.
1764 Example::
1766 patt = Word(alphas)[1, ...]
1767 patt.parse_string('ablaj /* comment */ lskjd')
1768 # -> ['ablaj']
1770 patt.ignore(c_style_comment)
1771 patt.parse_string('ablaj /* comment */ lskjd')
1772 # -> ['ablaj', 'lskjd']
1773 """
1774 import typing
1776 if isinstance(other, str_type):
1777 other = Suppress(other)
1779 if isinstance(other, Suppress):
1780 if other not in self.ignoreExprs:
1781 self.ignoreExprs.append(other)
1782 else:
1783 self.ignoreExprs.append(Suppress(other.copy()))
1784 return self
1786 def set_debug_actions(
1787 self,
1788 start_action: DebugStartAction,
1789 success_action: DebugSuccessAction,
1790 exception_action: DebugExceptionAction,
1791 ) -> "ParserElement":
1792 """
1793 Customize display of debugging messages while doing pattern matching:
1795 - ``start_action`` - method to be called when an expression is about to be parsed;
1796 should have the signature ``fn(input_string: str, location: int, expression: ParserElement, cache_hit: bool)``
1798 - ``success_action`` - method to be called when an expression has successfully parsed;
1799 should have the signature ``fn(input_string: str, start_location: int, end_location: int, expression: ParserELement, parsed_tokens: ParseResults, cache_hit: bool)``
1801 - ``exception_action`` - method to be called when expression fails to parse;
1802 should have the signature ``fn(input_string: str, location: int, expression: ParserElement, exception: Exception, cache_hit: bool)``
1803 """
1804 self.debugActions = self.DebugActions(
1805 start_action or _default_start_debug_action, # type: ignore[truthy-function]
1806 success_action or _default_success_debug_action, # type: ignore[truthy-function]
1807 exception_action or _default_exception_debug_action, # type: ignore[truthy-function]
1808 )
1809 self.debug = True
1810 return self
1812 def set_debug(self, flag: bool = True, recurse: bool = False) -> "ParserElement":
1813 """
1814 Enable display of debugging messages while doing pattern matching.
1815 Set ``flag`` to ``True`` to enable, ``False`` to disable.
1816 Set ``recurse`` to ``True`` to set the debug flag on this expression and all sub-expressions.
1818 Example::
1820 wd = Word(alphas).set_name("alphaword")
1821 integer = Word(nums).set_name("numword")
1822 term = wd | integer
1824 # turn on debugging for wd
1825 wd.set_debug()
1827 term[1, ...].parse_string("abc 123 xyz 890")
1829 prints::
1831 Match alphaword at loc 0(1,1)
1832 Matched alphaword -> ['abc']
1833 Match alphaword at loc 3(1,4)
1834 Exception raised:Expected alphaword (at char 4), (line:1, col:5)
1835 Match alphaword at loc 7(1,8)
1836 Matched alphaword -> ['xyz']
1837 Match alphaword at loc 11(1,12)
1838 Exception raised:Expected alphaword (at char 12), (line:1, col:13)
1839 Match alphaword at loc 15(1,16)
1840 Exception raised:Expected alphaword (at char 15), (line:1, col:16)
1842 The output shown is that produced by the default debug actions - custom debug actions can be
1843 specified using :class:`set_debug_actions`. Prior to attempting
1844 to match the ``wd`` expression, the debugging message ``"Match <exprname> at loc <n>(<line>,<col>)"``
1845 is shown. Then if the parse succeeds, a ``"Matched"`` message is shown, or an ``"Exception raised"``
1846 message is shown. Also note the use of :class:`set_name` to assign a human-readable name to the expression,
1847 which makes debugging and exception messages easier to understand - for instance, the default
1848 name created for the :class:`Word` expression without calling ``set_name`` is ``"W:(A-Za-z)"``.
1849 """
1850 if recurse:
1851 for expr in self.visit_all():
1852 expr.set_debug(flag, recurse=False)
1853 return self
1855 if flag:
1856 self.set_debug_actions(
1857 _default_start_debug_action,
1858 _default_success_debug_action,
1859 _default_exception_debug_action,
1860 )
1861 else:
1862 self.debug = False
1863 return self
1865 @property
1866 def default_name(self) -> str:
1867 if self._defaultName is None:
1868 self._defaultName = self._generateDefaultName()
1869 return self._defaultName
1871 @abstractmethod
1872 def _generateDefaultName(self) -> str:
1873 """
1874 Child classes must define this method, which defines how the ``default_name`` is set.
1875 """
1877 def set_name(self, name: str) -> "ParserElement":
1878 """
1879 Define name for this expression, makes debugging and exception messages clearer.
1881 Example::
1883 Word(nums).parse_string("ABC") # -> Exception: Expected W:(0-9) (at char 0), (line:1, col:1)
1884 Word(nums).set_name("integer").parse_string("ABC") # -> Exception: Expected integer (at char 0), (line:1, col:1)
1885 """
1886 self.customName = name
1887 self.errmsg = "Expected " + self.name
1888 if __diag__.enable_debug_on_named_expressions:
1889 self.set_debug()
1890 return self
1892 @property
1893 def name(self) -> str:
1894 # This will use a user-defined name if available, but otherwise defaults back to the auto-generated name
1895 return self.customName if self.customName is not None else self.default_name
1897 def __str__(self) -> str:
1898 return self.name
1900 def __repr__(self) -> str:
1901 return str(self)
1903 def streamline(self) -> "ParserElement":
1904 self.streamlined = True
1905 self._defaultName = None
1906 return self
1908 def recurse(self) -> List["ParserElement"]:
1909 return []
1911 def _checkRecursion(self, parseElementList):
1912 subRecCheckList = parseElementList[:] + [self]
1913 for e in self.recurse():
1914 e._checkRecursion(subRecCheckList)
1916 def validate(self, validateTrace=None) -> None:
1917 """
1918 Check defined expressions for valid structure, check for infinite recursive definitions.
1919 """
1920 warnings.warn(
1921 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
1922 DeprecationWarning,
1923 stacklevel=2,
1924 )
1925 self._checkRecursion([])
1927 def parse_file(
1928 self,
1929 file_or_filename: Union[str, Path, TextIO],
1930 encoding: str = "utf-8",
1931 parse_all: bool = False,
1932 *,
1933 parseAll: bool = False,
1934 ) -> ParseResults:
1935 """
1936 Execute the parse expression on the given file or filename.
1937 If a filename is specified (instead of a file object),
1938 the entire file is opened, read, and closed before parsing.
1939 """
1940 parseAll = parseAll or parse_all
1941 try:
1942 file_or_filename = typing.cast(TextIO, file_or_filename)
1943 file_contents = file_or_filename.read()
1944 except AttributeError:
1945 file_or_filename = typing.cast(str, file_or_filename)
1946 with open(file_or_filename, "r", encoding=encoding) as f:
1947 file_contents = f.read()
1948 try:
1949 return self.parse_string(file_contents, parseAll)
1950 except ParseBaseException as exc:
1951 if ParserElement.verbose_stacktrace:
1952 raise
1953 else:
1954 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1955 raise exc.with_traceback(None)
1957 def __eq__(self, other):
1958 if self is other:
1959 return True
1960 elif isinstance(other, str_type):
1961 return self.matches(other, parse_all=True)
1962 elif isinstance(other, ParserElement):
1963 return vars(self) == vars(other)
1964 return False
1966 def __hash__(self):
1967 return id(self)
1969 def matches(
1970 self, test_string: str, parse_all: bool = True, *, parseAll: bool = True
1971 ) -> bool:
1972 """
1973 Method for quick testing of a parser against a test string. Good for simple
1974 inline microtests of sub expressions while building up larger parser.
1976 Parameters:
1978 - ``test_string`` - to test against this expression for a match
1979 - ``parse_all`` - (default= ``True``) - flag to pass to :class:`parse_string` when running tests
1981 Example::
1983 expr = Word(nums)
1984 assert expr.matches("100")
1985 """
1986 parseAll = parseAll and parse_all
1987 try:
1988 self.parse_string(str(test_string), parse_all=parseAll)
1989 return True
1990 except ParseBaseException:
1991 return False
1993 def run_tests(
1994 self,
1995 tests: Union[str, List[str]],
1996 parse_all: bool = True,
1997 comment: typing.Optional[Union["ParserElement", str]] = "#",
1998 full_dump: bool = True,
1999 print_results: bool = True,
2000 failure_tests: bool = False,
2001 post_parse: typing.Optional[Callable[[str, ParseResults], str]] = None,
2002 file: typing.Optional[TextIO] = None,
2003 with_line_numbers: bool = False,
2004 *,
2005 parseAll: bool = True,
2006 fullDump: bool = True,
2007 printResults: bool = True,
2008 failureTests: bool = False,
2009 postParse: typing.Optional[Callable[[str, ParseResults], str]] = None,
2010 ) -> Tuple[bool, List[Tuple[str, Union[ParseResults, Exception]]]]:
2011 """
2012 Execute the parse expression on a series of test strings, showing each
2013 test, the parsed results or where the parse failed. Quick and easy way to
2014 run a parse expression against a list of sample strings.
2016 Parameters:
2018 - ``tests`` - a list of separate test strings, or a multiline string of test strings
2019 - ``parse_all`` - (default= ``True``) - flag to pass to :class:`parse_string` when running tests
2020 - ``comment`` - (default= ``'#'``) - expression for indicating embedded comments in the test
2021 string; pass None to disable comment filtering
2022 - ``full_dump`` - (default= ``True``) - dump results as list followed by results names in nested outline;
2023 if False, only dump nested list
2024 - ``print_results`` - (default= ``True``) prints test output to stdout
2025 - ``failure_tests`` - (default= ``False``) indicates if these tests are expected to fail parsing
2026 - ``post_parse`` - (default= ``None``) optional callback for successful parse results; called as
2027 `fn(test_string, parse_results)` and returns a string to be added to the test output
2028 - ``file`` - (default= ``None``) optional file-like object to which test output will be written;
2029 if None, will default to ``sys.stdout``
2030 - ``with_line_numbers`` - default= ``False``) show test strings with line and column numbers
2032 Returns: a (success, results) tuple, where success indicates that all tests succeeded
2033 (or failed if ``failure_tests`` is True), and the results contain a list of lines of each
2034 test's output
2036 Example::
2038 number_expr = pyparsing_common.number.copy()
2040 result = number_expr.run_tests('''
2041 # unsigned integer
2042 100
2043 # negative integer
2044 -100
2045 # float with scientific notation
2046 6.02e23
2047 # integer with scientific notation
2048 1e-12
2049 ''')
2050 print("Success" if result[0] else "Failed!")
2052 result = number_expr.run_tests('''
2053 # stray character
2054 100Z
2055 # missing leading digit before '.'
2056 -.100
2057 # too many '.'
2058 3.14.159
2059 ''', failure_tests=True)
2060 print("Success" if result[0] else "Failed!")
2062 prints::
2064 # unsigned integer
2065 100
2066 [100]
2068 # negative integer
2069 -100
2070 [-100]
2072 # float with scientific notation
2073 6.02e23
2074 [6.02e+23]
2076 # integer with scientific notation
2077 1e-12
2078 [1e-12]
2080 Success
2082 # stray character
2083 100Z
2084 ^
2085 FAIL: Expected end of text (at char 3), (line:1, col:4)
2087 # missing leading digit before '.'
2088 -.100
2089 ^
2090 FAIL: Expected {real number with scientific notation | real number | signed integer} (at char 0), (line:1, col:1)
2092 # too many '.'
2093 3.14.159
2094 ^
2095 FAIL: Expected end of text (at char 4), (line:1, col:5)
2097 Success
2099 Each test string must be on a single line. If you want to test a string that spans multiple
2100 lines, create a test like this::
2102 expr.run_tests(r"this is a test\\n of strings that spans \\n 3 lines")
2104 (Note that this is a raw string literal, you must include the leading ``'r'``.)
2105 """
2106 from .testing import pyparsing_test
2108 parseAll = parseAll and parse_all
2109 fullDump = fullDump and full_dump
2110 printResults = printResults and print_results
2111 failureTests = failureTests or failure_tests
2112 postParse = postParse or post_parse
2113 if isinstance(tests, str_type):
2114 tests = typing.cast(str, tests)
2115 line_strip = type(tests).strip
2116 tests = [line_strip(test_line) for test_line in tests.rstrip().splitlines()]
2117 comment_specified = comment is not None
2118 if comment_specified:
2119 if isinstance(comment, str_type):
2120 comment = typing.cast(str, comment)
2121 comment = Literal(comment)
2122 comment = typing.cast(ParserElement, comment)
2123 if file is None:
2124 file = sys.stdout
2125 print_ = file.write
2127 result: Union[ParseResults, Exception]
2128 allResults: List[Tuple[str, Union[ParseResults, Exception]]] = []
2129 comments: List[str] = []
2130 success = True
2131 NL = Literal(r"\n").add_parse_action(replace_with("\n")).ignore(quoted_string)
2132 BOM = "\ufeff"
2133 for t in tests:
2134 if comment_specified and comment.matches(t, False) or comments and not t:
2135 comments.append(
2136 pyparsing_test.with_line_numbers(t) if with_line_numbers else t
2137 )
2138 continue
2139 if not t:
2140 continue
2141 out = [
2142 "\n" + "\n".join(comments) if comments else "",
2143 pyparsing_test.with_line_numbers(t) if with_line_numbers else t,
2144 ]
2145 comments = []
2146 try:
2147 # convert newline marks to actual newlines, and strip leading BOM if present
2148 t = NL.transform_string(t.lstrip(BOM))
2149 result = self.parse_string(t, parse_all=parseAll)
2150 except ParseBaseException as pe:
2151 fatal = "(FATAL)" if isinstance(pe, ParseFatalException) else ""
2152 out.append(pe.explain())
2153 out.append("FAIL: " + str(pe))
2154 if ParserElement.verbose_stacktrace:
2155 out.extend(traceback.format_tb(pe.__traceback__))
2156 success = success and failureTests
2157 result = pe
2158 except Exception as exc:
2159 out.append(f"FAIL-EXCEPTION: {type(exc).__name__}: {exc}")
2160 if ParserElement.verbose_stacktrace:
2161 out.extend(traceback.format_tb(exc.__traceback__))
2162 success = success and failureTests
2163 result = exc
2164 else:
2165 success = success and not failureTests
2166 if postParse is not None:
2167 try:
2168 pp_value = postParse(t, result)
2169 if pp_value is not None:
2170 if isinstance(pp_value, ParseResults):
2171 out.append(pp_value.dump())
2172 else:
2173 out.append(str(pp_value))
2174 else:
2175 out.append(result.dump())
2176 except Exception as e:
2177 out.append(result.dump(full=fullDump))
2178 out.append(
2179 f"{postParse.__name__} failed: {type(e).__name__}: {e}"
2180 )
2181 else:
2182 out.append(result.dump(full=fullDump))
2183 out.append("")
2185 if printResults:
2186 print_("\n".join(out))
2188 allResults.append((t, result))
2190 return success, allResults
2192 def create_diagram(
2193 self,
2194 output_html: Union[TextIO, Path, str],
2195 vertical: int = 3,
2196 show_results_names: bool = False,
2197 show_groups: bool = False,
2198 embed: bool = False,
2199 **kwargs,
2200 ) -> None:
2201 """
2202 Create a railroad diagram for the parser.
2204 Parameters:
2206 - ``output_html`` (str or file-like object) - output target for generated
2207 diagram HTML
2208 - ``vertical`` (int) - threshold for formatting multiple alternatives vertically
2209 instead of horizontally (default=3)
2210 - ``show_results_names`` - bool flag whether diagram should show annotations for
2211 defined results names
2212 - ``show_groups`` - bool flag whether groups should be highlighted with an unlabeled surrounding box
2213 - ``embed`` - bool flag whether generated HTML should omit <HEAD>, <BODY>, and <DOCTYPE> tags to embed
2214 the resulting HTML in an enclosing HTML source
2215 - ``head`` - str containing additional HTML to insert into the <HEAD> section of the generated code;
2216 can be used to insert custom CSS styling
2217 - ``body`` - str containing additional HTML to insert at the beginning of the <BODY> section of the
2218 generated code
2220 Additional diagram-formatting keyword arguments can also be included;
2221 see railroad.Diagram class.
2222 """
2224 try:
2225 from .diagram import to_railroad, railroad_to_html
2226 except ImportError as ie:
2227 raise Exception(
2228 "must ``pip install pyparsing[diagrams]`` to generate parser railroad diagrams"
2229 ) from ie
2231 self.streamline()
2233 railroad = to_railroad(
2234 self,
2235 vertical=vertical,
2236 show_results_names=show_results_names,
2237 show_groups=show_groups,
2238 diagram_kwargs=kwargs,
2239 )
2240 if isinstance(output_html, (str, Path)):
2241 with open(output_html, "w", encoding="utf-8") as diag_file:
2242 diag_file.write(railroad_to_html(railroad, embed=embed, **kwargs))
2243 else:
2244 # we were passed a file-like object, just write to it
2245 output_html.write(railroad_to_html(railroad, embed=embed, **kwargs))
2247 # Compatibility synonyms
2248 # fmt: off
2249 @staticmethod
2250 @replaced_by_pep8(inline_literals_using)
2251 def inlineLiteralsUsing(): ...
2253 @staticmethod
2254 @replaced_by_pep8(set_default_whitespace_chars)
2255 def setDefaultWhitespaceChars(): ...
2257 @replaced_by_pep8(set_results_name)
2258 def setResultsName(self): ...
2260 @replaced_by_pep8(set_break)
2261 def setBreak(self): ...
2263 @replaced_by_pep8(set_parse_action)
2264 def setParseAction(self): ...
2266 @replaced_by_pep8(add_parse_action)
2267 def addParseAction(self): ...
2269 @replaced_by_pep8(add_condition)
2270 def addCondition(self): ...
2272 @replaced_by_pep8(set_fail_action)
2273 def setFailAction(self): ...
2275 @replaced_by_pep8(try_parse)
2276 def tryParse(self): ...
2278 @staticmethod
2279 @replaced_by_pep8(enable_left_recursion)
2280 def enableLeftRecursion(): ...
2282 @staticmethod
2283 @replaced_by_pep8(enable_packrat)
2284 def enablePackrat(): ...
2286 @replaced_by_pep8(parse_string)
2287 def parseString(self): ...
2289 @replaced_by_pep8(scan_string)
2290 def scanString(self): ...
2292 @replaced_by_pep8(transform_string)
2293 def transformString(self): ...
2295 @replaced_by_pep8(search_string)
2296 def searchString(self): ...
2298 @replaced_by_pep8(ignore_whitespace)
2299 def ignoreWhitespace(self): ...
2301 @replaced_by_pep8(leave_whitespace)
2302 def leaveWhitespace(self): ...
2304 @replaced_by_pep8(set_whitespace_chars)
2305 def setWhitespaceChars(self): ...
2307 @replaced_by_pep8(parse_with_tabs)
2308 def parseWithTabs(self): ...
