Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/pyparsing/core.py: 43%
Shortcuts on this page
r m x toggle line displays
j k next/prev highlighted chunk
0 (zero) top of page
1 (one) first highlighted chunk
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)
56if sys.version_info >= (3, 7):
57 _RePattern = re.Pattern
58else:
59 _RePattern = typing.Pattern
61#
62# Copyright (c) 2003-2022 Paul T. McGuire
63#
64# Permission is hereby granted, free of charge, to any person obtaining
65# a copy of this software and associated documentation files (the
66# "Software"), to deal in the Software without restriction, including
67# without limitation the rights to use, copy, modify, merge, publish,
68# distribute, sublicense, and/or sell copies of the Software, and to
69# permit persons to whom the Software is furnished to do so, subject to
70# the following conditions:
71#
72# The above copyright notice and this permission notice shall be
73# included in all copies or substantial portions of the Software.
74#
75# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
76# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
77# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
78# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
79# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
80# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
81# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
82#
85if sys.version_info >= (3, 8):
86 from functools import cached_property
87else:
89 class cached_property:
90 def __init__(self, func):
91 self._func = func
93 def __get__(self, instance, owner=None):
94 ret = instance.__dict__[self._func.__name__] = self._func(instance)
95 return ret
98class __compat__(__config_flags):
99 """
100 A cross-version compatibility configuration for pyparsing features that will be
101 released in a future version. By setting values in this configuration to True,
102 those features can be enabled in prior versions for compatibility development
103 and testing.
105 - ``collect_all_And_tokens`` - flag to enable fix for Issue #63 that fixes erroneous grouping
106 of results names when an :class:`And` expression is nested within an :class:`Or` or :class:`MatchFirst`;
107 maintained for compatibility, but setting to ``False`` no longer restores pre-2.3.1
108 behavior
109 """
111 _type_desc = "compatibility"
113 collect_all_And_tokens = True
115 _all_names = [__ for __ in locals() if not __.startswith("_")]
116 _fixed_names = """
117 collect_all_And_tokens
118 """.split()
121class __diag__(__config_flags):
122 _type_desc = "diagnostic"
124 warn_multiple_tokens_in_named_alternation = False
125 warn_ungrouped_named_tokens_in_collection = False
126 warn_name_set_on_empty_Forward = False
127 warn_on_parse_using_empty_Forward = False
128 warn_on_assignment_to_Forward = False
129 warn_on_multiple_string_args_to_oneof = False
130 warn_on_match_first_with_lshift_operator = False
131 enable_debug_on_named_expressions = False
133 _all_names = [__ for __ in locals() if not __.startswith("_")]
134 _warning_names = [name for name in _all_names if name.startswith("warn")]
135 _debug_names = [name for name in _all_names if name.startswith("enable_debug")]
137 @classmethod
138 def enable_all_warnings(cls) -> None:
139 for name in cls._warning_names:
140 cls.enable(name)
143class Diagnostics(Enum):
144 """
145 Diagnostic configuration (all default to disabled)
147 - ``warn_multiple_tokens_in_named_alternation`` - flag to enable warnings when a results
148 name is defined on a :class:`MatchFirst` or :class:`Or` expression with one or more :class:`And` subexpressions
149 - ``warn_ungrouped_named_tokens_in_collection`` - flag to enable warnings when a results
150 name is defined on a containing expression with ungrouped subexpressions that also
151 have results names
152 - ``warn_name_set_on_empty_Forward`` - flag to enable warnings when a :class:`Forward` is defined
153 with a results name, but has no contents defined
154 - ``warn_on_parse_using_empty_Forward`` - flag to enable warnings when a :class:`Forward` is
155 defined in a grammar but has never had an expression attached to it
156 - ``warn_on_assignment_to_Forward`` - flag to enable warnings when a :class:`Forward` is defined
157 but is overwritten by assigning using ``'='`` instead of ``'<<='`` or ``'<<'``
158 - ``warn_on_multiple_string_args_to_oneof`` - flag to enable warnings when :class:`one_of` is
159 incorrectly called with multiple str arguments
160 - ``enable_debug_on_named_expressions`` - flag to auto-enable debug on all subsequent
161 calls to :class:`ParserElement.set_name`
163 Diagnostics are enabled/disabled by calling :class:`enable_diag` and :class:`disable_diag`.
164 All warnings can be enabled by calling :class:`enable_all_warnings`.
165 """
167 warn_multiple_tokens_in_named_alternation = 0
168 warn_ungrouped_named_tokens_in_collection = 1
169 warn_name_set_on_empty_Forward = 2
170 warn_on_parse_using_empty_Forward = 3
171 warn_on_assignment_to_Forward = 4
172 warn_on_multiple_string_args_to_oneof = 5
173 warn_on_match_first_with_lshift_operator = 6
174 enable_debug_on_named_expressions = 7
177def enable_diag(diag_enum: Diagnostics) -> None:
178 """
179 Enable a global pyparsing diagnostic flag (see :class:`Diagnostics`).
180 """
181 __diag__.enable(diag_enum.name)
184def disable_diag(diag_enum: Diagnostics) -> None:
185 """
186 Disable a global pyparsing diagnostic flag (see :class:`Diagnostics`).
187 """
188 __diag__.disable(diag_enum.name)
191def enable_all_warnings() -> None:
192 """
193 Enable all global pyparsing diagnostic warnings (see :class:`Diagnostics`).
194 """
195 __diag__.enable_all_warnings()
198# hide abstract class
199del __config_flags
202def _should_enable_warnings(
203 cmd_line_warn_options: typing.Iterable[str], warn_env_var: typing.Optional[str]
204) -> bool:
205 enable = bool(warn_env_var)
206 for warn_opt in cmd_line_warn_options:
207 w_action, w_message, w_category, w_module, w_line = (warn_opt + "::::").split(
208 ":"
209 )[:5]
210 if not w_action.lower().startswith("i") and (
211 not (w_message or w_category or w_module) or w_module == "pyparsing"
212 ):
213 enable = True
214 elif w_action.lower().startswith("i") and w_module in ("pyparsing", ""):
215 enable = False
216 return enable
219if _should_enable_warnings(
220 sys.warnoptions, os.environ.get("PYPARSINGENABLEALLWARNINGS")
221):
222 enable_all_warnings()
225# build list of single arg builtins, that can be used as parse actions
226# fmt: off
227_single_arg_builtins = {
228 sum, len, sorted, reversed, list, tuple, set, any, all, min, max
229}
230# fmt: on
232_generatorType = types.GeneratorType
233ParseImplReturnType = Tuple[int, Any]
234PostParseReturnType = Union[ParseResults, Sequence[ParseResults]]
235ParseAction = Union[
236 Callable[[], Any],
237 Callable[[ParseResults], Any],
238 Callable[[int, ParseResults], Any],
239 Callable[[str, int, ParseResults], Any],
240]
241ParseCondition = Union[
242 Callable[[], bool],
243 Callable[[ParseResults], bool],
244 Callable[[int, ParseResults], bool],
245 Callable[[str, int, ParseResults], bool],
246]
247ParseFailAction = Callable[[str, int, "ParserElement", Exception], None]
248DebugStartAction = Callable[[str, int, "ParserElement", bool], None]
249DebugSuccessAction = Callable[
250 [str, int, int, "ParserElement", ParseResults, bool], None
251]
252DebugExceptionAction = Callable[[str, int, "ParserElement", Exception, bool], None]
255alphas: str = string.ascii_uppercase + string.ascii_lowercase
256identchars: str = pyparsing_unicode.Latin1.identchars
257identbodychars: str = pyparsing_unicode.Latin1.identbodychars
258nums: str = "0123456789"
259hexnums: str = nums + "ABCDEFabcdef"
260alphanums: str = alphas + nums
261printables: str = "".join([c for c in string.printable if c not in string.whitespace])
263_trim_arity_call_line: traceback.StackSummary = None # type: ignore[assignment]
266def _trim_arity(func, max_limit=3):
267 """decorator to trim function calls to match the arity of the target"""
268 global _trim_arity_call_line
270 if func in _single_arg_builtins:
271 return lambda s, l, t: func(t)
273 limit = 0
274 found_arity = False
276 # synthesize what would be returned by traceback.extract_stack at the call to
277 # user's parse action 'func', so that we don't incur call penalty at parse time
279 # fmt: off
280 LINE_DIFF = 9
281 # IF ANY CODE CHANGES, EVEN JUST COMMENTS OR BLANK LINES, BETWEEN THE NEXT LINE AND
282 # THE CALL TO FUNC INSIDE WRAPPER, LINE_DIFF MUST BE MODIFIED!!!!
283 _trim_arity_call_line = (_trim_arity_call_line or traceback.extract_stack(limit=2)[-1])
284 pa_call_line_synth = (_trim_arity_call_line[0], _trim_arity_call_line[1] + LINE_DIFF)
286 def wrapper(*args):
287 nonlocal found_arity, limit
288 if found_arity:
289 return func(*args[limit:])
290 while 1:
291 try:
292 ret = func(*args[limit:])
293 found_arity = True
294 return ret
295 except TypeError as te:
296 # re-raise TypeErrors if they did not come from our arity testing
297 if found_arity:
298 raise
299 else:
300 tb = te.__traceback__
301 frames = traceback.extract_tb(tb, limit=2)
302 frame_summary = frames[-1]
303 trim_arity_type_error = (
304 [frame_summary[:2]][-1][:2] == pa_call_line_synth
305 )
306 del tb
308 if trim_arity_type_error:
309 if limit < max_limit:
310 limit += 1
311 continue
313 raise
314 # fmt: on
316 # copy func name to wrapper for sensible debug output
317 # (can't use functools.wraps, since that messes with function signature)
318 func_name = getattr(func, "__name__", getattr(func, "__class__").__name__)
319 wrapper.__name__ = func_name
320 wrapper.__doc__ = func.__doc__
322 return wrapper
325def condition_as_parse_action(
326 fn: ParseCondition, message: typing.Optional[str] = None, fatal: bool = False
327) -> ParseAction:
328 """
329 Function to convert a simple predicate function that returns ``True`` or ``False``
330 into a parse action. Can be used in places when a parse action is required
331 and :class:`ParserElement.add_condition` cannot be used (such as when adding a condition
332 to an operator level in :class:`infix_notation`).
334 Optional keyword arguments:
336 - ``message`` - define a custom message to be used in the raised exception
337 - ``fatal`` - if True, will raise :class:`ParseFatalException` to stop parsing immediately;
338 otherwise will raise :class:`ParseException`
340 """
341 msg = message if message is not None else "failed user-defined condition"
342 exc_type = ParseFatalException if fatal else ParseException
343 fn = _trim_arity(fn)
345 @wraps(fn)
346 def pa(s, l, t):
347 if not bool(fn(s, l, t)):
348 raise exc_type(s, l, msg)
350 return pa
353def _default_start_debug_action(
354 instring: str, loc: int, expr: "ParserElement", cache_hit: bool = False
355):
356 cache_hit_str = "*" if cache_hit else ""
357 print(
358 (
359 f"{cache_hit_str}Match {expr} at loc {loc}({lineno(loc, instring)},{col(loc, instring)})\n"
360 f" {line(loc, instring)}\n"
361 f" {' ' * (col(loc, instring) - 1)}^"
362 )
363 )
366def _default_success_debug_action(
367 instring: str,
368 startloc: int,
369 endloc: int,
370 expr: "ParserElement",
371 toks: ParseResults,
372 cache_hit: bool = False,
373):
374 cache_hit_str = "*" if cache_hit else ""
375 print(f"{cache_hit_str}Matched {expr} -> {toks.as_list()}")
378def _default_exception_debug_action(
379 instring: str,
380 loc: int,
381 expr: "ParserElement",
382 exc: Exception,
383 cache_hit: bool = False,
384):
385 cache_hit_str = "*" if cache_hit else ""
386 print(f"{cache_hit_str}Match {expr} failed, {type(exc).__name__} raised: {exc}")
389def null_debug_action(*args):
390 """'Do-nothing' debug action, to suppress debugging output during parsing."""
393class ParserElement(ABC):
394 """Abstract base level parser element class."""
396 DEFAULT_WHITE_CHARS: str = " \n\t\r"
397 verbose_stacktrace: bool = False
398 _literalStringClass: type = None # type: ignore[assignment]
400 @staticmethod
401 def set_default_whitespace_chars(chars: str) -> None:
402 r"""
403 Overrides the default whitespace chars
405 Example::
407 # default whitespace chars are space, <TAB> and newline
408 Word(alphas)[1, ...].parse_string("abc def\nghi jkl") # -> ['abc', 'def', 'ghi', 'jkl']
410 # change to just treat newline as significant
411 ParserElement.set_default_whitespace_chars(" \t")
412 Word(alphas)[1, ...].parse_string("abc def\nghi jkl") # -> ['abc', 'def']
413 """
414 ParserElement.DEFAULT_WHITE_CHARS = chars
416 # update whitespace all parse expressions defined in this module
417 for expr in _builtin_exprs:
418 if expr.copyDefaultWhiteChars:
419 expr.whiteChars = set(chars)
421 @staticmethod
422 def inline_literals_using(cls: type) -> None:
423 """
424 Set class to be used for inclusion of string literals into a parser.
426 Example::
428 # default literal class used is Literal
429 integer = Word(nums)
430 date_str = integer("year") + '/' + integer("month") + '/' + integer("day")
432 date_str.parse_string("1999/12/31") # -> ['1999', '/', '12', '/', '31']
435 # change to Suppress
436 ParserElement.inline_literals_using(Suppress)
437 date_str = integer("year") + '/' + integer("month") + '/' + integer("day")
439 date_str.parse_string("1999/12/31") # -> ['1999', '12', '31']
440 """
441 ParserElement._literalStringClass = cls
443 @classmethod
444 def using_each(cls, seq, **class_kwargs):
445 """
446 Yields a sequence of class(obj, **class_kwargs) for obj in seq.
448 Example::
450 LPAR, RPAR, LBRACE, RBRACE, SEMI = Suppress.using_each("(){};")
452 """
453 yield from (cls(obj, **class_kwargs) for obj in seq)
455 class DebugActions(NamedTuple):
456 debug_try: typing.Optional[DebugStartAction]
457 debug_match: typing.Optional[DebugSuccessAction]
458 debug_fail: typing.Optional[DebugExceptionAction]
460 def __init__(self, savelist: bool = False):
461 self.parseAction: List[ParseAction] = list()
462 self.failAction: typing.Optional[ParseFailAction] = None
463 self.customName: str = None # type: ignore[assignment]
464 self._defaultName: typing.Optional[str] = None
465 self.resultsName: str = None # type: ignore[assignment]
466 self.saveAsList = savelist
467 self.skipWhitespace = True
468 self.whiteChars = set(ParserElement.DEFAULT_WHITE_CHARS)
469 self.copyDefaultWhiteChars = True
470 # used when checking for left-recursion
471 self.mayReturnEmpty = False
472 self.keepTabs = False
473 self.ignoreExprs: List["ParserElement"] = list()
474 self.debug = False
475 self.streamlined = False
476 # optimize exception handling for subclasses that don't advance parse index
477 self.mayIndexError = True
478 self.errmsg: Union[str, None] = ""
479 # mark results names as modal (report only last) or cumulative (list all)
480 self.modalResults = True
481 # custom debug actions
482 self.debugActions = self.DebugActions(None, None, None)
483 # avoid redundant calls to preParse
484 self.callPreparse = True
485 self.callDuringTry = False
486 self.suppress_warnings_: List[Diagnostics] = []
488 def suppress_warning(self, warning_type: Diagnostics) -> "ParserElement":
489 """
490 Suppress warnings emitted for a particular diagnostic on this expression.
492 Example::
494 base = pp.Forward()
495 base.suppress_warning(Diagnostics.warn_on_parse_using_empty_Forward)
497 # statement would normally raise a warning, but is now suppressed
498 print(base.parse_string("x"))
500 """
501 self.suppress_warnings_.append(warning_type)
502 return self
504 def visit_all(self):
505 """General-purpose method to yield all expressions and sub-expressions
506 in a grammar. Typically just for internal use.
507 """
508 to_visit = deque([self])
509 seen = set()
510 while to_visit:
511 cur = to_visit.popleft()
513 # guard against looping forever through recursive grammars
514 if cur in seen:
515 continue
516 seen.add(cur)
518 to_visit.extend(cur.recurse())
519 yield cur
521 def copy(self) -> "ParserElement":
522 """
523 Make a copy of this :class:`ParserElement`. Useful for defining
524 different parse actions for the same parsing pattern, using copies of
525 the original parse element.
527 Example::
529 integer = Word(nums).set_parse_action(lambda toks: int(toks[0]))
530 integerK = integer.copy().add_parse_action(lambda toks: toks[0] * 1024) + Suppress("K")
531 integerM = integer.copy().add_parse_action(lambda toks: toks[0] * 1024 * 1024) + Suppress("M")
533 print((integerK | integerM | integer)[1, ...].parse_string("5K 100 640K 256M"))
535 prints::
537 [5120, 100, 655360, 268435456]
539 Equivalent form of ``expr.copy()`` is just ``expr()``::
541 integerM = integer().add_parse_action(lambda toks: toks[0] * 1024 * 1024) + Suppress("M")
542 """
543 cpy = copy.copy(self)
544 cpy.parseAction = self.parseAction[:]
545 cpy.ignoreExprs = self.ignoreExprs[:]
546 if self.copyDefaultWhiteChars:
547 cpy.whiteChars = set(ParserElement.DEFAULT_WHITE_CHARS)
548 return cpy
550 def set_results_name(
551 self, name: str, list_all_matches: bool = False, *, listAllMatches: bool = False
552 ) -> "ParserElement":
553 """
554 Define name for referencing matching tokens as a nested attribute
555 of the returned parse results.
557 Normally, results names are assigned as you would assign keys in a dict:
558 any existing value is overwritten by later values. If it is necessary to
559 keep all values captured for a particular results name, call ``set_results_name``
560 with ``list_all_matches`` = True.
562 NOTE: ``set_results_name`` returns a *copy* of the original :class:`ParserElement` object;
563 this is so that the client can define a basic element, such as an
564 integer, and reference it in multiple places with different names.
566 You can also set results names using the abbreviated syntax,
567 ``expr("name")`` in place of ``expr.set_results_name("name")``
568 - see :class:`__call__`. If ``list_all_matches`` is required, use
569 ``expr("name*")``.
571 Example::
573 integer = Word(nums)
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, list_all_matches=False) -> "ParserElement":
585 if name is None:
586 return self
587 newself = self.copy()
588 if name.endswith("*"):
589 name = name[:-1]
590 list_all_matches = True
591 newself.resultsName = name
592 newself.modalResults = not list_all_matches
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, do_actions=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, do_actions, callPreParse)
611 breaker._originalParseMethod = _parseMethod # type: ignore [attr-defined]
612 self._parse = breaker # type: ignore [assignment]
613 elif hasattr(self._parse, "_originalParseMethod"):
614 self._parse = self._parse._originalParseMethod # type: ignore [attr-defined, assignment]
615 return self
617 def set_parse_action(self, *fns: ParseAction, **kwargs: Any) -> "ParserElement":
618 """
619 Define one or more actions to perform when successfully matching parse element definition.
621 Parse actions can be called to perform data conversions, do extra validation,
622 update external data structures, or enhance or replace the parsed tokens.
623 Each parse action ``fn`` is a callable method with 0-3 arguments, called as
624 ``fn(s, loc, toks)`` , ``fn(loc, toks)`` , ``fn(toks)`` , or just ``fn()`` , where:
626 - ``s`` = the original string being parsed (see note below)
627 - ``loc`` = the location of the matching substring
628 - ``toks`` = a list of the matched tokens, packaged as a :class:`ParseResults` object
630 The parsed tokens are passed to the parse action as ParseResults. They can be
631 modified in place using list-style append, extend, and pop operations to update
632 the parsed list elements; and with dictionary-style item set and del operations
633 to add, update, or remove any named results. If the tokens are modified in place,
634 it is not necessary to return them with a return statement.
636 Parse actions can also completely replace the given tokens, with another ``ParseResults``
637 object, or with some entirely different object (common for parse actions that perform data
638 conversions). A convenient way to build a new parse result is to define the values
639 using a dict, and then create the return value using :class:`ParseResults.from_dict`.
641 If None is passed as the ``fn`` parse action, all previously added parse actions for this
642 expression are cleared.
644 Optional keyword arguments:
646 - ``call_during_try`` = (default= ``False``) indicate if parse action should be run during
647 lookaheads and alternate testing. For parse actions that have side effects, it is
648 important to only call the parse action once it is determined that it is being
649 called as part of a successful parse. For parse actions that perform additional
650 validation, then call_during_try should be passed as True, so that the validation
651 code is included in the preliminary "try" parses.
653 Note: the default parsing behavior is to expand tabs in the input string
654 before starting the parsing process. See :class:`parse_string` for more
655 information on parsing strings containing ``<TAB>`` s, and suggested
656 methods to maintain a consistent view of the parsed string, the parse
657 location, and line and column positions within the parsed string.
659 Example::
661 # parse dates in the form YYYY/MM/DD
663 # use parse action to convert toks from str to int at parse time
664 def convert_to_int(toks):
665 return int(toks[0])
667 # use a parse action to verify that the date is a valid date
668 def is_valid_date(instring, loc, toks):
669 from datetime import date
670 year, month, day = toks[::2]
671 try:
672 date(year, month, day)
673 except ValueError:
674 raise ParseException(instring, loc, "invalid date given")
676 integer = Word(nums)
677 date_str = integer + '/' + integer + '/' + integer
679 # add parse actions
680 integer.set_parse_action(convert_to_int)
681 date_str.set_parse_action(is_valid_date)
683 # note that integer fields are now ints, not strings
684 date_str.run_tests('''
685 # successful parse - note that integer fields were converted to ints
686 1999/12/31
688 # fail - invalid date
689 1999/13/31
690 ''')
691 """
692 if list(fns) == [None]:
693 self.parseAction.clear()
694 return self
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 )
703 return self
705 def add_parse_action(self, *fns: ParseAction, **kwargs: Any) -> "ParserElement":
706 """
707 Add one or more parse actions to expression's list of parse actions. See :class:`set_parse_action`.
709 See examples in :class:`copy`.
710 """
711 self.parseAction += [_trim_arity(fn) for fn in fns]
712 self.callDuringTry = self.callDuringTry or kwargs.get(
713 "call_during_try", kwargs.get("callDuringTry", False)
714 )
715 return self
717 def add_condition(self, *fns: ParseCondition, **kwargs: Any) -> "ParserElement":
718 """Add a boolean predicate function to expression's list of parse actions. See
719 :class:`set_parse_action` for function call signatures. Unlike ``set_parse_action``,
720 functions passed to ``add_condition`` need to return boolean success/fail of the condition.
722 Optional keyword arguments:
724 - ``message`` = define a custom message to be used in the raised exception
725 - ``fatal`` = if True, will raise ParseFatalException to stop parsing immediately; otherwise will raise
726 ParseException
727 - ``call_during_try`` = boolean to indicate if this method should be called during internal tryParse calls,
728 default=False
730 Example::
732 integer = Word(nums).set_parse_action(lambda toks: int(toks[0]))
733 year_int = integer.copy()
734 year_int.add_condition(lambda toks: toks[0] >= 2000, message="Only support years 2000 and later")
735 date_str = year_int + '/' + integer + '/' + integer
737 result = date_str.parse_string("1999/12/31") # -> Exception: Only support years 2000 and later (at char 0),
738 (line:1, col:1)
739 """
740 for fn in fns:
741 self.parseAction.append(
742 condition_as_parse_action(
743 fn,
744 message=str(kwargs.get("message")),
745 fatal=bool(kwargs.get("fatal", False)),
746 )
747 )
749 self.callDuringTry = self.callDuringTry or kwargs.get(
750 "call_during_try", kwargs.get("callDuringTry", False)
751 )
752 return self
754 def set_fail_action(self, fn: ParseFailAction) -> "ParserElement":
755 """
756 Define action to perform if parsing fails at this expression.
757 Fail acton fn is a callable function that takes the arguments
758 ``fn(s, loc, expr, err)`` where:
760 - ``s`` = string being parsed
761 - ``loc`` = location where expression match was attempted and failed
762 - ``expr`` = the parse expression that failed
763 - ``err`` = the exception thrown
765 The function returns no value. It may throw :class:`ParseFatalException`
766 if it is desired to stop parsing immediately."""