2310 @replaced_by_pep8(set_debug_actions)
2311 def setDebugActions(self): ...
2313 @replaced_by_pep8(set_debug)
2314 def setDebug(self): ...
2316 @replaced_by_pep8(set_name)
2317 def setName(self): ...
2319 @replaced_by_pep8(parse_file)
2320 def parseFile(self): ...
2322 @replaced_by_pep8(run_tests)
2323 def runTests(self): ...
2325 canParseNext = can_parse_next
2326 resetCache = reset_cache
2327 defaultName = default_name
2328 # fmt: on
2331class _PendingSkip(ParserElement):
2332 # internal placeholder class to hold a place were '...' is added to a parser element,
2333 # once another ParserElement is added, this placeholder will be replaced with a SkipTo
2334 def __init__(self, expr: ParserElement, must_skip: bool = False):
2335 super().__init__()
2336 self.anchor = expr
2337 self.must_skip = must_skip
2339 def _generateDefaultName(self) -> str:
2340 return str(self.anchor + Empty()).replace("Empty", "...")
2342 def __add__(self, other) -> "ParserElement":
2343 skipper = SkipTo(other).set_name("...")("_skipped*")
2344 if self.must_skip:
2346 def must_skip(t):
2347 if not t._skipped or t._skipped.as_list() == [""]:
2348 del t[0]
2349 t.pop("_skipped", None)
2351 def show_skip(t):
2352 if t._skipped.as_list()[-1:] == [""]:
2353 t.pop("_skipped")
2354 t["_skipped"] = "missing <" + repr(self.anchor) + ">"
2356 return (
2357 self.anchor + skipper().add_parse_action(must_skip)
2358 | skipper().add_parse_action(show_skip)
2359 ) + other
2361 return self.anchor + skipper + other
2363 def __repr__(self):
2364 return self.defaultName
2366 def parseImpl(self, *args):
2367 raise Exception(
2368 "use of `...` expression without following SkipTo target expression"
2369 )
2372class Token(ParserElement):
2373 """Abstract :class:`ParserElement` subclass, for defining atomic
2374 matching patterns.
2375 """
2377 def __init__(self):
2378 super().__init__(savelist=False)
2380 def _generateDefaultName(self) -> str:
2381 return type(self).__name__
2384class NoMatch(Token):
2385 """
2386 A token that will never match.
2387 """
2389 def __init__(self):
2390 super().__init__()
2391 self.mayReturnEmpty = True
2392 self.mayIndexError = False
2393 self.errmsg = "Unmatchable token"
2395 def parseImpl(self, instring, loc, doActions=True):
2396 raise ParseException(instring, loc, self.errmsg, self)
2399class Literal(Token):
2400 """
2401 Token to exactly match a specified string.
2403 Example::
2405 Literal('blah').parse_string('blah') # -> ['blah']
2406 Literal('blah').parse_string('blahfooblah') # -> ['blah']
2407 Literal('blah').parse_string('bla') # -> Exception: Expected "blah"
2409 For case-insensitive matching, use :class:`CaselessLiteral`.
2411 For keyword matching (force word break before and after the matched string),
2412 use :class:`Keyword` or :class:`CaselessKeyword`.
2413 """
2415 def __new__(cls, match_string: str = "", *, matchString: str = ""):
2416 # Performance tuning: select a subclass with optimized parseImpl
2417 if cls is Literal:
2418 match_string = matchString or match_string
2419 if not match_string:
2420 return super().__new__(Empty)
2421 if len(match_string) == 1:
2422 return super().__new__(_SingleCharLiteral)
2424 # Default behavior
2425 return super().__new__(cls)
2427 # Needed to make copy.copy() work correctly if we customize __new__
2428 def __getnewargs__(self):
2429 return (self.match,)
2431 def __init__(self, match_string: str = "", *, matchString: str = ""):
2432 super().__init__()
2433 match_string = matchString or match_string
2434 self.match = match_string
2435 self.matchLen = len(match_string)
2436 self.firstMatchChar = match_string[:1]
2437 self.errmsg = "Expected " + self.name
2438 self.mayReturnEmpty = False
2439 self.mayIndexError = False
2441 def _generateDefaultName(self) -> str:
2442 return repr(self.match)
2444 def parseImpl(self, instring, loc, doActions=True):
2445 if instring[loc] == self.firstMatchChar and instring.startswith(
2446 self.match, loc
2447 ):
2448 return loc + self.matchLen, self.match
2449 raise ParseException(instring, loc, self.errmsg, self)
2452class Empty(Literal):
2453 """
2454 An empty token, will always match.
2455 """
2457 def __init__(self, match_string="", *, matchString=""):
2458 super().__init__("")
2459 self.mayReturnEmpty = True
2460 self.mayIndexError = False
2462 def _generateDefaultName(self) -> str:
2463 return "Empty"
2465 def parseImpl(self, instring, loc, doActions=True):
2466 return loc, []
2469class _SingleCharLiteral(Literal):
2470 def parseImpl(self, instring, loc, doActions=True):
2471 if instring[loc] == self.firstMatchChar:
2472 return loc + 1, self.match
2473 raise ParseException(instring, loc, self.errmsg, self)
2476ParserElement._literalStringClass = Literal
2479class Keyword(Token):
2480 """
2481 Token to exactly match a specified string as a keyword, that is,
2482 it must be immediately preceded and followed by whitespace or
2483 non-keyword characters. Compare with :class:`Literal`:
2485 - ``Literal("if")`` will match the leading ``'if'`` in
2486 ``'ifAndOnlyIf'``.
2487 - ``Keyword("if")`` will not; it will only match the leading
2488 ``'if'`` in ``'if x=1'``, or ``'if(y==2)'``
2490 Accepts two optional constructor arguments in addition to the
2491 keyword string:
2493 - ``ident_chars`` is a string of characters that would be valid
2494 identifier characters, defaulting to all alphanumerics + "_" and
2495 "$"
2496 - ``caseless`` allows case-insensitive matching, default is ``False``.
2498 Example::
2500 Keyword("start").parse_string("start") # -> ['start']
2501 Keyword("start").parse_string("starting") # -> Exception
2503 For case-insensitive matching, use :class:`CaselessKeyword`.
2504 """
2506 DEFAULT_KEYWORD_CHARS = alphanums + "_$"
2508 def __init__(
2509 self,
2510 match_string: str = "",
2511 ident_chars: typing.Optional[str] = None,
2512 caseless: bool = False,
2513 *,
2514 matchString: str = "",
2515 identChars: typing.Optional[str] = None,
2516 ):
2517 super().__init__()
2518 identChars = identChars or ident_chars
2519 if identChars is None:
2520 identChars = Keyword.DEFAULT_KEYWORD_CHARS
2521 match_string = matchString or match_string
2522 self.match = match_string
2523 self.matchLen = len(match_string)
2524 try:
2525 self.firstMatchChar = match_string[0]
2526 except IndexError:
2527 raise ValueError("null string passed to Keyword; use Empty() instead")
2528 self.errmsg = f"Expected {type(self).__name__} {self.name}"
2529 self.mayReturnEmpty = False
2530 self.mayIndexError = False
2531 self.caseless = caseless
2532 if caseless:
2533 self.caselessmatch = match_string.upper()
2534 identChars = identChars.upper()
2535 self.identChars = set(identChars)
2537 def _generateDefaultName(self) -> str:
2538 return repr(self.match)
2540 def parseImpl(self, instring, loc, doActions=True):
2541 errmsg = self.errmsg
2542 errloc = loc
2543 if self.caseless:
2544 if instring[loc : loc + self.matchLen].upper() == self.caselessmatch:
2545 if loc == 0 or instring[loc - 1].upper() not in self.identChars:
2546 if (
2547 loc >= len(instring) - self.matchLen
2548 or instring[loc + self.matchLen].upper() not in self.identChars
2549 ):
2550 return loc + self.matchLen, self.match
2551 else:
2552 # followed by keyword char
2553 errmsg += ", was immediately followed by keyword character"
2554 errloc = loc + self.matchLen
2555 else:
2556 # preceded by keyword char
2557 errmsg += ", keyword was immediately preceded by keyword character"
2558 errloc = loc - 1
2559 # else no match just raise plain exception
2561 else:
2562 if (
2563 instring[loc] == self.firstMatchChar
2564 and self.matchLen == 1
2565 or instring.startswith(self.match, loc)
2566 ):
2567 if loc == 0 or instring[loc - 1] not in self.identChars:
2568 if (
2569 loc >= len(instring) - self.matchLen
2570 or instring[loc + self.matchLen] not in self.identChars
2571 ):
2572 return loc + self.matchLen, self.match
2573 else:
2574 # followed by keyword char
2575 errmsg += (
2576 ", keyword was immediately followed by keyword character"
2577 )
2578 errloc = loc + self.matchLen
2579 else:
2580 # preceded by keyword char
2581 errmsg += ", keyword was immediately preceded by keyword character"
2582 errloc = loc - 1
2583 # else no match just raise plain exception
2585 raise ParseException(instring, errloc, errmsg, self)
2587 @staticmethod
2588 def set_default_keyword_chars(chars) -> None:
2589 """
2590 Overrides the default characters used by :class:`Keyword` expressions.
2591 """
2592 Keyword.DEFAULT_KEYWORD_CHARS = chars
2594 setDefaultKeywordChars = set_default_keyword_chars
2597class CaselessLiteral(Literal):
2598 """
2599 Token to match a specified string, ignoring case of letters.
2600 Note: the matched results will always be in the case of the given
2601 match string, NOT the case of the input text.
2603 Example::
2605 CaselessLiteral("CMD")[1, ...].parse_string("cmd CMD Cmd10")
2606 # -> ['CMD', 'CMD', 'CMD']
2608 (Contrast with example for :class:`CaselessKeyword`.)
2609 """
2611 def __init__(self, match_string: str = "", *, matchString: str = ""):
2612 match_string = matchString or match_string
2613 super().__init__(match_string.upper())
2614 # Preserve the defining literal.
2615 self.returnString = match_string
2616 self.errmsg = "Expected " + self.name
2618 def parseImpl(self, instring, loc, doActions=True):
2619 if instring[loc : loc + self.matchLen].upper() == self.match:
2620 return loc + self.matchLen, self.returnString
2621 raise ParseException(instring, loc, self.errmsg, self)
2624class CaselessKeyword(Keyword):
2625 """
2626 Caseless version of :class:`Keyword`.
2628 Example::
2630 CaselessKeyword("CMD")[1, ...].parse_string("cmd CMD Cmd10")
2631 # -> ['CMD', 'CMD']
2633 (Contrast with example for :class:`CaselessLiteral`.)
2634 """
2636 def __init__(
2637 self,
2638 match_string: str = "",
2639 ident_chars: typing.Optional[str] = None,
2640 *,
2641 matchString: str = "",
2642 identChars: typing.Optional[str] = None,
2643 ):
2644 identChars = identChars or ident_chars
2645 match_string = matchString or match_string
2646 super().__init__(match_string, identChars, caseless=True)
2649class CloseMatch(Token):
2650 """A variation on :class:`Literal` which matches "close" matches,
2651 that is, strings with at most 'n' mismatching characters.
2652 :class:`CloseMatch` takes parameters:
2654 - ``match_string`` - string to be matched
2655 - ``caseless`` - a boolean indicating whether to ignore casing when comparing characters
2656 - ``max_mismatches`` - (``default=1``) maximum number of
2657 mismatches allowed to count as a match
2659 The results from a successful parse will contain the matched text
2660 from the input string and the following named results:
2662 - ``mismatches`` - a list of the positions within the
2663 match_string where mismatches were found
2664 - ``original`` - the original match_string used to compare
2665 against the input string
2667 If ``mismatches`` is an empty list, then the match was an exact
2668 match.
2670 Example::
2672 patt = CloseMatch("ATCATCGAATGGA")
2673 patt.parse_string("ATCATCGAAXGGA") # -> (['ATCATCGAAXGGA'], {'mismatches': [[9]], 'original': ['ATCATCGAATGGA']})
2674 patt.parse_string("ATCAXCGAAXGGA") # -> Exception: Expected 'ATCATCGAATGGA' (with up to 1 mismatches) (at char 0), (line:1, col:1)
2676 # exact match
2677 patt.parse_string("ATCATCGAATGGA") # -> (['ATCATCGAATGGA'], {'mismatches': [[]], 'original': ['ATCATCGAATGGA']})
2679 # close match allowing up to 2 mismatches
2680 patt = CloseMatch("ATCATCGAATGGA", max_mismatches=2)
2681 patt.parse_string("ATCAXCGAAXGGA") # -> (['ATCAXCGAAXGGA'], {'mismatches': [[4, 9]], 'original': ['ATCATCGAATGGA']})
2682 """
2684 def __init__(
2685 self,
2686 match_string: str,
2687 max_mismatches: typing.Optional[int] = None,
2688 *,
2689 maxMismatches: int = 1,
2690 caseless=False,
2691 ):
2692 maxMismatches = max_mismatches if max_mismatches is not None else maxMismatches
2693 super().__init__()
2694 self.match_string = match_string
2695 self.maxMismatches = maxMismatches
2696 self.errmsg = f"Expected {self.match_string!r} (with up to {self.maxMismatches} mismatches)"
2697 self.caseless = caseless
2698 self.mayIndexError = False
2699 self.mayReturnEmpty = False
2701 def _generateDefaultName(self) -> str:
2702 return f"{type(self).__name__}:{self.match_string!r}"
2704 def parseImpl(self, instring, loc, doActions=True):
2705 start = loc
2706 instrlen = len(instring)
2707 maxloc = start + len(self.match_string)
2709 if maxloc <= instrlen:
2710 match_string = self.match_string
2711 match_stringloc = 0
2712 mismatches = []
2713 maxMismatches = self.maxMismatches
2715 for match_stringloc, s_m in enumerate(
2716 zip(instring[loc:maxloc], match_string)
2717 ):
2718 src, mat = s_m
2719 if self.caseless:
2720 src, mat = src.lower(), mat.lower()
2722 if src != mat:
2723 mismatches.append(match_stringloc)
2724 if len(mismatches) > maxMismatches:
2725 break
2726 else:
2727 loc = start + match_stringloc + 1
2728 results = ParseResults([instring[start:loc]])
2729 results["original"] = match_string
2730 results["mismatches"] = mismatches
2731 return loc, results
2733 raise ParseException(instring, loc, self.errmsg, self)
2736class Word(Token):
2737 """Token for matching words composed of allowed character sets.
2739 Parameters:
2741 - ``init_chars`` - string of all characters that should be used to
2742 match as a word; "ABC" will match "AAA", "ABAB", "CBAC", etc.;
2743 if ``body_chars`` is also specified, then this is the string of
2744 initial characters
2745 - ``body_chars`` - string of characters that
2746 can be used for matching after a matched initial character as
2747 given in ``init_chars``; if omitted, same as the initial characters
2748 (default=``None``)
2749 - ``min`` - minimum number of characters to match (default=1)
2750 - ``max`` - maximum number of characters to match (default=0)
2751 - ``exact`` - exact number of characters to match (default=0)
2752 - ``as_keyword`` - match as a keyword (default=``False``)
2753 - ``exclude_chars`` - characters that might be
2754 found in the input ``body_chars`` string but which should not be
2755 accepted for matching ;useful to define a word of all
2756 printables except for one or two characters, for instance
2757 (default=``None``)
2759 :class:`srange` is useful for defining custom character set strings
2760 for defining :class:`Word` expressions, using range notation from
2761 regular expression character sets.
2763 A common mistake is to use :class:`Word` to match a specific literal
2764 string, as in ``Word("Address")``. Remember that :class:`Word`
2765 uses the string argument to define *sets* of matchable characters.
2766 This expression would match "Add", "AAA", "dAred", or any other word
2767 made up of the characters 'A', 'd', 'r', 'e', and 's'. To match an
2768 exact literal string, use :class:`Literal` or :class:`Keyword`.
2770 pyparsing includes helper strings for building Words:
2772 - :class:`alphas`
2773 - :class:`nums`
2774 - :class:`alphanums`
2775 - :class:`hexnums`
2776 - :class:`alphas8bit` (alphabetic characters in ASCII range 128-255
2777 - accented, tilded, umlauted, etc.)
2778 - :class:`punc8bit` (non-alphabetic characters in ASCII range
2779 128-255 - currency, symbols, superscripts, diacriticals, etc.)
2780 - :class:`printables` (any non-whitespace character)
2782 ``alphas``, ``nums``, and ``printables`` are also defined in several
2783 Unicode sets - see :class:`pyparsing_unicode``.
2785 Example::
2787 # a word composed of digits
2788 integer = Word(nums) # equivalent to Word("0123456789") or Word(srange("0-9"))
2790 # a word with a leading capital, and zero or more lowercase
2791 capital_word = Word(alphas.upper(), alphas.lower())
2793 # hostnames are alphanumeric, with leading alpha, and '-'
2794 hostname = Word(alphas, alphanums + '-')
2796 # roman numeral (not a strict parser, accepts invalid mix of characters)
2797 roman = Word("IVXLCDM")
2799 # any string of non-whitespace characters, except for ','
2800 csv_value = Word(printables, exclude_chars=",")
2801 """
2803 def __init__(
2804 self,
2805 init_chars: str = "",
2806 body_chars: typing.Optional[str] = None,
2807 min: int = 1,
2808 max: int = 0,
2809 exact: int = 0,
2810 as_keyword: bool = False,
2811 exclude_chars: typing.Optional[str] = None,
2812 *,
2813 initChars: typing.Optional[str] = None,
2814 bodyChars: typing.Optional[str] = None,
2815 asKeyword: bool = False,
2816 excludeChars: typing.Optional[str] = None,
2817 ):
2818 initChars = initChars or init_chars
2819 bodyChars = bodyChars or body_chars
2820 asKeyword = asKeyword or as_keyword
2821 excludeChars = excludeChars or exclude_chars
2822 super().__init__()
2823 if not initChars:
2824 raise ValueError(
2825 f"invalid {type(self).__name__}, initChars cannot be empty string"
2826 )
2828 initChars_set = set(initChars)
2829 if excludeChars:
2830 excludeChars_set = set(excludeChars)
2831 initChars_set -= excludeChars_set
2832 if bodyChars:
2833 bodyChars = "".join(set(bodyChars) - excludeChars_set)
2834 self.initChars = initChars_set
2835 self.initCharsOrig = "".join(sorted(initChars_set))
2837 if bodyChars:
2838 self.bodyChars = set(bodyChars)
2839 self.bodyCharsOrig = "".join(sorted(bodyChars))
2840 else:
2841 self.bodyChars = initChars_set
2842 self.bodyCharsOrig = self.initCharsOrig
2844 self.maxSpecified = max > 0
2846 if min < 1:
2847 raise ValueError(
2848 "cannot specify a minimum length < 1; use Opt(Word()) if zero-length word is permitted"
2849 )
2851 if self.maxSpecified and min > max:
2852 raise ValueError(
2853 f"invalid args, if min and max both specified min must be <= max (min={min}, max={max})"
2854 )
2856 self.minLen = min
2858 if max > 0:
2859 self.maxLen = max
2860 else:
2861 self.maxLen = _MAX_INT
2863 if exact > 0:
2864 min = max = exact
2865 self.maxLen = exact
2866 self.minLen = exact
2868 self.errmsg = "Expected " + self.name
2869 self.mayIndexError = False
2870 self.asKeyword = asKeyword
2871 if self.asKeyword:
2872 self.errmsg += " as a keyword"
2874 # see if we can make a regex for this Word
2875 if " " not in (self.initChars | self.bodyChars):
2876 if len(self.initChars) == 1:
2877 re_leading_fragment = re.escape(self.initCharsOrig)
2878 else:
2879 re_leading_fragment = f"[{_collapse_string_to_ranges(self.initChars)}]"
2881 if self.bodyChars == self.initChars:
2882 if max == 0:
2883 repeat = "+"
2884 elif max == 1:
2885 repeat = ""
2886 else:
2887 if self.minLen != self.maxLen:
2888 repeat = f"{{{self.minLen},{'' if self.maxLen == _MAX_INT else self.maxLen}}}"
2889 else:
2890 repeat = f"{{{self.minLen}}}"
2891 self.reString = f"{re_leading_fragment}{repeat}"
2892 else:
2893 if max == 1:
2894 re_body_fragment = ""
2895 repeat = ""
2896 else:
2897 re_body_fragment = f"[{_collapse_string_to_ranges(self.bodyChars)}]"
2898 if max == 0:
2899 repeat = "*"
2900 elif max == 2:
2901 repeat = "?" if min <= 1 else ""
2902 else:
2903 if min != max:
2904 repeat = f"{{{min - 1 if min > 0 else 0},{max - 1}}}"
2905 else:
2906 repeat = f"{{{min - 1 if min > 0 else 0}}}"
2908 self.reString = (
2909 f"{re_leading_fragment}" f"{re_body_fragment}" f"{repeat}"
2910 )
2912 if self.asKeyword:
2913 self.reString = rf"\b{self.reString}\b"
2915 try:
2916 self.re = re.compile(self.reString)
2917 except re.error:
2918 self.re = None # type: ignore[assignment]
2919 else:
2920 self.re_match = self.re.match
2921 self.parseImpl = self.parseImpl_regex # type: ignore[assignment]
2923 def _generateDefaultName(self) -> str:
2924 def charsAsStr(s):
2925 max_repr_len = 16
2926 s = _collapse_string_to_ranges(s, re_escape=False)
2927 if len(s) > max_repr_len:
2928 return s[: max_repr_len - 3] + "..."
2929 else:
2930 return s
2932 if self.initChars != self.bodyChars:
2933 base = f"W:({charsAsStr(self.initChars)}, {charsAsStr(self.bodyChars)})"
2934 else:
2935 base = f"W:({charsAsStr(self.initChars)})"
2937 # add length specification
2938 if self.minLen > 1 or self.maxLen != _MAX_INT:
2939 if self.minLen == self.maxLen:
2940 if self.minLen == 1:
2941 return base[2:]
2942 else:
2943 return base + f"{{{self.minLen}}}"
2944 elif self.maxLen == _MAX_INT:
2945 return base + f"{{{self.minLen},...}}"
2946 else:
2947 return base + f"{{{self.minLen},{self.maxLen}}}"
2948 return base
2950 def parseImpl(self, instring, loc, doActions=True):
2951 if instring[loc] not in self.initChars:
2952 raise ParseException(instring, loc, self.errmsg, self)
2954 start = loc
2955 loc += 1
2956 instrlen = len(instring)
2957 bodychars = self.bodyChars
2958 maxloc = start + self.maxLen
2959 maxloc = min(maxloc, instrlen)
2960 while loc < maxloc and instring[loc] in bodychars:
2961 loc += 1
2963 throwException = False
2964 if loc - start < self.minLen:
2965 throwException = True
2966 elif self.maxSpecified and loc < instrlen and instring[loc] in bodychars:
2967 throwException = True
2968 elif self.asKeyword:
2969 if (
2970 start > 0
2971 and instring[start - 1] in bodychars
2972 or loc < instrlen
2973 and instring[loc] in bodychars
2974 ):
2975 throwException = True
2977 if throwException:
2978 raise ParseException(instring, loc, self.errmsg, self)
2980 return loc, instring[start:loc]
2982 def parseImpl_regex(self, instring, loc, doActions=True):
2983 result = self.re_match(instring, loc)
2984 if not result:
2985 raise ParseException(instring, loc, self.errmsg, self)
2987 loc = result.end()
2988 return loc, result.group()
2991class Char(Word):
2992 """A short-cut class for defining :class:`Word` ``(characters, exact=1)``,
2993 when defining a match of any single character in a string of
2994 characters.
2995 """
2997 def __init__(
2998 self,
2999 charset: str,
3000 as_keyword: bool = False,
3001 exclude_chars: typing.Optional[str] = None,
3002 *,
3003 asKeyword: bool = False,
3004 excludeChars: typing.Optional[str] = None,
3005 ):
3006 asKeyword = asKeyword or as_keyword
3007 excludeChars = excludeChars or exclude_chars
3008 super().__init__(
3009 charset, exact=1, as_keyword=asKeyword, exclude_chars=excludeChars
3010 )
3013class Regex(Token):
3014 r"""Token for matching strings that match a given regular
3015 expression. Defined with string specifying the regular expression in
3016 a form recognized by the stdlib Python `re module <https://docs.python.org/3/library/re.html>`_.
3017 If the given regex contains named groups (defined using ``(?P<name>...)``),
3018 these will be preserved as named :class:`ParseResults`.
3020 If instead of the Python stdlib ``re`` module you wish to use a different RE module
3021 (such as the ``regex`` module), you can do so by building your ``Regex`` object with
3022 a compiled RE that was compiled using ``regex``.
3024 Example::
3026 realnum = Regex(r"[+-]?\d+\.\d*")
3027 # ref: https://stackoverflow.com/questions/267399/how-do-you-match-only-valid-roman-numerals-with-a-regular-expression
3028 roman = Regex(r"M{0,4}(CM|CD|D?{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})")
3030 # named fields in a regex will be returned as named results
3031 date = Regex(r'(?P<year>\d{4})-(?P<month>\d\d?)-(?P<day>\d\d?)')
3033 # the Regex class will accept re's compiled using the regex module
3034 import regex
3035 parser = pp.Regex(regex.compile(r'[0-9]'))
3036 """
3038 def __init__(
3039 self,
3040 pattern: Any,
3041 flags: Union[re.RegexFlag, int] = 0,
3042 as_group_list: bool = False,
3043 as_match: bool = False,
3044 *,
3045 asGroupList: bool = False,
3046 asMatch: bool = False,
3047 ):
3048 """The parameters ``pattern`` and ``flags`` are passed
3049 to the ``re.compile()`` function as-is. See the Python
3050 `re module <https://docs.python.org/3/library/re.html>`_ module for an
3051 explanation of the acceptable patterns and flags.
3052 """
3053 super().__init__()
3054 asGroupList = asGroupList or as_group_list
3055 asMatch = asMatch or as_match
3057 if isinstance(pattern, str_type):
3058 if not pattern:
3059 raise ValueError("null string passed to Regex; use Empty() instead")
3061 self._re = None
3062 self.reString = self.pattern = pattern
3063 self.flags = flags
3065 elif hasattr(pattern, "pattern") and hasattr(pattern, "match"):
3066 self._re = pattern
3067 self.pattern = self.reString = pattern.pattern
3068 self.flags = flags
3070 else:
3071 raise TypeError(
3072 "Regex may only be constructed with a string or a compiled RE object"
3073 )
3075 self.errmsg = "Expected " + self.name
3076 self.mayIndexError = False
3077 self.asGroupList = asGroupList
3078 self.asMatch = asMatch
3079 if self.asGroupList:
3080 self.parseImpl = self.parseImplAsGroupList # type: ignore [assignment]
3081 if self.asMatch:
3082 self.parseImpl = self.parseImplAsMatch # type: ignore [assignment]
3084 @cached_property
3085 def re(self):
3086 if self._re:
3087 return self._re
3088 else:
3089 try:
3090 return re.compile(self.pattern, self.flags)
3091 except re.error:
3092 raise ValueError(f"invalid pattern ({self.pattern!r}) passed to Regex")
3094 @cached_property
3095 def re_match(self):
3096 return self.re.match
3098 @cached_property
3099 def mayReturnEmpty(self):
3100 return self.re_match("") is not None
3102 def _generateDefaultName(self) -> str:
3103 return "Re:({})".format(repr(self.pattern).replace("\\\\", "\\"))
3105 def parseImpl(self, instring, loc, doActions=True):
3106 result = self.re_match(instring, loc)
3107 if not result:
3108 raise ParseException(instring, loc, self.errmsg, self)
3110 loc = result.end()
3111 ret = ParseResults(result.group())
3112 d = result.groupdict()
3113 if d:
3114 for k, v in d.items():
3115 ret[k] = v
3116 return loc, ret
3118 def parseImplAsGroupList(self, instring, loc, doActions=True):
3119 result = self.re_match(instring, loc)
3120 if not result:
3121 raise ParseException(instring, loc, self.errmsg, self)
3123 loc = result.end()
3124 ret = result.groups()
3125 return loc, ret
3127 def parseImplAsMatch(self, instring, loc, doActions=True):
3128 result = self.re_match(instring, loc)
3129 if not result:
3130 raise ParseException(instring, loc, self.errmsg, self)
3132 loc = result.end()
3133 ret = result
3134 return loc, ret
3136 def sub(self, repl: str) -> ParserElement:
3137 r"""
3138 Return :class:`Regex` with an attached parse action to transform the parsed
3139 result as if called using `re.sub(expr, repl, string) <https://docs.python.org/3/library/re.html#re.sub>`_.
3141 Example::
3143 make_html = Regex(r"(\w+):(.*?):").sub(r"<\1>\2</\1>")
3144 print(make_html.transform_string("h1:main title:"))
3145 # prints "<h1>main title</h1>"
3146 """
3147 if self.asGroupList:
3148 raise TypeError("cannot use sub() with Regex(as_group_list=True)")
3150 if self.asMatch and callable(repl):
3151 raise TypeError(
3152 "cannot use sub() with a callable with Regex(as_match=True)"
3153 )
3155 if self.asMatch:
3157 def pa(tokens):
3158 return tokens[0].expand(repl)
3160 else:
3162 def pa(tokens):
3163 return self.re.sub(repl, tokens[0])
3165 return self.add_parse_action(pa)
3168class QuotedString(Token):
3169 r"""
3170 Token for matching strings that are delimited by quoting characters.
3172 Defined with the following parameters:
3174 - ``quote_char`` - string of one or more characters defining the
3175 quote delimiting string
3176 - ``esc_char`` - character to re_escape quotes, typically backslash
3177 (default= ``None``)
3178 - ``esc_quote`` - special quote sequence to re_escape an embedded quote
3179 string (such as SQL's ``""`` to re_escape an embedded ``"``)
3180 (default= ``None``)
3181 - ``multiline`` - boolean indicating whether quotes can span
3182 multiple lines (default= ``False``)
3183 - ``unquote_results`` - boolean indicating whether the matched text
3184 should be unquoted (default= ``True``)
3185 - ``end_quote_char`` - string of one or more characters defining the
3186 end of the quote delimited string (default= ``None`` => same as
3187 quote_char)
3188 - ``convert_whitespace_escapes`` - convert escaped whitespace
3189 (``'\t'``, ``'\n'``, etc.) to actual whitespace
3190 (default= ``True``)
3192 Example::
3194 qs = QuotedString('"')
3195 print(qs.search_string('lsjdf "This is the quote" sldjf'))
3196 complex_qs = QuotedString('{{', end_quote_char='}}')
3197 print(complex_qs.search_string('lsjdf {{This is the "quote"}} sldjf'))
3198 sql_qs = QuotedString('"', esc_quote='""')
3199 print(sql_qs.search_string('lsjdf "This is the quote with ""embedded"" quotes" sldjf'))
3201 prints::
3203 [['This is the quote']]
3204 [['This is the "quote"']]
3205 [['This is the quote with "embedded" quotes']]
3206 """
3207 ws_map = dict(((r"\t", "\t"), (r"\n", "\n"), (r"\f", "\f"), (r"\r", "\r")))
3209 def __init__(
3210 self,
3211 quote_char: str = "",
3212 esc_char: typing.Optional[str] = None,
3213 esc_quote: typing.Optional[str] = None,
3214 multiline: bool = False,
3215 unquote_results: bool = True,
3216 end_quote_char: typing.Optional[str] = None,
3217 convert_whitespace_escapes: bool = True,
3218 *,
3219 quoteChar: str = "",
3220 escChar: typing.Optional[str] = None,
3221 escQuote: typing.Optional[str] = None,
3222 unquoteResults: bool = True,
3223 endQuoteChar: typing.Optional[str] = None,
3224 convertWhitespaceEscapes: bool = True,
3225 ):
3226 super().__init__()
3227 escChar = escChar or esc_char
3228 escQuote = escQuote or esc_quote
3229 unquoteResults = unquoteResults and unquote_results
3230 endQuoteChar = endQuoteChar or end_quote_char
3231 convertWhitespaceEscapes = (
3232 convertWhitespaceEscapes and convert_whitespace_escapes
3233 )
3234 quote_char = quoteChar or quote_char
3236 # remove white space from quote chars - wont work anyway
3237 quote_char = quote_char.strip()
3238 if not quote_char:
3239 raise ValueError("quote_char cannot be the empty string")
3241 if endQuoteChar is None:
3242 endQuoteChar = quote_char
3243 else:
3244 endQuoteChar = endQuoteChar.strip()
3245 if not endQuoteChar:
3246 raise ValueError("end_quote_char cannot be the empty string")
3248 self.quoteChar: str = quote_char
3249 self.quoteCharLen: int = len(quote_char)
3250 self.firstQuoteChar: str = quote_char[0]
3251 self.endQuoteChar: str = endQuoteChar
3252 self.endQuoteCharLen: int = len(endQuoteChar)
3253 self.escChar: str = escChar or ""
3254 self.escQuote: str = escQuote or ""
3255 self.unquoteResults: bool = unquoteResults
3256 self.convertWhitespaceEscapes: bool = convertWhitespaceEscapes
3257 self.multiline = multiline
3259 sep = ""
3260 inner_pattern = ""
3262 if escQuote:
3263 inner_pattern += rf"{sep}(?:{re.escape(escQuote)})"
3264 sep = "|"
3266 if escChar:
3267 inner_pattern += rf"{sep}(?:{re.escape(escChar)}.)"
3268 sep = "|"
3269 self.escCharReplacePattern = re.escape(escChar) + "(.)"
3271 if len(self.endQuoteChar) > 1:
3272 inner_pattern += (
3273 f"{sep}(?:"
3274 + "|".join(
3275 f"(?:{re.escape(self.endQuoteChar[:i])}(?!{re.escape(self.endQuoteChar[i:])}))"
3276 for i in range(len(self.endQuoteChar) - 1, 0, -1)
3277 )
3278 + ")"
3279 )
3280 sep = "|"
3282 self.flags = re.RegexFlag(0)
3284 if multiline:
3285 self.flags = re.MULTILINE | re.DOTALL
3286 inner_pattern += (
3287 rf"{sep}(?:[^{_escape_regex_range_chars(self.endQuoteChar[0])}"
3288 rf"{(_escape_regex_range_chars(escChar) if escChar is not None else '')}])"
3289 )
3290 else:
3291 inner_pattern += (
3292 rf"{sep}(?:[^{_escape_regex_range_chars(self.endQuoteChar[0])}\n\r"
3293 rf"{(_escape_regex_range_chars(escChar) if escChar is not None else '')}])"
3294 )
3296 self.pattern = "".join(
3297 [
3298 re.escape(self.quoteChar),
3299 "(?:",
3300 inner_pattern,
3301 ")*",
3302 re.escape(self.endQuoteChar),
3303 ]
3304 )
3306 if self.unquoteResults:
3307 if self.convertWhitespaceEscapes:
3308 self.unquote_scan_re = re.compile(
3309 rf"({'|'.join(re.escape(k) for k in self.ws_map)})|({re.escape(self.escChar)}.)|(\n|.)",
3310 flags=self.flags,
3311 )
3312 else:
3313 self.unquote_scan_re = re.compile(
3314 rf"({re.escape(self.escChar)}.)|(\n|.)", flags=self.flags
3315 )
3317 try:
3318 self.re = re.compile(self.pattern, self.flags)
3319 self.reString = self.pattern
3320 self.re_match = self.re.match
3321 except re.error:
3322 raise ValueError(f"invalid pattern {self.pattern!r} passed to Regex")
3324 self.errmsg = "Expected " + self.name
3325 self.mayIndexError = False
3326 self.mayReturnEmpty = True
3328 def _generateDefaultName(self) -> str:
3329 if self.quoteChar == self.endQuoteChar and isinstance(self.quoteChar, str_type):
3330 return f"string enclosed in {self.quoteChar!r}"
3332 return f"quoted string, starting with {self.quoteChar} ending with {self.endQuoteChar}"
3334 def parseImpl(self, instring, loc, doActions=True):
3335 result = (
3336 instring[loc] == self.firstQuoteChar
3337 and self.re_match(instring, loc)
3338 or None
3339 )
3340 if not result:
3341 raise ParseException(instring, loc, self.errmsg, self)
3343 loc = result.end()
3344 ret = result.group()
3346 if self.unquoteResults:
3347 # strip off quotes
3348 ret = ret[self.quoteCharLen : -self.endQuoteCharLen]
3350 if isinstance(ret, str_type):
3351 if self.convertWhitespaceEscapes:
3352 ret = "".join(
3353 self.ws_map[match.group(1)]
3354 if match.group(1)
3355 else match.group(2)[-1]
3356 if match.group(2)
3357 else match.group(3)
3358 for match in self.unquote_scan_re.finditer(ret)
3359 )
3360 else:
3361 ret = "".join(
3362 match.group(1)[-1] if match.group(1) else match.group(2)
3363 for match in self.unquote_scan_re.finditer(ret)
3364 )
3366 # replace escaped quotes
3367 if self.escQuote:
3368 ret = ret.replace(self.escQuote, self.endQuoteChar)
3370 return loc, ret
3373class CharsNotIn(Token):
3374 """Token for matching words composed of characters *not* in a given
3375 set (will include whitespace in matched characters if not listed in
3376 the provided exclusion set - see example). Defined with string
3377 containing all disallowed characters, and an optional minimum,
3378 maximum, and/or exact length. The default value for ``min`` is
3379 1 (a minimum value < 1 is not valid); the default values for
3380 ``max`` and ``exact`` are 0, meaning no maximum or exact
3381 length restriction.