767 self.failAction = fn
768 return self
770 def _skipIgnorables(self, instring: str, loc: int) -> int:
771 if not self.ignoreExprs:
772 return loc
773 exprsFound = True
774 ignore_expr_fns = [e._parse for e in self.ignoreExprs]
775 last_loc = loc
776 while exprsFound:
777 exprsFound = False
778 for ignore_fn in ignore_expr_fns:
779 try:
780 while 1:
781 loc, dummy = ignore_fn(instring, loc)
782 exprsFound = True
783 except ParseException:
784 pass
785 # check if all ignore exprs matched but didn't actually advance the parse location
786 if loc == last_loc:
787 break
788 last_loc = loc
789 return loc
791 def preParse(self, instring: str, loc: int) -> int:
792 if self.ignoreExprs:
793 loc = self._skipIgnorables(instring, loc)
795 if self.skipWhitespace:
796 instrlen = len(instring)
797 white_chars = self.whiteChars
798 while loc < instrlen and instring[loc] in white_chars:
799 loc += 1
801 return loc
803 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
804 return loc, []
806 def postParse(self, instring, loc, tokenlist):
807 return tokenlist
809 # @profile
810 def _parseNoCache(
811 self, instring, loc, do_actions=True, callPreParse=True
812 ) -> Tuple[int, ParseResults]:
813 TRY, MATCH, FAIL = 0, 1, 2
814 debugging = self.debug # and do_actions)
815 len_instring = len(instring)
817 if debugging or self.failAction:
818 # print("Match {} at loc {}({}, {})".format(self, loc, lineno(loc, instring), col(loc, instring)))
819 try:
820 if callPreParse and self.callPreparse:
821 pre_loc = self.preParse(instring, loc)
822 else:
823 pre_loc = loc
824 tokens_start = pre_loc
825 if self.debugActions.debug_try:
826 self.debugActions.debug_try(instring, tokens_start, self, False)
827 if self.mayIndexError or pre_loc >= len_instring:
828 try:
829 loc, tokens = self.parseImpl(instring, pre_loc, do_actions)
830 except IndexError:
831 raise ParseException(instring, len_instring, self.errmsg, self)
832 else:
833 loc, tokens = self.parseImpl(instring, pre_loc, do_actions)
834 except Exception as err:
835 # print("Exception raised:", err)
836 if self.debugActions.debug_fail:
837 self.debugActions.debug_fail(
838 instring, tokens_start, self, err, False
839 )
840 if self.failAction:
841 self.failAction(instring, tokens_start, self, err)
842 raise
843 else:
844 if callPreParse and self.callPreparse:
845 pre_loc = self.preParse(instring, loc)
846 else:
847 pre_loc = loc
848 tokens_start = pre_loc
849 if self.mayIndexError or pre_loc >= len_instring:
850 try:
851 loc, tokens = self.parseImpl(instring, pre_loc, do_actions)
852 except IndexError:
853 raise ParseException(instring, len_instring, self.errmsg, self)
854 else:
855 loc, tokens = self.parseImpl(instring, pre_loc, do_actions)
857 tokens = self.postParse(instring, loc, tokens)
859 ret_tokens = ParseResults(
860 tokens, self.resultsName, asList=self.saveAsList, modal=self.modalResults
861 )
862 if self.parseAction and (do_actions or self.callDuringTry):
863 if debugging:
864 try:
865 for fn in self.parseAction:
866 try:
867 tokens = fn(instring, tokens_start, ret_tokens) # type: ignore [call-arg, arg-type]
868 except IndexError as parse_action_exc:
869 exc = ParseException("exception raised in parse action")
870 raise exc from parse_action_exc
872 if tokens is not None and tokens is not ret_tokens:
873 ret_tokens = ParseResults(
874 tokens,
875 self.resultsName,
876 asList=self.saveAsList
877 and isinstance(tokens, (ParseResults, list)),
878 modal=self.modalResults,
879 )
880 except Exception as err:
881 # print "Exception raised in user parse action:", err
882 if self.debugActions.debug_fail:
883 self.debugActions.debug_fail(
884 instring, tokens_start, self, err, False
885 )
886 raise
887 else:
888 for fn in self.parseAction:
889 try:
890 tokens = fn(instring, tokens_start, ret_tokens) # type: ignore [call-arg, arg-type]
891 except IndexError as parse_action_exc:
892 exc = ParseException("exception raised in parse action")
893 raise exc from parse_action_exc
895 if tokens is not None and tokens is not ret_tokens:
896 ret_tokens = ParseResults(
897 tokens,
898 self.resultsName,
899 asList=self.saveAsList
900 and isinstance(tokens, (ParseResults, list)),
901 modal=self.modalResults,
902 )
903 if debugging:
904 # print("Matched", self, "->", ret_tokens.as_list())
905 if self.debugActions.debug_match:
906 self.debugActions.debug_match(
907 instring, tokens_start, loc, self, ret_tokens, False
908 )
910 return loc, ret_tokens
912 def try_parse(
913 self,
914 instring: str,
915 loc: int,
916 *,
917 raise_fatal: bool = False,
918 do_actions: bool = False,
919 ) -> int:
920 try:
921 return self._parse(instring, loc, do_actions=do_actions)[0]
922 except ParseFatalException:
923 if raise_fatal:
924 raise
925 raise ParseException(instring, loc, self.errmsg, self)
927 def can_parse_next(self, instring: str, loc: int, do_actions: bool = False) -> bool:
928 try:
929 self.try_parse(instring, loc, do_actions=do_actions)
930 except (ParseException, IndexError):
931 return False
932 else:
933 return True
935 # cache for left-recursion in Forward references
936 recursion_lock = RLock()
937 recursion_memos: typing.Dict[
938 Tuple[int, "Forward", bool], Tuple[int, Union[ParseResults, Exception]]
939 ] = {}
941 class _CacheType(dict):
942 """
943 class to help type checking
944 """
946 not_in_cache: bool
948 def get(self, *args): ...
950 def set(self, *args): ...
952 # argument cache for optimizing repeated calls when backtracking through recursive expressions
953 packrat_cache = (
954 _CacheType()
955 ) # set later by enable_packrat(); this is here so that reset_cache() doesn't fail
956 packrat_cache_lock = RLock()
957 packrat_cache_stats = [0, 0]
959 # this method gets repeatedly called during backtracking with the same arguments -
960 # we can cache these arguments and save ourselves the trouble of re-parsing the contained expression
961 def _parseCache(
962 self, instring, loc, do_actions=True, callPreParse=True
963 ) -> Tuple[int, ParseResults]:
964 HIT, MISS = 0, 1
965 TRY, MATCH, FAIL = 0, 1, 2
966 lookup = (self, instring, loc, callPreParse, do_actions)
967 with ParserElement.packrat_cache_lock:
968 cache = ParserElement.packrat_cache
969 value = cache.get(lookup)
970 if value is cache.not_in_cache:
971 ParserElement.packrat_cache_stats[MISS] += 1
972 try:
973 value = self._parseNoCache(instring, loc, do_actions, callPreParse)
974 except ParseBaseException as pe:
975 # cache a copy of the exception, without the traceback
976 cache.set(lookup, pe.__class__(*pe.args))
977 raise
978 else:
979 cache.set(lookup, (value[0], value[1].copy(), loc))
980 return value
981 else:
982 ParserElement.packrat_cache_stats[HIT] += 1
983 if self.debug and self.debugActions.debug_try:
984 try:
985 self.debugActions.debug_try(instring, loc, self, cache_hit=True) # type: ignore [call-arg]
986 except TypeError:
987 pass
988 if isinstance(value, Exception):
989 if self.debug and self.debugActions.debug_fail:
990 try:
991 self.debugActions.debug_fail(
992 instring, loc, self, value, cache_hit=True # type: ignore [call-arg]
993 )
994 except TypeError:
995 pass
996 raise value
998 value = cast(Tuple[int, ParseResults, int], value)
999 loc_, result, endloc = value[0], value[1].copy(), value[2]
1000 if self.debug and self.debugActions.debug_match:
1001 try:
1002 self.debugActions.debug_match(
1003 instring, loc_, endloc, self, result, cache_hit=True # type: ignore [call-arg]
1004 )
1005 except TypeError:
1006 pass
1008 return loc_, result
1010 _parse = _parseNoCache
1012 @staticmethod
1013 def reset_cache() -> None:
1014 ParserElement.packrat_cache.clear()
1015 ParserElement.packrat_cache_stats[:] = [0] * len(
1016 ParserElement.packrat_cache_stats
1017 )
1018 ParserElement.recursion_memos.clear()
1020 _packratEnabled = False
1021 _left_recursion_enabled = False
1023 @staticmethod
1024 def disable_memoization() -> None:
1025 """
1026 Disables active Packrat or Left Recursion parsing and their memoization
1028 This method also works if neither Packrat nor Left Recursion are enabled.
1029 This makes it safe to call before activating Packrat nor Left Recursion
1030 to clear any previous settings.
1031 """
1032 ParserElement.reset_cache()
1033 ParserElement._left_recursion_enabled = False
1034 ParserElement._packratEnabled = False
1035 ParserElement._parse = ParserElement._parseNoCache
1037 @staticmethod
1038 def enable_left_recursion(
1039 cache_size_limit: typing.Optional[int] = None, *, force=False
1040 ) -> None:
1041 """
1042 Enables "bounded recursion" parsing, which allows for both direct and indirect
1043 left-recursion. During parsing, left-recursive :class:`Forward` elements are
1044 repeatedly matched with a fixed recursion depth that is gradually increased
1045 until finding the longest match.
1047 Example::
1049 import pyparsing as pp
1050 pp.ParserElement.enable_left_recursion()
1052 E = pp.Forward("E")
1053 num = pp.Word(pp.nums)
1054 # match `num`, or `num '+' num`, or `num '+' num '+' num`, ...
1055 E <<= E + '+' - num | num
1057 print(E.parse_string("1+2+3"))
1059 Recursion search naturally memoizes matches of ``Forward`` elements and may
1060 thus skip reevaluation of parse actions during backtracking. This may break
1061 programs with parse actions which rely on strict ordering of side-effects.
1063 Parameters:
1065 - ``cache_size_limit`` - (default=``None``) - memoize at most this many
1066 ``Forward`` elements during matching; if ``None`` (the default),
1067 memoize all ``Forward`` elements.
1069 Bounded Recursion parsing works similar but not identical to Packrat parsing,
1070 thus the two cannot be used together. Use ``force=True`` to disable any
1071 previous, conflicting settings.
1072 """
1073 if force:
1074 ParserElement.disable_memoization()
1075 elif ParserElement._packratEnabled:
1076 raise RuntimeError("Packrat and Bounded Recursion are not compatible")
1077 if cache_size_limit is None:
1078 ParserElement.recursion_memos = _UnboundedMemo() # type: ignore[assignment]
1079 elif cache_size_limit > 0:
1080 ParserElement.recursion_memos = _LRUMemo(capacity=cache_size_limit) # type: ignore[assignment]
1081 else:
1082 raise NotImplementedError(f"Memo size of {cache_size_limit}")
1083 ParserElement._left_recursion_enabled = True
1085 @staticmethod
1086 def enable_packrat(
1087 cache_size_limit: Union[int, None] = 128, *, force: bool = False
1088 ) -> None:
1089 """
1090 Enables "packrat" parsing, which adds memoizing to the parsing logic.
1091 Repeated parse attempts at the same string location (which happens
1092 often in many complex grammars) can immediately return a cached value,
1093 instead of re-executing parsing/validating code. Memoizing is done of
1094 both valid results and parsing exceptions.
1096 Parameters:
1098 - ``cache_size_limit`` - (default= ``128``) - if an integer value is provided
1099 will limit the size of the packrat cache; if None is passed, then
1100 the cache size will be unbounded; if 0 is passed, the cache will
1101 be effectively disabled.
1103 This speedup may break existing programs that use parse actions that
1104 have side-effects. For this reason, packrat parsing is disabled when
1105 you first import pyparsing. To activate the packrat feature, your
1106 program must call the class method :class:`ParserElement.enable_packrat`.
1107 For best results, call ``enable_packrat()`` immediately after
1108 importing pyparsing.
1110 Example::
1112 import pyparsing
1113 pyparsing.ParserElement.enable_packrat()
1115 Packrat parsing works similar but not identical to Bounded Recursion parsing,
1116 thus the two cannot be used together. Use ``force=True`` to disable any
1117 previous, conflicting settings.
1118 """
1119 if force:
1120 ParserElement.disable_memoization()
1121 elif ParserElement._left_recursion_enabled:
1122 raise RuntimeError("Packrat and Bounded Recursion are not compatible")
1124 if ParserElement._packratEnabled:
1125 return
1127 ParserElement._packratEnabled = True
1128 if cache_size_limit is None:
1129 ParserElement.packrat_cache = _UnboundedCache()
1130 else:
1131 ParserElement.packrat_cache = _FifoCache(cache_size_limit) # type: ignore[assignment]
1132 ParserElement._parse = ParserElement._parseCache
1134 def parse_string(
1135 self, instring: str, parse_all: bool = False, *, parseAll: bool = False
1136 ) -> ParseResults:
1137 """
1138 Parse a string with respect to the parser definition. This function is intended as the primary interface to the
1139 client code.
1141 :param instring: The input string to be parsed.
1142 :param parse_all: If set, the entire input string must match the grammar.
1143 :param parseAll: retained for pre-PEP8 compatibility, will be removed in a future release.
1144 :raises ParseException: Raised if ``parse_all`` is set and the input string does not match the whole grammar.
1145 :returns: the parsed data as a :class:`ParseResults` object, which may be accessed as a `list`, a `dict`, or
1146 an object with attributes if the given parser includes results names.
1148 If the input string is required to match the entire grammar, ``parse_all`` flag must be set to ``True``. This
1149 is also equivalent to ending the grammar with :class:`StringEnd`\\ ().
1151 To report proper column numbers, ``parse_string`` operates on a copy of the input string where all tabs are
1152 converted to spaces (8 spaces per tab, as per the default in ``string.expandtabs``). If the input string
1153 contains tabs and the grammar uses parse actions that use the ``loc`` argument to index into the string
1154 being parsed, one can ensure a consistent view of the input string by doing one of the following:
1156 - calling ``parse_with_tabs`` on your grammar before calling ``parse_string`` (see :class:`parse_with_tabs`),
1157 - define your parse action using the full ``(s,loc,toks)`` signature, and reference the input string using the
1158 parse action's ``s`` argument, or
1159 - explicitly expand the tabs in your input string before calling ``parse_string``.
1161 Examples:
1163 By default, partial matches are OK.
1165 >>> res = Word('a').parse_string('aaaaabaaa')
1166 >>> print(res)
1167 ['aaaaa']
1169 The parsing behavior varies by the inheriting class of this abstract class. Please refer to the children
1170 directly to see more examples.
1172 It raises an exception if parse_all flag is set and instring does not match the whole grammar.
1174 >>> res = Word('a').parse_string('aaaaabaaa', parse_all=True)
1175 Traceback (most recent call last):
1176 ...
1177 pyparsing.ParseException: Expected end of text, found 'b' (at char 5), (line:1, col:6)
1178 """
1179 parseAll = parse_all or parseAll
1181 ParserElement.reset_cache()
1182 if not self.streamlined:
1183 self.streamline()
1184 for e in self.ignoreExprs:
1185 e.streamline()
1186 if not self.keepTabs:
1187 instring = instring.expandtabs()
1188 try:
1189 loc, tokens = self._parse(instring, 0)
1190 if parseAll:
1191 loc = self.preParse(instring, loc)
1192 se = Empty() + StringEnd().set_debug(False)
1193 se._parse(instring, loc)
1194 except ParseBaseException as exc:
1195 if ParserElement.verbose_stacktrace:
1196 raise
1197 else:
1198 # catch and re-raise exception from here, clearing out pyparsing internal stack trace
1199 raise exc.with_traceback(None)
1200 else:
1201 return tokens
1203 def scan_string(
1204 self,
1205 instring: str,
1206 max_matches: int = _MAX_INT,
1207 overlap: bool = False,
1208 *,
1209 debug: bool = False,
1210 maxMatches: int = _MAX_INT,
1211 ) -> Generator[Tuple[ParseResults, int, int], None, None]:
1212 """
1213 Scan the input string for expression matches. Each match will return the
1214 matching tokens, start location, and end location. May be called with optional
1215 ``max_matches`` argument, to clip scanning after 'n' matches are found. If
1216 ``overlap`` is specified, then overlapping matches will be reported.
1218 Note that the start and end locations are reported relative to the string
1219 being parsed. See :class:`parse_string` for more information on parsing
1220 strings with embedded tabs.
1222 Example::
1224 source = "sldjf123lsdjjkf345sldkjf879lkjsfd987"
1225 print(source)
1226 for tokens, start, end in Word(alphas).scan_string(source):
1227 print(' '*start + '^'*(end-start))
1228 print(' '*start + tokens[0])
1230 prints::
1232 sldjf123lsdjjkf345sldkjf879lkjsfd987
1233 ^^^^^
1234 sldjf
1235 ^^^^^^^
1236 lsdjjkf
1237 ^^^^^^
1238 sldkjf
1239 ^^^^^^
1240 lkjsfd
1241 """
1242 maxMatches = min(maxMatches, max_matches)
1243 if not self.streamlined:
1244 self.streamline()
1245 for e in self.ignoreExprs:
1246 e.streamline()
1248 if not self.keepTabs:
1249 instring = str(instring).expandtabs()
1250 instrlen = len(instring)
1251 loc = 0
1252 preparseFn = self.preParse
1253 parseFn = self._parse
1254 ParserElement.resetCache()
1255 matches = 0
1256 try:
1257 while loc <= instrlen and matches < maxMatches:
1258 try:
1259 preloc: int = preparseFn(instring, loc)
1260 nextLoc: int
1261 tokens: ParseResults
1262 nextLoc, tokens = parseFn(instring, preloc, callPreParse=False)
1263 except ParseException:
1264 loc = preloc + 1
1265 else:
1266 if nextLoc > loc:
1267 matches += 1
1268 if debug:
1269 print(
1270 {
1271 "tokens": tokens.asList(),
1272 "start": preloc,
1273 "end": nextLoc,
1274 }
1275 )
1276 yield tokens, preloc, nextLoc
1277 if overlap:
1278 nextloc = preparseFn(instring, loc)
1279 if nextloc > loc:
1280 loc = nextLoc
1281 else:
1282 loc += 1
1283 else:
1284 loc = nextLoc
1285 else:
1286 loc = preloc + 1
1287 except ParseBaseException as exc:
1288 if ParserElement.verbose_stacktrace:
1289 raise
1291 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1292 raise exc.with_traceback(None)
1294 def transform_string(self, instring: str, *, debug: bool = False) -> str:
1295 """
1296 Extension to :class:`scan_string`, to modify matching text with modified tokens that may
1297 be returned from a parse action. To use ``transform_string``, define a grammar and
1298 attach a parse action to it that modifies the returned token list.
1299 Invoking ``transform_string()`` on a target string will then scan for matches,
1300 and replace the matched text patterns according to the logic in the parse
1301 action. ``transform_string()`` returns the resulting transformed string.
1303 Example::
1305 wd = Word(alphas)
1306 wd.set_parse_action(lambda toks: toks[0].title())
1308 print(wd.transform_string("now is the winter of our discontent made glorious summer by this sun of york."))
1310 prints::
1312 Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York.
1313 """
1314 out: List[str] = []
1315 lastE = 0
1316 # force preservation of <TAB>s, to minimize unwanted transformation of string, and to
1317 # keep string locs straight between transform_string and scan_string
1318 self.keepTabs = True
1319 try:
1320 for t, s, e in self.scan_string(instring, debug=debug):
1321 out.append(instring[lastE:s])
1322 lastE = e
1324 if not t:
1325 continue
1327 if isinstance(t, ParseResults):
1328 out += t.as_list()
1329 elif isinstance(t, Iterable) and not isinstance(t, str_type):
1330 out.extend(t)
1331 else:
1332 out.append(t)
1334 out.append(instring[lastE:])
1335 out = [o for o in out if o]
1336 return "".join([str(s) for s in _flatten(out)])
1337 except ParseBaseException as exc:
1338 if ParserElement.verbose_stacktrace:
1339 raise
1341 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1342 raise exc.with_traceback(None)
1344 def search_string(
1345 self,
1346 instring: str,
1347 max_matches: int = _MAX_INT,
1348 *,
1349 debug: bool = False,
1350 maxMatches: int = _MAX_INT,
1351 ) -> ParseResults:
1352 """
1353 Another extension to :class:`scan_string`, simplifying the access to the tokens found
1354 to match the given parse expression. May be called with optional
1355 ``max_matches`` argument, to clip searching after 'n' matches are found.
1357 Example::
1359 # a capitalized word starts with an uppercase letter, followed by zero or more lowercase letters
1360 cap_word = Word(alphas.upper(), alphas.lower())
1362 print(cap_word.search_string("More than Iron, more than Lead, more than Gold I need Electricity"))
1364 # the sum() builtin can be used to merge results into a single ParseResults object
1365 print(sum(cap_word.search_string("More than Iron, more than Lead, more than Gold I need Electricity")))
1367 prints::
1369 [['More'], ['Iron'], ['Lead'], ['Gold'], ['I'], ['Electricity']]
1370 ['More', 'Iron', 'Lead', 'Gold', 'I', 'Electricity']
1371 """
1372 maxMatches = min(maxMatches, max_matches)
1373 try:
1374 return ParseResults(
1375 [t for t, s, e in self.scan_string(instring, maxMatches, debug=debug)]
1376 )
1377 except ParseBaseException as exc:
1378 if ParserElement.verbose_stacktrace:
1379 raise
1381 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1382 raise exc.with_traceback(None)
1384 def split(
1385 self,
1386 instring: str,
1387 maxsplit: int = _MAX_INT,
1388 include_separators: bool = False,
1389 *,
1390 includeSeparators=False,
1391 ) -> Generator[str, None, None]:
1392 """
1393 Generator method to split a string using the given expression as a separator.
1394 May be called with optional ``maxsplit`` argument, to limit the number of splits;
1395 and the optional ``include_separators`` argument (default= ``False``), if the separating
1396 matching text should be included in the split results.
1398 Example::
1400 punc = one_of(list(".,;:/-!?"))
1401 print(list(punc.split("This, this?, this sentence, is badly punctuated!")))
1403 prints::
1405 ['This', ' this', '', ' this sentence', ' is badly punctuated', '']
1406 """
1407 includeSeparators = includeSeparators or include_separators
1408 last = 0
1409 for t, s, e in self.scan_string(instring, max_matches=maxsplit):
1410 yield instring[last:s]
1411 if includeSeparators:
1412 yield t[0]
1413 last = e
1414 yield instring[last:]
1416 def __add__(self, other) -> "ParserElement":
1417 """
1418 Implementation of ``+`` operator - returns :class:`And`. Adding strings to a :class:`ParserElement`
1419 converts them to :class:`Literal`\\ s by default.
1421 Example::
1423 greet = Word(alphas) + "," + Word(alphas) + "!"
1424 hello = "Hello, World!"
1425 print(hello, "->", greet.parse_string(hello))
1427 prints::
1429 Hello, World! -> ['Hello', ',', 'World', '!']
1431 ``...`` may be used as a parse expression as a short form of :class:`SkipTo`::
1433 Literal('start') + ... + Literal('end')
1435 is equivalent to::
1437 Literal('start') + SkipTo('end')("_skipped*") + Literal('end')
1439 Note that the skipped text is returned with '_skipped' as a results name,
1440 and to support having multiple skips in the same parser, the value returned is
1441 a list of all skipped text.
1442 """
1443 if other is Ellipsis:
1444 return _PendingSkip(self)
1446 if isinstance(other, str_type):
1447 other = self._literalStringClass(other)
1448 if not isinstance(other, ParserElement):
1449 return NotImplemented
1450 return And([self, other])
1452 def __radd__(self, other) -> "ParserElement":
1453 """
1454 Implementation of ``+`` operator when left operand is not a :class:`ParserElement`
1455 """
1456 if other is Ellipsis:
1457 return SkipTo(self)("_skipped*") + self
1459 if isinstance(other, str_type):
1460 other = self._literalStringClass(other)
1461 if not isinstance(other, ParserElement):
1462 return NotImplemented
1463 return other + self
1465 def __sub__(self, other) -> "ParserElement":
1466 """
1467 Implementation of ``-`` operator, returns :class:`And` with error stop
1468 """
1469 if isinstance(other, str_type):
1470 other = self._literalStringClass(other)
1471 if not isinstance(other, ParserElement):
1472 return NotImplemented
1473 return self + And._ErrorStop() + other
1475 def __rsub__(self, other) -> "ParserElement":
1476 """
1477 Implementation of ``-`` operator when left operand is not a :class:`ParserElement`
1478 """
1479 if isinstance(other, str_type):
1480 other = self._literalStringClass(other)
1481 if not isinstance(other, ParserElement):
1482 return NotImplemented
1483 return other - self
1485 def __mul__(self, other) -> "ParserElement":
1486 """
1487 Implementation of ``*`` operator, allows use of ``expr * 3`` in place of
1488 ``expr + expr + expr``. Expressions may also be multiplied by a 2-integer
1489 tuple, similar to ``{min, max}`` multipliers in regular expressions. Tuples
1490 may also include ``None`` as in:
1492 - ``expr*(n, None)`` or ``expr*(n, )`` is equivalent
1493 to ``expr*n + ZeroOrMore(expr)``
1494 (read as "at least n instances of ``expr``")
1495 - ``expr*(None, n)`` is equivalent to ``expr*(0, n)``
1496 (read as "0 to n instances of ``expr``")
1497 - ``expr*(None, None)`` is equivalent to ``ZeroOrMore(expr)``
1498 - ``expr*(1, None)`` is equivalent to ``OneOrMore(expr)``
1500 Note that ``expr*(None, n)`` does not raise an exception if
1501 more than n exprs exist in the input stream; that is,
1502 ``expr*(None, n)`` does not enforce a maximum number of expr
1503 occurrences. If this behavior is desired, then write
1504 ``expr*(None, n) + ~expr``
1505 """
1506 if other is Ellipsis:
1507 other = (0, None)
1508 elif isinstance(other, tuple) and other[:1] == (Ellipsis,):
1509 other = ((0,) + other[1:] + (None,))[:2]
1511 if not isinstance(other, (int, tuple)):
1512 return NotImplemented
1514 if isinstance(other, int):
1515 minElements, optElements = other, 0
1516 else:
1517 other = tuple(o if o is not Ellipsis else None for o in other)
1518 other = (other + (None, None))[:2]
1519 if other[0] is None:
1520 other = (0, other[1])
1521 if isinstance(other[0], int) and other[1] is None:
1522 if other[0] == 0:
1523 return ZeroOrMore(self)
1524 if other[0] == 1:
1525 return OneOrMore(self)
1526 else:
1527 return self * other[0] + ZeroOrMore(self)
1528 elif isinstance(other[0], int) and isinstance(other[1], int):
1529 minElements, optElements = other
1530 optElements -= minElements
1531 else:
1532 return NotImplemented
1534 if minElements < 0:
1535 raise ValueError("cannot multiply ParserElement by negative value")
1536 if optElements < 0:
1537 raise ValueError(
1538 "second tuple value must be greater or equal to first tuple value"
1539 )
1540 if minElements == optElements == 0:
1541 return And([])
1543 if optElements:
1545 def makeOptionalList(n):
1546 if n > 1:
1547 return Opt(self + makeOptionalList(n - 1))
1548 else:
1549 return Opt(self)
1551 if minElements:
1552 if minElements == 1:
1553 ret = self + makeOptionalList(optElements)
1554 else:
1555 ret = And([self] * minElements) + makeOptionalList(optElements)
1556 else:
1557 ret = makeOptionalList(optElements)
1558 else:
1559 if minElements == 1:
1560 ret = self
1561 else:
1562 ret = And([self] * minElements)
1563 return ret
1565 def __rmul__(self, other) -> "ParserElement":
1566 return self.__mul__(other)
1568 def __or__(self, other) -> "ParserElement":
1569 """
1570 Implementation of ``|`` operator - returns :class:`MatchFirst`
1571 """
1572 if other is Ellipsis:
1573 return _PendingSkip(self, must_skip=True)
1575 if isinstance(other, str_type):
1576 # `expr | ""` is equivalent to `Opt(expr)`
1577 if other == "":
1578 return Opt(self)
1579 other = self._literalStringClass(other)
1580 if not isinstance(other, ParserElement):
1581 return NotImplemented
1582 return MatchFirst([self, other])
1584 def __ror__(self, other) -> "ParserElement":
1585 """
1586 Implementation of ``|`` operator when left operand is not a :class:`ParserElement`
1587 """
1588 if isinstance(other, str_type):
1589 other = self._literalStringClass(other)
1590 if not isinstance(other, ParserElement):
1591 return NotImplemented
1592 return other | self
1594 def __xor__(self, other) -> "ParserElement":
1595 """
1596 Implementation of ``^`` operator - returns :class:`Or`
1597 """
1598 if isinstance(other, str_type):
1599 other = self._literalStringClass(other)
1600 if not isinstance(other, ParserElement):
1601 return NotImplemented
1602 return Or([self, other])
1604 def __rxor__(self, other) -> "ParserElement":
1605 """
1606 Implementation of ``^`` operator when left operand is not a :class:`ParserElement`
1607 """
1608 if isinstance(other, str_type):
1609 other = self._literalStringClass(other)
1610 if not isinstance(other, ParserElement):
1611 return NotImplemented
1612 return other ^ self
1614 def __and__(self, other) -> "ParserElement":
1615 """
1616 Implementation of ``&`` operator - returns :class:`Each`
1617 """
1618 if isinstance(other, str_type):
1619 other = self._literalStringClass(other)
1620 if not isinstance(other, ParserElement):
1621 return NotImplemented
1622 return Each([self, other])
1624 def __rand__(self, other) -> "ParserElement":
1625 """
1626 Implementation of ``&`` operator when left operand is not a :class:`ParserElement`
1627 """
1628 if isinstance(other, str_type):
1629 other = self._literalStringClass(other)
1630 if not isinstance(other, ParserElement):
1631 return NotImplemented
1632 return other & self
1634 def __invert__(self) -> "ParserElement":
1635 """
1636 Implementation of ``~`` operator - returns :class:`NotAny`
1637 """
1638 return NotAny(self)
1640 # disable __iter__ to override legacy use of sequential access to __getitem__ to
1641 # iterate over a sequence
1642 __iter__ = None
1644 def __getitem__(self, key):
1645 """
1646 use ``[]`` indexing notation as a short form for expression repetition:
1648 - ``expr[n]`` is equivalent to ``expr*n``
1649 - ``expr[m, n]`` is equivalent to ``expr*(m, n)``
1650 - ``expr[n, ...]`` or ``expr[n,]`` is equivalent
1651 to ``expr*n + ZeroOrMore(expr)``
1652 (read as "at least n instances of ``expr``")
1653 - ``expr[..., n]`` is equivalent to ``expr*(0, n)``
1654 (read as "0 to n instances of ``expr``")
1655 - ``expr[...]`` and ``expr[0, ...]`` are equivalent to ``ZeroOrMore(expr)``
1656 - ``expr[1, ...]`` is equivalent to ``OneOrMore(expr)``
1658 ``None`` may be used in place of ``...``.
1660 Note that ``expr[..., n]`` and ``expr[m, n]`` do not raise an exception
1661 if more than ``n`` ``expr``\\ s exist in the input stream. If this behavior is
1662 desired, then write ``expr[..., n] + ~expr``.
1664 For repetition with a stop_on expression, use slice notation:
1666 - ``expr[...: end_expr]`` and ``expr[0, ...: end_expr]`` are equivalent to ``ZeroOrMore(expr, stop_on=end_expr)``
1667 - ``expr[1, ...: end_expr]`` is equivalent to ``OneOrMore(expr, stop_on=end_expr)``
1669 """
1671 stop_on_defined = False
1672 stop_on = NoMatch()
1673 if isinstance(key, slice):
1674 key, stop_on = key.start, key.stop
1675 if key is None:
1676 key = ...
1677 stop_on_defined = True
1678 elif isinstance(key, tuple) and isinstance(key[-1], slice):
1679 key, stop_on = (key[0], key[1].start), key[1].stop
1680 stop_on_defined = True
1682 # convert single arg keys to tuples
1683 if isinstance(key, str_type):
1684 key = (key,)
1685 try:
1686 iter(key)
1687 except TypeError:
1688 key = (key, key)
1690 if len(key) > 2:
1691 raise TypeError(
1692 f"only 1 or 2 index arguments supported ({key[:5]}{f'... [{len(key)}]' if len(key) > 5 else ''})"
1693 )
1695 # clip to 2 elements
1696 ret = self * tuple(key[:2])
1697 ret = typing.cast(_MultipleMatch, ret)
1699 if stop_on_defined:
1700 ret.stopOn(stop_on)
1702 return ret
1704 def __call__(self, name: typing.Optional[str] = None) -> "ParserElement":
1705 """
1706 Shortcut for :class:`set_results_name`, with ``list_all_matches=False``.
1708 If ``name`` is given with a trailing ``'*'`` character, then ``list_all_matches`` will be
1709 passed as ``True``.
1711 If ``name`` is omitted, same as calling :class:`copy`.
1713 Example::
1715 # these are equivalent
1716 userdata = Word(alphas).set_results_name("name") + Word(nums + "-").set_results_name("socsecno")
1717 userdata = Word(alphas)("name") + Word(nums + "-")("socsecno")
1718 """
1719 if name is not None:
1720 return self._setResultsName(name)
1722 return self.copy()
1724 def suppress(self) -> "ParserElement":
1725 """
1726 Suppresses the output of this :class:`ParserElement`; useful to keep punctuation from
1727 cluttering up returned output.
1728 """
1729 return Suppress(self)
1731 def ignore_whitespace(self, recursive: bool = True) -> "ParserElement":
1732 """
1733 Enables the skipping of whitespace before matching the characters in the
1734 :class:`ParserElement`'s defined pattern.
1736 :param recursive: If ``True`` (the default), also enable whitespace skipping in child elements (if any)
1737 """
1738 self.skipWhitespace = True
1739 return self
1741 def leave_whitespace(self, recursive: bool = True) -> "ParserElement":
1742 """
1743 Disables the skipping of whitespace before matching the characters in the
1744 :class:`ParserElement`'s defined pattern. This is normally only used internally by
1745 the pyparsing module, but may be needed in some whitespace-sensitive grammars.
1747 :param recursive: If true (the default), also disable whitespace skipping in child elements (if any)
1748 """
1749 self.skipWhitespace = False
1750 return self
1752 def set_whitespace_chars(
1753 self, chars: Union[Set[str], str], copy_defaults: bool = False
1754 ) -> "ParserElement":
1755 """
1756 Overrides the default whitespace chars
1757 """
1758 self.skipWhitespace = True
1759 self.whiteChars = set(chars)
1760 self.copyDefaultWhiteChars = copy_defaults
1761 return self
1763 def parse_with_tabs(self) -> "ParserElement":
1764 """
1765 Overrides default behavior to expand ``<TAB>`` s to spaces before parsing the input string.
1766 Must be called before ``parse_string`` when the input grammar contains elements that
1767 match ``<TAB>`` characters.
1768 """
1769 self.keepTabs = True
1770 return self
1772 def ignore(self, other: "ParserElement") -> "ParserElement":
1773 """
1774 Define expression to be ignored (e.g., comments) while doing pattern
1775 matching; may be called repeatedly, to define multiple comment or other
1776 ignorable patterns.
1778 Example::
1780 patt = Word(alphas)[...]
1781 patt.parse_string('ablaj /* comment */ lskjd')
1782 # -> ['ablaj']
1784 patt.ignore(c_style_comment)
1785 patt.parse_string('ablaj /* comment */ lskjd')
1786 # -> ['ablaj', 'lskjd']
1787 """
1788 if isinstance(other, str_type):
1789 other = Suppress(other)
1791 if isinstance(other, Suppress):
1792 if other not in self.ignoreExprs:
1793 self.ignoreExprs.append(other)
1794 else:
1795 self.ignoreExprs.append(Suppress(other.copy()))
1796 return self
1798 def set_debug_actions(
1799 self,
1800 start_action: DebugStartAction,
1801 success_action: DebugSuccessAction,
1802 exception_action: DebugExceptionAction,
1803 ) -> "ParserElement":
1804 """
1805 Customize display of debugging messages while doing pattern matching:
1807 - ``start_action`` - method to be called when an expression is about to be parsed;
1808 should have the signature ``fn(input_string: str, location: int, expression: ParserElement, cache_hit: bool)``
1810 - ``success_action`` - method to be called when an expression has successfully parsed;
1811 should have the signature ``fn(input_string: str, start_location: int, end_location: int, expression: ParserELement, parsed_tokens: ParseResults, cache_hit: bool)``
1813 - ``exception_action`` - method to be called when expression fails to parse;
1814 should have the signature ``fn(input_string: str, location: int, expression: ParserElement, exception: Exception, cache_hit: bool)``
1815 """
1816 self.debugActions = self.DebugActions(
1817 start_action or _default_start_debug_action, # type: ignore[truthy-function]
1818 success_action or _default_success_debug_action, # type: ignore[truthy-function]
1819 exception_action or _default_exception_debug_action, # type: ignore[truthy-function]
1820 )
1821 self.debug = True
1822 return self
1824 def set_debug(self, flag: bool = True, recurse: bool = False) -> "ParserElement":
1825 """
1826 Enable display of debugging messages while doing pattern matching.
1827 Set ``flag`` to ``True`` to enable, ``False`` to disable.
1828 Set ``recurse`` to ``True`` to set the debug flag on this expression and all sub-expressions.
1830 Example::
1832 wd = Word(alphas).set_name("alphaword")
1833 integer = Word(nums).set_name("numword")
1834 term = wd | integer
1836 # turn on debugging for wd
1837 wd.set_debug()
1839 term[1, ...].parse_string("abc 123 xyz 890")
1841 prints::
1843 Match alphaword at loc 0(1,1)
1844 Matched alphaword -> ['abc']
1845 Match alphaword at loc 3(1,4)
1846 Exception raised:Expected alphaword (at char 4), (line:1, col:5)
1847 Match alphaword at loc 7(1,8)
1848 Matched alphaword -> ['xyz']
1849 Match alphaword at loc 11(1,12)
1850 Exception raised:Expected alphaword (at char 12), (line:1, col:13)
1851 Match alphaword at loc 15(1,16)
1852 Exception raised:Expected alphaword (at char 15), (line:1, col:16)
1854 The output shown is that produced by the default debug actions - custom debug actions can be
1855 specified using :class:`set_debug_actions`. Prior to attempting
1856 to match the ``wd`` expression, the debugging message ``"Match <exprname> at loc <n>(<line>,<col>)"``
1857 is shown. Then if the parse succeeds, a ``"Matched"`` message is shown, or an ``"Exception raised"``
1858 message is shown. Also note the use of :class:`set_name` to assign a human-readable name to the expression,
1859 which makes debugging and exception messages easier to understand - for instance, the default
1860 name created for the :class:`Word` expression without calling ``set_name`` is ``"W:(A-Za-z)"``.
1861 """
1862 if recurse:
1863 for expr in self.visit_all():
1864 expr.set_debug(flag, recurse=False)
1865 return self
1867 if flag:
1868 self.set_debug_actions(
1869 _default_start_debug_action,
1870 _default_success_debug_action,
1871 _default_exception_debug_action,
1872 )
1873 else:
1874 self.debug = False
1875 return self
1877 @property
1878 def default_name(self) -> str:
1879 if self._defaultName is None:
1880 self._defaultName = self._generateDefaultName()
1881 return self._defaultName
1883 @abstractmethod
1884 def _generateDefaultName(self) -> str:
1885 """
1886 Child classes must define this method, which defines how the ``default_name`` is set.
1887 """
1889 def set_name(self, name: typing.Optional[str]) -> "ParserElement":
1890 """
1891 Define name for this expression, makes debugging and exception messages clearer. If
1892 `__diag__.enable_debug_on_named_expressions` is set to True, setting a name will also
1893 enable debug for this expression.
1895 If `name` is None, clears any custom name for this expression, and clears the
1896 debug flag is it was enabled via `__diag__.enable_debug_on_named_expressions`.
1898 Example::
1900 integer = Word(nums)
1901 integer.parse_string("ABC") # -> Exception: Expected W:(0-9) (at char 0), (line:1, col:1)
1903 integer.set_name("integer")
1904 integer.parse_string("ABC") # -> Exception: Expected integer (at char 0), (line:1, col:1)
1905 """
1906 self.customName = name
1907 self.errmsg = f"Expected {str(self)}"
1909 if __diag__.enable_debug_on_named_expressions:
1910 self.set_debug(name is not None)
1912 return self
1914 @property
1915 def name(self) -> str:
1916 # This will use a user-defined name if available, but otherwise defaults back to the auto-generated name
1917 return self.customName if self.customName is not None else self.default_name
1919 @name.setter
1920 def name(self, new_name) -> None:
1921 self.set_name(new_name)
1923 def __str__(self) -> str:
1924 return self.name
1926 def __repr__(self) -> str:
1927 return str(self)
1929 def streamline(self) -> "ParserElement":
1930 self.streamlined = True
1931 self._defaultName = None
1932 return self
1934 def recurse(self) -> List["ParserElement"]:
1935 return []
1937 def _checkRecursion(self, parseElementList):
1938 subRecCheckList = parseElementList[:] + [self]
1939 for e in self.recurse():
1940 e._checkRecursion(subRecCheckList)
1942 def validate(self, validateTrace=None) -> None:
1943 """
1944 Check defined expressions for valid structure, check for infinite recursive definitions.
1945 """
1946 warnings.warn(
1947 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
1948 DeprecationWarning,
1949 stacklevel=2,
1950 )
1951 self._checkRecursion([])
1953 def parse_file(
1954 self,
1955 file_or_filename: Union[str, Path, TextIO],
1956 encoding: str = "utf-8",
1957 parse_all: bool = False,
1958 *,
1959 parseAll: bool = False,
1960 ) -> ParseResults:
1961 """
1962 Execute the parse expression on the given file or filename.
1963 If a filename is specified (instead of a file object),
1964 the entire file is opened, read, and closed before parsing.
1965 """
1966 parseAll = parseAll or parse_all
1967 try:
1968 file_or_filename = typing.cast(TextIO, file_or_filename)
1969 file_contents = file_or_filename.read()
1970 except AttributeError:
1971 file_or_filename = typing.cast(str, file_or_filename)
1972 with open(file_or_filename, "r", encoding=encoding) as f:
1973 file_contents = f.read()
1974 try:
1975 return self.parse_string(file_contents, parseAll)
1976 except ParseBaseException as exc:
1977 if ParserElement.verbose_stacktrace:
1978 raise
1980 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1981 raise exc.with_traceback(None)
1983 def __eq__(self, other):
1984 if self is other:
1985 return True
1986 elif isinstance(other, str_type):
1987 return self.matches(other, parse_all=True)
1988 elif isinstance(other, ParserElement):
1989 return vars(self) == vars(other)
1990 return False
1992 def __hash__(self):
1993 return id(self)
1995 def matches(
1996 self, test_string: str, parse_all: bool = True, *, parseAll: bool = True
1997 ) -> bool:
1998 """
1999 Method for quick testing of a parser against a test string. Good for simple
2000 inline microtests of sub expressions while building up larger parser.
2002 Parameters:
2004 - ``test_string`` - to test against this expression for a match
2005 - ``parse_all`` - (default= ``True``) - flag to pass to :class:`parse_string` when running tests
2007 Example::
2009 expr = Word(nums)
2010 assert expr.matches("100")
2011 """
2012 parseAll = parseAll and parse_all
2013 try:
2014 self.parse_string(str(test_string), parse_all=parseAll)
2015 return True
2016 except ParseBaseException:
2017 return False
2019 def run_tests(
2020 self,
2021 tests: Union[str, List[str]],
2022 parse_all: bool = True,
2023 comment: typing.Optional[Union["ParserElement", str]] = "#",
2024 full_dump: bool = True,
2025 print_results: bool = True,
2026 failure_tests: bool = False,
2027 post_parse: typing.Optional[
2028 Callable[[str, ParseResults], typing.Optional[str]]
2029 ] = None,
2030 file: typing.Optional[TextIO] = None,
2031 with_line_numbers: bool = False,
2032 *,
2033 parseAll: bool = True,
2034 fullDump: bool = True,
2035 printResults: bool = True,
2036 failureTests: bool = False,
2037 postParse: typing.Optional[
2038 Callable[[str, ParseResults], typing.Optional[str]]
2039 ] = None,
2040 ) -> Tuple[bool, List[Tuple[str, Union[ParseResults, Exception]]]]:
2041 """
2042 Execute the parse expression on a series of test strings, showing each
2043 test, the parsed results or where the parse failed. Quick and easy way to
2044 run a parse expression against a list of sample strings.
2046 Parameters:
2048 - ``tests`` - a list of separate test strings, or a multiline string of test strings
2049 - ``parse_all`` - (default= ``True``) - flag to pass to :class:`parse_string` when running tests
2050 - ``comment`` - (default= ``'#'``) - expression for indicating embedded comments in the test
2051 string; pass None to disable comment filtering
2052 - ``full_dump`` - (default= ``True``) - dump results as list followed by results names in nested outline;
2053 if False, only dump nested list
2054 - ``print_results`` - (default= ``True``) prints test output to stdout
2055 - ``failure_tests`` - (default= ``False``) indicates if these tests are expected to fail parsing
2056 - ``post_parse`` - (default= ``None``) optional callback for successful parse results; called as
2057 `fn(test_string, parse_results)` and returns a string to be added to the test output
2058 - ``file`` - (default= ``None``) optional file-like object to which test output will be written;
2059 if None, will default to ``sys.stdout``
2060 - ``with_line_numbers`` - default= ``False``) show test strings with line and column numbers
2062 Returns: a (success, results) tuple, where success indicates that all tests succeeded
2063 (or failed if ``failure_tests`` is True), and the results contain a list of lines of each
2064 test's output
2066 Example::
2068 number_expr = pyparsing_common.number.copy()
2070 result = number_expr.run_tests('''
2071 # unsigned integer
2072 100
2073 # negative integer
2074 -100
2075 # float with scientific notation
2076 6.02e23
2077 # integer with scientific notation
2078 1e-12
2079 ''')
2080 print("Success" if result[0] else "Failed!")
2082 result = number_expr.run_tests('''
2083 # stray character
2084 100Z
2085 # missing leading digit before '.'
2086 -.100
2087 # too many '.'
2088 3.14.159
2089 ''', failure_tests=True)
2090 print("Success" if result[0] else "Failed!")
2092 prints::
2094 # unsigned integer
2095 100
2096 [100]
2098 # negative integer
2099 -100
2100 [-100]
2102 # float with scientific notation
2103 6.02e23
2104 [6.02e+23]
2106 # integer with scientific notation
2107 1e-12
2108 [1e-12]
2110 Success
2112 # stray character
2113 100Z
2114 ^
2115 FAIL: Expected end of text (at char 3), (line:1, col:4)
2117 # missing leading digit before '.'
2118 -.100
2119 ^
2120 FAIL: Expected {real number with scientific notation | real number | signed integer} (at char 0), (line:1, col:1)
2122 # too many '.'
2123 3.14.159
2124 ^
2125 FAIL: Expected end of text (at char 4), (line:1, col:5)
2127 Success
2129 Each test string must be on a single line. If you want to test a string that spans multiple
2130 lines, create a test like this::
2132 expr.run_tests(r"this is a test\\n of strings that spans \\n 3 lines")
2134 (Note that this is a raw string literal, you must include the leading ``'r'``.)
2135 """
2136 from .testing import pyparsing_test
2138 parseAll = parseAll and parse_all
2139 fullDump = fullDump and full_dump
2140 printResults = printResults and print_results
2141 failureTests = failureTests or failure_tests
2142 postParse = postParse or post_parse
2143 if isinstance(tests, str_type):
2144 tests = typing.cast(str, tests)
2145 line_strip = type(tests).strip
2146 tests = [line_strip(test_line) for test_line in tests.rstrip().splitlines()]
2147 comment_specified = comment is not None
2148 if comment_specified:
2149 if isinstance(comment, str_type):
2150 comment = typing.cast(str, comment)
2151 comment = Literal(comment)
2152 comment = typing.cast(ParserElement, comment)
2153 if file is None:
2154 file = sys.stdout
2155 print_ = file.write
2157 result: Union[ParseResults, Exception]
2158 allResults: List[Tuple[str, Union[ParseResults, Exception]]] = []
2159 comments: List[str] = []
2160 success = True
2161 NL = Literal(r"\n").add_parse_action(replace_with("\n")).ignore(quoted_string)
2162 BOM = "\ufeff"
2163 nlstr = "\n"
2164 for t in tests:
2165 if comment_specified and comment.matches(t, False) or comments and not t:
2166 comments.append(
2167 pyparsing_test.with_line_numbers(t) if with_line_numbers else t
2168 )
2169 continue
2170 if not t:
2171 continue
2172 out = [
2173 f"{nlstr}{nlstr.join(comments) if comments else ''}",
2174 pyparsing_test.with_line_numbers(t) if with_line_numbers else t,
2175 ]
2176 comments.clear()
2177 try:
2178 # convert newline marks to actual newlines, and strip leading BOM if present
2179 t = NL.transform_string(t.lstrip(BOM))
2180 result = self.parse_string(t, parse_all=parseAll)
2181 except ParseBaseException as pe:
2182 fatal = "(FATAL) " if isinstance(pe, ParseFatalException) else ""
2183 out.append(pe.explain())
2184 out.append(f"FAIL: {fatal}{pe}")
2185 if ParserElement.verbose_stacktrace:
2186 out.extend(traceback.format_tb(pe.__traceback__))
2187 success = success and failureTests
2188 result = pe
2189 except Exception as exc:
2190 out.append(f"FAIL-EXCEPTION: {type(exc).__name__}: {exc}")
2191 if ParserElement.verbose_stacktrace:
2192 out.extend(traceback.format_tb(exc.__traceback__))
2193 success = success and failureTests
2194 result = exc
2195 else:
2196 success = success and not failureTests
2197 if postParse is not None:
2198 try:
2199 pp_value = postParse(t, result)
2200 if pp_value is not None:
2201 if isinstance(pp_value, ParseResults):
2202 out.append(pp_value.dump())
2203 else:
2204 out.append(str(pp_value))
2205 else:
2206 out.append(result.dump())
2207 except Exception as e:
2208 out.append(result.dump(full=fullDump))
2209 out.append(
2210 f"{postParse.__name__} failed: {type(e).__name__}: {e}"
2211 )
2212 else:
2213 out.append(result.dump(full=fullDump))
2214 out.append("")
2216 if printResults:
2217 print_("\n".join(out))
2219 allResults.append((t, result))
2221 return success, allResults
2223 def create_diagram(
2224 self,
2225 output_html: Union[TextIO, Path, str],
2226 vertical: int = 3,
2227 show_results_names: bool = False,
2228 show_groups: bool = False,
2229 embed: bool = False,
2230 **kwargs,
2231 ) -> None:
2232 """
2233 Create a railroad diagram for the parser.
2235 Parameters:
2237 - ``output_html`` (str or file-like object) - output target for generated
2238 diagram HTML
2239 - ``vertical`` (int) - threshold for formatting multiple alternatives vertically
2240 instead of horizontally (default=3)
2241 - ``show_results_names`` - bool flag whether diagram should show annotations for
2242 defined results names
2243 - ``show_groups`` - bool flag whether groups should be highlighted with an unlabeled surrounding box
2244 - ``embed`` - bool flag whether generated HTML should omit <HEAD>, <BODY>, and <DOCTYPE> tags to embed
2245 the resulting HTML in an enclosing HTML source
2246 - ``head`` - str containing additional HTML to insert into the <HEAD> section of the generated code;
2247 can be used to insert custom CSS styling
2248 - ``body`` - str containing additional HTML to insert at the beginning of the <BODY> section of the
2249 generated code
2251 Additional diagram-formatting keyword arguments can also be included;
2252 see railroad.Diagram class.