3383 Example::
3385 # define a comma-separated-value as anything that is not a ','
3386 csv_value = CharsNotIn(',')
3387 print(DelimitedList(csv_value).parse_string("dkls,lsdkjf,s12 34,@!#,213"))
3389 prints::
3391 ['dkls', 'lsdkjf', 's12 34', '@!#', '213']
3392 """
3394 def __init__(
3395 self,
3396 not_chars: str = "",
3397 min: int = 1,
3398 max: int = 0,
3399 exact: int = 0,
3400 *,
3401 notChars: str = "",
3402 ):
3403 super().__init__()
3404 self.skipWhitespace = False
3405 self.notChars = not_chars or notChars
3406 self.notCharsSet = set(self.notChars)
3408 if min < 1:
3409 raise ValueError(
3410 "cannot specify a minimum length < 1; use "
3411 "Opt(CharsNotIn()) if zero-length char group is permitted"
3412 )
3414 self.minLen = min
3416 if max > 0:
3417 self.maxLen = max
3418 else:
3419 self.maxLen = _MAX_INT
3421 if exact > 0:
3422 self.maxLen = exact
3423 self.minLen = exact
3425 self.errmsg = "Expected " + self.name
3426 self.mayReturnEmpty = self.minLen == 0
3427 self.mayIndexError = False
3429 def _generateDefaultName(self) -> str:
3430 not_chars_str = _collapse_string_to_ranges(self.notChars)
3431 if len(not_chars_str) > 16:
3432 return f"!W:({self.notChars[: 16 - 3]}...)"
3433 else:
3434 return f"!W:({self.notChars})"
3436 def parseImpl(self, instring, loc, doActions=True):
3437 notchars = self.notCharsSet
3438 if instring[loc] in notchars:
3439 raise ParseException(instring, loc, self.errmsg, self)
3441 start = loc
3442 loc += 1
3443 maxlen = min(start + self.maxLen, len(instring))
3444 while loc < maxlen and instring[loc] not in notchars:
3445 loc += 1
3447 if loc - start < self.minLen:
3448 raise ParseException(instring, loc, self.errmsg, self)
3450 return loc, instring[start:loc]
3453class White(Token):
3454 """Special matching class for matching whitespace. Normally,
3455 whitespace is ignored by pyparsing grammars. This class is included
3456 when some whitespace structures are significant. Define with
3457 a string containing the whitespace characters to be matched; default
3458 is ``" \\t\\r\\n"``. Also takes optional ``min``,
3459 ``max``, and ``exact`` arguments, as defined for the
3460 :class:`Word` class.
3461 """
3463 whiteStrs = {
3464 " ": "<SP>",
3465 "\t": "<TAB>",
3466 "\n": "<LF>",
3467 "\r": "<CR>",
3468 "\f": "<FF>",
3469 "\u00A0": "<NBSP>",
3470 "\u1680": "<OGHAM_SPACE_MARK>",
3471 "\u180E": "<MONGOLIAN_VOWEL_SEPARATOR>",
3472 "\u2000": "<EN_QUAD>",
3473 "\u2001": "<EM_QUAD>",
3474 "\u2002": "<EN_SPACE>",
3475 "\u2003": "<EM_SPACE>",
3476 "\u2004": "<THREE-PER-EM_SPACE>",
3477 "\u2005": "<FOUR-PER-EM_SPACE>",
3478 "\u2006": "<SIX-PER-EM_SPACE>",
3479 "\u2007": "<FIGURE_SPACE>",
3480 "\u2008": "<PUNCTUATION_SPACE>",
3481 "\u2009": "<THIN_SPACE>",
3482 "\u200A": "<HAIR_SPACE>",
3483 "\u200B": "<ZERO_WIDTH_SPACE>",
3484 "\u202F": "<NNBSP>",
3485 "\u205F": "<MMSP>",
3486 "\u3000": "<IDEOGRAPHIC_SPACE>",
3487 }
3489 def __init__(self, ws: str = " \t\r\n", min: int = 1, max: int = 0, exact: int = 0):
3490 super().__init__()
3491 self.matchWhite = ws
3492 self.set_whitespace_chars(
3493 "".join(c for c in self.whiteStrs if c not in self.matchWhite),
3494 copy_defaults=True,
3495 )
3496 # self.leave_whitespace()
3497 self.mayReturnEmpty = True
3498 self.errmsg = "Expected " + self.name
3500 self.minLen = min
3502 if max > 0:
3503 self.maxLen = max
3504 else:
3505 self.maxLen = _MAX_INT
3507 if exact > 0:
3508 self.maxLen = exact
3509 self.minLen = exact
3511 def _generateDefaultName(self) -> str:
3512 return "".join(White.whiteStrs[c] for c in self.matchWhite)
3514 def parseImpl(self, instring, loc, doActions=True):
3515 if instring[loc] not in self.matchWhite:
3516 raise ParseException(instring, loc, self.errmsg, self)
3517 start = loc
3518 loc += 1
3519 maxloc = start + self.maxLen
3520 maxloc = min(maxloc, len(instring))
3521 while loc < maxloc and instring[loc] in self.matchWhite:
3522 loc += 1
3524 if loc - start < self.minLen:
3525 raise ParseException(instring, loc, self.errmsg, self)
3527 return loc, instring[start:loc]
3530class PositionToken(Token):
3531 def __init__(self):
3532 super().__init__()
3533 self.mayReturnEmpty = True
3534 self.mayIndexError = False
3537class GoToColumn(PositionToken):
3538 """Token to advance to a specific column of input text; useful for
3539 tabular report scraping.
3540 """
3542 def __init__(self, colno: int):
3543 super().__init__()
3544 self.col = colno
3546 def preParse(self, instring: str, loc: int) -> int:
3547 if col(loc, instring) != self.col:
3548 instrlen = len(instring)
3549 if self.ignoreExprs:
3550 loc = self._skipIgnorables(instring, loc)
3551 while (
3552 loc < instrlen
3553 and instring[loc].isspace()
3554 and col(loc, instring) != self.col
3555 ):
3556 loc += 1
3557 return loc
3559 def parseImpl(self, instring, loc, doActions=True):
3560 thiscol = col(loc, instring)
3561 if thiscol > self.col:
3562 raise ParseException(instring, loc, "Text not in expected column", self)
3563 newloc = loc + self.col - thiscol
3564 ret = instring[loc:newloc]
3565 return newloc, ret
3568class LineStart(PositionToken):
3569 r"""Matches if current position is at the beginning of a line within
3570 the parse string
3572 Example::
3574 test = '''\
3575 AAA this line
3576 AAA and this line
3577 AAA but not this one
3578 B AAA and definitely not this one
3579 '''
3581 for t in (LineStart() + 'AAA' + rest_of_line).search_string(test):
3582 print(t)
3584 prints::
3586 ['AAA', ' this line']
3587 ['AAA', ' and this line']
3589 """
3591 def __init__(self):
3592 super().__init__()
3593 self.leave_whitespace()
3594 self.orig_whiteChars = set() | self.whiteChars
3595 self.whiteChars.discard("\n")
3596 self.skipper = Empty().set_whitespace_chars(self.whiteChars)
3597 self.errmsg = "Expected start of line"
3599 def preParse(self, instring: str, loc: int) -> int:
3600 if loc == 0:
3601 return loc
3602 else:
3603 ret = self.skipper.preParse(instring, loc)
3604 if "\n" in self.orig_whiteChars:
3605 while instring[ret : ret + 1] == "\n":
3606 ret = self.skipper.preParse(instring, ret + 1)
3607 return ret
3609 def parseImpl(self, instring, loc, doActions=True):
3610 if col(loc, instring) == 1:
3611 return loc, []
3612 raise ParseException(instring, loc, self.errmsg, self)
3615class LineEnd(PositionToken):
3616 """Matches if current position is at the end of a line within the
3617 parse string
3618 """
3620 def __init__(self):
3621 super().__init__()
3622 self.whiteChars.discard("\n")
3623 self.set_whitespace_chars(self.whiteChars, copy_defaults=False)
3624 self.errmsg = "Expected end of line"
3626 def parseImpl(self, instring, loc, doActions=True):
3627 if loc < len(instring):
3628 if instring[loc] == "\n":
3629 return loc + 1, "\n"
3630 else:
3631 raise ParseException(instring, loc, self.errmsg, self)
3632 elif loc == len(instring):
3633 return loc + 1, []
3634 else:
3635 raise ParseException(instring, loc, self.errmsg, self)
3638class StringStart(PositionToken):
3639 """Matches if current position is at the beginning of the parse
3640 string
3641 """
3643 def __init__(self):
3644 super().__init__()
3645 self.errmsg = "Expected start of text"
3647 def parseImpl(self, instring, loc, doActions=True):
3648 if loc != 0:
3649 # see if entire string up to here is just whitespace and ignoreables
3650 if loc != self.preParse(instring, 0):
3651 raise ParseException(instring, loc, self.errmsg, self)
3652 return loc, []
3655class StringEnd(PositionToken):
3656 """
3657 Matches if current position is at the end of the parse string
3658 """
3660 def __init__(self):
3661 super().__init__()
3662 self.errmsg = "Expected end of text"
3664 def parseImpl(self, instring, loc, doActions=True):
3665 if loc < len(instring):
3666 raise ParseException(instring, loc, self.errmsg, self)
3667 elif loc == len(instring):
3668 return loc + 1, []
3669 elif loc > len(instring):
3670 return loc, []
3671 else:
3672 raise ParseException(instring, loc, self.errmsg, self)
3675class WordStart(PositionToken):
3676 """Matches if the current position is at the beginning of a
3677 :class:`Word`, and is not preceded by any character in a given
3678 set of ``word_chars`` (default= ``printables``). To emulate the
3679 ``\b`` behavior of regular expressions, use
3680 ``WordStart(alphanums)``. ``WordStart`` will also match at
3681 the beginning of the string being parsed, or at the beginning of
3682 a line.
3683 """
3685 def __init__(self, word_chars: str = printables, *, wordChars: str = printables):
3686 wordChars = word_chars if wordChars == printables else wordChars
3687 super().__init__()
3688 self.wordChars = set(wordChars)
3689 self.errmsg = "Not at the start of a word"
3691 def parseImpl(self, instring, loc, doActions=True):
3692 if loc != 0:
3693 if (
3694 instring[loc - 1] in self.wordChars
3695 or instring[loc] not in self.wordChars
3696 ):
3697 raise ParseException(instring, loc, self.errmsg, self)
3698 return loc, []
3701class WordEnd(PositionToken):
3702 """Matches if the current position is at the end of a :class:`Word`,
3703 and is not followed by any character in a given set of ``word_chars``
3704 (default= ``printables``). To emulate the ``\b`` behavior of
3705 regular expressions, use ``WordEnd(alphanums)``. ``WordEnd``
3706 will also match at the end of the string being parsed, or at the end
3707 of a line.
3708 """
3710 def __init__(self, word_chars: str = printables, *, wordChars: str = printables):
3711 wordChars = word_chars if wordChars == printables else wordChars
3712 super().__init__()
3713 self.wordChars = set(wordChars)
3714 self.skipWhitespace = False
3715 self.errmsg = "Not at the end of a word"
3717 def parseImpl(self, instring, loc, doActions=True):
3718 instrlen = len(instring)
3719 if instrlen > 0 and loc < instrlen:
3720 if (
3721 instring[loc] in self.wordChars
3722 or instring[loc - 1] not in self.wordChars
3723 ):
3724 raise ParseException(instring, loc, self.errmsg, self)
3725 return loc, []
3728class ParseExpression(ParserElement):
3729 """Abstract subclass of ParserElement, for combining and
3730 post-processing parsed tokens.
3731 """
3733 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = False):
3734 super().__init__(savelist)
3735 self.exprs: List[ParserElement]
3736 if isinstance(exprs, _generatorType):
3737 exprs = list(exprs)
3739 if isinstance(exprs, str_type):
3740 self.exprs = [self._literalStringClass(exprs)]
3741 elif isinstance(exprs, ParserElement):
3742 self.exprs = [exprs]
3743 elif isinstance(exprs, Iterable):
3744 exprs = list(exprs)
3745 # if sequence of strings provided, wrap with Literal
3746 if any(isinstance(expr, str_type) for expr in exprs):
3747 exprs = (
3748 self._literalStringClass(e) if isinstance(e, str_type) else e
3749 for e in exprs
3750 )
3751 self.exprs = list(exprs)
3752 else:
3753 try:
3754 self.exprs = list(exprs)
3755 except TypeError:
3756 self.exprs = [exprs]
3757 self.callPreparse = False
3759 def recurse(self) -> List[ParserElement]:
3760 return self.exprs[:]
3762 def append(self, other) -> ParserElement:
3763 self.exprs.append(other)
3764 self._defaultName = None
3765 return self
3767 def leave_whitespace(self, recursive: bool = True) -> ParserElement:
3768 """
3769 Extends ``leave_whitespace`` defined in base class, and also invokes ``leave_whitespace`` on
3770 all contained expressions.
3771 """
3772 super().leave_whitespace(recursive)
3774 if recursive:
3775 self.exprs = [e.copy() for e in self.exprs]
3776 for e in self.exprs:
3777 e.leave_whitespace(recursive)
3778 return self
3780 def ignore_whitespace(self, recursive: bool = True) -> ParserElement:
3781 """
3782 Extends ``ignore_whitespace`` defined in base class, and also invokes ``leave_whitespace`` on
3783 all contained expressions.
3784 """
3785 super().ignore_whitespace(recursive)
3786 if recursive:
3787 self.exprs = [e.copy() for e in self.exprs]
3788 for e in self.exprs:
3789 e.ignore_whitespace(recursive)
3790 return self
3792 def ignore(self, other) -> ParserElement:
3793 if isinstance(other, Suppress):
3794 if other not in self.ignoreExprs:
3795 super().ignore(other)
3796 for e in self.exprs:
3797 e.ignore(self.ignoreExprs[-1])
3798 else:
3799 super().ignore(other)
3800 for e in self.exprs:
3801 e.ignore(self.ignoreExprs[-1])
3802 return self
3804 def _generateDefaultName(self) -> str:
3805 return f"{self.__class__.__name__}:({str(self.exprs)})"
3807 def streamline(self) -> ParserElement:
3808 if self.streamlined:
3809 return self
3811 super().streamline()
3813 for e in self.exprs:
3814 e.streamline()
3816 # collapse nested :class:`And`'s of the form ``And(And(And(a, b), c), d)`` to ``And(a, b, c, d)``
3817 # but only if there are no parse actions or resultsNames on the nested And's
3818 # (likewise for :class:`Or`'s and :class:`MatchFirst`'s)
3819 if len(self.exprs) == 2:
3820 other = self.exprs[0]
3821 if (
3822 isinstance(other, self.__class__)
3823 and not other.parseAction
3824 and other.resultsName is None
3825 and not other.debug
3826 ):
3827 self.exprs = other.exprs[:] + [self.exprs[1]]
3828 self._defaultName = None
3829 self.mayReturnEmpty |= other.mayReturnEmpty
3830 self.mayIndexError |= other.mayIndexError
3832 other = self.exprs[-1]
3833 if (
3834 isinstance(other, self.__class__)
3835 and not other.parseAction
3836 and other.resultsName is None
3837 and not other.debug
3838 ):
3839 self.exprs = self.exprs[:-1] + other.exprs[:]
3840 self._defaultName = None
3841 self.mayReturnEmpty |= other.mayReturnEmpty
3842 self.mayIndexError |= other.mayIndexError
3844 self.errmsg = "Expected " + str(self)
3846 return self
3848 def validate(self, validateTrace=None) -> None:
3849 warnings.warn(
3850 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
3851 DeprecationWarning,
3852 stacklevel=2,
3853 )
3854 tmp = (validateTrace if validateTrace is not None else [])[:] + [self]
3855 for e in self.exprs:
3856 e.validate(tmp)
3857 self._checkRecursion([])
3859 def copy(self) -> ParserElement:
3860 ret = super().copy()
3861 ret = typing.cast(ParseExpression, ret)
3862 ret.exprs = [e.copy() for e in self.exprs]
3863 return ret
3865 def _setResultsName(self, name, listAllMatches=False):
3866 if (
3867 __diag__.warn_ungrouped_named_tokens_in_collection
3868 and Diagnostics.warn_ungrouped_named_tokens_in_collection
3869 not in self.suppress_warnings_
3870 ):
3871 for e in self.exprs:
3872 if (
3873 isinstance(e, ParserElement)
3874 and e.resultsName
3875 and Diagnostics.warn_ungrouped_named_tokens_in_collection
3876 not in e.suppress_warnings_
3877 ):
3878 warnings.warn(
3879 "{}: setting results name {!r} on {} expression "
3880 "collides with {!r} on contained expression".format(
3881 "warn_ungrouped_named_tokens_in_collection",
3882 name,
3883 type(self).__name__,
3884 e.resultsName,
3885 ),
3886 stacklevel=3,
3887 )
3889 return super()._setResultsName(name, listAllMatches)
3891 # Compatibility synonyms
3892 # fmt: off
3893 @replaced_by_pep8(leave_whitespace)
3894 def leaveWhitespace(self): ...
3896 @replaced_by_pep8(ignore_whitespace)
3897 def ignoreWhitespace(self): ...
3898 # fmt: on
3901class And(ParseExpression):
3902 """
3903 Requires all given :class:`ParseExpression` s to be found in the given order.
3904 Expressions may be separated by whitespace.
3905 May be constructed using the ``'+'`` operator.
3906 May also be constructed using the ``'-'`` operator, which will
3907 suppress backtracking.
3909 Example::
3911 integer = Word(nums)
3912 name_expr = Word(alphas)[1, ...]
3914 expr = And([integer("id"), name_expr("name"), integer("age")])
3915 # more easily written as:
3916 expr = integer("id") + name_expr("name") + integer("age")
3917 """
3919 class _ErrorStop(Empty):
3920 def __init__(self, *args, **kwargs):
3921 super().__init__(*args, **kwargs)
3922 self.leave_whitespace()
3924 def _generateDefaultName(self) -> str:
3925 return "-"
3927 def __init__(
3928 self, exprs_arg: typing.Iterable[ParserElement], savelist: bool = True
3929 ):
3930 exprs: List[ParserElement] = list(exprs_arg)
3931 if exprs and Ellipsis in exprs:
3932 tmp = []
3933 for i, expr in enumerate(exprs):
3934 if expr is Ellipsis:
3935 if i < len(exprs) - 1:
3936 skipto_arg: ParserElement = typing.cast(
3937 ParseExpression, (Empty() + exprs[i + 1])
3938 ).exprs[-1]
3939 tmp.append(SkipTo(skipto_arg)("_skipped*"))
3940 else:
3941 raise Exception(
3942 "cannot construct And with sequence ending in ..."
3943 )
3944 else:
3945 tmp.append(expr)
3946 exprs[:] = tmp
3947 super().__init__(exprs, savelist)
3948 if self.exprs:
3949 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
3950 if not isinstance(self.exprs[0], White):
3951 self.set_whitespace_chars(
3952 self.exprs[0].whiteChars,
3953 copy_defaults=self.exprs[0].copyDefaultWhiteChars,
3954 )
3955 self.skipWhitespace = self.exprs[0].skipWhitespace
3956 else:
3957 self.skipWhitespace = False
3958 else:
3959 self.mayReturnEmpty = True
3960 self.callPreparse = True
3962 def streamline(self) -> ParserElement:
3963 # collapse any _PendingSkip's
3964 if self.exprs:
3965 if any(
3966 isinstance(e, ParseExpression)
3967 and e.exprs
3968 and isinstance(e.exprs[-1], _PendingSkip)
3969 for e in self.exprs[:-1]
3970 ):
3971 deleted_expr_marker = NoMatch()
3972 for i, e in enumerate(self.exprs[:-1]):
3973 if e is deleted_expr_marker:
3974 continue
3975 if (
3976 isinstance(e, ParseExpression)
3977 and e.exprs
3978 and isinstance(e.exprs[-1], _PendingSkip)
3979 ):
3980 e.exprs[-1] = e.exprs[-1] + self.exprs[i + 1]
3981 self.exprs[i + 1] = deleted_expr_marker
3982 self.exprs = [e for e in self.exprs if e is not deleted_expr_marker]
3984 super().streamline()
3986 # link any IndentedBlocks to the prior expression
3987 prev: ParserElement
3988 cur: ParserElement
3989 for prev, cur in zip(self.exprs, self.exprs[1:]):
3990 # traverse cur or any first embedded expr of cur looking for an IndentedBlock
3991 # (but watch out for recursive grammar)
3992 seen = set()
3993 while True:
3994 if id(cur) in seen:
3995 break
3996 seen.add(id(cur))
3997 if isinstance(cur, IndentedBlock):
3998 prev.add_parse_action(
3999 lambda s, l, t, cur_=cur: setattr(
4000 cur_, "parent_anchor", col(l, s)
4001 )
4002 )
4003 break
4004 subs = cur.recurse()
4005 next_first = next(iter(subs), None)
4006 if next_first is None:
4007 break
4008 cur = typing.cast(ParserElement, next_first)
4010 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
4011 return self
4013 def parseImpl(self, instring, loc, doActions=True):
4014 # pass False as callPreParse arg to _parse for first element, since we already
4015 # pre-parsed the string as part of our And pre-parsing
4016 loc, resultlist = self.exprs[0]._parse(
4017 instring, loc, doActions, callPreParse=False
4018 )
4019 errorStop = False
4020 for e in self.exprs[1:]:
4021 # if isinstance(e, And._ErrorStop):
4022 if type(e) is And._ErrorStop:
4023 errorStop = True
4024 continue
4025 if errorStop:
4026 try:
4027 loc, exprtokens = e._parse(instring, loc, doActions)
4028 except ParseSyntaxException:
4029 raise
4030 except ParseBaseException as pe:
4031 pe.__traceback__ = None
4032 raise ParseSyntaxException._from_exception(pe)
4033 except IndexError:
4034 raise ParseSyntaxException(
4035 instring, len(instring), self.errmsg, self
4036 )
4037 else:
4038 loc, exprtokens = e._parse(instring, loc, doActions)
4039 resultlist += exprtokens
4040 return loc, resultlist
4042 def __iadd__(self, other):
4043 if isinstance(other, str_type):
4044 other = self._literalStringClass(other)
4045 if not isinstance(other, ParserElement):
4046 return NotImplemented
4047 return self.append(other) # And([self, other])
4049 def _checkRecursion(self, parseElementList):
4050 subRecCheckList = parseElementList[:] + [self]
4051 for e in self.exprs:
4052 e._checkRecursion(subRecCheckList)
4053 if not e.mayReturnEmpty:
4054 break
4056 def _generateDefaultName(self) -> str:
4057 inner = " ".join(str(e) for e in self.exprs)
4058 # strip off redundant inner {}'s
4059 while len(inner) > 1 and inner[0 :: len(inner) - 1] == "{}":
4060 inner = inner[1:-1]
4061 return "{" + inner + "}"
4064class Or(ParseExpression):
4065 """Requires that at least one :class:`ParseExpression` is found. If
4066 two expressions match, the expression that matches the longest
4067 string will be used. May be constructed using the ``'^'``
4068 operator.
4070 Example::
4072 # construct Or using '^' operator
4074 number = Word(nums) ^ Combine(Word(nums) + '.' + Word(nums))
4075 print(number.search_string("123 3.1416 789"))
4077 prints::
4079 [['123'], ['3.1416'], ['789']]
4080 """
4082 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = False):
4083 super().__init__(exprs, savelist)
4084 if self.exprs:
4085 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4086 self.skipWhitespace = all(e.skipWhitespace for e in self.exprs)
4087 else:
4088 self.mayReturnEmpty = True
4090 def streamline(self) -> ParserElement:
4091 super().streamline()
4092 if self.exprs:
4093 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4094 self.saveAsList = any(e.saveAsList for e in self.exprs)
4095 self.skipWhitespace = all(
4096 e.skipWhitespace and not isinstance(e, White) for e in self.exprs
4097 )
4098 else:
4099 self.saveAsList = False
4100 return self
4102 def parseImpl(self, instring, loc, doActions=True):
4103 maxExcLoc = -1
4104 maxException = None
4105 matches = []
4106 fatals = []
4107 if all(e.callPreparse for e in self.exprs):
4108 loc = self.preParse(instring, loc)
4109 for e in self.exprs:
4110 try:
4111 loc2 = e.try_parse(instring, loc, raise_fatal=True)
4112 except ParseFatalException as pfe:
4113 pfe.__traceback__ = None
4114 pfe.parser_element = e
4115 fatals.append(pfe)
4116 maxException = None
4117 maxExcLoc = -1
4118 except ParseException as err:
4119 if not fatals:
4120 err.__traceback__ = None
4121 if err.loc > maxExcLoc:
4122 maxException = err
4123 maxExcLoc = err.loc
4124 except IndexError:
4125 if len(instring) > maxExcLoc:
4126 maxException = ParseException(
4127 instring, len(instring), e.errmsg, self
4128 )
4129 maxExcLoc = len(instring)
4130 else:
4131 # save match among all matches, to retry longest to shortest
4132 matches.append((loc2, e))
4134 if matches:
4135 # re-evaluate all matches in descending order of length of match, in case attached actions
4136 # might change whether or how much they match of the input.
4137 matches.sort(key=itemgetter(0), reverse=True)
4139 if not doActions:
4140 # no further conditions or parse actions to change the selection of
4141 # alternative, so the first match will be the best match
4142 best_expr = matches[0][1]
4143 return best_expr._parse(instring, loc, doActions)
4145 longest = -1, None
4146 for loc1, expr1 in matches:
4147 if loc1 <= longest[0]:
4148 # already have a longer match than this one will deliver, we are done
4149 return longest
4151 try:
4152 loc2, toks = expr1._parse(instring, loc, doActions)
4153 except ParseException as err:
4154 err.__traceback__ = None
4155 if err.loc > maxExcLoc:
4156 maxException = err
4157 maxExcLoc = err.loc
4158 else:
4159 if loc2 >= loc1:
4160 return loc2, toks
4161 # didn't match as much as before
4162 elif loc2 > longest[0]:
4163 longest = loc2, toks
4165 if longest != (-1, None):
4166 return longest
4168 if fatals:
4169 if len(fatals) > 1:
4170 fatals.sort(key=lambda e: -e.loc)
4171 if fatals[0].loc == fatals[1].loc:
4172 fatals.sort(key=lambda e: (-e.loc, -len(str(e.parser_element))))
4173 max_fatal = fatals[0]
4174 raise max_fatal
4176 if maxException is not None:
4177 # infer from this check that all alternatives failed at the current position
4178 # so emit this collective error message instead of any single error message
4179 if maxExcLoc == loc:
4180 maxException.msg = self.errmsg
4181 raise maxException
4182 else:
4183 raise ParseException(
4184 instring, loc, "no defined alternatives to match", self
4185 )
4187 def __ixor__(self, other):
4188 if isinstance(other, str_type):
4189 other = self._literalStringClass(other)
4190 if not isinstance(other, ParserElement):
4191 return NotImplemented
4192 return self.append(other) # Or([self, other])
4194 def _generateDefaultName(self) -> str:
4195 return "{" + " ^ ".join(str(e) for e in self.exprs) + "}"
4197 def _setResultsName(self, name, listAllMatches=False):
4198 if (
4199 __diag__.warn_multiple_tokens_in_named_alternation
4200 and Diagnostics.warn_multiple_tokens_in_named_alternation
4201 not in self.suppress_warnings_
4202 ):
4203 if any(
4204 isinstance(e, And)
4205 and Diagnostics.warn_multiple_tokens_in_named_alternation
4206 not in e.suppress_warnings_
4207 for e in self.exprs
4208 ):
4209 warnings.warn(
4210 "{}: setting results name {!r} on {} expression "
4211 "will return a list of all parsed tokens in an And alternative, "
4212 "in prior versions only the first token was returned; enclose "
4213 "contained argument in Group".format(
4214 "warn_multiple_tokens_in_named_alternation",
4215 name,
4216 type(self).__name__,
4217 ),
4218 stacklevel=3,
4219 )
4221 return super()._setResultsName(name, listAllMatches)
4224class MatchFirst(ParseExpression):
4225 """Requires that at least one :class:`ParseExpression` is found. If
4226 more than one expression matches, the first one listed is the one that will
4227 match. May be constructed using the ``'|'`` operator.
4229 Example::
4231 # construct MatchFirst using '|' operator
4233 # watch the order of expressions to match
4234 number = Word(nums) | Combine(Word(nums) + '.' + Word(nums))
4235 print(number.search_string("123 3.1416 789")) # Fail! -> [['123'], ['3'], ['1416'], ['789']]
4237 # put more selective expression first
4238 number = Combine(Word(nums) + '.' + Word(nums)) | Word(nums)
4239 print(number.search_string("123 3.1416 789")) # Better -> [['123'], ['3.1416'], ['789']]
4240 """
4242 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = False):
4243 super().__init__(exprs, savelist)
4244 if self.exprs:
4245 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4246 self.skipWhitespace = all(e.skipWhitespace for e in self.exprs)
4247 else:
4248 self.mayReturnEmpty = True
4250 def streamline(self) -> ParserElement:
4251 if self.streamlined:
4252 return self
4254 super().streamline()
4255 if self.exprs:
4256 self.saveAsList = any(e.saveAsList for e in self.exprs)
4257 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4258 self.skipWhitespace = all(
4259 e.skipWhitespace and not isinstance(e, White) for e in self.exprs
4260 )
4261 else:
4262 self.saveAsList = False
4263 self.mayReturnEmpty = True
4264 return self
4266 def parseImpl(self, instring, loc, doActions=True):
4267 maxExcLoc = -1
4268 maxException = None
4270 for e in self.exprs:
4271 try:
4272 return e._parse(
4273 instring,
4274 loc,
4275 doActions,
4276 )
4277 except ParseFatalException as pfe:
4278 pfe.__traceback__ = None
4279 pfe.parser_element = e
4280 raise
4281 except ParseException as err:
4282 if err.loc > maxExcLoc:
4283 maxException = err
4284 maxExcLoc = err.loc
4285 except IndexError:
4286 if len(instring) > maxExcLoc:
4287 maxException = ParseException(
4288 instring, len(instring), e.errmsg, self
4289 )
4290 maxExcLoc = len(instring)
4292 if maxException is not None:
4293 # infer from this check that all alternatives failed at the current position
4294 # so emit this collective error message instead of any individual error message
4295 if maxExcLoc == loc:
4296 maxException.msg = self.errmsg
4297 raise maxException
4298 else:
4299 raise ParseException(
4300 instring, loc, "no defined alternatives to match", self
4301 )
4303 def __ior__(self, other):
4304 if isinstance(other, str_type):
4305 other = self._literalStringClass(other)
4306 if not isinstance(other, ParserElement):
4307 return NotImplemented
4308 return self.append(other) # MatchFirst([self, other])
4310 def _generateDefaultName(self) -> str:
4311 return "{" + " | ".join(str(e) for e in self.exprs) + "}"
4313 def _setResultsName(self, name, listAllMatches=False):
4314 if (
4315 __diag__.warn_multiple_tokens_in_named_alternation
4316 and Diagnostics.warn_multiple_tokens_in_named_alternation
4317 not in self.suppress_warnings_
4318 ):
4319 if any(
4320 isinstance(e, And)
4321 and Diagnostics.warn_multiple_tokens_in_named_alternation
4322 not in e.suppress_warnings_
4323 for e in self.exprs
4324 ):
4325 warnings.warn(
4326 "{}: setting results name {!r} on {} expression "
4327 "will return a list of all parsed tokens in an And alternative, "
4328 "in prior versions only the first token was returned; enclose "
4329 "contained argument in Group".format(
4330 "warn_multiple_tokens_in_named_alternation",
4331 name,
4332 type(self).__name__,
4333 ),
4334 stacklevel=3,
4335 )
4337 return super()._setResultsName(name, listAllMatches)
4340class Each(ParseExpression):
4341 """Requires all given :class:`ParseExpression` s to be found, but in
4342 any order. Expressions may be separated by whitespace.
4344 May be constructed using the ``'&'`` operator.
4346 Example::
4348 color = one_of("RED ORANGE YELLOW GREEN BLUE PURPLE BLACK WHITE BROWN")
4349 shape_type = one_of("SQUARE CIRCLE TRIANGLE STAR HEXAGON OCTAGON")
4350 integer = Word(nums)
4351 shape_attr = "shape:" + shape_type("shape")
4352 posn_attr = "posn:" + Group(integer("x") + ',' + integer("y"))("posn")
4353 color_attr = "color:" + color("color")
4354 size_attr = "size:" + integer("size")
4356 # use Each (using operator '&') to accept attributes in any order
4357 # (shape and posn are required, color and size are optional)
4358 shape_spec = shape_attr & posn_attr & Opt(color_attr) & Opt(size_attr)
4360 shape_spec.run_tests('''
4361 shape: SQUARE color: BLACK posn: 100, 120
4362 shape: CIRCLE size: 50 color: BLUE posn: 50,80
4363 color:GREEN size:20 shape:TRIANGLE posn:20,40
4364 '''
4365 )
4367 prints::
4369 shape: SQUARE color: BLACK posn: 100, 120
4370 ['shape:', 'SQUARE', 'color:', 'BLACK', 'posn:', ['100', ',', '120']]
4371 - color: BLACK
4372 - posn: ['100', ',', '120']
4373 - x: 100
4374 - y: 120
4375 - shape: SQUARE
4378 shape: CIRCLE size: 50 color: BLUE posn: 50,80
4379 ['shape:', 'CIRCLE', 'size:', '50', 'color:', 'BLUE', 'posn:', ['50', ',', '80']]
4380 - color: BLUE
4381 - posn: ['50', ',', '80']
4382 - x: 50
4383 - y: 80
4384 - shape: CIRCLE
4385 - size: 50
4388 color: GREEN size: 20 shape: TRIANGLE posn: 20,40
4389 ['color:', 'GREEN', 'size:', '20', 'shape:', 'TRIANGLE', 'posn:', ['20', ',', '40']]
4390 - color: GREEN
4391 - posn: ['20', ',', '40']
4392 - x: 20
4393 - y: 40
4394 - shape: TRIANGLE
4395 - size: 20
4396 """
4398 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = True):
4399 super().__init__(exprs, savelist)
4400 if self.exprs:
4401 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
4402 else:
4403 self.mayReturnEmpty = True
4404 self.skipWhitespace = True
4405 self.initExprGroups = True
4406 self.saveAsList = True
4408 def __iand__(self, other):
4409 if isinstance(other, str_type):
4410 other = self._literalStringClass(other)
4411 if not isinstance(other, ParserElement):
4412 return NotImplemented
4413 return self.append(other) # Each([self, other])
4415 def streamline(self) -> ParserElement:
4416 super().streamline()
4417 if self.exprs:
4418 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
4419 else:
4420 self.mayReturnEmpty = True
4421 return self
4423 def parseImpl(self, instring, loc, doActions=True):
4424 if self.initExprGroups:
4425 self.opt1map = dict(
4426 (id(e.expr), e) for e in self.exprs if isinstance(e, Opt)
4427 )
4428 opt1 = [e.expr for e in self.exprs if isinstance(e, Opt)]
4429 opt2 = [
4430 e
4431 for e in self.exprs
4432 if e.mayReturnEmpty and not isinstance(e, (Opt, Regex, ZeroOrMore))
4433 ]
4434 self.optionals = opt1 + opt2
4435 self.multioptionals = [
4436 e.expr.set_results_name(e.resultsName, list_all_matches=True)
4437 for e in self.exprs
4438 if isinstance(e, _MultipleMatch)
4439 ]
4440 self.multirequired = [
4441 e.expr.set_results_name(e.resultsName, list_all_matches=True)
4442 for e in self.exprs
4443 if isinstance(e, OneOrMore)
4444 ]
4445 self.required = [
4446 e for e in self.exprs if not isinstance(e, (Opt, ZeroOrMore, OneOrMore))
4447 ]
4448 self.required += self.multirequired
4449 self.initExprGroups = False
4451 tmpLoc = loc
4452 tmpReqd = self.required[:]
4453 tmpOpt = self.optionals[:]
4454 multis = self.multioptionals[:]
4455 matchOrder = []
4457 keepMatching = True
4458 failed = []
4459 fatals = []
4460 while keepMatching:
4461 tmpExprs = tmpReqd + tmpOpt + multis
4462 failed.clear()
4463 fatals.clear()
4464 for e in tmpExprs:
4465 try:
4466 tmpLoc = e.try_parse(instring, tmpLoc, raise_fatal=True)
4467 except ParseFatalException as pfe:
4468 pfe.__traceback__ = None
4469 pfe.parser_element = e
4470 fatals.append(pfe)
4471 failed.append(e)
4472 except ParseException:
4473 failed.append(e)
4474 else:
4475 matchOrder.append(self.opt1map.get(id(e), e))
4476 if e in tmpReqd:
4477 tmpReqd.remove(e)
4478 elif e in tmpOpt:
4479 tmpOpt.remove(e)
4480 if len(failed) == len(tmpExprs):
4481 keepMatching = False
4483 # look for any ParseFatalExceptions
4484 if fatals:
4485 if len(fatals) > 1:
4486 fatals.sort(key=lambda e: -e.loc)
4487 if fatals[0].loc == fatals[1].loc:
4488 fatals.sort(key=lambda e: (-e.loc, -len(str(e.parser_element))))
4489 max_fatal = fatals[0]
4490 raise max_fatal
4492 if tmpReqd:
4493 missing = ", ".join([str(e) for e in tmpReqd])
4494 raise ParseException(
4495 instring,
4496 loc,
4497 f"Missing one or more required elements ({missing})",
4498 )
4500 # add any unmatched Opts, in case they have default values defined
4501 matchOrder += [e for e in self.exprs if isinstance(e, Opt) and e.expr in tmpOpt]
4503 total_results = ParseResults([])
4504 for e in matchOrder:
4505 loc, results = e._parse(instring, loc, doActions)
4506 total_results += results
4508 return loc, total_results
4510 def _generateDefaultName(self) -> str:
4511 return "{" + " & ".join(str(e) for e in self.exprs) + "}"
4514class ParseElementEnhance(ParserElement):
4515 """Abstract subclass of :class:`ParserElement`, for combining and
4516 post-processing parsed tokens.