2253 """
2255 try:
2256 from .diagram import to_railroad, railroad_to_html
2257 except ImportError as ie:
2258 raise Exception(
2259 "must ``pip install pyparsing[diagrams]`` to generate parser railroad diagrams"
2260 ) from ie
2262 self.streamline()
2264 railroad = to_railroad(
2265 self,
2266 vertical=vertical,
2267 show_results_names=show_results_names,
2268 show_groups=show_groups,
2269 diagram_kwargs=kwargs,
2270 )
2271 if not isinstance(output_html, (str, Path)):
2272 # we were passed a file-like object, just write to it
2273 output_html.write(railroad_to_html(railroad, embed=embed, **kwargs))
2274 return
2276 with open(output_html, "w", encoding="utf-8") as diag_file:
2277 diag_file.write(railroad_to_html(railroad, embed=embed, **kwargs))
2279 # Compatibility synonyms
2280 # fmt: off
2281 inlineLiteralsUsing = staticmethod(replaced_by_pep8("inlineLiteralsUsing", inline_literals_using))
2282 setDefaultWhitespaceChars = staticmethod(replaced_by_pep8(
2283 "setDefaultWhitespaceChars", set_default_whitespace_chars
2284 ))
2285 disableMemoization = staticmethod(replaced_by_pep8("disableMemoization", disable_memoization))
2286 enableLeftRecursion = staticmethod(replaced_by_pep8("enableLeftRecursion", enable_left_recursion))
2287 enablePackrat = staticmethod(replaced_by_pep8("enablePackrat", enable_packrat))
2288 resetCache = staticmethod(replaced_by_pep8("resetCache", reset_cache))
2290 setResultsName = replaced_by_pep8("setResultsName", set_results_name)
2291 setBreak = replaced_by_pep8("setBreak", set_break)
2292 setParseAction = replaced_by_pep8("setParseAction", set_parse_action)
2293 addParseAction = replaced_by_pep8("addParseAction", add_parse_action)
2294 addCondition = replaced_by_pep8("addCondition", add_condition)
2295 setFailAction = replaced_by_pep8("setFailAction", set_fail_action)
2296 tryParse = replaced_by_pep8("tryParse", try_parse)
2297 parseString = replaced_by_pep8("parseString", parse_string)
2298 scanString = replaced_by_pep8("scanString", scan_string)
2299 transformString = replaced_by_pep8("transformString", transform_string)
2300 searchString = replaced_by_pep8("searchString", search_string)
2301 ignoreWhitespace = replaced_by_pep8("ignoreWhitespace", ignore_whitespace)
2302 leaveWhitespace = replaced_by_pep8("leaveWhitespace", leave_whitespace)
2303 setWhitespaceChars = replaced_by_pep8("setWhitespaceChars", set_whitespace_chars)
2304 parseWithTabs = replaced_by_pep8("parseWithTabs", parse_with_tabs)
2305 setDebugActions = replaced_by_pep8("setDebugActions", set_debug_actions)
2306 setDebug = replaced_by_pep8("setDebug", set_debug)
2307 setName = replaced_by_pep8("setName", set_name)
2308 parseFile = replaced_by_pep8("parseFile", parse_file)
2309 runTests = replaced_by_pep8("runTests", run_tests)
2310 canParseNext = replaced_by_pep8("canParseNext", can_parse_next)
2311 defaultName = default_name
2312 # fmt: on
2315class _PendingSkip(ParserElement):
2316 # internal placeholder class to hold a place were '...' is added to a parser element,
2317 # once another ParserElement is added, this placeholder will be replaced with a SkipTo
2318 def __init__(self, expr: ParserElement, must_skip: bool = False):
2319 super().__init__()
2320 self.anchor = expr
2321 self.must_skip = must_skip
2323 def _generateDefaultName(self) -> str:
2324 return str(self.anchor + Empty()).replace("Empty", "...")
2326 def __add__(self, other) -> "ParserElement":
2327 skipper = SkipTo(other).set_name("...")("_skipped*")
2328 if self.must_skip:
2330 def must_skip(t):
2331 if not t._skipped or t._skipped.as_list() == [""]:
2332 del t[0]
2333 t.pop("_skipped", None)
2335 def show_skip(t):
2336 if t._skipped.as_list()[-1:] == [""]:
2337 t.pop("_skipped")
2338 t["_skipped"] = f"missing <{self.anchor!r}>"
2340 return (
2341 self.anchor + skipper().add_parse_action(must_skip)
2342 | skipper().add_parse_action(show_skip)
2343 ) + other
2345 return self.anchor + skipper + other
2347 def __repr__(self):
2348 return self.defaultName
2350 def parseImpl(self, *args) -> ParseImplReturnType:
2351 raise Exception(
2352 "use of `...` expression without following SkipTo target expression"
2353 )
2356class Token(ParserElement):
2357 """Abstract :class:`ParserElement` subclass, for defining atomic
2358 matching patterns.
2359 """
2361 def __init__(self):
2362 super().__init__(savelist=False)
2364 def _generateDefaultName(self) -> str:
2365 return type(self).__name__
2368class NoMatch(Token):
2369 """
2370 A token that will never match.
2371 """
2373 def __init__(self):
2374 super().__init__()
2375 self.mayReturnEmpty = True
2376 self.mayIndexError = False
2377 self.errmsg = "Unmatchable token"
2379 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2380 raise ParseException(instring, loc, self.errmsg, self)
2383class Literal(Token):
2384 """
2385 Token to exactly match a specified string.
2387 Example::
2389 Literal('abc').parse_string('abc') # -> ['abc']
2390 Literal('abc').parse_string('abcdef') # -> ['abc']
2391 Literal('abc').parse_string('ab') # -> Exception: Expected "abc"
2393 For case-insensitive matching, use :class:`CaselessLiteral`.
2395 For keyword matching (force word break before and after the matched string),
2396 use :class:`Keyword` or :class:`CaselessKeyword`.
2397 """
2399 def __new__(cls, match_string: str = "", *, matchString: str = ""):
2400 # Performance tuning: select a subclass with optimized parseImpl
2401 if cls is Literal:
2402 match_string = matchString or match_string
2403 if not match_string:
2404 return super().__new__(Empty)
2405 if len(match_string) == 1:
2406 return super().__new__(_SingleCharLiteral)
2408 # Default behavior
2409 return super().__new__(cls)
2411 # Needed to make copy.copy() work correctly if we customize __new__
2412 def __getnewargs__(self):
2413 return (self.match,)
2415 def __init__(self, match_string: str = "", *, matchString: str = ""):
2416 super().__init__()
2417 match_string = matchString or match_string
2418 self.match = match_string
2419 self.matchLen = len(match_string)
2420 self.firstMatchChar = match_string[:1]
2421 self.errmsg = f"Expected {self.name}"
2422 self.mayReturnEmpty = False
2423 self.mayIndexError = False
2425 def _generateDefaultName(self) -> str:
2426 return repr(self.match)
2428 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2429 if instring[loc] == self.firstMatchChar and instring.startswith(
2430 self.match, loc
2431 ):
2432 return loc + self.matchLen, self.match
2433 raise ParseException(instring, loc, self.errmsg, self)
2436class Empty(Literal):
2437 """
2438 An empty token, will always match.
2439 """
2441 def __init__(self, match_string="", *, matchString=""):
2442 super().__init__("")
2443 self.mayReturnEmpty = True
2444 self.mayIndexError = False
2446 def _generateDefaultName(self) -> str:
2447 return "Empty"
2449 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2450 return loc, []
2453class _SingleCharLiteral(Literal):
2454 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2455 if instring[loc] == self.firstMatchChar:
2456 return loc + 1, self.match
2457 raise ParseException(instring, loc, self.errmsg, self)
2460ParserElement._literalStringClass = Literal
2463class Keyword(Token):
2464 """
2465 Token to exactly match a specified string as a keyword, that is,
2466 it must be immediately preceded and followed by whitespace or
2467 non-keyword characters. Compare with :class:`Literal`:
2469 - ``Literal("if")`` will match the leading ``'if'`` in
2470 ``'ifAndOnlyIf'``.
2471 - ``Keyword("if")`` will not; it will only match the leading
2472 ``'if'`` in ``'if x=1'``, or ``'if(y==2)'``
2474 Accepts two optional constructor arguments in addition to the
2475 keyword string:
2477 - ``ident_chars`` is a string of characters that would be valid
2478 identifier characters, defaulting to all alphanumerics + "_" and
2479 "$"
2480 - ``caseless`` allows case-insensitive matching, default is ``False``.
2482 Example::
2484 Keyword("start").parse_string("start") # -> ['start']
2485 Keyword("start").parse_string("starting") # -> Exception
2487 For case-insensitive matching, use :class:`CaselessKeyword`.
2488 """
2490 DEFAULT_KEYWORD_CHARS = alphanums + "_$"
2492 def __init__(
2493 self,
2494 match_string: str = "",
2495 ident_chars: typing.Optional[str] = None,
2496 caseless: bool = False,
2497 *,
2498 matchString: str = "",
2499 identChars: typing.Optional[str] = None,
2500 ):
2501 super().__init__()
2502 identChars = identChars or ident_chars
2503 if identChars is None:
2504 identChars = Keyword.DEFAULT_KEYWORD_CHARS
2505 match_string = matchString or match_string
2506 self.match = match_string
2507 self.matchLen = len(match_string)
2508 try:
2509 self.firstMatchChar = match_string[0]
2510 except IndexError:
2511 raise ValueError("null string passed to Keyword; use Empty() instead")
2512 self.errmsg = f"Expected {type(self).__name__} {self.name}"
2513 self.mayReturnEmpty = False
2514 self.mayIndexError = False
2515 self.caseless = caseless
2516 if caseless:
2517 self.caselessmatch = match_string.upper()
2518 identChars = identChars.upper()
2519 self.identChars = set(identChars)
2521 def _generateDefaultName(self) -> str:
2522 return repr(self.match)
2524 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2525 errmsg = self.errmsg
2526 errloc = loc
2527 if self.caseless:
2528 if instring[loc : loc + self.matchLen].upper() == self.caselessmatch:
2529 if loc == 0 or instring[loc - 1].upper() not in self.identChars:
2530 if (
2531 loc >= len(instring) - self.matchLen
2532 or instring[loc + self.matchLen].upper() not in self.identChars
2533 ):
2534 return loc + self.matchLen, self.match
2536 # followed by keyword char
2537 errmsg += ", was immediately followed by keyword character"
2538 errloc = loc + self.matchLen
2539 else:
2540 # preceded by keyword char
2541 errmsg += ", keyword was immediately preceded by keyword character"
2542 errloc = loc - 1
2543 # else no match just raise plain exception
2545 elif (
2546 instring[loc] == self.firstMatchChar
2547 and self.matchLen == 1
2548 or instring.startswith(self.match, loc)
2549 ):
2550 if loc == 0 or instring[loc - 1] not in self.identChars:
2551 if (
2552 loc >= len(instring) - self.matchLen
2553 or instring[loc + self.matchLen] not in self.identChars
2554 ):
2555 return loc + self.matchLen, self.match
2557 # followed by keyword char
2558 errmsg += ", keyword was immediately followed by keyword character"
2559 errloc = loc + self.matchLen
2560 else:
2561 # preceded by keyword char
2562 errmsg += ", keyword was immediately preceded by keyword character"
2563 errloc = loc - 1
2564 # else no match just raise plain exception
2566 raise ParseException(instring, errloc, errmsg, self)
2568 @staticmethod
2569 def set_default_keyword_chars(chars) -> None:
2570 """
2571 Overrides the default characters used by :class:`Keyword` expressions.
2572 """
2573 Keyword.DEFAULT_KEYWORD_CHARS = chars
2575 # Compatibility synonyms
2576 setDefaultKeywordChars = staticmethod(
2577 replaced_by_pep8("setDefaultKeywordChars", set_default_keyword_chars)
2578 )
2581class CaselessLiteral(Literal):
2582 """
2583 Token to match a specified string, ignoring case of letters.
2584 Note: the matched results will always be in the case of the given
2585 match string, NOT the case of the input text.
2587 Example::
2589 CaselessLiteral("CMD")[1, ...].parse_string("cmd CMD Cmd10")
2590 # -> ['CMD', 'CMD', 'CMD']
2592 (Contrast with example for :class:`CaselessKeyword`.)
2593 """
2595 def __init__(self, match_string: str = "", *, matchString: str = ""):
2596 match_string = matchString or match_string
2597 super().__init__(match_string.upper())
2598 # Preserve the defining literal.
2599 self.returnString = match_string
2600 self.errmsg = f"Expected {self.name}"
2602 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2603 if instring[loc : loc + self.matchLen].upper() == self.match:
2604 return loc + self.matchLen, self.returnString
2605 raise ParseException(instring, loc, self.errmsg, self)
2608class CaselessKeyword(Keyword):
2609 """
2610 Caseless version of :class:`Keyword`.
2612 Example::
2614 CaselessKeyword("CMD")[1, ...].parse_string("cmd CMD Cmd10")
2615 # -> ['CMD', 'CMD']
2617 (Contrast with example for :class:`CaselessLiteral`.)
2618 """
2620 def __init__(
2621 self,
2622 match_string: str = "",
2623 ident_chars: typing.Optional[str] = None,
2624 *,
2625 matchString: str = "",
2626 identChars: typing.Optional[str] = None,
2627 ):
2628 identChars = identChars or ident_chars
2629 match_string = matchString or match_string
2630 super().__init__(match_string, identChars, caseless=True)
2633class CloseMatch(Token):
2634 """A variation on :class:`Literal` which matches "close" matches,
2635 that is, strings with at most 'n' mismatching characters.
2636 :class:`CloseMatch` takes parameters:
2638 - ``match_string`` - string to be matched
2639 - ``caseless`` - a boolean indicating whether to ignore casing when comparing characters
2640 - ``max_mismatches`` - (``default=1``) maximum number of
2641 mismatches allowed to count as a match
2643 The results from a successful parse will contain the matched text
2644 from the input string and the following named results:
2646 - ``mismatches`` - a list of the positions within the
2647 match_string where mismatches were found
2648 - ``original`` - the original match_string used to compare
2649 against the input string
2651 If ``mismatches`` is an empty list, then the match was an exact
2652 match.
2654 Example::
2656 patt = CloseMatch("ATCATCGAATGGA")
2657 patt.parse_string("ATCATCGAAXGGA") # -> (['ATCATCGAAXGGA'], {'mismatches': [[9]], 'original': ['ATCATCGAATGGA']})
2658 patt.parse_string("ATCAXCGAAXGGA") # -> Exception: Expected 'ATCATCGAATGGA' (with up to 1 mismatches) (at char 0), (line:1, col:1)
2660 # exact match
2661 patt.parse_string("ATCATCGAATGGA") # -> (['ATCATCGAATGGA'], {'mismatches': [[]], 'original': ['ATCATCGAATGGA']})
2663 # close match allowing up to 2 mismatches
2664 patt = CloseMatch("ATCATCGAATGGA", max_mismatches=2)
2665 patt.parse_string("ATCAXCGAAXGGA") # -> (['ATCAXCGAAXGGA'], {'mismatches': [[4, 9]], 'original': ['ATCATCGAATGGA']})
2666 """
2668 def __init__(
2669 self,
2670 match_string: str,
2671 max_mismatches: typing.Optional[int] = None,
2672 *,
2673 maxMismatches: int = 1,
2674 caseless=False,
2675 ):
2676 maxMismatches = max_mismatches if max_mismatches is not None else maxMismatches
2677 super().__init__()
2678 self.match_string = match_string
2679 self.maxMismatches = maxMismatches
2680 self.errmsg = f"Expected {self.match_string!r} (with up to {self.maxMismatches} mismatches)"
2681 self.caseless = caseless
2682 self.mayIndexError = False
2683 self.mayReturnEmpty = False
2685 def _generateDefaultName(self) -> str:
2686 return f"{type(self).__name__}:{self.match_string!r}"
2688 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2689 start = loc
2690 instrlen = len(instring)
2691 maxloc = start + len(self.match_string)
2693 if maxloc <= instrlen:
2694 match_string = self.match_string
2695 match_stringloc = 0
2696 mismatches = []
2697 maxMismatches = self.maxMismatches
2699 for match_stringloc, s_m in enumerate(
2700 zip(instring[loc:maxloc], match_string)
2701 ):
2702 src, mat = s_m
2703 if self.caseless:
2704 src, mat = src.lower(), mat.lower()
2706 if src != mat:
2707 mismatches.append(match_stringloc)
2708 if len(mismatches) > maxMismatches:
2709 break
2710 else:
2711 loc = start + match_stringloc + 1
2712 results = ParseResults([instring[start:loc]])
2713 results["original"] = match_string
2714 results["mismatches"] = mismatches
2715 return loc, results
2717 raise ParseException(instring, loc, self.errmsg, self)
2720class Word(Token):
2721 """Token for matching words composed of allowed character sets.
2723 Parameters:
2725 - ``init_chars`` - string of all characters that should be used to
2726 match as a word; "ABC" will match "AAA", "ABAB", "CBAC", etc.;
2727 if ``body_chars`` is also specified, then this is the string of
2728 initial characters
2729 - ``body_chars`` - string of characters that
2730 can be used for matching after a matched initial character as
2731 given in ``init_chars``; if omitted, same as the initial characters
2732 (default=``None``)
2733 - ``min`` - minimum number of characters to match (default=1)
2734 - ``max`` - maximum number of characters to match (default=0)
2735 - ``exact`` - exact number of characters to match (default=0)
2736 - ``as_keyword`` - match as a keyword (default=``False``)
2737 - ``exclude_chars`` - characters that might be
2738 found in the input ``body_chars`` string but which should not be
2739 accepted for matching ;useful to define a word of all
2740 printables except for one or two characters, for instance
2741 (default=``None``)
2743 :class:`srange` is useful for defining custom character set strings
2744 for defining :class:`Word` expressions, using range notation from
2745 regular expression character sets.
2747 A common mistake is to use :class:`Word` to match a specific literal
2748 string, as in ``Word("Address")``. Remember that :class:`Word`
2749 uses the string argument to define *sets* of matchable characters.
2750 This expression would match "Add", "AAA", "dAred", or any other word
2751 made up of the characters 'A', 'd', 'r', 'e', and 's'. To match an
2752 exact literal string, use :class:`Literal` or :class:`Keyword`.
2754 pyparsing includes helper strings for building Words:
2756 - :class:`alphas`
2757 - :class:`nums`
2758 - :class:`alphanums`
2759 - :class:`hexnums`
2760 - :class:`alphas8bit` (alphabetic characters in ASCII range 128-255
2761 - accented, tilded, umlauted, etc.)
2762 - :class:`punc8bit` (non-alphabetic characters in ASCII range
2763 128-255 - currency, symbols, superscripts, diacriticals, etc.)
2764 - :class:`printables` (any non-whitespace character)
2766 ``alphas``, ``nums``, and ``printables`` are also defined in several
2767 Unicode sets - see :class:`pyparsing_unicode``.
2769 Example::
2771 # a word composed of digits
2772 integer = Word(nums) # equivalent to Word("0123456789") or Word(srange("0-9"))
2774 # a word with a leading capital, and zero or more lowercase
2775 capitalized_word = Word(alphas.upper(), alphas.lower())
2777 # hostnames are alphanumeric, with leading alpha, and '-'
2778 hostname = Word(alphas, alphanums + '-')
2780 # roman numeral (not a strict parser, accepts invalid mix of characters)
2781 roman = Word("IVXLCDM")
2783 # any string of non-whitespace characters, except for ','
2784 csv_value = Word(printables, exclude_chars=",")
2785 """
2787 def __init__(
2788 self,
2789 init_chars: str = "",
2790 body_chars: typing.Optional[str] = None,
2791 min: int = 1,
2792 max: int = 0,
2793 exact: int = 0,
2794 as_keyword: bool = False,
2795 exclude_chars: typing.Optional[str] = None,
2796 *,
2797 initChars: typing.Optional[str] = None,
2798 bodyChars: typing.Optional[str] = None,
2799 asKeyword: bool = False,
2800 excludeChars: typing.Optional[str] = None,
2801 ):
2802 initChars = initChars or init_chars
2803 bodyChars = bodyChars or body_chars
2804 asKeyword = asKeyword or as_keyword
2805 excludeChars = excludeChars or exclude_chars
2806 super().__init__()
2807 if not initChars:
2808 raise ValueError(
2809 f"invalid {type(self).__name__}, initChars cannot be empty string"
2810 )
2812 initChars_set = set(initChars)
2813 if excludeChars:
2814 excludeChars_set = set(excludeChars)
2815 initChars_set -= excludeChars_set
2816 if bodyChars:
2817 bodyChars = "".join(set(bodyChars) - excludeChars_set)
2818 self.initChars = initChars_set
2819 self.initCharsOrig = "".join(sorted(initChars_set))
2821 if bodyChars:
2822 self.bodyChars = set(bodyChars)
2823 self.bodyCharsOrig = "".join(sorted(bodyChars))
2824 else:
2825 self.bodyChars = initChars_set
2826 self.bodyCharsOrig = self.initCharsOrig
2828 self.maxSpecified = max > 0
2830 if min < 1:
2831 raise ValueError(
2832 "cannot specify a minimum length < 1; use Opt(Word()) if zero-length word is permitted"
2833 )
2835 if self.maxSpecified and min > max:
2836 raise ValueError(
2837 f"invalid args, if min and max both specified min must be <= max (min={min}, max={max})"
2838 )
2840 self.minLen = min
2842 if max > 0:
2843 self.maxLen = max
2844 else:
2845 self.maxLen = _MAX_INT
2847 if exact > 0:
2848 min = max = exact
2849 self.maxLen = exact
2850 self.minLen = exact
2852 self.errmsg = f"Expected {self.name}"
2853 self.mayIndexError = False
2854 self.asKeyword = asKeyword
2855 if self.asKeyword:
2856 self.errmsg += " as a keyword"
2858 # see if we can make a regex for this Word
2859 if " " not in (self.initChars | self.bodyChars):
2860 if len(self.initChars) == 1:
2861 re_leading_fragment = re.escape(self.initCharsOrig)
2862 else:
2863 re_leading_fragment = f"[{_collapse_string_to_ranges(self.initChars)}]"
2865 if self.bodyChars == self.initChars:
2866 if max == 0 and self.minLen == 1:
2867 repeat = "+"
2868 elif max == 1:
2869 repeat = ""
2870 else:
2871 if self.minLen != self.maxLen:
2872 repeat = f"{{{self.minLen},{'' if self.maxLen == _MAX_INT else self.maxLen}}}"
2873 else:
2874 repeat = f"{{{self.minLen}}}"
2875 self.reString = f"{re_leading_fragment}{repeat}"
2876 else:
2877 if max == 1:
2878 re_body_fragment = ""
2879 repeat = ""
2880 else:
2881 re_body_fragment = f"[{_collapse_string_to_ranges(self.bodyChars)}]"
2882 if max == 0 and self.minLen == 1:
2883 repeat = "*"
2884 elif max == 2:
2885 repeat = "?" if min <= 1 else ""
2886 else:
2887 if min != max:
2888 repeat = f"{{{min - 1 if min > 0 else ''},{max - 1 if max > 0 else ''}}}"
2889 else:
2890 repeat = f"{{{min - 1 if min > 0 else ''}}}"
2892 self.reString = f"{re_leading_fragment}{re_body_fragment}{repeat}"
2894 if self.asKeyword:
2895 self.reString = rf"\b{self.reString}\b"
2897 try:
2898 self.re = re.compile(self.reString)
2899 except re.error:
2900 self.re = None # type: ignore[assignment]
2901 else:
2902 self.re_match = self.re.match
2903 self.parseImpl = self.parseImpl_regex # type: ignore[assignment]
2905 def _generateDefaultName(self) -> str:
2906 def charsAsStr(s):
2907 max_repr_len = 16
2908 s = _collapse_string_to_ranges(s, re_escape=False)
2910 if len(s) > max_repr_len:
2911 return s[: max_repr_len - 3] + "..."
2913 return s
2915 if self.initChars != self.bodyChars:
2916 base = f"W:({charsAsStr(self.initChars)}, {charsAsStr(self.bodyChars)})"
2917 else:
2918 base = f"W:({charsAsStr(self.initChars)})"
2920 # add length specification
2921 if self.minLen > 1 or self.maxLen != _MAX_INT:
2922 if self.minLen == self.maxLen:
2923 if self.minLen == 1:
2924 return base[2:]
2925 else:
2926 return base + f"{{{self.minLen}}}"
2927 elif self.maxLen == _MAX_INT:
2928 return base + f"{{{self.minLen},...}}"
2929 else:
2930 return base + f"{{{self.minLen},{self.maxLen}}}"
2931 return base
2933 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2934 if instring[loc] not in self.initChars:
2935 raise ParseException(instring, loc, self.errmsg, self)
2937 start = loc
2938 loc += 1
2939 instrlen = len(instring)
2940 body_chars: set[str] = self.bodyChars
2941 maxloc = start + self.maxLen
2942 maxloc = min(maxloc, instrlen)
2943 while loc < maxloc and instring[loc] in body_chars:
2944 loc += 1
2946 throw_exception = False
2947 if loc - start < self.minLen:
2948 throw_exception = True
2949 elif self.maxSpecified and loc < instrlen and instring[loc] in body_chars:
2950 throw_exception = True
2951 elif self.asKeyword and (
2952 (start > 0 and instring[start - 1] in body_chars)
2953 or (loc < instrlen and instring[loc] in body_chars)
2954 ):
2955 throw_exception = True
2957 if throw_exception:
2958 raise ParseException(instring, loc, self.errmsg, self)
2960 return loc, instring[start:loc]
2962 def parseImpl_regex(self, instring, loc, do_actions=True) -> ParseImplReturnType:
2963 result = self.re_match(instring, loc)
2964 if not result:
2965 raise ParseException(instring, loc, self.errmsg, self)
2967 loc = result.end()
2968 return loc, result.group()
2971class Char(Word):
2972 """A short-cut class for defining :class:`Word` ``(characters, exact=1)``,
2973 when defining a match of any single character in a string of
2974 characters.