4517 """
4519 def __init__(self, expr: Union[ParserElement, str], savelist: bool = False):
4520 super().__init__(savelist)
4521 if isinstance(expr, str_type):
4522 expr_str = typing.cast(str, expr)
4523 if issubclass(self._literalStringClass, Token):
4524 expr = self._literalStringClass(expr_str) # type: ignore[call-arg]
4525 elif issubclass(type(self), self._literalStringClass):
4526 expr = Literal(expr_str)
4527 else:
4528 expr = self._literalStringClass(Literal(expr_str)) # type: ignore[assignment, call-arg]
4529 expr = typing.cast(ParserElement, expr)
4530 self.expr = expr
4531 if expr is not None:
4532 self.mayIndexError = expr.mayIndexError
4533 self.mayReturnEmpty = expr.mayReturnEmpty
4534 self.set_whitespace_chars(
4535 expr.whiteChars, copy_defaults=expr.copyDefaultWhiteChars
4536 )
4537 self.skipWhitespace = expr.skipWhitespace
4538 self.saveAsList = expr.saveAsList
4539 self.callPreparse = expr.callPreparse
4540 self.ignoreExprs.extend(expr.ignoreExprs)
4542 def recurse(self) -> List[ParserElement]:
4543 return [self.expr] if self.expr is not None else []
4545 def parseImpl(self, instring, loc, doActions=True):
4546 if self.expr is not None:
4547 try:
4548 return self.expr._parse(instring, loc, doActions, callPreParse=False)
4549 except ParseBaseException as pbe:
4550 pbe.msg = self.errmsg
4551 raise
4552 else:
4553 raise ParseException(instring, loc, "No expression defined", self)
4555 def leave_whitespace(self, recursive: bool = True) -> ParserElement:
4556 super().leave_whitespace(recursive)
4558 if recursive:
4559 if self.expr is not None:
4560 self.expr = self.expr.copy()
4561 self.expr.leave_whitespace(recursive)
4562 return self
4564 def ignore_whitespace(self, recursive: bool = True) -> ParserElement:
4565 super().ignore_whitespace(recursive)
4567 if recursive:
4568 if self.expr is not None:
4569 self.expr = self.expr.copy()
4570 self.expr.ignore_whitespace(recursive)
4571 return self
4573 def ignore(self, other) -> ParserElement:
4574 if isinstance(other, Suppress):
4575 if other not in self.ignoreExprs:
4576 super().ignore(other)
4577 if self.expr is not None:
4578 self.expr.ignore(self.ignoreExprs[-1])
4579 else:
4580 super().ignore(other)
4581 if self.expr is not None:
4582 self.expr.ignore(self.ignoreExprs[-1])
4583 return self
4585 def streamline(self) -> ParserElement:
4586 super().streamline()
4587 if self.expr is not None:
4588 self.expr.streamline()
4589 return self
4591 def _checkRecursion(self, parseElementList):
4592 if self in parseElementList:
4593 raise RecursiveGrammarException(parseElementList + [self])
4594 subRecCheckList = parseElementList[:] + [self]
4595 if self.expr is not None:
4596 self.expr._checkRecursion(subRecCheckList)
4598 def validate(self, validateTrace=None) -> None:
4599 warnings.warn(
4600 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
4601 DeprecationWarning,
4602 stacklevel=2,
4603 )
4604 if validateTrace is None:
4605 validateTrace = []
4606 tmp = validateTrace[:] + [self]
4607 if self.expr is not None:
4608 self.expr.validate(tmp)
4609 self._checkRecursion([])
4611 def _generateDefaultName(self) -> str:
4612 return f"{self.__class__.__name__}:({str(self.expr)})"
4614 # Compatibility synonyms
4615 # fmt: off
4616 @replaced_by_pep8(leave_whitespace)
4617 def leaveWhitespace(self): ...
4619 @replaced_by_pep8(ignore_whitespace)
4620 def ignoreWhitespace(self): ...
4621 # fmt: on
4624class IndentedBlock(ParseElementEnhance):
4625 """
4626 Expression to match one or more expressions at a given indentation level.
4627 Useful for parsing text where structure is implied by indentation (like Python source code).
4628 """
4630 class _Indent(Empty):
4631 def __init__(self, ref_col: int):
4632 super().__init__()
4633 self.errmsg = f"expected indent at column {ref_col}"
4634 self.add_condition(lambda s, l, t: col(l, s) == ref_col)
4636 class _IndentGreater(Empty):
4637 def __init__(self, ref_col: int):
4638 super().__init__()
4639 self.errmsg = f"expected indent at column greater than {ref_col}"
4640 self.add_condition(lambda s, l, t: col(l, s) > ref_col)
4642 def __init__(
4643 self, expr: ParserElement, *, recursive: bool = False, grouped: bool = True
4644 ):
4645 super().__init__(expr, savelist=True)
4646 # if recursive:
4647 # raise NotImplementedError("IndentedBlock with recursive is not implemented")
4648 self._recursive = recursive
4649 self._grouped = grouped
4650 self.parent_anchor = 1
4652 def parseImpl(self, instring, loc, doActions=True):
4653 # advance parse position to non-whitespace by using an Empty()
4654 # this should be the column to be used for all subsequent indented lines
4655 anchor_loc = Empty().preParse(instring, loc)
4657 # see if self.expr matches at the current location - if not it will raise an exception
4658 # and no further work is necessary
4659 self.expr.try_parse(instring, anchor_loc, do_actions=doActions)
4661 indent_col = col(anchor_loc, instring)
4662 peer_detect_expr = self._Indent(indent_col)
4664 inner_expr = Empty() + peer_detect_expr + self.expr
4665 if self._recursive:
4666 sub_indent = self._IndentGreater(indent_col)
4667 nested_block = IndentedBlock(
4668 self.expr, recursive=self._recursive, grouped=self._grouped
4669 )
4670 nested_block.set_debug(self.debug)
4671 nested_block.parent_anchor = indent_col
4672 inner_expr += Opt(sub_indent + nested_block)
4674 inner_expr.set_name(f"inner {hex(id(inner_expr))[-4:].upper()}@{indent_col}")
4675 block = OneOrMore(inner_expr)
4677 trailing_undent = self._Indent(self.parent_anchor) | StringEnd()
4679 if self._grouped:
4680 wrapper = Group
4681 else:
4682 wrapper = lambda expr: expr
4683 return (wrapper(block) + Optional(trailing_undent)).parseImpl(
4684 instring, anchor_loc, doActions
4685 )
4688class AtStringStart(ParseElementEnhance):
4689 """Matches if expression matches at the beginning of the parse
4690 string::
4692 AtStringStart(Word(nums)).parse_string("123")
4693 # prints ["123"]
4695 AtStringStart(Word(nums)).parse_string(" 123")
4696 # raises ParseException
4697 """
4699 def __init__(self, expr: Union[ParserElement, str]):
4700 super().__init__(expr)
4701 self.callPreparse = False
4703 def parseImpl(self, instring, loc, doActions=True):
4704 if loc != 0:
4705 raise ParseException(instring, loc, "not found at string start")
4706 return super().parseImpl(instring, loc, doActions)
4709class AtLineStart(ParseElementEnhance):
4710 r"""Matches if an expression matches at the beginning of a line within
4711 the parse string
4713 Example::
4715 test = '''\
4716 AAA this line
4717 AAA and this line
4718 AAA but not this one
4719 B AAA and definitely not this one
4720 '''
4722 for t in (AtLineStart('AAA') + rest_of_line).search_string(test):
4723 print(t)
4725 prints::
4727 ['AAA', ' this line']
4728 ['AAA', ' and this line']
4730 """
4732 def __init__(self, expr: Union[ParserElement, str]):
4733 super().__init__(expr)
4734 self.callPreparse = False
4736 def parseImpl(self, instring, loc, doActions=True):
4737 if col(loc, instring) != 1:
4738 raise ParseException(instring, loc, "not found at line start")
4739 return super().parseImpl(instring, loc, doActions)
4742class FollowedBy(ParseElementEnhance):
4743 """Lookahead matching of the given parse expression.
4744 ``FollowedBy`` does *not* advance the parsing position within
4745 the input string, it only verifies that the specified parse
4746 expression matches at the current position. ``FollowedBy``
4747 always returns a null token list. If any results names are defined
4748 in the lookahead expression, those *will* be returned for access by
4749 name.
4751 Example::
4753 # use FollowedBy to match a label only if it is followed by a ':'
4754 data_word = Word(alphas)
4755 label = data_word + FollowedBy(':')
4756 attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stop_on=label).set_parse_action(' '.join))
4758 attr_expr[1, ...].parse_string("shape: SQUARE color: BLACK posn: upper left").pprint()
4760 prints::
4762 [['shape', 'SQUARE'], ['color', 'BLACK'], ['posn', 'upper left']]
4763 """
4765 def __init__(self, expr: Union[ParserElement, str]):
4766 super().__init__(expr)
4767 self.mayReturnEmpty = True
4769 def parseImpl(self, instring, loc, doActions=True):
4770 # by using self._expr.parse and deleting the contents of the returned ParseResults list
4771 # we keep any named results that were defined in the FollowedBy expression
4772 _, ret = self.expr._parse(instring, loc, doActions=doActions)
4773 del ret[:]
4775 return loc, ret
4778class PrecededBy(ParseElementEnhance):
4779 """Lookbehind matching of the given parse expression.
4780 ``PrecededBy`` does not advance the parsing position within the
4781 input string, it only verifies that the specified parse expression
4782 matches prior to the current position. ``PrecededBy`` always
4783 returns a null token list, but if a results name is defined on the
4784 given expression, it is returned.
4786 Parameters:
4788 - ``expr`` - expression that must match prior to the current parse
4789 location
4790 - ``retreat`` - (default= ``None``) - (int) maximum number of characters
4791 to lookbehind prior to the current parse location
4793 If the lookbehind expression is a string, :class:`Literal`,
4794 :class:`Keyword`, or a :class:`Word` or :class:`CharsNotIn`
4795 with a specified exact or maximum length, then the retreat
4796 parameter is not required. Otherwise, retreat must be specified to
4797 give a maximum number of characters to look back from
4798 the current parse position for a lookbehind match.
4800 Example::
4802 # VB-style variable names with type prefixes
4803 int_var = PrecededBy("#") + pyparsing_common.identifier
4804 str_var = PrecededBy("$") + pyparsing_common.identifier
4806 """
4808 def __init__(
4809 self, expr: Union[ParserElement, str], retreat: typing.Optional[int] = None
4810 ):
4811 super().__init__(expr)
4812 self.expr = self.expr().leave_whitespace()
4813 self.mayReturnEmpty = True
4814 self.mayIndexError = False
4815 self.exact = False
4816 if isinstance(expr, str_type):
4817 expr = typing.cast(str, expr)
4818 retreat = len(expr)
4819 self.exact = True
4820 elif isinstance(expr, (Literal, Keyword)):
4821 retreat = expr.matchLen
4822 self.exact = True
4823 elif isinstance(expr, (Word, CharsNotIn)) and expr.maxLen != _MAX_INT:
4824 retreat = expr.maxLen
4825 self.exact = True
4826 elif isinstance(expr, PositionToken):
4827 retreat = 0
4828 self.exact = True
4829 self.retreat = retreat
4830 self.errmsg = "not preceded by " + str(expr)
4831 self.skipWhitespace = False
4832 self.parseAction.append(lambda s, l, t: t.__delitem__(slice(None, None)))
4834 def parseImpl(self, instring, loc=0, doActions=True):
4835 if self.exact:
4836 if loc < self.retreat:
4837 raise ParseException(instring, loc, self.errmsg)
4838 start = loc - self.retreat
4839 _, ret = self.expr._parse(instring, start)
4840 else:
4841 # retreat specified a maximum lookbehind window, iterate
4842 test_expr = self.expr + StringEnd()
4843 instring_slice = instring[max(0, loc - self.retreat) : loc]
4844 last_expr = ParseException(instring, loc, self.errmsg)
4845 for offset in range(1, min(loc, self.retreat + 1) + 1):
4846 try:
4847 # print('trying', offset, instring_slice, repr(instring_slice[loc - offset:]))
4848 _, ret = test_expr._parse(
4849 instring_slice, len(instring_slice) - offset
4850 )
4851 except ParseBaseException as pbe:
4852 last_expr = pbe
4853 else:
4854 break
4855 else:
4856 raise last_expr
4857 return loc, ret
4860class Located(ParseElementEnhance):
4861 """
4862 Decorates a returned token with its starting and ending
4863 locations in the input string.
4865 This helper adds the following results names:
4867 - ``locn_start`` - location where matched expression begins
4868 - ``locn_end`` - location where matched expression ends
4869 - ``value`` - the actual parsed results
4871 Be careful if the input text contains ``<TAB>`` characters, you
4872 may want to call :class:`ParserElement.parse_with_tabs`
4874 Example::
4876 wd = Word(alphas)
4877 for match in Located(wd).search_string("ljsdf123lksdjjf123lkkjj1222"):
4878 print(match)
4880 prints::
4882 [0, ['ljsdf'], 5]
4883 [8, ['lksdjjf'], 15]
4884 [18, ['lkkjj'], 23]
4886 """
4888 def parseImpl(self, instring, loc, doActions=True):
4889 start = loc
4890 loc, tokens = self.expr._parse(instring, start, doActions, callPreParse=False)
4891 ret_tokens = ParseResults([start, tokens, loc])
4892 ret_tokens["locn_start"] = start
4893 ret_tokens["value"] = tokens
4894 ret_tokens["locn_end"] = loc
4895 if self.resultsName:
4896 # must return as a list, so that the name will be attached to the complete group
4897 return loc, [ret_tokens]
4898 else:
4899 return loc, ret_tokens
4902class NotAny(ParseElementEnhance):
4903 """
4904 Lookahead to disallow matching with the given parse expression.
4905 ``NotAny`` does *not* advance the parsing position within the
4906 input string, it only verifies that the specified parse expression
4907 does *not* match at the current position. Also, ``NotAny`` does
4908 *not* skip over leading whitespace. ``NotAny`` always returns
4909 a null token list. May be constructed using the ``'~'`` operator.
4911 Example::
4913 AND, OR, NOT = map(CaselessKeyword, "AND OR NOT".split())
4915 # take care not to mistake keywords for identifiers
4916 ident = ~(AND | OR | NOT) + Word(alphas)
4917 boolean_term = Opt(NOT) + ident
4919 # very crude boolean expression - to support parenthesis groups and
4920 # operation hierarchy, use infix_notation
4921 boolean_expr = boolean_term + ((AND | OR) + boolean_term)[...]
4923 # integers that are followed by "." are actually floats
4924 integer = Word(nums) + ~Char(".")
4925 """
4927 def __init__(self, expr: Union[ParserElement, str]):
4928 super().__init__(expr)
4929 # do NOT use self.leave_whitespace(), don't want to propagate to exprs
4930 # self.leave_whitespace()
4931 self.skipWhitespace = False
4933 self.mayReturnEmpty = True
4934 self.errmsg = "Found unwanted token, " + str(self.expr)
4936 def parseImpl(self, instring, loc, doActions=True):
4937 if self.expr.can_parse_next(instring, loc, do_actions=doActions):
4938 raise ParseException(instring, loc, self.errmsg, self)
4939 return loc, []
4941 def _generateDefaultName(self) -> str:
4942 return "~{" + str(self.expr) + "}"
4945class _MultipleMatch(ParseElementEnhance):
4946 def __init__(
4947 self,
4948 expr: Union[str, ParserElement],
4949 stop_on: typing.Optional[Union[ParserElement, str]] = None,
4950 *,
4951 stopOn: typing.Optional[Union[ParserElement, str]] = None,
4952 ):
4953 super().__init__(expr)
4954 stopOn = stopOn or stop_on
4955 self.saveAsList = True
4956 ender = stopOn
4957 if isinstance(ender, str_type):
4958 ender = self._literalStringClass(ender)
4959 self.stopOn(ender)
4961 def stopOn(self, ender) -> ParserElement:
4962 if isinstance(ender, str_type):
4963 ender = self._literalStringClass(ender)
4964 self.not_ender = ~ender if ender is not None else None
4965 return self
4967 def parseImpl(self, instring, loc, doActions=True):
4968 self_expr_parse = self.expr._parse
4969 self_skip_ignorables = self._skipIgnorables
4970 check_ender = self.not_ender is not None
4971 if check_ender:
4972 try_not_ender = self.not_ender.try_parse
4974 # must be at least one (but first see if we are the stopOn sentinel;
4975 # if so, fail)
4976 if check_ender:
4977 try_not_ender(instring, loc)
4978 loc, tokens = self_expr_parse(instring, loc, doActions)
4979 try:
4980 hasIgnoreExprs = not not self.ignoreExprs
4981 while 1:
4982 if check_ender:
4983 try_not_ender(instring, loc)
4984 if hasIgnoreExprs:
4985 preloc = self_skip_ignorables(instring, loc)
4986 else:
4987 preloc = loc
4988 loc, tmptokens = self_expr_parse(instring, preloc, doActions)
4989 tokens += tmptokens
4990 except (ParseException, IndexError):
4991 pass
4993 return loc, tokens
4995 def _setResultsName(self, name, listAllMatches=False):
4996 if (
4997 __diag__.warn_ungrouped_named_tokens_in_collection
4998 and Diagnostics.warn_ungrouped_named_tokens_in_collection
4999 not in self.suppress_warnings_
5000 ):
5001 for e in [self.expr] + self.expr.recurse():
5002 if (
5003 isinstance(e, ParserElement)
5004 and e.resultsName
5005 and Diagnostics.warn_ungrouped_named_tokens_in_collection
5006 not in e.suppress_warnings_
5007 ):
5008 warnings.warn(
5009 "{}: setting results name {!r} on {} expression "
5010 "collides with {!r} on contained expression".format(
5011 "warn_ungrouped_named_tokens_in_collection",
5012 name,
5013 type(self).__name__,
5014 e.resultsName,
5015 ),
5016 stacklevel=3,
5017 )
5019 return super()._setResultsName(name, listAllMatches)
5022class OneOrMore(_MultipleMatch):
5023 """
5024 Repetition of one or more of the given expression.
5026 Parameters:
5028 - ``expr`` - expression that must match one or more times
5029 - ``stop_on`` - (default= ``None``) - expression for a terminating sentinel
5030 (only required if the sentinel would ordinarily match the repetition
5031 expression)
5033 Example::
5035 data_word = Word(alphas)
5036 label = data_word + FollowedBy(':')
5037 attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).set_parse_action(' '.join))
5039 text = "shape: SQUARE posn: upper left color: BLACK"
5040 attr_expr[1, ...].parse_string(text).pprint() # Fail! read 'color' as data instead of next label -> [['shape', 'SQUARE color']]
5042 # use stop_on attribute for OneOrMore to avoid reading label string as part of the data
5043 attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stop_on=label).set_parse_action(' '.join))
5044 OneOrMore(attr_expr).parse_string(text).pprint() # Better -> [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'BLACK']]
5046 # could also be written as
5047 (attr_expr * (1,)).parse_string(text).pprint()
5048 """
5050 def _generateDefaultName(self) -> str:
5051 return "{" + str(self.expr) + "}..."
5054class ZeroOrMore(_MultipleMatch):
5055 """
5056 Optional repetition of zero or more of the given expression.
5058 Parameters:
5060 - ``expr`` - expression that must match zero or more times
5061 - ``stop_on`` - expression for a terminating sentinel
5062 (only required if the sentinel would ordinarily match the repetition
5063 expression) - (default= ``None``)
5065 Example: similar to :class:`OneOrMore`
5066 """
5068 def __init__(
5069 self,
5070 expr: Union[str, ParserElement],
5071 stop_on: typing.Optional[Union[ParserElement, str]] = None,
5072 *,
5073 stopOn: typing.Optional[Union[ParserElement, str]] = None,
5074 ):
5075 super().__init__(expr, stopOn=stopOn or stop_on)
5076 self.mayReturnEmpty = True
5078 def parseImpl(self, instring, loc, doActions=True):
5079 try:
5080 return super().parseImpl(instring, loc, doActions)
5081 except (ParseException, IndexError):
5082 return loc, ParseResults([], name=self.resultsName)
5084 def _generateDefaultName(self) -> str:
5085 return "[" + str(self.expr) + "]..."
5088class DelimitedList(ParseElementEnhance):
5089 def __init__(
5090 self,
5091 expr: Union[str, ParserElement],
5092 delim: Union[str, ParserElement] = ",",
5093 combine: bool = False,
5094 min: typing.Optional[int] = None,
5095 max: typing.Optional[int] = None,
5096 *,
5097 allow_trailing_delim: bool = False,
5098 ):
5099 """Helper to define a delimited list of expressions - the delimiter
5100 defaults to ','. By default, the list elements and delimiters can
5101 have intervening whitespace, and comments, but this can be
5102 overridden by passing ``combine=True`` in the constructor. If
5103 ``combine`` is set to ``True``, the matching tokens are
5104 returned as a single token string, with the delimiters included;
5105 otherwise, the matching tokens are returned as a list of tokens,
5106 with the delimiters suppressed.
5108 If ``allow_trailing_delim`` is set to True, then the list may end with
5109 a delimiter.
5111 Example::
5113 DelimitedList(Word(alphas)).parse_string("aa,bb,cc") # -> ['aa', 'bb', 'cc']
5114 DelimitedList(Word(hexnums), delim=':', combine=True).parse_string("AA:BB:CC:DD:EE") # -> ['AA:BB:CC:DD:EE']
5115 """
5116 if isinstance(expr, str_type):
5117 expr = ParserElement._literalStringClass(expr)
5118 expr = typing.cast(ParserElement, expr)
5120 if min is not None:
5121 if min < 1:
5122 raise ValueError("min must be greater than 0")
5123 if max is not None:
5124 if min is not None and max < min:
5125 raise ValueError("max must be greater than, or equal to min")
5127 self.content = expr
5128 self.raw_delim = str(delim)
5129 self.delim = delim
5130 self.combine = combine
5131 if not combine:
5132 self.delim = Suppress(delim)
5133 self.min = min or 1
5134 self.max = max
5135 self.allow_trailing_delim = allow_trailing_delim
5137 delim_list_expr = self.content + (self.delim + self.content) * (
5138 self.min - 1,
5139 None if self.max is None else self.max - 1,
5140 )
5141 if self.allow_trailing_delim:
5142 delim_list_expr += Opt(self.delim)
5144 if self.combine:
5145 delim_list_expr = Combine(delim_list_expr)
5147 super().__init__(delim_list_expr, savelist=True)
5149 def _generateDefaultName(self) -> str:
5150 return "{0} [{1} {0}]...".format(self.content.streamline(), self.raw_delim)
5153class _NullToken:
5154 def __bool__(self):
5155 return False
5157 def __str__(self):
5158 return ""
5161class Opt(ParseElementEnhance):
5162 """
5163 Optional matching of the given expression.
5165 Parameters:
5167 - ``expr`` - expression that must match zero or more times
5168 - ``default`` (optional) - value to be returned if the optional expression is not found.
5170 Example::
5172 # US postal code can be a 5-digit zip, plus optional 4-digit qualifier
5173 zip = Combine(Word(nums, exact=5) + Opt('-' + Word(nums, exact=4)))
5174 zip.run_tests('''
5175 # traditional ZIP code
5176 12345
5178 # ZIP+4 form
5179 12101-0001
5181 # invalid ZIP
5182 98765-
5183 ''')
5185 prints::
5187 # traditional ZIP code
5188 12345
5189 ['12345']
5191 # ZIP+4 form
5192 12101-0001
5193 ['12101-0001']
5195 # invalid ZIP
5196 98765-
5197 ^
5198 FAIL: Expected end of text (at char 5), (line:1, col:6)
5199 """
5201 __optionalNotMatched = _NullToken()
5203 def __init__(
5204 self, expr: Union[ParserElement, str], default: Any = __optionalNotMatched
5205 ):
5206 super().__init__(expr, savelist=False)
5207 self.saveAsList = self.expr.saveAsList
5208 self.defaultValue = default
5209 self.mayReturnEmpty = True
5211 def parseImpl(self, instring, loc, doActions=True):
5212 self_expr = self.expr
5213 try:
5214 loc, tokens = self_expr._parse(instring, loc, doActions, callPreParse=False)
5215 except (ParseException, IndexError):
5216 default_value = self.defaultValue
5217 if default_value is not self.__optionalNotMatched:
5218 if self_expr.resultsName:
5219 tokens = ParseResults([default_value])
5220 tokens[self_expr.resultsName] = default_value
5221 else:
5222 tokens = [default_value]
5223 else:
5224 tokens = []
5225 return loc, tokens
5227 def _generateDefaultName(self) -> str:
5228 inner = str(self.expr)
5229 # strip off redundant inner {}'s
5230 while len(inner) > 1 and inner[0 :: len(inner) - 1] == "{}":
5231 inner = inner[1:-1]
5232 return "[" + inner + "]"
5235Optional = Opt
5238class SkipTo(ParseElementEnhance):
5239 """
5240 Token for skipping over all undefined text until the matched
5241 expression is found.
5243 Parameters:
5245 - ``expr`` - target expression marking the end of the data to be skipped
5246 - ``include`` - if ``True``, the target expression is also parsed
5247 (the skipped text and target expression are returned as a 2-element
5248 list) (default= ``False``).
5249 - ``ignore`` - (default= ``None``) used to define grammars (typically quoted strings and
5250 comments) that might contain false matches to the target expression
5251 - ``fail_on`` - (default= ``None``) define expressions that are not allowed to be
5252 included in the skipped test; if found before the target expression is found,
5253 the :class:`SkipTo` is not a match
5255 Example::
5257 report = '''
5258 Outstanding Issues Report - 1 Jan 2000
5260 # | Severity | Description | Days Open
5261 -----+----------+-------------------------------------------+-----------
5262 101 | Critical | Intermittent system crash | 6
5263 94 | Cosmetic | Spelling error on Login ('log|n') | 14
5264 79 | Minor | System slow when running too many reports | 47
5265 '''
5266 integer = Word(nums)
5267 SEP = Suppress('|')
5268 # use SkipTo to simply match everything up until the next SEP
5269 # - ignore quoted strings, so that a '|' character inside a quoted string does not match
5270 # - parse action will call token.strip() for each matched token, i.e., the description body
5271 string_data = SkipTo(SEP, ignore=quoted_string)
5272 string_data.set_parse_action(token_map(str.strip))
5273 ticket_expr = (integer("issue_num") + SEP
5274 + string_data("sev") + SEP
5275 + string_data("desc") + SEP
5276 + integer("days_open"))
5278 for tkt in ticket_expr.search_string(report):
5279 print tkt.dump()
5281 prints::
5283 ['101', 'Critical', 'Intermittent system crash', '6']
5284 - days_open: '6'
5285 - desc: 'Intermittent system crash'
5286 - issue_num: '101'
5287 - sev: 'Critical'
5288 ['94', 'Cosmetic', "Spelling error on Login ('log|n')", '14']
5289 - days_open: '14'
5290 - desc: "Spelling error on Login ('log|n')"
5291 - issue_num: '94'
5292 - sev: 'Cosmetic'
5293 ['79', 'Minor', 'System slow when running too many reports', '47']
5294 - days_open: '47'
5295 - desc: 'System slow when running too many reports'
5296 - issue_num: '79'
5297 - sev: 'Minor'
5298 """
5300 def __init__(
5301 self,
5302 other: Union[ParserElement, str],
5303 include: bool = False,
5304 ignore: typing.Optional[Union[ParserElement, str]] = None,
5305 fail_on: typing.Optional[Union[ParserElement, str]] = None,
5306 *,
5307 failOn: typing.Optional[Union[ParserElement, str]] = None,
5308 ):
5309 super().__init__(other)
5310 failOn = failOn or fail_on
5311 if ignore is not None:
5312 self.ignore(ignore)
5313 self.mayReturnEmpty = True
5314 self.mayIndexError = False
5315 self.includeMatch = include
5316 self.saveAsList = False
5317 if isinstance(failOn, str_type):
5318 self.failOn = self._literalStringClass(failOn)
5319 else:
5320 self.failOn = failOn
5321 self.errmsg = "No match found for " + str(self.expr)
5323 def parseImpl(self, instring, loc, doActions=True):
5324 startloc = loc
5325 instrlen = len(instring)
5326 self_expr_parse = self.expr._parse
5327 self_failOn_canParseNext = (
5328 self.failOn.canParseNext if self.failOn is not None else None
5329 )
5330 self_preParse = self.preParse if self.callPreparse else None
5332 tmploc = loc
5333 while tmploc <= instrlen:
5334 if self_failOn_canParseNext is not None:
5335 # break if failOn expression matches
5336 if self_failOn_canParseNext(instring, tmploc):
5337 break
5339 if self_preParse is not None:
5340 # skip grammar-ignored expressions
5341 tmploc = self_preParse(instring, tmploc)
5343 try:
5344 self_expr_parse(instring, tmploc, doActions=False, callPreParse=False)
5345 except (ParseException, IndexError):
5346 # no match, advance loc in string
5347 tmploc += 1
5348 else:
5349 # matched skipto expr, done
5350 break
5352 else:
5353 # ran off the end of the input string without matching skipto expr, fail
5354 raise ParseException(instring, loc, self.errmsg, self)
5356 # build up return values
5357 loc = tmploc
5358 skiptext = instring[startloc:loc]
5359 skipresult = ParseResults(skiptext)
5361 if self.includeMatch:
5362 loc, mat = self_expr_parse(instring, loc, doActions, callPreParse=False)
5363 skipresult += mat
5365 return loc, skipresult
5368class Forward(ParseElementEnhance):
5369 """
5370 Forward declaration of an expression to be defined later -
5371 used for recursive grammars, such as algebraic infix notation.
5372 When the expression is known, it is assigned to the ``Forward``
5373 variable using the ``'<<'`` operator.
5375 Note: take care when assigning to ``Forward`` not to overlook
5376 precedence of operators.
5378 Specifically, ``'|'`` has a lower precedence than ``'<<'``, so that::
5380 fwd_expr << a | b | c
5382 will actually be evaluated as::
5384 (fwd_expr << a) | b | c
5386 thereby leaving b and c out as parseable alternatives. It is recommended that you
5387 explicitly group the values inserted into the ``Forward``::
5389 fwd_expr << (a | b | c)
5391 Converting to use the ``'<<='`` operator instead will avoid this problem.
5393 See :class:`ParseResults.pprint` for an example of a recursive
5394 parser created using ``Forward``.
5395 """
5397 def __init__(self, other: typing.Optional[Union[ParserElement, str]] = None):
5398 self.caller_frame = traceback.extract_stack(limit=2)[0]
5399 super().__init__(other, savelist=False) # type: ignore[arg-type]
5400 self.lshift_line = None
5402 def __lshift__(self, other) -> "Forward":
5403 if hasattr(self, "caller_frame"):
5404 del self.caller_frame
5405 if isinstance(other, str_type):
5406 other = self._literalStringClass(other)
5408 if not isinstance(other, ParserElement):
5409 return NotImplemented
5411 self.expr = other
5412 self.streamlined = other.streamlined
5413 self.mayIndexError = self.expr.mayIndexError
5414 self.mayReturnEmpty = self.expr.mayReturnEmpty
5415 self.set_whitespace_chars(
5416 self.expr.whiteChars, copy_defaults=self.expr.copyDefaultWhiteChars
5417 )
5418 self.skipWhitespace = self.expr.skipWhitespace
5419 self.saveAsList = self.expr.saveAsList
5420 self.ignoreExprs.extend(self.expr.ignoreExprs)
5421 self.lshift_line = traceback.extract_stack(limit=2)[-2] # type: ignore[assignment]
5422 return self
5424 def __ilshift__(self, other) -> "Forward":
5425 if not isinstance(other, ParserElement):
5426 return NotImplemented
5428 return self << other
5430 def __or__(self, other) -> "ParserElement":
5431 caller_line = traceback.extract_stack(limit=2)[-2]
5432 if (
5433 __diag__.warn_on_match_first_with_lshift_operator
5434 and caller_line == self.lshift_line
5435 and Diagnostics.warn_on_match_first_with_lshift_operator
5436 not in self.suppress_warnings_
5437 ):
5438 warnings.warn(
5439 "using '<<' operator with '|' is probably an error, use '<<='",
5440 stacklevel=2,
5441 )
5442 ret = super().__or__(other)
5443 return ret
5445 def __del__(self):
5446 # see if we are getting dropped because of '=' reassignment of var instead of '<<=' or '<<'
5447 if (
5448 self.expr is None
5449 and __diag__.warn_on_assignment_to_Forward
5450 and Diagnostics.warn_on_assignment_to_Forward not in self.suppress_warnings_
5451 ):
5452 warnings.warn_explicit(
5453 "Forward defined here but no expression attached later using '<<=' or '<<'",
5454 UserWarning,
5455 filename=self.caller_frame.filename,
5456 lineno=self.caller_frame.lineno,
5457 )
5459 def parseImpl(self, instring, loc, doActions=True):
5460 if (
5461 self.expr is None
5462 and __diag__.warn_on_parse_using_empty_Forward
5463 and Diagnostics.warn_on_parse_using_empty_Forward
5464 not in self.suppress_warnings_
5465 ):
5466 # walk stack until parse_string, scan_string, search_string, or transform_string is found
5467 parse_fns = (
5468 "parse_string",
5469 "scan_string",
5470 "search_string",
5471 "transform_string",
5472 )
5473 tb = traceback.extract_stack(limit=200)
5474 for i, frm in enumerate(reversed(tb), start=1):
5475 if frm.name in parse_fns:
5476 stacklevel = i + 1
5477 break
5478 else:
5479 stacklevel = 2
5480 warnings.warn(
5481 "Forward expression was never assigned a value, will not parse any input",
5482 stacklevel=stacklevel,
5483 )
5484 if not ParserElement._left_recursion_enabled:
5485 return super().parseImpl(instring, loc, doActions)
5486 # ## Bounded Recursion algorithm ##
5487 # Recursion only needs to be processed at ``Forward`` elements, since they are
5488 # the only ones that can actually refer to themselves. The general idea is
5489 # to handle recursion stepwise: We start at no recursion, then recurse once,
5490 # recurse twice, ..., until more recursion offers no benefit (we hit the bound).
5491 #
5492 # The "trick" here is that each ``Forward`` gets evaluated in two contexts
5493 # - to *match* a specific recursion level, and
5494 # - to *search* the bounded recursion level
5495 # and the two run concurrently. The *search* must *match* each recursion level
5496 # to find the best possible match. This is handled by a memo table, which
5497 # provides the previous match to the next level match attempt.
5498 #
5499 # See also "Left Recursion in Parsing Expression Grammars", Medeiros et al.
5500 #
5501 # There is a complication since we not only *parse* but also *transform* via
5502 # actions: We do not want to run the actions too often while expanding. Thus,
5503 # we expand using `doActions=False` and only run `doActions=True` if the next
5504 # recursion level is acceptable.
5505 with ParserElement.recursion_lock:
5506 memo = ParserElement.recursion_memos
5507 try:
5508 # we are parsing at a specific recursion expansion - use it as-is
5509 prev_loc, prev_result = memo[loc, self, doActions]
5510 if isinstance(prev_result, Exception):
5511 raise prev_result
5512 return prev_loc, prev_result.copy()
5513 except KeyError:
5514 act_key = (loc, self, True)
5515 peek_key = (loc, self, False)
5516 # we are searching for the best recursion expansion - keep on improving
5517 # both `doActions` cases must be tracked separately here!
5518 prev_loc, prev_peek = memo[peek_key] = (
5519 loc - 1,
5520 ParseException(
5521 instring, loc, "Forward recursion without base case", self
5522 ),
5523 )
5524 if doActions:
5525 memo[act_key] = memo[peek_key]
5526 while True:
5527 try:
5528 new_loc, new_peek = super().parseImpl(instring, loc, False)
5529 except ParseException:
5530 # we failed before getting any match – do not hide the error
5531 if isinstance(prev_peek, Exception):
5532 raise
5533 new_loc, new_peek = prev_loc, prev_peek
5534 # the match did not get better: we are done
5535 if new_loc <= prev_loc:
5536 if doActions:
5537 # replace the match for doActions=False as well,
5538 # in case the action did backtrack
5539 prev_loc, prev_result = memo[peek_key] = memo[act_key]
5540 del memo[peek_key], memo[act_key]
5541 return prev_loc, prev_result.copy()
5542 del memo[peek_key]
5543 return prev_loc, prev_peek.copy()
5544 # the match did get better: see if we can improve further
5545 else:
5546 if doActions:
5547 try:
5548 memo[act_key] = super().parseImpl(instring, loc, True)
5549 except ParseException as e:
5550 memo[peek_key] = memo[act_key] = (new_loc, e)
5551 raise
5552 prev_loc, prev_peek = memo[peek_key] = new_loc, new_peek
5554 def leave_whitespace(self, recursive: bool = True) -> ParserElement:
5555 self.skipWhitespace = False
5556 return self
5558 def ignore_whitespace(self, recursive: bool = True) -> ParserElement:
5559 self.skipWhitespace = True
5560 return self
5562 def streamline(self) -> ParserElement:
5563 if not self.streamlined:
5564 self.streamlined = True
5565 if self.expr is not None:
5566 self.expr.streamline()
5567 return self
5569 def validate(self, validateTrace=None) -> None:
5570 warnings.warn(
5571 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
5572 DeprecationWarning,
5573 stacklevel=2,
5574 )
5575 if validateTrace is None:
5576 validateTrace = []
5578 if self not in validateTrace:
5579 tmp = validateTrace[:] + [self]
5580 if self.expr is not None:
5581 self.expr.validate(tmp)
5582 self._checkRecursion([])
5584 def _generateDefaultName(self) -> str:
5585 # Avoid infinite recursion by setting a temporary _defaultName
5586 self._defaultName = ": ..."