2975 """
2977 def __init__(
2978 self,
2979 charset: str,
2980 as_keyword: bool = False,
2981 exclude_chars: typing.Optional[str] = None,
2982 *,
2983 asKeyword: bool = False,
2984 excludeChars: typing.Optional[str] = None,
2985 ):
2986 asKeyword = asKeyword or as_keyword
2987 excludeChars = excludeChars or exclude_chars
2988 super().__init__(
2989 charset, exact=1, as_keyword=asKeyword, exclude_chars=excludeChars
2990 )
2993class Regex(Token):
2994 r"""Token for matching strings that match a given regular
2995 expression. Defined with string specifying the regular expression in
2996 a form recognized by the stdlib Python `re module <https://docs.python.org/3/library/re.html>`_.
2997 If the given regex contains named groups (defined using ``(?P<name>...)``),
2998 these will be preserved as named :class:`ParseResults`.
3000 If instead of the Python stdlib ``re`` module you wish to use a different RE module
3001 (such as the ``regex`` module), you can do so by building your ``Regex`` object with
3002 a compiled RE that was compiled using ``regex``.
3004 Example::
3006 realnum = Regex(r"[+-]?\d+\.\d*")
3007 # ref: https://stackoverflow.com/questions/267399/how-do-you-match-only-valid-roman-numerals-with-a-regular-expression
3008 roman = Regex(r"M{0,4}(CM|CD|D?{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})")
3010 # named fields in a regex will be returned as named results
3011 date = Regex(r'(?P<year>\d{4})-(?P<month>\d\d?)-(?P<day>\d\d?)')
3013 # the Regex class will accept re's compiled using the regex module
3014 import regex
3015 parser = pp.Regex(regex.compile(r'[0-9]'))
3016 """
3018 def __init__(
3019 self,
3020 pattern: Any,
3021 flags: Union[re.RegexFlag, int] = 0,
3022 as_group_list: bool = False,
3023 as_match: bool = False,
3024 *,
3025 asGroupList: bool = False,
3026 asMatch: bool = False,
3027 ):
3028 """The parameters ``pattern`` and ``flags`` are passed
3029 to the ``re.compile()`` function as-is. See the Python
3030 `re module <https://docs.python.org/3/library/re.html>`_ module for an
3031 explanation of the acceptable patterns and flags.
3032 """
3033 super().__init__()
3034 asGroupList = asGroupList or as_group_list
3035 asMatch = asMatch or as_match
3037 if isinstance(pattern, str_type):
3038 if not pattern:
3039 raise ValueError("null string passed to Regex; use Empty() instead")
3041 self._re = None
3042 self.reString = self.pattern = pattern
3043 self.flags = flags
3045 elif hasattr(pattern, "pattern") and hasattr(pattern, "match"):
3046 self._re = pattern
3047 self.pattern = self.reString = pattern.pattern
3048 self.flags = flags
3050 else:
3051 raise TypeError(
3052 "Regex may only be constructed with a string or a compiled RE object"
3053 )
3055 self.errmsg = f"Expected {self.name}"
3056 self.mayIndexError = False
3057 self.asGroupList = asGroupList
3058 self.asMatch = asMatch
3059 if self.asGroupList:
3060 self.parseImpl = self.parseImplAsGroupList # type: ignore [assignment]
3061 if self.asMatch:
3062 self.parseImpl = self.parseImplAsMatch # type: ignore [assignment]
3064 @cached_property
3065 def re(self) -> _RePattern:
3066 if self._re:
3067 return self._re
3069 try:
3070 return re.compile(self.pattern, self.flags)
3071 except re.error:
3072 raise ValueError(f"invalid pattern ({self.pattern!r}) passed to Regex")
3074 @cached_property
3075 def re_match(self) -> Callable[[str], Any]:
3076 return self.re.match
3078 @cached_property
3079 def mayReturnEmpty(self) -> bool:
3080 return self.re_match("") is not None
3082 def _generateDefaultName(self) -> str:
3083 unescaped = self.pattern.replace("\\\\", "\\")
3084 return f"Re:({unescaped!r})"
3086 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3087 result = self.re_match(instring, loc)
3088 if not result:
3089 raise ParseException(instring, loc, self.errmsg, self)
3091 loc = result.end()
3092 ret = ParseResults(result.group())
3093 d = result.groupdict()
3095 for k, v in d.items():
3096 ret[k] = v
3098 return loc, ret
3100 def parseImplAsGroupList(self, instring, loc, do_actions=True):
3101 result = self.re_match(instring, loc)
3102 if not result:
3103 raise ParseException(instring, loc, self.errmsg, self)
3105 loc = result.end()
3106 ret = result.groups()
3107 return loc, ret
3109 def parseImplAsMatch(self, instring, loc, do_actions=True):
3110 result = self.re_match(instring, loc)
3111 if not result:
3112 raise ParseException(instring, loc, self.errmsg, self)
3114 loc = result.end()
3115 ret = result
3116 return loc, ret
3118 def sub(self, repl: str) -> ParserElement:
3119 r"""
3120 Return :class:`Regex` with an attached parse action to transform the parsed
3121 result as if called using `re.sub(expr, repl, string) <https://docs.python.org/3/library/re.html#re.sub>`_.
3123 Example::
3125 make_html = Regex(r"(\w+):(.*?):").sub(r"<\1>\2</\1>")
3126 print(make_html.transform_string("h1:main title:"))
3127 # prints "<h1>main title</h1>"
3128 """
3129 if self.asGroupList:
3130 raise TypeError("cannot use sub() with Regex(as_group_list=True)")
3132 if self.asMatch and callable(repl):
3133 raise TypeError(
3134 "cannot use sub() with a callable with Regex(as_match=True)"
3135 )
3137 if self.asMatch:
3139 def pa(tokens):
3140 return tokens[0].expand(repl)
3142 else:
3144 def pa(tokens):
3145 return self.re.sub(repl, tokens[0])
3147 return self.add_parse_action(pa)
3150class QuotedString(Token):
3151 r"""
3152 Token for matching strings that are delimited by quoting characters.
3154 Defined with the following parameters:
3156 - ``quote_char`` - string of one or more characters defining the
3157 quote delimiting string
3158 - ``esc_char`` - character to re_escape quotes, typically backslash
3159 (default= ``None``)
3160 - ``esc_quote`` - special quote sequence to re_escape an embedded quote
3161 string (such as SQL's ``""`` to re_escape an embedded ``"``)
3162 (default= ``None``)
3163 - ``multiline`` - boolean indicating whether quotes can span
3164 multiple lines (default= ``False``)
3165 - ``unquote_results`` - boolean indicating whether the matched text
3166 should be unquoted (default= ``True``)
3167 - ``end_quote_char`` - string of one or more characters defining the
3168 end of the quote delimited string (default= ``None`` => same as
3169 quote_char)
3170 - ``convert_whitespace_escapes`` - convert escaped whitespace
3171 (``'\t'``, ``'\n'``, etc.) to actual whitespace
3172 (default= ``True``)
3174 Example::
3176 qs = QuotedString('"')
3177 print(qs.search_string('lsjdf "This is the quote" sldjf'))
3178 complex_qs = QuotedString('{{', end_quote_char='}}')
3179 print(complex_qs.search_string('lsjdf {{This is the "quote"}} sldjf'))
3180 sql_qs = QuotedString('"', esc_quote='""')
3181 print(sql_qs.search_string('lsjdf "This is the quote with ""embedded"" quotes" sldjf'))
3183 prints::
3185 [['This is the quote']]
3186 [['This is the "quote"']]
3187 [['This is the quote with "embedded" quotes']]
3188 """
3190 ws_map = dict(((r"\t", "\t"), (r"\n", "\n"), (r"\f", "\f"), (r"\r", "\r")))
3192 def __init__(
3193 self,
3194 quote_char: str = "",
3195 esc_char: typing.Optional[str] = None,
3196 esc_quote: typing.Optional[str] = None,
3197 multiline: bool = False,
3198 unquote_results: bool = True,
3199 end_quote_char: typing.Optional[str] = None,
3200 convert_whitespace_escapes: bool = True,
3201 *,
3202 quoteChar: str = "",
3203 escChar: typing.Optional[str] = None,
3204 escQuote: typing.Optional[str] = None,
3205 unquoteResults: bool = True,
3206 endQuoteChar: typing.Optional[str] = None,
3207 convertWhitespaceEscapes: bool = True,
3208 ):
3209 super().__init__()
3210 esc_char = escChar or esc_char
3211 esc_quote = escQuote or esc_quote
3212 unquote_results = unquoteResults and unquote_results
3213 end_quote_char = endQuoteChar or end_quote_char
3214 convert_whitespace_escapes = (
3215 convertWhitespaceEscapes and convert_whitespace_escapes
3216 )
3217 quote_char = quoteChar or quote_char
3219 # remove white space from quote chars
3220 quote_char = quote_char.strip()
3221 if not quote_char:
3222 raise ValueError("quote_char cannot be the empty string")
3224 if end_quote_char is None:
3225 end_quote_char = quote_char
3226 else:
3227 end_quote_char = end_quote_char.strip()
3228 if not end_quote_char:
3229 raise ValueError("end_quote_char cannot be the empty string")
3231 self.quote_char: str = quote_char
3232 self.quote_char_len: int = len(quote_char)
3233 self.first_quote_char: str = quote_char[0]
3234 self.end_quote_char: str = end_quote_char
3235 self.end_quote_char_len: int = len(end_quote_char)
3236 self.esc_char: str = esc_char or ""
3237 self.has_esc_char: bool = esc_char is not None
3238 self.esc_quote: str = esc_quote or ""
3239 self.unquote_results: bool = unquote_results
3240 self.convert_whitespace_escapes: bool = convert_whitespace_escapes
3241 self.multiline = multiline
3242 self.re_flags = re.RegexFlag(0)
3244 # fmt: off
3245 # build up re pattern for the content between the quote delimiters
3246 inner_pattern: List[str] = []
3248 if esc_quote:
3249 inner_pattern.append(rf"(?:{re.escape(esc_quote)})")
3251 if esc_char:
3252 inner_pattern.append(rf"(?:{re.escape(esc_char)}.)")
3254 if len(self.end_quote_char) > 1:
3255 inner_pattern.append(
3256 "(?:"
3257 + "|".join(
3258 f"(?:{re.escape(self.end_quote_char[:i])}(?!{re.escape(self.end_quote_char[i:])}))"
3259 for i in range(len(self.end_quote_char) - 1, 0, -1)
3260 )
3261 + ")"
3262 )
3264 if self.multiline:
3265 self.re_flags |= re.MULTILINE | re.DOTALL
3266 inner_pattern.append(
3267 rf"(?:[^{_escape_regex_range_chars(self.end_quote_char[0])}"
3268 rf"{(_escape_regex_range_chars(self.esc_char) if self.has_esc_char else '')}])"
3269 )
3270 else:
3271 inner_pattern.append(
3272 rf"(?:[^{_escape_regex_range_chars(self.end_quote_char[0])}\n\r"
3273 rf"{(_escape_regex_range_chars(self.esc_char) if self.has_esc_char else '')}])"
3274 )
3276 self.pattern = "".join(
3277 [
3278 re.escape(self.quote_char),
3279 "(?:",
3280 '|'.join(inner_pattern),
3281 ")*",
3282 re.escape(self.end_quote_char),
3283 ]
3284 )
3286 if self.unquote_results:
3287 if self.convert_whitespace_escapes:
3288 self.unquote_scan_re = re.compile(
3289 rf"({'|'.join(re.escape(k) for k in self.ws_map)})"
3290 rf"|({re.escape(self.esc_char)}.)"
3291 rf"|(\n|.)",
3292 flags=self.re_flags,
3293 )
3294 else:
3295 self.unquote_scan_re = re.compile(
3296 rf"({re.escape(self.esc_char)}.)"
3297 rf"|(\n|.)",
3298 flags=self.re_flags
3299 )
3300 # fmt: on
3302 try:
3303 self.re = re.compile(self.pattern, self.re_flags)
3304 self.reString = self.pattern
3305 self.re_match = self.re.match
3306 except re.error:
3307 raise ValueError(f"invalid pattern {self.pattern!r} passed to Regex")
3309 self.errmsg = f"Expected {self.name}"
3310 self.mayIndexError = False
3311 self.mayReturnEmpty = True
3313 def _generateDefaultName(self) -> str:
3314 if self.quote_char == self.end_quote_char and isinstance(
3315 self.quote_char, str_type
3316 ):
3317 return f"string enclosed in {self.quote_char!r}"
3319 return f"quoted string, starting with {self.quote_char} ending with {self.end_quote_char}"
3321 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3322 # check first character of opening quote to see if that is a match
3323 # before doing the more complicated regex match
3324 result = (
3325 instring[loc] == self.first_quote_char
3326 and self.re_match(instring, loc)
3327 or None
3328 )
3329 if not result:
3330 raise ParseException(instring, loc, self.errmsg, self)
3332 # get ending loc and matched string from regex matching result
3333 loc = result.end()
3334 ret = result.group()
3336 if self.unquote_results:
3337 # strip off quotes
3338 ret = ret[self.quote_char_len : -self.end_quote_char_len]
3340 if isinstance(ret, str_type):
3341 # fmt: off
3342 if self.convert_whitespace_escapes:
3343 # as we iterate over matches in the input string,
3344 # collect from whichever match group of the unquote_scan_re
3345 # regex matches (only 1 group will match at any given time)
3346 ret = "".join(
3347 # match group 1 matches \t, \n, etc.
3348 self.ws_map[match.group(1)] if match.group(1)
3349 # match group 2 matches escaped characters
3350 else match.group(2)[-1] if match.group(2)
3351 # match group 3 matches any character
3352 else match.group(3)
3353 for match in self.unquote_scan_re.finditer(ret)
3354 )
3355 else:
3356 ret = "".join(
3357 # match group 1 matches escaped characters
3358 match.group(1)[-1] if match.group(1)
3359 # match group 2 matches any character
3360 else match.group(2)
3361 for match in self.unquote_scan_re.finditer(ret)
3362 )
3363 # fmt: on
3365 # replace escaped quotes
3366 if self.esc_quote:
3367 ret = ret.replace(self.esc_quote, self.end_quote_char)
3369 return loc, ret
3372class CharsNotIn(Token):
3373 """Token for matching words composed of characters *not* in a given
3374 set (will include whitespace in matched characters if not listed in
3375 the provided exclusion set - see example). Defined with string
3376 containing all disallowed characters, and an optional minimum,
3377 maximum, and/or exact length. The default value for ``min`` is
3378 1 (a minimum value < 1 is not valid); the default values for
3379 ``max`` and ``exact`` are 0, meaning no maximum or exact
3380 length restriction.
3382 Example::
3384 # define a comma-separated-value as anything that is not a ','
3385 csv_value = CharsNotIn(',')
3386 print(DelimitedList(csv_value).parse_string("dkls,lsdkjf,s12 34,@!#,213"))
3388 prints::
3390 ['dkls', 'lsdkjf', 's12 34', '@!#', '213']
3391 """
3393 def __init__(
3394 self,
3395 not_chars: str = "",
3396 min: int = 1,
3397 max: int = 0,
3398 exact: int = 0,
3399 *,
3400 notChars: str = "",
3401 ):
3402 super().__init__()
3403 self.skipWhitespace = False
3404 self.notChars = not_chars or notChars
3405 self.notCharsSet = set(self.notChars)
3407 if min < 1:
3408 raise ValueError(
3409 "cannot specify a minimum length < 1; use"
3410 " Opt(CharsNotIn()) if zero-length char group is permitted"
3411 )
3413 self.minLen = min
3415 if max > 0:
3416 self.maxLen = max
3417 else:
3418 self.maxLen = _MAX_INT
3420 if exact > 0:
3421 self.maxLen = exact
3422 self.minLen = exact
3424 self.errmsg = f"Expected {self.name}"
3425 self.mayReturnEmpty = self.minLen == 0
3426 self.mayIndexError = False
3428 def _generateDefaultName(self) -> str:
3429 not_chars_str = _collapse_string_to_ranges(self.notChars)
3430 if len(not_chars_str) > 16:
3431 return f"!W:({self.notChars[: 16 - 3]}...)"
3432 else:
3433 return f"!W:({self.notChars})"
3435 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3436 notchars = self.notCharsSet
3437 if instring[loc] in notchars:
3438 raise ParseException(instring, loc, self.errmsg, self)
3440 start = loc
3441 loc += 1
3442 maxlen = min(start + self.maxLen, len(instring))
3443 while loc < maxlen and instring[loc] not in notchars:
3444 loc += 1
3446 if loc - start < self.minLen:
3447 raise ParseException(instring, loc, self.errmsg, self)
3449 return loc, instring[start:loc]
3452class White(Token):
3453 """Special matching class for matching whitespace. Normally,
3454 whitespace is ignored by pyparsing grammars. This class is included
3455 when some whitespace structures are significant. Define with
3456 a string containing the whitespace characters to be matched; default
3457 is ``" \\t\\r\\n"``. Also takes optional ``min``,
3458 ``max``, and ``exact`` arguments, as defined for the
3459 :class:`Word` class.
3460 """
3462 whiteStrs = {
3463 " ": "<SP>",
3464 "\t": "<TAB>",
3465 "\n": "<LF>",
3466 "\r": "<CR>",
3467 "\f": "<FF>",
3468 "\u00A0": "<NBSP>",
3469 "\u1680": "<OGHAM_SPACE_MARK>",
3470 "\u180E": "<MONGOLIAN_VOWEL_SEPARATOR>",
3471 "\u2000": "<EN_QUAD>",
3472 "\u2001": "<EM_QUAD>",
3473 "\u2002": "<EN_SPACE>",
3474 "\u2003": "<EM_SPACE>",
3475 "\u2004": "<THREE-PER-EM_SPACE>",
3476 "\u2005": "<FOUR-PER-EM_SPACE>",
3477 "\u2006": "<SIX-PER-EM_SPACE>",
3478 "\u2007": "<FIGURE_SPACE>",
3479 "\u2008": "<PUNCTUATION_SPACE>",
3480 "\u2009": "<THIN_SPACE>",
3481 "\u200A": "<HAIR_SPACE>",
3482 "\u200B": "<ZERO_WIDTH_SPACE>",
3483 "\u202F": "<NNBSP>",
3484 "\u205F": "<MMSP>",
3485 "\u3000": "<IDEOGRAPHIC_SPACE>",
3486 }
3488 def __init__(self, ws: str = " \t\r\n", min: int = 1, max: int = 0, exact: int = 0):
3489 super().__init__()
3490 self.matchWhite = ws
3491 self.set_whitespace_chars(
3492 "".join(c for c in self.whiteStrs if c not in self.matchWhite),
3493 copy_defaults=True,
3494 )
3495 # self.leave_whitespace()
3496 self.mayReturnEmpty = True
3497 self.errmsg = f"Expected {self.name}"
3499 self.minLen = min
3501 if max > 0:
3502 self.maxLen = max
3503 else:
3504 self.maxLen = _MAX_INT
3506 if exact > 0:
3507 self.maxLen = exact
3508 self.minLen = exact
3510 def _generateDefaultName(self) -> str:
3511 return "".join(White.whiteStrs[c] for c in self.matchWhite)
3513 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3514 if instring[loc] not in self.matchWhite:
3515 raise ParseException(instring, loc, self.errmsg, self)
3516 start = loc
3517 loc += 1
3518 maxloc = start + self.maxLen
3519 maxloc = min(maxloc, len(instring))
3520 while loc < maxloc and instring[loc] in self.matchWhite:
3521 loc += 1
3523 if loc - start < self.minLen:
3524 raise ParseException(instring, loc, self.errmsg, self)
3526 return loc, instring[start:loc]
3529class PositionToken(Token):
3530 def __init__(self):
3531 super().__init__()
3532 self.mayReturnEmpty = True
3533 self.mayIndexError = False
3536class GoToColumn(PositionToken):
3537 """Token to advance to a specific column of input text; useful for
3538 tabular report scraping.
3539 """
3541 def __init__(self, colno: int):
3542 super().__init__()
3543 self.col = colno
3545 def preParse(self, instring: str, loc: int) -> int:
3546 if col(loc, instring) == self.col:
3547 return loc
3549 instrlen = len(instring)
3550 if self.ignoreExprs:
3551 loc = self._skipIgnorables(instring, loc)
3552 while (
3553 loc < instrlen
3554 and instring[loc].isspace()
3555 and col(loc, instring) != self.col
3556 ):
3557 loc += 1
3559 return loc
3561 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3562 thiscol = col(loc, instring)
3563 if thiscol > self.col:
3564 raise ParseException(instring, loc, "Text not in expected column", self)
3565 newloc = loc + self.col - thiscol
3566 ret = instring[loc:newloc]
3567 return newloc, ret
3570class LineStart(PositionToken):
3571 r"""Matches if current position is at the beginning of a line within
3572 the parse string
3574 Example::
3576 test = '''\
3577 AAA this line
3578 AAA and this line
3579 AAA but not this one
3580 B AAA and definitely not this one
3581 '''
3583 for t in (LineStart() + 'AAA' + rest_of_line).search_string(test):
3584 print(t)
3586 prints::
3588 ['AAA', ' this line']
3589 ['AAA', ' and this line']
3591 """
3593 def __init__(self):
3594 super().__init__()
3595 self.leave_whitespace()
3596 self.orig_whiteChars = set() | self.whiteChars
3597 self.whiteChars.discard("\n")
3598 self.skipper = Empty().set_whitespace_chars(self.whiteChars)
3599 self.set_name("start of line")
3601 def preParse(self, instring: str, loc: int) -> int:
3602 if loc == 0:
3603 return loc
3605 ret = self.skipper.preParse(instring, loc)
3607 if "\n" in self.orig_whiteChars:
3608 while instring[ret : ret + 1] == "\n":
3609 ret = self.skipper.preParse(instring, ret + 1)
3611 return ret
3613 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3614 if col(loc, instring) == 1:
3615 return loc, []
3616 raise ParseException(instring, loc, self.errmsg, self)
3619class LineEnd(PositionToken):
3620 """Matches if current position is at the end of a line within the
3621 parse string
3622 """
3624 def __init__(self):
3625 super().__init__()
3626 self.whiteChars.discard("\n")
3627 self.set_whitespace_chars(self.whiteChars, copy_defaults=False)
3628 self.set_name("end of line")
3630 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3631 if loc < len(instring):
3632 if instring[loc] == "\n":
3633 return loc + 1, "\n"
3634 else:
3635 raise ParseException(instring, loc, self.errmsg, self)
3636 elif loc == len(instring):
3637 return loc + 1, []
3638 else:
3639 raise ParseException(instring, loc, self.errmsg, self)
3642class StringStart(PositionToken):
3643 """Matches if current position is at the beginning of the parse
3644 string
3645 """
3647 def __init__(self):
3648 super().__init__()
3649 self.set_name("start of text")
3651 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3652 # see if entire string up to here is just whitespace and ignoreables
3653 if loc != 0 and loc != self.preParse(instring, 0):
3654 raise ParseException(instring, loc, self.errmsg, self)
3656 return loc, []
3659class StringEnd(PositionToken):
3660 """
3661 Matches if current position is at the end of the parse string
3662 """
3664 def __init__(self):
3665 super().__init__()
3666 self.set_name("end of text")
3668 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3669 if loc < len(instring):
3670 raise ParseException(instring, loc, self.errmsg, self)
3671 if loc == len(instring):
3672 return loc + 1, []
3673 if loc > len(instring):
3674 return loc, []
3676 raise ParseException(instring, loc, self.errmsg, self)
3679class WordStart(PositionToken):
3680 """Matches if the current position is at the beginning of a
3681 :class:`Word`, and is not preceded by any character in a given
3682 set of ``word_chars`` (default= ``printables``). To emulate the
3683 ``\b`` behavior of regular expressions, use
3684 ``WordStart(alphanums)``. ``WordStart`` will also match at
3685 the beginning of the string being parsed, or at the beginning of
3686 a line.
3687 """
3689 def __init__(self, word_chars: str = printables, *, wordChars: str = printables):
3690 wordChars = word_chars if wordChars == printables else wordChars
3691 super().__init__()
3692 self.wordChars = set(wordChars)
3693 self.set_name("start of a word")
3695 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3696 if loc != 0:
3697 if (
3698 instring[loc - 1] in self.wordChars
3699 or instring[loc] not in self.wordChars
3700 ):
3701 raise ParseException(instring, loc, self.errmsg, self)
3702 return loc, []
3705class WordEnd(PositionToken):
3706 """Matches if the current position is at the end of a :class:`Word`,
3707 and is not followed by any character in a given set of ``word_chars``
3708 (default= ``printables``). To emulate the ``\b`` behavior of
3709 regular expressions, use ``WordEnd(alphanums)``. ``WordEnd``
3710 will also match at the end of the string being parsed, or at the end
3711 of a line.