5588 # Use the string representation of main expression.
5589 retString = "..."
5590 try:
5591 if self.expr is not None:
5592 retString = str(self.expr)[:1000]
5593 else:
5594 retString = "None"
5595 finally:
5596 return self.__class__.__name__ + ": " + retString
5598 def copy(self) -> ParserElement:
5599 if self.expr is not None:
5600 return super().copy()
5601 else:
5602 ret = Forward()
5603 ret <<= self
5604 return ret
5606 def _setResultsName(self, name, list_all_matches=False):
5607 if (
5608 __diag__.warn_name_set_on_empty_Forward
5609 and Diagnostics.warn_name_set_on_empty_Forward
5610 not in self.suppress_warnings_
5611 ):
5612 if self.expr is None:
5613 warnings.warn(
5614 "{}: setting results name {!r} on {} expression "
5615 "that has no contained expression".format(
5616 "warn_name_set_on_empty_Forward", name, type(self).__name__
5617 ),
5618 stacklevel=3,
5619 )
5621 return super()._setResultsName(name, list_all_matches)
5623 # Compatibility synonyms
5624 # fmt: off
5625 @replaced_by_pep8(leave_whitespace)
5626 def leaveWhitespace(self): ...
5628 @replaced_by_pep8(ignore_whitespace)
5629 def ignoreWhitespace(self): ...
5630 # fmt: on
5633class TokenConverter(ParseElementEnhance):
5634 """
5635 Abstract subclass of :class:`ParseExpression`, for converting parsed results.
5636 """
5638 def __init__(self, expr: Union[ParserElement, str], savelist=False):
5639 super().__init__(expr) # , savelist)
5640 self.saveAsList = False
5643class Combine(TokenConverter):
5644 """Converter to concatenate all matching tokens to a single string.
5645 By default, the matching patterns must also be contiguous in the
5646 input string; this can be disabled by specifying
5647 ``'adjacent=False'`` in the constructor.
5649 Example::
5651 real = Word(nums) + '.' + Word(nums)
5652 print(real.parse_string('3.1416')) # -> ['3', '.', '1416']
5653 # will also erroneously match the following
5654 print(real.parse_string('3. 1416')) # -> ['3', '.', '1416']
5656 real = Combine(Word(nums) + '.' + Word(nums))
5657 print(real.parse_string('3.1416')) # -> ['3.1416']
5658 # no match when there are internal spaces
5659 print(real.parse_string('3. 1416')) # -> Exception: Expected W:(0123...)
5660 """
5662 def __init__(
5663 self,
5664 expr: ParserElement,
5665 join_string: str = "",
5666 adjacent: bool = True,
5667 *,
5668 joinString: typing.Optional[str] = None,
5669 ):
5670 super().__init__(expr)
5671 joinString = joinString if joinString is not None else join_string
5672 # suppress whitespace-stripping in contained parse expressions, but re-enable it on the Combine itself
5673 if adjacent:
5674 self.leave_whitespace()
5675 self.adjacent = adjacent
5676 self.skipWhitespace = True
5677 self.joinString = joinString
5678 self.callPreparse = True
5680 def ignore(self, other) -> ParserElement:
5681 if self.adjacent:
5682 ParserElement.ignore(self, other)
5683 else:
5684 super().ignore(other)
5685 return self
5687 def postParse(self, instring, loc, tokenlist):
5688 retToks = tokenlist.copy()
5689 del retToks[:]
5690 retToks += ParseResults(
5691 ["".join(tokenlist._asStringList(self.joinString))], modal=self.modalResults
5692 )
5694 if self.resultsName and retToks.haskeys():
5695 return [retToks]
5696 else:
5697 return retToks
5700class Group(TokenConverter):
5701 """Converter to return the matched tokens as a list - useful for
5702 returning tokens of :class:`ZeroOrMore` and :class:`OneOrMore` expressions.
5704 The optional ``aslist`` argument when set to True will return the
5705 parsed tokens as a Python list instead of a pyparsing ParseResults.
5707 Example::
5709 ident = Word(alphas)
5710 num = Word(nums)
5711 term = ident | num
5712 func = ident + Opt(DelimitedList(term))
5713 print(func.parse_string("fn a, b, 100"))
5714 # -> ['fn', 'a', 'b', '100']
5716 func = ident + Group(Opt(DelimitedList(term)))
5717 print(func.parse_string("fn a, b, 100"))
5718 # -> ['fn', ['a', 'b', '100']]
5719 """
5721 def __init__(self, expr: ParserElement, aslist: bool = False):
5722 super().__init__(expr)
5723 self.saveAsList = True
5724 self._asPythonList = aslist
5726 def postParse(self, instring, loc, tokenlist):
5727 if self._asPythonList:
5728 return ParseResults.List(
5729 tokenlist.asList()
5730 if isinstance(tokenlist, ParseResults)
5731 else list(tokenlist)
5732 )
5733 else:
5734 return [tokenlist]
5737class Dict(TokenConverter):
5738 """Converter to return a repetitive expression as a list, but also
5739 as a dictionary. Each element can also be referenced using the first
5740 token in the expression as its key. Useful for tabular report
5741 scraping when the first column can be used as a item key.
5743 The optional ``asdict`` argument when set to True will return the
5744 parsed tokens as a Python dict instead of a pyparsing ParseResults.
5746 Example::
5748 data_word = Word(alphas)
5749 label = data_word + FollowedBy(':')
5751 text = "shape: SQUARE posn: upper left color: light blue texture: burlap"
5752 attr_expr = (label + Suppress(':') + OneOrMore(data_word, stop_on=label).set_parse_action(' '.join))
5754 # print attributes as plain groups
5755 print(attr_expr[1, ...].parse_string(text).dump())
5757 # instead of OneOrMore(expr), parse using Dict(Group(expr)[1, ...]) - Dict will auto-assign names
5758 result = Dict(Group(attr_expr)[1, ...]).parse_string(text)
5759 print(result.dump())
5761 # access named fields as dict entries, or output as dict
5762 print(result['shape'])
5763 print(result.as_dict())
5765 prints::
5767 ['shape', 'SQUARE', 'posn', 'upper left', 'color', 'light blue', 'texture', 'burlap']
5768 [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']]
5769 - color: 'light blue'
5770 - posn: 'upper left'
5771 - shape: 'SQUARE'
5772 - texture: 'burlap'
5773 SQUARE
5774 {'color': 'light blue', 'posn': 'upper left', 'texture': 'burlap', 'shape': 'SQUARE'}
5776 See more examples at :class:`ParseResults` of accessing fields by results name.
5777 """
5779 def __init__(self, expr: ParserElement, asdict: bool = False):
5780 super().__init__(expr)
5781 self.saveAsList = True
5782 self._asPythonDict = asdict
5784 def postParse(self, instring, loc, tokenlist):
5785 for i, tok in enumerate(tokenlist):
5786 if len(tok) == 0:
5787 continue
5789 ikey = tok[0]
5790 if isinstance(ikey, int):
5791 ikey = str(ikey).strip()
5793 if len(tok) == 1:
5794 tokenlist[ikey] = _ParseResultsWithOffset("", i)
5796 elif len(tok) == 2 and not isinstance(tok[1], ParseResults):
5797 tokenlist[ikey] = _ParseResultsWithOffset(tok[1], i)
5799 else:
5800 try:
5801 dictvalue = tok.copy() # ParseResults(i)
5802 except Exception:
5803 exc = TypeError(
5804 "could not extract dict values from parsed results"
5805 " - Dict expression must contain Grouped expressions"
5806 )
5807 raise exc from None
5809 del dictvalue[0]
5811 if len(dictvalue) != 1 or (
5812 isinstance(dictvalue, ParseResults) and dictvalue.haskeys()
5813 ):
5814 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue, i)
5815 else:
5816 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue[0], i)
5818 if self._asPythonDict:
5819 return [tokenlist.as_dict()] if self.resultsName else tokenlist.as_dict()
5820 else:
5821 return [tokenlist] if self.resultsName else tokenlist
5824class Suppress(TokenConverter):
5825 """Converter for ignoring the results of a parsed expression.
5827 Example::
5829 source = "a, b, c,d"
5830 wd = Word(alphas)
5831 wd_list1 = wd + (',' + wd)[...]
5832 print(wd_list1.parse_string(source))
5834 # often, delimiters that are useful during parsing are just in the
5835 # way afterward - use Suppress to keep them out of the parsed output
5836 wd_list2 = wd + (Suppress(',') + wd)[...]
5837 print(wd_list2.parse_string(source))
5839 # Skipped text (using '...') can be suppressed as well
5840 source = "lead in START relevant text END trailing text"
5841 start_marker = Keyword("START")
5842 end_marker = Keyword("END")
5843 find_body = Suppress(...) + start_marker + ... + end_marker
5844 print(find_body.parse_string(source)
5846 prints::
5848 ['a', ',', 'b', ',', 'c', ',', 'd']
5849 ['a', 'b', 'c', 'd']
5850 ['START', 'relevant text ', 'END']
5852 (See also :class:`DelimitedList`.)
5853 """
5855 def __init__(self, expr: Union[ParserElement, str], savelist: bool = False):
5856 if expr is ...:
5857 expr = _PendingSkip(NoMatch())
5858 super().__init__(expr)
5860 def __add__(self, other) -> "ParserElement":
5861 if isinstance(self.expr, _PendingSkip):
5862 return Suppress(SkipTo(other)) + other
5863 else:
5864 return super().__add__(other)
5866 def __sub__(self, other) -> "ParserElement":
5867 if isinstance(self.expr, _PendingSkip):
5868 return Suppress(SkipTo(other)) - other
5869 else:
5870 return super().__sub__(other)
5872 def postParse(self, instring, loc, tokenlist):
5873 return []
5875 def suppress(self) -> ParserElement:
5876 return self
5879def trace_parse_action(f: ParseAction) -> ParseAction:
5880 """Decorator for debugging parse actions.
5882 When the parse action is called, this decorator will print
5883 ``">> entering method-name(line:<current_source_line>, <parse_location>, <matched_tokens>)"``.
5884 When the parse action completes, the decorator will print
5885 ``"<<"`` followed by the returned value, or any exception that the parse action raised.
5887 Example::
5889 wd = Word(alphas)
5891 @trace_parse_action
5892 def remove_duplicate_chars(tokens):
5893 return ''.join(sorted(set(''.join(tokens))))
5895 wds = wd[1, ...].set_parse_action(remove_duplicate_chars)
5896 print(wds.parse_string("slkdjs sld sldd sdlf sdljf"))
5898 prints::
5900 >>entering remove_duplicate_chars(line: 'slkdjs sld sldd sdlf sdljf', 0, (['slkdjs', 'sld', 'sldd', 'sdlf', 'sdljf'], {}))
5901 <<leaving remove_duplicate_chars (ret: 'dfjkls')
5902 ['dfjkls']
5903 """
5904 f = _trim_arity(f)
5906 def z(*paArgs):
5907 thisFunc = f.__name__
5908 s, l, t = paArgs[-3:]
5909 if len(paArgs) > 3:
5910 thisFunc = paArgs[0].__class__.__name__ + "." + thisFunc
5911 sys.stderr.write(f">>entering {thisFunc}(line: {line(l, s)!r}, {l}, {t!r})\n")
5912 try:
5913 ret = f(*paArgs)
5914 except Exception as exc:
5915 sys.stderr.write(f"<<leaving {thisFunc} (exception: {exc})\n")
5916 raise
5917 sys.stderr.write(f"<<leaving {thisFunc} (ret: {ret!r})\n")
5918 return ret
5920 z.__name__ = f.__name__
5921 return z
5924# convenience constants for positional expressions
5925empty = Empty().set_name("empty")
5926line_start = LineStart().set_name("line_start")
5927line_end = LineEnd().set_name("line_end")
5928string_start = StringStart().set_name("string_start")
5929string_end = StringEnd().set_name("string_end")
5931_escapedPunc = Regex(r"\\[\\[\]\/\-\*\.\$\+\^\?()~ ]").set_parse_action(
5932 lambda s, l, t: t[0][1]
5933)
5934_escapedHexChar = Regex(r"\\0?[xX][0-9a-fA-F]+").set_parse_action(
5935 lambda s, l, t: chr(int(t[0].lstrip(r"\0x"), 16))
5936)
5937_escapedOctChar = Regex(r"\\0[0-7]+").set_parse_action(
5938 lambda s, l, t: chr(int(t[0][1:], 8))
5939)
5940_singleChar = (
5941 _escapedPunc | _escapedHexChar | _escapedOctChar | CharsNotIn(r"\]", exact=1)
5942)
5943_charRange = Group(_singleChar + Suppress("-") + _singleChar)
5944_reBracketExpr = (
5945 Literal("[")
5946 + Opt("^").set_results_name("negate")
5947 + Group(OneOrMore(_charRange | _singleChar)).set_results_name("body")
5948 + Literal("]")
5949)
5952def srange(s: str) -> str:
5953 r"""Helper to easily define string ranges for use in :class:`Word`
5954 construction. Borrows syntax from regexp ``'[]'`` string range
5955 definitions::
5957 srange("[0-9]") -> "0123456789"
5958 srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz"
5959 srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_"
5961 The input string must be enclosed in []'s, and the returned string
5962 is the expanded character set joined into a single string. The
5963 values enclosed in the []'s may be:
5965 - a single character
5966 - an escaped character with a leading backslash (such as ``\-``
5967 or ``\]``)
5968 - an escaped hex character with a leading ``'\x'``
5969 (``\x21``, which is a ``'!'`` character) (``\0x##``
5970 is also supported for backwards compatibility)
5971 - an escaped octal character with a leading ``'\0'``
5972 (``\041``, which is a ``'!'`` character)
5973 - a range of any of the above, separated by a dash (``'a-z'``,
5974 etc.)
5975 - any combination of the above (``'aeiouy'``,
5976 ``'a-zA-Z0-9_$'``, etc.)
5977 """
5978 _expanded = (
5979 lambda p: p
5980 if not isinstance(p, ParseResults)
5981 else "".join(chr(c) for c in range(ord(p[0]), ord(p[1]) + 1))
5982 )
5983 try:
5984 return "".join(_expanded(part) for part in _reBracketExpr.parse_string(s).body)
5985 except Exception as e:
5986 return ""
5989def token_map(func, *args) -> ParseAction:
5990 """Helper to define a parse action by mapping a function to all
5991 elements of a :class:`ParseResults` list. If any additional args are passed,
5992 they are forwarded to the given function as additional arguments
5993 after the token, as in
5994 ``hex_integer = Word(hexnums).set_parse_action(token_map(int, 16))``,
5995 which will convert the parsed data to an integer using base 16.
5997 Example (compare the last to example in :class:`ParserElement.transform_string`::
5999 hex_ints = Word(hexnums)[1, ...].set_parse_action(token_map(int, 16))
6000 hex_ints.run_tests('''
6001 00 11 22 aa FF 0a 0d 1a
6002 ''')
6004 upperword = Word(alphas).set_parse_action(token_map(str.upper))
6005 upperword[1, ...].run_tests('''
6006 my kingdom for a horse
6007 ''')
6009 wd = Word(alphas).set_parse_action(token_map(str.title))
6010 wd[1, ...].set_parse_action(' '.join).run_tests('''
6011 now is the winter of our discontent made glorious summer by this sun of york
6012 ''')
6014 prints::
6016 00 11 22 aa FF 0a 0d 1a
6017 [0, 17, 34, 170, 255, 10, 13, 26]
6019 my kingdom for a horse
6020 ['MY', 'KINGDOM', 'FOR', 'A', 'HORSE']
6022 now is the winter of our discontent made glorious summer by this sun of york
6023 ['Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York']
6024 """
6026 def pa(s, l, t):
6027 return [func(tokn, *args) for tokn in t]
6029 func_name = getattr(func, "__name__", getattr(func, "__class__").__name__)
6030 pa.__name__ = func_name
6032 return pa
6035def autoname_elements() -> None:
6036 """
6037 Utility to simplify mass-naming of parser elements, for
6038 generating railroad diagram with named subdiagrams.
6039 """
6040 calling_frame = sys._getframe().f_back
6041 if calling_frame is None:
6042 return
6043 calling_frame = typing.cast(types.FrameType, calling_frame)
6044 for name, var in calling_frame.f_locals.items():
6045 if isinstance(var, ParserElement) and not var.customName:
6046 var.set_name(name)
6049dbl_quoted_string = Combine(
6050 Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"'
6051).set_name("string enclosed in double quotes")
6053sgl_quoted_string = Combine(
6054 Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'"
6055).set_name("string enclosed in single quotes")
6057quoted_string = Combine(
6058 (Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"').set_name(
6059 "double quoted string"
6060 )
6061 | (Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'").set_name(
6062 "single quoted string"
6063 )
6064).set_name("quoted string using single or double quotes")
6066python_quoted_string = Combine(
6067 (Regex(r'"""(?:[^"\\]|""(?!")|"(?!"")|\\.)*', flags=re.MULTILINE) + '"""').set_name(
6068 "multiline double quoted string"
6069 )
6070 ^ (
6071 Regex(r"'''(?:[^'\\]|''(?!')|'(?!'')|\\.)*", flags=re.MULTILINE) + "'''"
6072 ).set_name("multiline single quoted string")
6073 ^ (Regex(r'"(?:[^"\n\r\\]|(?:\\")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"').set_name(
6074 "double quoted string"
6075 )
6076 ^ (Regex(r"'(?:[^'\n\r\\]|(?:\\')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'").set_name(
6077 "single quoted string"
6078 )
6079).set_name("Python quoted string")
6081unicode_string = Combine("u" + quoted_string.copy()).set_name("unicode string literal")
6084alphas8bit = srange(r"[\0xc0-\0xd6\0xd8-\0xf6\0xf8-\0xff]")
6085punc8bit = srange(r"[\0xa1-\0xbf\0xd7\0xf7]")
6087# build list of built-in expressions, for future reference if a global default value
6088# gets updated
6089_builtin_exprs: List[ParserElement] = [
6090 v for v in vars().values() if isinstance(v, ParserElement)
6091]
6093# backward compatibility names
6094# fmt: off
6095sglQuotedString = sgl_quoted_string
6096dblQuotedString = dbl_quoted_string
6097quotedString = quoted_string
6098unicodeString = unicode_string
6099lineStart = line_start
6100lineEnd = line_end
6101stringStart = string_start
6102stringEnd = string_end
6104@replaced_by_pep8(null_debug_action)
6105def nullDebugAction(): ...
6107@replaced_by_pep8(trace_parse_action)
6108def traceParseAction(): ...
6110@replaced_by_pep8(condition_as_parse_action)
6111def conditionAsParseAction(): ...
6113@replaced_by_pep8(token_map)
6114def tokenMap(): ...
6115# fmt: on