3712 """
3714 def __init__(self, word_chars: str = printables, *, wordChars: str = printables):
3715 wordChars = word_chars if wordChars == printables else wordChars
3716 super().__init__()
3717 self.wordChars = set(wordChars)
3718 self.skipWhitespace = False
3719 self.set_name("end of a word")
3721 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
3722 instrlen = len(instring)
3723 if instrlen > 0 and loc < instrlen:
3724 if (
3725 instring[loc] in self.wordChars
3726 or instring[loc - 1] not in self.wordChars
3727 ):
3728 raise ParseException(instring, loc, self.errmsg, self)
3729 return loc, []
3732class Tag(Token):
3733 """
3734 A meta-element for inserting a named result into the parsed
3735 tokens that may be checked later in a parse action or while
3736 processing the parsed results. Accepts an optional tag value,
3737 defaulting to `True`.
3739 Example::
3741 end_punc = "." | ("!" + Tag("enthusiastic")))
3742 greeting = "Hello," + Word(alphas) + end_punc
3744 result = greeting.parse_string("Hello, World.")
3745 print(result.dump())
3747 result = greeting.parse_string("Hello, World!")
3748 print(result.dump())
3750 prints::
3752 ['Hello,', 'World', '.']
3754 ['Hello,', 'World', '!']
3755 - enthusiastic: True
3756 """
3757 def __init__(self, tag_name: str, value: Any = True):
3758 super().__init__()
3759 self.mayReturnEmpty = True
3760 self.mayIndexError = False
3761 self.leave_whitespace()
3762 self.tag_name = tag_name
3763 self.tag_value = value
3764 self.add_parse_action(self._add_tag)
3766 def _add_tag(self, tokens: ParseResults):
3767 tokens[self.tag_name] = self.tag_value
3769 def _generateDefaultName(self) -> str:
3770 return f"{type(self).__name__}:{self.tag_name}={self.tag_value!r}"
3773class ParseExpression(ParserElement):
3774 """Abstract subclass of ParserElement, for combining and
3775 post-processing parsed tokens.
3776 """
3778 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = False):
3779 super().__init__(savelist)
3780 self.exprs: List[ParserElement]
3781 if isinstance(exprs, _generatorType):
3782 exprs = list(exprs)
3784 if isinstance(exprs, str_type):
3785 self.exprs = [self._literalStringClass(exprs)]
3786 elif isinstance(exprs, ParserElement):
3787 self.exprs = [exprs]
3788 elif isinstance(exprs, Iterable):
3789 exprs = list(exprs)
3790 # if sequence of strings provided, wrap with Literal
3791 if any(isinstance(expr, str_type) for expr in exprs):
3792 exprs = (
3793 self._literalStringClass(e) if isinstance(e, str_type) else e
3794 for e in exprs
3795 )
3796 self.exprs = list(exprs)
3797 else:
3798 try:
3799 self.exprs = list(exprs)
3800 except TypeError:
3801 self.exprs = [exprs]
3802 self.callPreparse = False
3804 def recurse(self) -> List[ParserElement]:
3805 return self.exprs[:]
3807 def append(self, other) -> ParserElement:
3808 self.exprs.append(other)
3809 self._defaultName = None
3810 return self
3812 def leave_whitespace(self, recursive: bool = True) -> ParserElement:
3813 """
3814 Extends ``leave_whitespace`` defined in base class, and also invokes ``leave_whitespace`` on
3815 all contained expressions.
3816 """
3817 super().leave_whitespace(recursive)
3819 if recursive:
3820 self.exprs = [e.copy() for e in self.exprs]
3821 for e in self.exprs:
3822 e.leave_whitespace(recursive)
3823 return self
3825 def ignore_whitespace(self, recursive: bool = True) -> ParserElement:
3826 """
3827 Extends ``ignore_whitespace`` defined in base class, and also invokes ``leave_whitespace`` on
3828 all contained expressions.
3829 """
3830 super().ignore_whitespace(recursive)
3831 if recursive:
3832 self.exprs = [e.copy() for e in self.exprs]
3833 for e in self.exprs:
3834 e.ignore_whitespace(recursive)
3835 return self
3837 def ignore(self, other) -> ParserElement:
3838 if isinstance(other, Suppress):
3839 if other not in self.ignoreExprs:
3840 super().ignore(other)
3841 for e in self.exprs:
3842 e.ignore(self.ignoreExprs[-1])
3843 else:
3844 super().ignore(other)
3845 for e in self.exprs:
3846 e.ignore(self.ignoreExprs[-1])
3847 return self
3849 def _generateDefaultName(self) -> str:
3850 return f"{type(self).__name__}:({self.exprs})"
3852 def streamline(self) -> ParserElement:
3853 if self.streamlined:
3854 return self
3856 super().streamline()
3858 for e in self.exprs:
3859 e.streamline()
3861 # collapse nested :class:`And`'s of the form ``And(And(And(a, b), c), d)`` to ``And(a, b, c, d)``
3862 # but only if there are no parse actions or resultsNames on the nested And's
3863 # (likewise for :class:`Or`'s and :class:`MatchFirst`'s)
3864 if len(self.exprs) == 2:
3865 other = self.exprs[0]
3866 if (
3867 isinstance(other, self.__class__)
3868 and not other.parseAction
3869 and other.resultsName is None
3870 and not other.debug
3871 ):
3872 self.exprs = other.exprs[:] + [self.exprs[1]]
3873 self._defaultName = None
3874 self.mayReturnEmpty |= other.mayReturnEmpty
3875 self.mayIndexError |= other.mayIndexError
3877 other = self.exprs[-1]
3878 if (
3879 isinstance(other, self.__class__)
3880 and not other.parseAction
3881 and other.resultsName is None
3882 and not other.debug
3883 ):
3884 self.exprs = self.exprs[:-1] + other.exprs[:]
3885 self._defaultName = None
3886 self.mayReturnEmpty |= other.mayReturnEmpty
3887 self.mayIndexError |= other.mayIndexError
3889 self.errmsg = f"Expected {self}"
3891 return self
3893 def validate(self, validateTrace=None) -> None:
3894 warnings.warn(
3895 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
3896 DeprecationWarning,
3897 stacklevel=2,
3898 )
3899 tmp = (validateTrace if validateTrace is not None else [])[:] + [self]
3900 for e in self.exprs:
3901 e.validate(tmp)
3902 self._checkRecursion([])
3904 def copy(self) -> ParserElement:
3905 ret = super().copy()
3906 ret = typing.cast(ParseExpression, ret)
3907 ret.exprs = [e.copy() for e in self.exprs]
3908 return ret
3910 def _setResultsName(self, name, list_all_matches=False) -> ParserElement:
3911 if not (
3912 __diag__.warn_ungrouped_named_tokens_in_collection
3913 and Diagnostics.warn_ungrouped_named_tokens_in_collection
3914 not in self.suppress_warnings_
3915 ):
3916 return super()._setResultsName(name, list_all_matches)
3918 for e in self.exprs:
3919 if (
3920 isinstance(e, ParserElement)
3921 and e.resultsName
3922 and (
3923 Diagnostics.warn_ungrouped_named_tokens_in_collection
3924 not in e.suppress_warnings_
3925 )
3926 ):
3927 warning = (
3928 "warn_ungrouped_named_tokens_in_collection:"
3929 f" setting results name {name!r} on {type(self).__name__} expression"
3930 f" collides with {e.resultsName!r} on contained expression"
3931 )
3932 warnings.warn(warning, stacklevel=3)
3933 break
3935 return super()._setResultsName(name, list_all_matches)
3937 # Compatibility synonyms
3938 # fmt: off
3939 leaveWhitespace = replaced_by_pep8("leaveWhitespace", leave_whitespace)
3940 ignoreWhitespace = replaced_by_pep8("ignoreWhitespace", ignore_whitespace)
3941 # fmt: on
3944class And(ParseExpression):
3945 """
3946 Requires all given :class:`ParseExpression` s to be found in the given order.
3947 Expressions may be separated by whitespace.
3948 May be constructed using the ``'+'`` operator.
3949 May also be constructed using the ``'-'`` operator, which will
3950 suppress backtracking.
3952 Example::
3954 integer = Word(nums)
3955 name_expr = Word(alphas)[1, ...]
3957 expr = And([integer("id"), name_expr("name"), integer("age")])
3958 # more easily written as:
3959 expr = integer("id") + name_expr("name") + integer("age")
3960 """
3962 class _ErrorStop(Empty):
3963 def __init__(self, *args, **kwargs):
3964 super().__init__(*args, **kwargs)
3965 self.leave_whitespace()
3967 def _generateDefaultName(self) -> str:
3968 return "-"
3970 def __init__(
3971 self, exprs_arg: typing.Iterable[ParserElement], savelist: bool = True
3972 ):
3973 exprs: List[ParserElement] = list(exprs_arg)
3974 if exprs and Ellipsis in exprs:
3975 tmp: List[ParserElement] = []
3976 for i, expr in enumerate(exprs):
3977 if expr is not Ellipsis:
3978 tmp.append(expr)
3979 continue
3981 if i < len(exprs) - 1:
3982 skipto_arg: ParserElement = typing.cast(
3983 ParseExpression, (Empty() + exprs[i + 1])
3984 ).exprs[-1]
3985 tmp.append(SkipTo(skipto_arg)("_skipped*"))
3986 continue
3988 raise Exception("cannot construct And with sequence ending in ...")
3989 exprs[:] = tmp
3990 super().__init__(exprs, savelist)
3991 if self.exprs:
3992 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
3993 if not isinstance(self.exprs[0], White):
3994 self.set_whitespace_chars(
3995 self.exprs[0].whiteChars,
3996 copy_defaults=self.exprs[0].copyDefaultWhiteChars,
3997 )
3998 self.skipWhitespace = self.exprs[0].skipWhitespace
3999 else:
4000 self.skipWhitespace = False
4001 else:
4002 self.mayReturnEmpty = True
4003 self.callPreparse = True
4005 def streamline(self) -> ParserElement:
4006 # collapse any _PendingSkip's
4007 if self.exprs and any(
4008 isinstance(e, ParseExpression)
4009 and e.exprs
4010 and isinstance(e.exprs[-1], _PendingSkip)
4011 for e in self.exprs[:-1]
4012 ):
4013 deleted_expr_marker = NoMatch()
4014 for i, e in enumerate(self.exprs[:-1]):
4015 if e is deleted_expr_marker:
4016 continue
4017 if (
4018 isinstance(e, ParseExpression)
4019 and e.exprs
4020 and isinstance(e.exprs[-1], _PendingSkip)
4021 ):
4022 e.exprs[-1] = e.exprs[-1] + self.exprs[i + 1]
4023 self.exprs[i + 1] = deleted_expr_marker
4024 self.exprs = [e for e in self.exprs if e is not deleted_expr_marker]
4026 super().streamline()
4028 # link any IndentedBlocks to the prior expression
4029 prev: ParserElement
4030 cur: ParserElement
4031 for prev, cur in zip(self.exprs, self.exprs[1:]):
4032 # traverse cur or any first embedded expr of cur looking for an IndentedBlock
4033 # (but watch out for recursive grammar)
4034 seen = set()
4035 while True:
4036 if id(cur) in seen:
4037 break
4038 seen.add(id(cur))
4039 if isinstance(cur, IndentedBlock):
4040 prev.add_parse_action(
4041 lambda s, l, t, cur_=cur: setattr(
4042 cur_, "parent_anchor", col(l, s)
4043 )
4044 )
4045 break
4046 subs = cur.recurse()
4047 next_first = next(iter(subs), None)
4048 if next_first is None:
4049 break
4050 cur = typing.cast(ParserElement, next_first)
4052 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
4053 return self
4055 def parseImpl(self, instring, loc, do_actions=True):
4056 # pass False as callPreParse arg to _parse for first element, since we already
4057 # pre-parsed the string as part of our And pre-parsing
4058 loc, resultlist = self.exprs[0]._parse(
4059 instring, loc, do_actions, callPreParse=False
4060 )
4061 errorStop = False
4062 for e in self.exprs[1:]:
4063 # if isinstance(e, And._ErrorStop):
4064 if type(e) is And._ErrorStop:
4065 errorStop = True
4066 continue
4067 if errorStop:
4068 try:
4069 loc, exprtokens = e._parse(instring, loc, do_actions)
4070 except ParseSyntaxException:
4071 raise
4072 except ParseBaseException as pe:
4073 pe.__traceback__ = None
4074 raise ParseSyntaxException._from_exception(pe)
4075 except IndexError:
4076 raise ParseSyntaxException(
4077 instring, len(instring), self.errmsg, self
4078 )
4079 else:
4080 loc, exprtokens = e._parse(instring, loc, do_actions)
4081 resultlist += exprtokens
4082 return loc, resultlist
4084 def __iadd__(self, other):
4085 if isinstance(other, str_type):
4086 other = self._literalStringClass(other)
4087 if not isinstance(other, ParserElement):
4088 return NotImplemented
4089 return self.append(other) # And([self, other])
4091 def _checkRecursion(self, parseElementList):
4092 subRecCheckList = parseElementList[:] + [self]
4093 for e in self.exprs:
4094 e._checkRecursion(subRecCheckList)
4095 if not e.mayReturnEmpty:
4096 break
4098 def _generateDefaultName(self) -> str:
4099 inner = " ".join(str(e) for e in self.exprs)
4100 # strip off redundant inner {}'s
4101 while len(inner) > 1 and inner[0 :: len(inner) - 1] == "{}":
4102 inner = inner[1:-1]
4103 return f"{{{inner}}}"
4106class Or(ParseExpression):
4107 """Requires that at least one :class:`ParseExpression` is found. If
4108 two expressions match, the expression that matches the longest
4109 string will be used. May be constructed using the ``'^'``
4110 operator.
4112 Example::
4114 # construct Or using '^' operator
4116 number = Word(nums) ^ Combine(Word(nums) + '.' + Word(nums))
4117 print(number.search_string("123 3.1416 789"))
4119 prints::
4121 [['123'], ['3.1416'], ['789']]
4122 """
4124 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = False):
4125 super().__init__(exprs, savelist)
4126 if self.exprs:
4127 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4128 self.skipWhitespace = all(e.skipWhitespace for e in self.exprs)
4129 else:
4130 self.mayReturnEmpty = True
4132 def streamline(self) -> ParserElement:
4133 super().streamline()
4134 if self.exprs:
4135 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4136 self.saveAsList = any(e.saveAsList for e in self.exprs)
4137 self.skipWhitespace = all(
4138 e.skipWhitespace and not isinstance(e, White) for e in self.exprs
4139 )
4140 else:
4141 self.saveAsList = False
4142 return self
4144 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4145 maxExcLoc = -1
4146 maxException = None
4147 matches: List[Tuple[int, ParserElement]] = []
4148 fatals: List[ParseFatalException] = []
4149 if all(e.callPreparse for e in self.exprs):
4150 loc = self.preParse(instring, loc)
4151 for e in self.exprs:
4152 try:
4153 loc2 = e.try_parse(instring, loc, raise_fatal=True)
4154 except ParseFatalException as pfe:
4155 pfe.__traceback__ = None
4156 pfe.parser_element = e
4157 fatals.append(pfe)
4158 maxException = None
4159 maxExcLoc = -1
4160 except ParseException as err:
4161 if not fatals:
4162 err.__traceback__ = None
4163 if err.loc > maxExcLoc:
4164 maxException = err
4165 maxExcLoc = err.loc
4166 except IndexError:
4167 if len(instring) > maxExcLoc:
4168 maxException = ParseException(
4169 instring, len(instring), e.errmsg, self
4170 )
4171 maxExcLoc = len(instring)
4172 else:
4173 # save match among all matches, to retry longest to shortest
4174 matches.append((loc2, e))
4176 if matches:
4177 # re-evaluate all matches in descending order of length of match, in case attached actions
4178 # might change whether or how much they match of the input.
4179 matches.sort(key=itemgetter(0), reverse=True)
4181 if not do_actions:
4182 # no further conditions or parse actions to change the selection of
4183 # alternative, so the first match will be the best match
4184 best_expr = matches[0][1]
4185 return best_expr._parse(instring, loc, do_actions)
4187 longest = -1, None
4188 for loc1, expr1 in matches:
4189 if loc1 <= longest[0]:
4190 # already have a longer match than this one will deliver, we are done
4191 return longest
4193 try:
4194 loc2, toks = expr1._parse(instring, loc, do_actions)
4195 except ParseException as err:
4196 err.__traceback__ = None
4197 if err.loc > maxExcLoc:
4198 maxException = err
4199 maxExcLoc = err.loc
4200 else:
4201 if loc2 >= loc1:
4202 return loc2, toks
4203 # didn't match as much as before
4204 elif loc2 > longest[0]:
4205 longest = loc2, toks
4207 if longest != (-1, None):
4208 return longest
4210 if fatals:
4211 if len(fatals) > 1:
4212 fatals.sort(key=lambda e: -e.loc)
4213 if fatals[0].loc == fatals[1].loc:
4214 fatals.sort(key=lambda e: (-e.loc, -len(str(e.parser_element))))
4215 max_fatal = fatals[0]
4216 raise max_fatal
4218 if maxException is not None:
4219 # infer from this check that all alternatives failed at the current position
4220 # so emit this collective error message instead of any single error message
4221 if maxExcLoc == loc:
4222 maxException.msg = self.errmsg
4223 raise maxException
4225 raise ParseException(instring, loc, "no defined alternatives to match", self)
4227 def __ixor__(self, other):
4228 if isinstance(other, str_type):
4229 other = self._literalStringClass(other)
4230 if not isinstance(other, ParserElement):
4231 return NotImplemented
4232 return self.append(other) # Or([self, other])
4234 def _generateDefaultName(self) -> str:
4235 return f"{{{' ^ '.join(str(e) for e in self.exprs)}}}"
4237 def _setResultsName(self, name, list_all_matches=False) -> ParserElement:
4238 if (
4239 __diag__.warn_multiple_tokens_in_named_alternation
4240 and Diagnostics.warn_multiple_tokens_in_named_alternation
4241 not in self.suppress_warnings_
4242 ):
4243 if any(
4244 isinstance(e, And)
4245 and Diagnostics.warn_multiple_tokens_in_named_alternation
4246 not in e.suppress_warnings_
4247 for e in self.exprs
4248 ):
4249 warning = (
4250 "warn_multiple_tokens_in_named_alternation:"
4251 f" setting results name {name!r} on {type(self).__name__} expression"
4252 " will return a list of all parsed tokens in an And alternative,"
4253 " in prior versions only the first token was returned; enclose"
4254 " contained argument in Group"
4255 )
4256 warnings.warn(warning, stacklevel=3)
4258 return super()._setResultsName(name, list_all_matches)
4261class MatchFirst(ParseExpression):
4262 """Requires that at least one :class:`ParseExpression` is found. If
4263 more than one expression matches, the first one listed is the one that will
4264 match. May be constructed using the ``'|'`` operator.
4266 Example::
4268 # construct MatchFirst using '|' operator
4270 # watch the order of expressions to match
4271 number = Word(nums) | Combine(Word(nums) + '.' + Word(nums))
4272 print(number.search_string("123 3.1416 789")) # Fail! -> [['123'], ['3'], ['1416'], ['789']]
4274 # put more selective expression first
4275 number = Combine(Word(nums) + '.' + Word(nums)) | Word(nums)
4276 print(number.search_string("123 3.1416 789")) # Better -> [['123'], ['3.1416'], ['789']]
4277 """
4279 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = False):
4280 super().__init__(exprs, savelist)
4281 if self.exprs:
4282 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4283 self.skipWhitespace = all(e.skipWhitespace for e in self.exprs)
4284 else:
4285 self.mayReturnEmpty = True
4287 def streamline(self) -> ParserElement:
4288 if self.streamlined:
4289 return self
4291 super().streamline()
4292 if self.exprs:
4293 self.saveAsList = any(e.saveAsList for e in self.exprs)
4294 self.mayReturnEmpty = any(e.mayReturnEmpty for e in self.exprs)
4295 self.skipWhitespace = all(
4296 e.skipWhitespace and not isinstance(e, White) for e in self.exprs
4297 )
4298 else:
4299 self.saveAsList = False
4300 self.mayReturnEmpty = True
4301 return self
4303 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4304 maxExcLoc = -1
4305 maxException = None
4307 for e in self.exprs:
4308 try:
4309 return e._parse(instring, loc, do_actions)
4310 except ParseFatalException as pfe:
4311 pfe.__traceback__ = None
4312 pfe.parser_element = e
4313 raise
4314 except ParseException as err:
4315 if err.loc > maxExcLoc:
4316 maxException = err
4317 maxExcLoc = err.loc
4318 except IndexError:
4319 if len(instring) > maxExcLoc:
4320 maxException = ParseException(
4321 instring, len(instring), e.errmsg, self
4322 )
4323 maxExcLoc = len(instring)
4325 if maxException is not None:
4326 # infer from this check that all alternatives failed at the current position
4327 # so emit this collective error message instead of any individual error message
4328 if maxExcLoc == loc:
4329 maxException.msg = self.errmsg
4330 raise maxException
4332 raise ParseException(instring, loc, "no defined alternatives to match", self)
4334 def __ior__(self, other):
4335 if isinstance(other, str_type):
4336 other = self._literalStringClass(other)
4337 if not isinstance(other, ParserElement):
4338 return NotImplemented
4339 return self.append(other) # MatchFirst([self, other])
4341 def _generateDefaultName(self) -> str:
4342 return f"{{{' | '.join(str(e) for e in self.exprs)}}}"
4344 def _setResultsName(self, name, list_all_matches=False) -> ParserElement:
4345 if (
4346 __diag__.warn_multiple_tokens_in_named_alternation
4347 and Diagnostics.warn_multiple_tokens_in_named_alternation
4348 not in self.suppress_warnings_
4349 ):
4350 if any(
4351 isinstance(e, And)
4352 and Diagnostics.warn_multiple_tokens_in_named_alternation
4353 not in e.suppress_warnings_
4354 for e in self.exprs
4355 ):
4356 warning = (
4357 "warn_multiple_tokens_in_named_alternation:"
4358 f" setting results name {name!r} on {type(self).__name__} expression"
4359 " will return a list of all parsed tokens in an And alternative,"
4360 " in prior versions only the first token was returned; enclose"
4361 " contained argument in Group"
4362 )
4363 warnings.warn(warning, stacklevel=3)
4365 return super()._setResultsName(name, list_all_matches)
4368class Each(ParseExpression):
4369 """Requires all given :class:`ParseExpression` s to be found, but in
4370 any order. Expressions may be separated by whitespace.
4372 May be constructed using the ``'&'`` operator.
4374 Example::
4376 color = one_of("RED ORANGE YELLOW GREEN BLUE PURPLE BLACK WHITE BROWN")
4377 shape_type = one_of("SQUARE CIRCLE TRIANGLE STAR HEXAGON OCTAGON")
4378 integer = Word(nums)
4379 shape_attr = "shape:" + shape_type("shape")
4380 posn_attr = "posn:" + Group(integer("x") + ',' + integer("y"))("posn")
4381 color_attr = "color:" + color("color")
4382 size_attr = "size:" + integer("size")
4384 # use Each (using operator '&') to accept attributes in any order
4385 # (shape and posn are required, color and size are optional)
4386 shape_spec = shape_attr & posn_attr & Opt(color_attr) & Opt(size_attr)
4388 shape_spec.run_tests('''
4389 shape: SQUARE color: BLACK posn: 100, 120
4390 shape: CIRCLE size: 50 color: BLUE posn: 50,80
4391 color:GREEN size:20 shape:TRIANGLE posn:20,40
4392 '''
4393 )
4395 prints::
4397 shape: SQUARE color: BLACK posn: 100, 120
4398 ['shape:', 'SQUARE', 'color:', 'BLACK', 'posn:', ['100', ',', '120']]
4399 - color: BLACK
4400 - posn: ['100', ',', '120']
4401 - x: 100
4402 - y: 120
4403 - shape: SQUARE
4406 shape: CIRCLE size: 50 color: BLUE posn: 50,80
4407 ['shape:', 'CIRCLE', 'size:', '50', 'color:', 'BLUE', 'posn:', ['50', ',', '80']]
4408 - color: BLUE
4409 - posn: ['50', ',', '80']
4410 - x: 50
4411 - y: 80
4412 - shape: CIRCLE
4413 - size: 50
4416 color: GREEN size: 20 shape: TRIANGLE posn: 20,40
4417 ['color:', 'GREEN', 'size:', '20', 'shape:', 'TRIANGLE', 'posn:', ['20', ',', '40']]
4418 - color: GREEN
4419 - posn: ['20', ',', '40']
4420 - x: 20
4421 - y: 40
4422 - shape: TRIANGLE
4423 - size: 20
4424 """
4426 def __init__(self, exprs: typing.Iterable[ParserElement], savelist: bool = True):
4427 super().__init__(exprs, savelist)
4428 if self.exprs:
4429 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
4430 else:
4431 self.mayReturnEmpty = True
4432 self.skipWhitespace = True
4433 self.initExprGroups = True
4434 self.saveAsList = True
4436 def __iand__(self, other):
4437 if isinstance(other, str_type):
4438 other = self._literalStringClass(other)
4439 if not isinstance(other, ParserElement):
4440 return NotImplemented
4441 return self.append(other) # Each([self, other])
4443 def streamline(self) -> ParserElement:
4444 super().streamline()
4445 if self.exprs:
4446 self.mayReturnEmpty = all(e.mayReturnEmpty for e in self.exprs)
4447 else:
4448 self.mayReturnEmpty = True
4449 return self
4451 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4452 if self.initExprGroups:
4453 self.opt1map = dict(
4454 (id(e.expr), e) for e in self.exprs if isinstance(e, Opt)
4455 )
4456 opt1 = [e.expr for e in self.exprs if isinstance(e, Opt)]
4457 opt2 = [
4458 e
4459 for e in self.exprs
4460 if e.mayReturnEmpty and not isinstance(e, (Opt, Regex, ZeroOrMore))
4461 ]
4462 self.optionals = opt1 + opt2
4463 self.multioptionals = [
4464 e.expr.set_results_name(e.resultsName, list_all_matches=True)
4465 for e in self.exprs
4466 if isinstance(e, _MultipleMatch)
4467 ]
4468 self.multirequired = [
4469 e.expr.set_results_name(e.resultsName, list_all_matches=True)
4470 for e in self.exprs
4471 if isinstance(e, OneOrMore)
4472 ]
4473 self.required = [
4474 e for e in self.exprs if not isinstance(e, (Opt, ZeroOrMore, OneOrMore))
4475 ]
4476 self.required += self.multirequired
4477 self.initExprGroups = False
4479 tmpLoc = loc
4480 tmpReqd = self.required[:]
4481 tmpOpt = self.optionals[:]
4482 multis = self.multioptionals[:]
4483 matchOrder: List[ParserElement] = []
4485 keepMatching = True
4486 failed: List[ParserElement] = []
4487 fatals: List[ParseFatalException] = []
4488 while keepMatching:
4489 tmpExprs = tmpReqd + tmpOpt + multis
4490 failed.clear()
4491 fatals.clear()
4492 for e in tmpExprs:
4493 try:
4494 tmpLoc = e.try_parse(instring, tmpLoc, raise_fatal=True)
4495 except ParseFatalException as pfe:
4496 pfe.__traceback__ = None
4497 pfe.parser_element = e
4498 fatals.append(pfe)
4499 failed.append(e)
4500 except ParseException:
4501 failed.append(e)
4502 else:
4503 matchOrder.append(self.opt1map.get(id(e), e))
4504 if e in tmpReqd:
4505 tmpReqd.remove(e)
4506 elif e in tmpOpt:
4507 tmpOpt.remove(e)
4508 if len(failed) == len(tmpExprs):
4509 keepMatching = False
4511 # look for any ParseFatalExceptions
4512 if fatals:
4513 if len(fatals) > 1:
4514 fatals.sort(key=lambda e: -e.loc)
4515 if fatals[0].loc == fatals[1].loc:
4516 fatals.sort(key=lambda e: (-e.loc, -len(str(e.parser_element))))
4517 max_fatal = fatals[0]
4518 raise max_fatal
4520 if tmpReqd:
4521 missing = ", ".join([str(e) for e in tmpReqd])
4522 raise ParseException(
4523 instring,
4524 loc,
4525 f"Missing one or more required elements ({missing})",
4526 )
4528 # add any unmatched Opts, in case they have default values defined
4529 matchOrder += [e for e in self.exprs if isinstance(e, Opt) and e.expr in tmpOpt]
4531 total_results = ParseResults([])
4532 for e in matchOrder:
4533 loc, results = e._parse(instring, loc, do_actions)
4534 total_results += results
4536 return loc, total_results
4538 def _generateDefaultName(self) -> str:
4539 return f"{{{' & '.join(str(e) for e in self.exprs)}}}"
4542class ParseElementEnhance(ParserElement):
4543 """Abstract subclass of :class:`ParserElement`, for combining and
4544 post-processing parsed tokens.
4545 """
4547 def __init__(self, expr: Union[ParserElement, str], savelist: bool = False):
4548 super().__init__(savelist)
4549 if isinstance(expr, str_type):
4550 expr_str = typing.cast(str, expr)
4551 if issubclass(self._literalStringClass, Token):
4552 expr = self._literalStringClass(expr_str) # type: ignore[call-arg]
4553 elif issubclass(type(self), self._literalStringClass):
4554 expr = Literal(expr_str)
4555 else:
4556 expr = self._literalStringClass(Literal(expr_str)) # type: ignore[assignment, call-arg]
4557 expr = typing.cast(ParserElement, expr)
4558 self.expr = expr
4559 if expr is not None:
4560 self.mayIndexError = expr.mayIndexError
4561 self.mayReturnEmpty = expr.mayReturnEmpty
4562 self.set_whitespace_chars(
4563 expr.whiteChars, copy_defaults=expr.copyDefaultWhiteChars
4564 )
4565 self.skipWhitespace = expr.skipWhitespace
4566 self.saveAsList = expr.saveAsList
4567 self.callPreparse = expr.callPreparse
4568 self.ignoreExprs.extend(expr.ignoreExprs)
4570 def recurse(self) -> List[ParserElement]:
4571 return [self.expr] if self.expr is not None else []
4573 def parseImpl(self, instring, loc, do_actions=True):
4574 if self.expr is None:
4575 raise ParseException(instring, loc, "No expression defined", self)
4577 try:
4578 return self.expr._parse(instring, loc, do_actions, callPreParse=False)
4579 except ParseSyntaxException:
4580 raise
4581 except ParseBaseException as pbe:
4582 if not isinstance(self, Forward) or self.customName is not None:
4583 if self.errmsg:
4584 pbe.msg = self.errmsg
4585 raise
4587 def leave_whitespace(self, recursive: bool = True) -> ParserElement:
4588 super().leave_whitespace(recursive)
4590 if recursive:
4591 if self.expr is not None:
4592 self.expr = self.expr.copy()
4593 self.expr.leave_whitespace(recursive)
4594 return self
4596 def ignore_whitespace(self, recursive: bool = True) -> ParserElement:
4597 super().ignore_whitespace(recursive)
4599 if recursive:
4600 if self.expr is not None:
4601 self.expr = self.expr.copy()
4602 self.expr.ignore_whitespace(recursive)
4603 return self
4605 def ignore(self, other) -> ParserElement:
4606 if not isinstance(other, Suppress) or other not in self.ignoreExprs:
4607 super().ignore(other)
4608 if self.expr is not None:
4609 self.expr.ignore(self.ignoreExprs[-1])
4611 return self
4613 def streamline(self) -> ParserElement:
4614 super().streamline()
4615 if self.expr is not None:
4616 self.expr.streamline()
4617 return self
4619 def _checkRecursion(self, parseElementList):
4620 if self in parseElementList:
4621 raise RecursiveGrammarException(parseElementList + [self])
4622 subRecCheckList = parseElementList[:] + [self]
4623 if self.expr is not None:
4624 self.expr._checkRecursion(subRecCheckList)
4626 def validate(self, validateTrace=None) -> None:
4627 warnings.warn(
4628 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
4629 DeprecationWarning,
4630 stacklevel=2,
4631 )
4632 if validateTrace is None:
4633 validateTrace = []
4634 tmp = validateTrace[:] + [self]
4635 if self.expr is not None:
4636 self.expr.validate(tmp)
4637 self._checkRecursion([])
4639 def _generateDefaultName(self) -> str:
4640 return f"{type(self).__name__}:({self.expr})"
4642 # Compatibility synonyms
4643 # fmt: off
4644 leaveWhitespace = replaced_by_pep8("leaveWhitespace", leave_whitespace)
4645 ignoreWhitespace = replaced_by_pep8("ignoreWhitespace", ignore_whitespace)
4646 # fmt: on
4649class IndentedBlock(ParseElementEnhance):
4650 """
4651 Expression to match one or more expressions at a given indentation level.
4652 Useful for parsing text where structure is implied by indentation (like Python source code).
4653 """
4655 class _Indent(Empty):
4656 def __init__(self, ref_col: int):
4657 super().__init__()
4658 self.errmsg = f"expected indent at column {ref_col}"
4659 self.add_condition(lambda s, l, t: col(l, s) == ref_col)
4661 class _IndentGreater(Empty):
4662 def __init__(self, ref_col: int):
4663 super().__init__()
4664 self.errmsg = f"expected indent at column greater than {ref_col}"
4665 self.add_condition(lambda s, l, t: col(l, s) > ref_col)
4667 def __init__(
4668 self, expr: ParserElement, *, recursive: bool = False, grouped: bool = True
4669 ):
4670 super().__init__(expr, savelist=True)
4671 # if recursive:
4672 # raise NotImplementedError("IndentedBlock with recursive is not implemented")
4673 self._recursive = recursive
4674 self._grouped = grouped
4675 self.parent_anchor = 1
4677 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4678 # advance parse position to non-whitespace by using an Empty()
4679 # this should be the column to be used for all subsequent indented lines
4680 anchor_loc = Empty().preParse(instring, loc)
4682 # see if self.expr matches at the current location - if not it will raise an exception
4683 # and no further work is necessary
4684 self.expr.try_parse(instring, anchor_loc, do_actions=do_actions)
4686 indent_col = col(anchor_loc, instring)
4687 peer_detect_expr = self._Indent(indent_col)
4689 inner_expr = Empty() + peer_detect_expr + self.expr
4690 if self._recursive:
4691 sub_indent = self._IndentGreater(indent_col)
4692 nested_block = IndentedBlock(
4693 self.expr, recursive=self._recursive, grouped=self._grouped
4694 )
4695 nested_block.set_debug(self.debug)
4696 nested_block.parent_anchor = indent_col
4697 inner_expr += Opt(sub_indent + nested_block)
4699 inner_expr.set_name(f"inner {hex(id(inner_expr))[-4:].upper()}@{indent_col}")
4700 block = OneOrMore(inner_expr)
4702 trailing_undent = self._Indent(self.parent_anchor) | StringEnd()
4704 if self._grouped:
4705 wrapper = Group
4706 else:
4707 wrapper = lambda expr: expr
4708 return (wrapper(block) + Optional(trailing_undent)).parseImpl(
4709 instring, anchor_loc, do_actions
4710 )
4713class AtStringStart(ParseElementEnhance):
4714 """Matches if expression matches at the beginning of the parse
4715 string::
4717 AtStringStart(Word(nums)).parse_string("123")
4718 # prints ["123"]
4720 AtStringStart(Word(nums)).parse_string(" 123")
4721 # raises ParseException
4722 """
4724 def __init__(self, expr: Union[ParserElement, str]):
4725 super().__init__(expr)
4726 self.callPreparse = False
4728 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4729 if loc != 0:
4730 raise ParseException(instring, loc, "not found at string start")
4731 return super().parseImpl(instring, loc, do_actions)
4734class AtLineStart(ParseElementEnhance):
4735 r"""Matches if an expression matches at the beginning of a line within
4736 the parse string
4738 Example::
4740 test = '''\
4741 AAA this line
4742 AAA and this line
4743 AAA but not this one
4744 B AAA and definitely not this one
4745 '''
4747 for t in (AtLineStart('AAA') + rest_of_line).search_string(test):
4748 print(t)
4750 prints::
4752 ['AAA', ' this line']
4753 ['AAA', ' and this line']
4755 """
4757 def __init__(self, expr: Union[ParserElement, str]):
4758 super().__init__(expr)
4759 self.callPreparse = False
4761 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4762 if col(loc, instring) != 1:
4763 raise ParseException(instring, loc, "not found at line start")
4764 return super().parseImpl(instring, loc, do_actions)
4767class FollowedBy(ParseElementEnhance):
4768 """Lookahead matching of the given parse expression.
4769 ``FollowedBy`` does *not* advance the parsing position within
4770 the input string, it only verifies that the specified parse
4771 expression matches at the current position. ``FollowedBy``
4772 always returns a null token list. If any results names are defined
4773 in the lookahead expression, those *will* be returned for access by
4774 name.
4776 Example::
4778 # use FollowedBy to match a label only if it is followed by a ':'
4779 data_word = Word(alphas)
4780 label = data_word + FollowedBy(':')
4781 attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stop_on=label).set_parse_action(' '.join))
4783 attr_expr[1, ...].parse_string("shape: SQUARE color: BLACK posn: upper left").pprint()
4785 prints::
4787 [['shape', 'SQUARE'], ['color', 'BLACK'], ['posn', 'upper left']]
4788 """
4790 def __init__(self, expr: Union[ParserElement, str]):
4791 super().__init__(expr)
4792 self.mayReturnEmpty = True
4794 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4795 # by using self._expr.parse and deleting the contents of the returned ParseResults list
4796 # we keep any named results that were defined in the FollowedBy expression
4797 _, ret = self.expr._parse(instring, loc, do_actions=do_actions)
4798 del ret[:]
4800 return loc, ret
4803class PrecededBy(ParseElementEnhance):
4804 """Lookbehind matching of the given parse expression.
4805 ``PrecededBy`` does not advance the parsing position within the
4806 input string, it only verifies that the specified parse expression
4807 matches prior to the current position. ``PrecededBy`` always
4808 returns a null token list, but if a results name is defined on the
4809 given expression, it is returned.
4811 Parameters:
4813 - ``expr`` - expression that must match prior to the current parse
4814 location
4815 - ``retreat`` - (default= ``None``) - (int) maximum number of characters
4816 to lookbehind prior to the current parse location
4818 If the lookbehind expression is a string, :class:`Literal`,
4819 :class:`Keyword`, or a :class:`Word` or :class:`CharsNotIn`
4820 with a specified exact or maximum length, then the retreat
4821 parameter is not required. Otherwise, retreat must be specified to
4822 give a maximum number of characters to look back from
4823 the current parse position for a lookbehind match.
4825 Example::
4827 # VB-style variable names with type prefixes
4828 int_var = PrecededBy("#") + pyparsing_common.identifier
4829 str_var = PrecededBy("$") + pyparsing_common.identifier
4831 """
4833 def __init__(
4834 self, expr: Union[ParserElement, str], retreat: typing.Optional[int] = None
4835 ):
4836 super().__init__(expr)
4837 self.expr = self.expr().leave_whitespace()
4838 self.mayReturnEmpty = True
4839 self.mayIndexError = False
4840 self.exact = False
4841 if isinstance(expr, str_type):
4842 expr = typing.cast(str, expr)
4843 retreat = len(expr)
4844 self.exact = True
4845 elif isinstance(expr, (Literal, Keyword)):
4846 retreat = expr.matchLen
4847 self.exact = True
4848 elif isinstance(expr, (Word, CharsNotIn)) and expr.maxLen != _MAX_INT:
4849 retreat = expr.maxLen
4850 self.exact = True
4851 elif isinstance(expr, PositionToken):
4852 retreat = 0
4853 self.exact = True
4854 self.retreat = retreat
4855 self.errmsg = f"not preceded by {expr}"
4856 self.skipWhitespace = False
4857 self.parseAction.append(lambda s, l, t: t.__delitem__(slice(None, None)))
4859 def parseImpl(self, instring, loc=0, do_actions=True) -> ParseImplReturnType:
4860 if self.exact:
4861 if loc < self.retreat:
4862 raise ParseException(instring, loc, self.errmsg)
4863 start = loc - self.retreat
4864 _, ret = self.expr._parse(instring, start)
4865 return loc, ret
4867 # retreat specified a maximum lookbehind window, iterate
4868 test_expr = self.expr + StringEnd()
4869 instring_slice = instring[max(0, loc - self.retreat) : loc]
4870 last_expr = ParseException(instring, loc, self.errmsg)
4872 for offset in range(1, min(loc, self.retreat + 1) + 1):
4873 try:
4874 # print('trying', offset, instring_slice, repr(instring_slice[loc - offset:]))
4875 _, ret = test_expr._parse(instring_slice, len(instring_slice) - offset)
4876 except ParseBaseException as pbe:
4877 last_expr = pbe
4878 else:
4879 break
4880 else:
4881 raise last_expr
4883 return loc, ret
4886class Located(ParseElementEnhance):
4887 """
4888 Decorates a returned token with its starting and ending
4889 locations in the input string.
4891 This helper adds the following results names:
4893 - ``locn_start`` - location where matched expression begins
4894 - ``locn_end`` - location where matched expression ends
4895 - ``value`` - the actual parsed results
4897 Be careful if the input text contains ``<TAB>`` characters, you
4898 may want to call :class:`ParserElement.parse_with_tabs`
4900 Example::
4902 wd = Word(alphas)
4903 for match in Located(wd).search_string("ljsdf123lksdjjf123lkkjj1222"):
4904 print(match)
4906 prints::
4908 [0, ['ljsdf'], 5]
4909 [8, ['lksdjjf'], 15]
4910 [18, ['lkkjj'], 23]
4912 """
4914 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4915 start = loc
4916 loc, tokens = self.expr._parse(instring, start, do_actions, callPreParse=False)
4917 ret_tokens = ParseResults([start, tokens, loc])
4918 ret_tokens["locn_start"] = start
4919 ret_tokens["value"] = tokens
4920 ret_tokens["locn_end"] = loc
4921 if self.resultsName:
4922 # must return as a list, so that the name will be attached to the complete group
4923 return loc, [ret_tokens]
4924 else:
4925 return loc, ret_tokens
4928class NotAny(ParseElementEnhance):
4929 """
4930 Lookahead to disallow matching with the given parse expression.
4931 ``NotAny`` does *not* advance the parsing position within the
4932 input string, it only verifies that the specified parse expression
4933 does *not* match at the current position. Also, ``NotAny`` does
4934 *not* skip over leading whitespace. ``NotAny`` always returns
4935 a null token list. May be constructed using the ``'~'`` operator.
4937 Example::
4939 AND, OR, NOT = map(CaselessKeyword, "AND OR NOT".split())
4941 # take care not to mistake keywords for identifiers
4942 ident = ~(AND | OR | NOT) + Word(alphas)
4943 boolean_term = Opt(NOT) + ident
4945 # very crude boolean expression - to support parenthesis groups and
4946 # operation hierarchy, use infix_notation
4947 boolean_expr = boolean_term + ((AND | OR) + boolean_term)[...]
4949 # integers that are followed by "." are actually floats
4950 integer = Word(nums) + ~Char(".")
4951 """
4953 def __init__(self, expr: Union[ParserElement, str]):
4954 super().__init__(expr)
4955 # do NOT use self.leave_whitespace(), don't want to propagate to exprs
4956 # self.leave_whitespace()
4957 self.skipWhitespace = False
4959 self.mayReturnEmpty = True
4960 self.errmsg = f"Found unwanted token, {self.expr}"
4962 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4963 if self.expr.can_parse_next(instring, loc, do_actions=do_actions):
4964 raise ParseException(instring, loc, self.errmsg, self)
4965 return loc, []
4967 def _generateDefaultName(self) -> str:
4968 return f"~{{{self.expr}}}"
4971class _MultipleMatch(ParseElementEnhance):
4972 def __init__(
4973 self,
4974 expr: Union[str, ParserElement],
4975 stop_on: typing.Optional[Union[ParserElement, str]] = None,
4976 *,
4977 stopOn: typing.Optional[Union[ParserElement, str]] = None,
4978 ):
4979 super().__init__(expr)
4980 stopOn = stopOn or stop_on
4981 self.saveAsList = True
4982 ender = stopOn
4983 if isinstance(ender, str_type):
4984 ender = self._literalStringClass(ender)
4985 self.stopOn(ender)
4987 def stopOn(self, ender) -> ParserElement:
4988 if isinstance(ender, str_type):
4989 ender = self._literalStringClass(ender)
4990 self.not_ender = ~ender if ender is not None else None
4991 return self
4993 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
4994 self_expr_parse = self.expr._parse
4995 self_skip_ignorables = self._skipIgnorables
4996 check_ender = self.not_ender is not None
4997 if check_ender:
4998 try_not_ender = self.not_ender.try_parse
5000 # must be at least one (but first see if we are the stopOn sentinel;
5001 # if so, fail)
5002 if check_ender:
5003 try_not_ender(instring, loc)
5004 loc, tokens = self_expr_parse(instring, loc, do_actions)
5005 try:
5006 hasIgnoreExprs = not not self.ignoreExprs
5007 while 1:
5008 if check_ender:
5009 try_not_ender(instring, loc)
5010 if hasIgnoreExprs:
5011 preloc = self_skip_ignorables(instring, loc)
5012 else:
5013 preloc = loc
5014 loc, tmptokens = self_expr_parse(instring, preloc, do_actions)
5015 tokens += tmptokens
5016 except (ParseException, IndexError):
5017 pass
5019 return loc, tokens
5021 def _setResultsName(self, name, list_all_matches=False) -> ParserElement:
5022 if (
5023 __diag__.warn_ungrouped_named_tokens_in_collection
5024 and Diagnostics.warn_ungrouped_named_tokens_in_collection
5025 not in self.suppress_warnings_
5026 ):
5027 for e in [self.expr] + self.expr.recurse():
5028 if (
5029 isinstance(e, ParserElement)
5030 and e.resultsName
5031 and (
5032 Diagnostics.warn_ungrouped_named_tokens_in_collection
5033 not in e.suppress_warnings_
5034 )
5035 ):
5036 warning = (
5037 "warn_ungrouped_named_tokens_in_collection:"
5038 f" setting results name {name!r} on {type(self).__name__} expression"
5039 f" collides with {e.resultsName!r} on contained expression"
5040 )
5041 warnings.warn(warning, stacklevel=3)
5042 break
5044 return super()._setResultsName(name, list_all_matches)
5047class OneOrMore(_MultipleMatch):
5048 """
5049 Repetition of one or more of the given expression.
5051 Parameters:
5053 - ``expr`` - expression that must match one or more times
5054 - ``stop_on`` - (default= ``None``) - expression for a terminating sentinel
5055 (only required if the sentinel would ordinarily match the repetition
5056 expression)
5058 Example::
5060 data_word = Word(alphas)
5061 label = data_word + FollowedBy(':')
5062 attr_expr = Group(label + Suppress(':') + OneOrMore(data_word).set_parse_action(' '.join))
5064 text = "shape: SQUARE posn: upper left color: BLACK"
5065 attr_expr[1, ...].parse_string(text).pprint() # Fail! read 'color' as data instead of next label -> [['shape', 'SQUARE color']]
5067 # use stop_on attribute for OneOrMore to avoid reading label string as part of the data
5068 attr_expr = Group(label + Suppress(':') + OneOrMore(data_word, stop_on=label).set_parse_action(' '.join))
5069 OneOrMore(attr_expr).parse_string(text).pprint() # Better -> [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'BLACK']]
5071 # could also be written as
5072 (attr_expr * (1,)).parse_string(text).pprint()
5073 """
5075 def _generateDefaultName(self) -> str:
5076 return f"{{{self.expr}}}..."
5079class ZeroOrMore(_MultipleMatch):
5080 """
5081 Optional repetition of zero or more of the given expression.
5083 Parameters:
5085 - ``expr`` - expression that must match zero or more times
5086 - ``stop_on`` - expression for a terminating sentinel
5087 (only required if the sentinel would ordinarily match the repetition
5088 expression) - (default= ``None``)
5090 Example: similar to :class:`OneOrMore`
5091 """
5093 def __init__(
5094 self,
5095 expr: Union[str, ParserElement],
5096 stop_on: typing.Optional[Union[ParserElement, str]] = None,
5097 *,
5098 stopOn: typing.Optional[Union[ParserElement, str]] = None,
5099 ):
5100 super().__init__(expr, stopOn=stopOn or stop_on)
5101 self.mayReturnEmpty = True
5103 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
5104 try:
5105 return super().parseImpl(instring, loc, do_actions)
5106 except (ParseException, IndexError):
5107 return loc, ParseResults([], name=self.resultsName)
5109 def _generateDefaultName(self) -> str:
5110 return f"[{self.expr}]..."
5113class DelimitedList(ParseElementEnhance):
5114 def __init__(
5115 self,
5116 expr: Union[str, ParserElement],
5117 delim: Union[str, ParserElement] = ",",
5118 combine: bool = False,
5119 min: typing.Optional[int] = None,
5120 max: typing.Optional[int] = None,
5121 *,
5122 allow_trailing_delim: bool = False,
5123 ):
5124 """Helper to define a delimited list of expressions - the delimiter
5125 defaults to ','. By default, the list elements and delimiters can
5126 have intervening whitespace, and comments, but this can be
5127 overridden by passing ``combine=True`` in the constructor. If
5128 ``combine`` is set to ``True``, the matching tokens are
5129 returned as a single token string, with the delimiters included;
5130 otherwise, the matching tokens are returned as a list of tokens,
5131 with the delimiters suppressed.
5133 If ``allow_trailing_delim`` is set to True, then the list may end with
5134 a delimiter.
5136 Example::
5138 DelimitedList(Word(alphas)).parse_string("aa,bb,cc") # -> ['aa', 'bb', 'cc']
5139 DelimitedList(Word(hexnums), delim=':', combine=True).parse_string("AA:BB:CC:DD:EE") # -> ['AA:BB:CC:DD:EE']
5140 """
5141 if isinstance(expr, str_type):
5142 expr = ParserElement._literalStringClass(expr)
5143 expr = typing.cast(ParserElement, expr)
5145 if min is not None and min < 1:
5146 raise ValueError("min must be greater than 0")
5148 if max is not None and min is not None and max < min:
5149 raise ValueError("max must be greater than, or equal to min")
5151 self.content = expr
5152 self.raw_delim = str(delim)
5153 self.delim = delim
5154 self.combine = combine
5155 if not combine:
5156 self.delim = Suppress(delim)
5157 self.min = min or 1
5158 self.max = max
5159 self.allow_trailing_delim = allow_trailing_delim
5161 delim_list_expr = self.content + (self.delim + self.content) * (
5162 self.min - 1,
5163 None if self.max is None else self.max - 1,
5164 )
5165 if self.allow_trailing_delim:
5166 delim_list_expr += Opt(self.delim)
5168 if self.combine:
5169 delim_list_expr = Combine(delim_list_expr)
5171 super().__init__(delim_list_expr, savelist=True)
5173 def _generateDefaultName(self) -> str:
5174 content_expr = self.content.streamline()
5175 return f"{content_expr} [{self.raw_delim} {content_expr}]..."
5178class _NullToken:
5179 def __bool__(self):
5180 return False
5182 def __str__(self):
5183 return ""
5186class Opt(ParseElementEnhance):
5187 """
5188 Optional matching of the given expression.
5190 Parameters:
5192 - ``expr`` - expression that must match zero or more times
5193 - ``default`` (optional) - value to be returned if the optional expression is not found.
5195 Example::
5197 # US postal code can be a 5-digit zip, plus optional 4-digit qualifier
5198 zip = Combine(Word(nums, exact=5) + Opt('-' + Word(nums, exact=4)))
5199 zip.run_tests('''
5200 # traditional ZIP code
5201 12345
5203 # ZIP+4 form
5204 12101-0001
5206 # invalid ZIP
5207 98765-
5208 ''')
5210 prints::
5212 # traditional ZIP code
5213 12345
5214 ['12345']
5216 # ZIP+4 form
5217 12101-0001
5218 ['12101-0001']
5220 # invalid ZIP
5221 98765-
5222 ^
5223 FAIL: Expected end of text (at char 5), (line:1, col:6)
5224 """
5226 __optionalNotMatched = _NullToken()
5228 def __init__(
5229 self, expr: Union[ParserElement, str], default: Any = __optionalNotMatched
5230 ):
5231 super().__init__(expr, savelist=False)
5232 self.saveAsList = self.expr.saveAsList
5233 self.defaultValue = default
5234 self.mayReturnEmpty = True
5236 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
5237 self_expr = self.expr
5238 try:
5239 loc, tokens = self_expr._parse(instring, loc, do_actions, callPreParse=False)
5240 except (ParseException, IndexError):
5241 default_value = self.defaultValue
5242 if default_value is not self.__optionalNotMatched:
5243 if self_expr.resultsName:
5244 tokens = ParseResults([default_value])
5245 tokens[self_expr.resultsName] = default_value
5246 else:
5247 tokens = [default_value]
5248 else:
5249 tokens = []
5250 return loc, tokens
5252 def _generateDefaultName(self) -> str:
5253 inner = str(self.expr)
5254 # strip off redundant inner {}'s
5255 while len(inner) > 1 and inner[0 :: len(inner) - 1] == "{}":
5256 inner = inner[1:-1]
5257 return f"[{inner}]"
5260Optional = Opt
5263class SkipTo(ParseElementEnhance):
5264 """
5265 Token for skipping over all undefined text until the matched
5266 expression is found.
5268 Parameters:
5270 - ``expr`` - target expression marking the end of the data to be skipped
5271 - ``include`` - if ``True``, the target expression is also parsed
5272 (the skipped text and target expression are returned as a 2-element
5273 list) (default= ``False``).
5274 - ``ignore`` - (default= ``None``) used to define grammars (typically quoted strings and
5275 comments) that might contain false matches to the target expression
5276 - ``fail_on`` - (default= ``None``) define expressions that are not allowed to be
5277 included in the skipped test; if found before the target expression is found,
5278 the :class:`SkipTo` is not a match
5280 Example::
5282 report = '''
5283 Outstanding Issues Report - 1 Jan 2000
5285 # | Severity | Description | Days Open
5286 -----+----------+-------------------------------------------+-----------
5287 101 | Critical | Intermittent system crash | 6
5288 94 | Cosmetic | Spelling error on Login ('log|n') | 14
5289 79 | Minor | System slow when running too many reports | 47
5290 '''
5291 integer = Word(nums)
5292 SEP = Suppress('|')
5293 # use SkipTo to simply match everything up until the next SEP
5294 # - ignore quoted strings, so that a '|' character inside a quoted string does not match
5295 # - parse action will call token.strip() for each matched token, i.e., the description body
5296 string_data = SkipTo(SEP, ignore=quoted_string)
5297 string_data.set_parse_action(token_map(str.strip))
5298 ticket_expr = (integer("issue_num") + SEP
5299 + string_data("sev") + SEP
5300 + string_data("desc") + SEP
5301 + integer("days_open"))
5303 for tkt in ticket_expr.search_string(report):
5304 print tkt.dump()
5306 prints::
5308 ['101', 'Critical', 'Intermittent system crash', '6']
5309 - days_open: '6'
5310 - desc: 'Intermittent system crash'
5311 - issue_num: '101'
5312 - sev: 'Critical'
5313 ['94', 'Cosmetic', "Spelling error on Login ('log|n')", '14']
5314 - days_open: '14'
5315 - desc: "Spelling error on Login ('log|n')"
5316 - issue_num: '94'
5317 - sev: 'Cosmetic'
5318 ['79', 'Minor', 'System slow when running too many reports', '47']
5319 - days_open: '47'
5320 - desc: 'System slow when running too many reports'
5321 - issue_num: '79'
5322 - sev: 'Minor'
5323 """
5325 def __init__(
5326 self,
5327 other: Union[ParserElement, str],
5328 include: bool = False,
5329 ignore: typing.Optional[Union[ParserElement, str]] = None,
5330 fail_on: typing.Optional[Union[ParserElement, str]] = None,
5331 *,
5332 failOn: typing.Optional[Union[ParserElement, str]] = None,
5333 ):
5334 super().__init__(other)
5335 failOn = failOn or fail_on
5336 self.ignoreExpr = ignore
5337 self.mayReturnEmpty = True
5338 self.mayIndexError = False
5339 self.includeMatch = include
5340 self.saveAsList = False
5341 if isinstance(failOn, str_type):
5342 self.failOn = self._literalStringClass(failOn)
5343 else:
5344 self.failOn = failOn
5345 self.errmsg = f"No match found for {self.expr}"
5346 self.ignorer = Empty().leave_whitespace()
5347 self._update_ignorer()
5349 def _update_ignorer(self):
5350 # rebuild internal ignore expr from current ignore exprs and assigned ignoreExpr
5351 self.ignorer.ignoreExprs.clear()
5352 for e in self.expr.ignoreExprs:
5353 self.ignorer.ignore(e)
5354 if self.ignoreExpr:
5355 self.ignorer.ignore(self.ignoreExpr)
5357 def ignore(self, expr):
5358 super().ignore(expr)
5359 self._update_ignorer()
5361 def parseImpl(self, instring, loc, do_actions=True):
5362 startloc = loc
5363 instrlen = len(instring)
5364 self_expr_parse = self.expr._parse
5365 self_failOn_canParseNext = (
5366 self.failOn.canParseNext if self.failOn is not None else None
5367 )
5368 ignorer_try_parse = self.ignorer.try_parse if self.ignorer.ignoreExprs else None
5370 tmploc = loc
5371 while tmploc <= instrlen:
5372 if self_failOn_canParseNext is not None:
5373 # break if failOn expression matches
5374 if self_failOn_canParseNext(instring, tmploc):
5375 break
5377 if ignorer_try_parse is not None:
5378 # advance past ignore expressions
5379 prev_tmploc = tmploc
5380 while 1:
5381 try:
5382 tmploc = ignorer_try_parse(instring, tmploc)
5383 except ParseBaseException:
5384 break
5385 # see if all ignorers matched, but didn't actually ignore anything
5386 if tmploc == prev_tmploc:
5387 break
5388 prev_tmploc = tmploc
5390 try:
5391 self_expr_parse(instring, tmploc, do_actions=False, callPreParse=False)
5392 except (ParseException, IndexError):
5393 # no match, advance loc in string
5394 tmploc += 1
5395 else:
5396 # matched skipto expr, done
5397 break
5399 else:
5400 # ran off the end of the input string without matching skipto expr, fail
5401 raise ParseException(instring, loc, self.errmsg, self)
5403 # build up return values
5404 loc = tmploc
5405 skiptext = instring[startloc:loc]
5406 skipresult = ParseResults(skiptext)
5408 if self.includeMatch:
5409 loc, mat = self_expr_parse(instring, loc, do_actions, callPreParse=False)
5410 skipresult += mat
5412 return loc, skipresult
5415class Forward(ParseElementEnhance):
5416 """
5417 Forward declaration of an expression to be defined later -
5418 used for recursive grammars, such as algebraic infix notation.
5419 When the expression is known, it is assigned to the ``Forward``
5420 variable using the ``'<<'`` operator.
5422 Note: take care when assigning to ``Forward`` not to overlook
5423 precedence of operators.
5425 Specifically, ``'|'`` has a lower precedence than ``'<<'``, so that::
5427 fwd_expr << a | b | c
5429 will actually be evaluated as::
5431 (fwd_expr << a) | b | c
5433 thereby leaving b and c out as parseable alternatives. It is recommended that you
5434 explicitly group the values inserted into the ``Forward``::
5436 fwd_expr << (a | b | c)
5438 Converting to use the ``'<<='`` operator instead will avoid this problem.
5440 See :class:`ParseResults.pprint` for an example of a recursive
5441 parser created using ``Forward``.
5442 """
5444 def __init__(self, other: typing.Optional[Union[ParserElement, str]] = None):
5445 self.caller_frame = traceback.extract_stack(limit=2)[0]
5446 super().__init__(other, savelist=False) # type: ignore[arg-type]
5447 self.lshift_line = None
5449 def __lshift__(self, other) -> "Forward":
5450 if hasattr(self, "caller_frame"):
5451 del self.caller_frame
5452 if isinstance(other, str_type):
5453 other = self._literalStringClass(other)
5455 if not isinstance(other, ParserElement):
5456 return NotImplemented
5458 self.expr = other
5459 self.streamlined = other.streamlined
5460 self.mayIndexError = self.expr.mayIndexError
5461 self.mayReturnEmpty = self.expr.mayReturnEmpty
5462 self.set_whitespace_chars(
5463 self.expr.whiteChars, copy_defaults=self.expr.copyDefaultWhiteChars
5464 )
5465 self.skipWhitespace = self.expr.skipWhitespace
5466 self.saveAsList = self.expr.saveAsList
5467 self.ignoreExprs.extend(self.expr.ignoreExprs)
5468 self.lshift_line = traceback.extract_stack(limit=2)[-2] # type: ignore[assignment]
5469 return self
5471 def __ilshift__(self, other) -> "Forward":
5472 if not isinstance(other, ParserElement):
5473 return NotImplemented
5475 return self << other
5477 def __or__(self, other) -> "ParserElement":
5478 caller_line = traceback.extract_stack(limit=2)[-2]
5479 if (
5480 __diag__.warn_on_match_first_with_lshift_operator
5481 and caller_line == self.lshift_line
5482 and Diagnostics.warn_on_match_first_with_lshift_operator
5483 not in self.suppress_warnings_
5484 ):
5485 warnings.warn(
5486 "using '<<' operator with '|' is probably an error, use '<<='",
5487 stacklevel=2,
5488 )
5489 ret = super().__or__(other)
5490 return ret
5492 def __del__(self):
5493 # see if we are getting dropped because of '=' reassignment of var instead of '<<=' or '<<'
5494 if (
5495 self.expr is None
5496 and __diag__.warn_on_assignment_to_Forward
5497 and Diagnostics.warn_on_assignment_to_Forward not in self.suppress_warnings_
5498 ):
5499 warnings.warn_explicit(
5500 "Forward defined here but no expression attached later using '<<=' or '<<'",
5501 UserWarning,
5502 filename=self.caller_frame.filename,
5503 lineno=self.caller_frame.lineno,
5504 )
5506 def parseImpl(self, instring, loc, do_actions=True) -> ParseImplReturnType:
5507 if (
5508 self.expr is None
5509 and __diag__.warn_on_parse_using_empty_Forward
5510 and Diagnostics.warn_on_parse_using_empty_Forward
5511 not in self.suppress_warnings_
5512 ):
5513 # walk stack until parse_string, scan_string, search_string, or transform_string is found
5514 parse_fns = (
5515 "parse_string",
5516 "scan_string",
5517 "search_string",
5518 "transform_string",
5519 )
5520 tb = traceback.extract_stack(limit=200)
5521 for i, frm in enumerate(reversed(tb), start=1):
5522 if frm.name in parse_fns:
5523 stacklevel = i + 1
5524 break
5525 else:
5526 stacklevel = 2
5527 warnings.warn(
5528 "Forward expression was never assigned a value, will not parse any input",
5529 stacklevel=stacklevel,
5530 )
5531 if not ParserElement._left_recursion_enabled:
5532 return super().parseImpl(instring, loc, do_actions)
5533 # ## Bounded Recursion algorithm ##
5534 # Recursion only needs to be processed at ``Forward`` elements, since they are
5535 # the only ones that can actually refer to themselves. The general idea is
5536 # to handle recursion stepwise: We start at no recursion, then recurse once,
5537 # recurse twice, ..., until more recursion offers no benefit (we hit the bound).
5538 #
5539 # The "trick" here is that each ``Forward`` gets evaluated in two contexts
5540 # - to *match* a specific recursion level, and
5541 # - to *search* the bounded recursion level
5542 # and the two run concurrently. The *search* must *match* each recursion level
5543 # to find the best possible match. This is handled by a memo table, which
5544 # provides the previous match to the next level match attempt.
5545 #
5546 # See also "Left Recursion in Parsing Expression Grammars", Medeiros et al.
5547 #
5548 # There is a complication since we not only *parse* but also *transform* via
5549 # actions: We do not want to run the actions too often while expanding. Thus,
5550 # we expand using `do_actions=False` and only run `do_actions=True` if the next
5551 # recursion level is acceptable.
5552 with ParserElement.recursion_lock:
5553 memo = ParserElement.recursion_memos
5554 try:
5555 # we are parsing at a specific recursion expansion - use it as-is
5556 prev_loc, prev_result = memo[loc, self, do_actions]
5557 if isinstance(prev_result, Exception):
5558 raise prev_result
5559 return prev_loc, prev_result.copy()
5560 except KeyError:
5561 act_key = (loc, self, True)
5562 peek_key = (loc, self, False)
5563 # we are searching for the best recursion expansion - keep on improving
5564 # both `do_actions` cases must be tracked separately here!
5565 prev_loc, prev_peek = memo[peek_key] = (
5566 loc - 1,
5567 ParseException(
5568 instring, loc, "Forward recursion without base case", self
5569 ),
5570 )
5571 if do_actions:
5572 memo[act_key] = memo[peek_key]
5573 while True:
5574 try:
5575 new_loc, new_peek = super().parseImpl(instring, loc, False)
5576 except ParseException:
5577 # we failed before getting any match – do not hide the error
5578 if isinstance(prev_peek, Exception):
5579 raise
5580 new_loc, new_peek = prev_loc, prev_peek
5581 # the match did not get better: we are done
5582 if new_loc <= prev_loc:
5583 if do_actions:
5584 # replace the match for do_actions=False as well,
5585 # in case the action did backtrack
5586 prev_loc, prev_result = memo[peek_key] = memo[act_key]
5587 del memo[peek_key], memo[act_key]
5588 return prev_loc, prev_result.copy()
5589 del memo[peek_key]
5590 return prev_loc, prev_peek.copy()
5591 # the match did get better: see if we can improve further
5592 if do_actions:
5593 try:
5594 memo[act_key] = super().parseImpl(instring, loc, True)
5595 except ParseException as e:
5596 memo[peek_key] = memo[act_key] = (new_loc, e)
5597 raise
5598 prev_loc, prev_peek = memo[peek_key] = new_loc, new_peek
5600 def leave_whitespace(self, recursive: bool = True) -> ParserElement:
5601 self.skipWhitespace = False
5602 return self
5604 def ignore_whitespace(self, recursive: bool = True) -> ParserElement:
5605 self.skipWhitespace = True
5606 return self
5608 def streamline(self) -> ParserElement:
5609 if not self.streamlined:
5610 self.streamlined = True
5611 if self.expr is not None:
5612 self.expr.streamline()
5613 return self
5615 def validate(self, validateTrace=None) -> None:
5616 warnings.warn(
5617 "ParserElement.validate() is deprecated, and should not be used to check for left recursion",
5618 DeprecationWarning,
5619 stacklevel=2,
5620 )
5621 if validateTrace is None:
5622 validateTrace = []
5624 if self not in validateTrace:
5625 tmp = validateTrace[:] + [self]
5626 if self.expr is not None:
5627 self.expr.validate(tmp)
5628 self._checkRecursion([])
5630 def _generateDefaultName(self) -> str:
5631 # Avoid infinite recursion by setting a temporary _defaultName
5632 self._defaultName = ": ..."
5634 # Use the string representation of main expression.
5635 retString = "..."
5636 try:
5637 if self.expr is not None:
5638 retString = str(self.expr)[:1000]
5639 else:
5640 retString = "None"
5641 finally:
5642 return f"{type(self).__name__}: {retString}"
5644 def copy(self) -> ParserElement:
5645 if self.expr is not None:
5646 return super().copy()
5647 else:
5648 ret = Forward()
5649 ret <<= self
5650 return ret
5652 def _setResultsName(self, name, list_all_matches=False) -> ParserElement:
5653 # fmt: off
5654 if (
5655 __diag__.warn_name_set_on_empty_Forward
5656 and Diagnostics.warn_name_set_on_empty_Forward not in self.suppress_warnings_
5657 and self.expr is None
5658 ):
5659 warning = (
5660 "warn_name_set_on_empty_Forward:"
5661 f" setting results name {name!r} on {type(self).__name__} expression"
5662 " that has no contained expression"
5663 )
5664 warnings.warn(warning, stacklevel=3)
5665 # fmt: on
5667 return super()._setResultsName(name, list_all_matches)
5669 # Compatibility synonyms
5670 # fmt: off
5671 leaveWhitespace = replaced_by_pep8("leaveWhitespace", leave_whitespace)
5672 ignoreWhitespace = replaced_by_pep8("ignoreWhitespace", ignore_whitespace)
5673 # fmt: on
5676class TokenConverter(ParseElementEnhance):
5677 """
5678 Abstract subclass of :class:`ParseExpression`, for converting parsed results.
5679 """
5681 def __init__(self, expr: Union[ParserElement, str], savelist=False):
5682 super().__init__(expr) # , savelist)
5683 self.saveAsList = False
5686class Combine(TokenConverter):
5687 """Converter to concatenate all matching tokens to a single string.
5688 By default, the matching patterns must also be contiguous in the
5689 input string; this can be disabled by specifying
5690 ``'adjacent=False'`` in the constructor.
5692 Example::
5694 real = Word(nums) + '.' + Word(nums)
5695 print(real.parse_string('3.1416')) # -> ['3', '.', '1416']
5696 # will also erroneously match the following
5697 print(real.parse_string('3. 1416')) # -> ['3', '.', '1416']
5699 real = Combine(Word(nums) + '.' + Word(nums))
5700 print(real.parse_string('3.1416')) # -> ['3.1416']
5701 # no match when there are internal spaces
5702 print(real.parse_string('3. 1416')) # -> Exception: Expected W:(0123...)
5703 """
5705 def __init__(
5706 self,
5707 expr: ParserElement,
5708 join_string: str = "",
5709 adjacent: bool = True,
5710 *,
5711 joinString: typing.Optional[str] = None,
5712 ):
5713 super().__init__(expr)
5714 joinString = joinString if joinString is not None else join_string
5715 # suppress whitespace-stripping in contained parse expressions, but re-enable it on the Combine itself
5716 if adjacent:
5717 self.leave_whitespace()
5718 self.adjacent = adjacent
5719 self.skipWhitespace = True
5720 self.joinString = joinString
5721 self.callPreparse = True
5723 def ignore(self, other) -> ParserElement:
5724 if self.adjacent:
5725 ParserElement.ignore(self, other)
5726 else:
5727 super().ignore(other)
5728 return self
5730 def postParse(self, instring, loc, tokenlist):
5731 retToks = tokenlist.copy()
5732 del retToks[:]
5733 retToks += ParseResults(
5734 ["".join(tokenlist._asStringList(self.joinString))], modal=self.modalResults
5735 )
5737 if self.resultsName and retToks.haskeys():
5738 return [retToks]
5739 else:
5740 return retToks
5743class Group(TokenConverter):
5744 """Converter to return the matched tokens as a list - useful for
5745 returning tokens of :class:`ZeroOrMore` and :class:`OneOrMore` expressions.
5747 The optional ``aslist`` argument when set to True will return the
5748 parsed tokens as a Python list instead of a pyparsing ParseResults.
5750 Example::
5752 ident = Word(alphas)
5753 num = Word(nums)
5754 term = ident | num
5755 func = ident + Opt(DelimitedList(term))
5756 print(func.parse_string("fn a, b, 100"))
5757 # -> ['fn', 'a', 'b', '100']
5759 func = ident + Group(Opt(DelimitedList(term)))
5760 print(func.parse_string("fn a, b, 100"))
5761 # -> ['fn', ['a', 'b', '100']]
5762 """
5764 def __init__(self, expr: ParserElement, aslist: bool = False):
5765 super().__init__(expr)
5766 self.saveAsList = True
5767 self._asPythonList = aslist
5769 def postParse(self, instring, loc, tokenlist):
5770 if self._asPythonList:
5771 return ParseResults.List(
5772 tokenlist.asList()
5773 if isinstance(tokenlist, ParseResults)
5774 else list(tokenlist)
5775 )
5777 return [tokenlist]
5780class Dict(TokenConverter):
5781 """Converter to return a repetitive expression as a list, but also
5782 as a dictionary. Each element can also be referenced using the first
5783 token in the expression as its key. Useful for tabular report
5784 scraping when the first column can be used as a item key.
5786 The optional ``asdict`` argument when set to True will return the
5787 parsed tokens as a Python dict instead of a pyparsing ParseResults.
5789 Example::
5791 data_word = Word(alphas)
5792 label = data_word + FollowedBy(':')
5794 text = "shape: SQUARE posn: upper left color: light blue texture: burlap"
5795 attr_expr = (label + Suppress(':') + OneOrMore(data_word, stop_on=label).set_parse_action(' '.join))
5797 # print attributes as plain groups
5798 print(attr_expr[1, ...].parse_string(text).dump())
5800 # instead of OneOrMore(expr), parse using Dict(Group(expr)[1, ...]) - Dict will auto-assign names
5801 result = Dict(Group(attr_expr)[1, ...]).parse_string(text)
5802 print(result.dump())
5804 # access named fields as dict entries, or output as dict
5805 print(result['shape'])
5806 print(result.as_dict())
5808 prints::
5810 ['shape', 'SQUARE', 'posn', 'upper left', 'color', 'light blue', 'texture', 'burlap']
5811 [['shape', 'SQUARE'], ['posn', 'upper left'], ['color', 'light blue'], ['texture', 'burlap']]
5812 - color: 'light blue'
5813 - posn: 'upper left'
5814 - shape: 'SQUARE'
5815 - texture: 'burlap'
5816 SQUARE
5817 {'color': 'light blue', 'posn': 'upper left', 'texture': 'burlap', 'shape': 'SQUARE'}
5819 See more examples at :class:`ParseResults` of accessing fields by results name.
5820 """
5822 def __init__(self, expr: ParserElement, asdict: bool = False):
5823 super().__init__(expr)
5824 self.saveAsList = True
5825 self._asPythonDict = asdict
5827 def postParse(self, instring, loc, tokenlist):
5828 for i, tok in enumerate(tokenlist):
5829 if len(tok) == 0:
5830 continue
5832 ikey = tok[0]
5833 if isinstance(ikey, int):
5834 ikey = str(ikey).strip()
5836 if len(tok) == 1:
5837 tokenlist[ikey] = _ParseResultsWithOffset("", i)
5839 elif len(tok) == 2 and not isinstance(tok[1], ParseResults):
5840 tokenlist[ikey] = _ParseResultsWithOffset(tok[1], i)
5842 else:
5843 try:
5844 dictvalue = tok.copy() # ParseResults(i)
5845 except Exception:
5846 exc = TypeError(
5847 "could not extract dict values from parsed results"
5848 " - Dict expression must contain Grouped expressions"
5849 )
5850 raise exc from None
5852 del dictvalue[0]
5854 if len(dictvalue) != 1 or (
5855 isinstance(dictvalue, ParseResults) and dictvalue.haskeys()
5856 ):
5857 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue, i)
5858 else:
5859 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue[0], i)
5861 if self._asPythonDict:
5862 return [tokenlist.as_dict()] if self.resultsName else tokenlist.as_dict()
5864 return [tokenlist] if self.resultsName else tokenlist
5867class Suppress(TokenConverter):
5868 """Converter for ignoring the results of a parsed expression.
5870 Example::
5872 source = "a, b, c,d"
5873 wd = Word(alphas)
5874 wd_list1 = wd + (',' + wd)[...]
5875 print(wd_list1.parse_string(source))
5877 # often, delimiters that are useful during parsing are just in the
5878 # way afterward - use Suppress to keep them out of the parsed output
5879 wd_list2 = wd + (Suppress(',') + wd)[...]
5880 print(wd_list2.parse_string(source))
5882 # Skipped text (using '...') can be suppressed as well
5883 source = "lead in START relevant text END trailing text"
5884 start_marker = Keyword("START")
5885 end_marker = Keyword("END")
5886 find_body = Suppress(...) + start_marker + ... + end_marker
5887 print(find_body.parse_string(source)
5889 prints::
5891 ['a', ',', 'b', ',', 'c', ',', 'd']
5892 ['a', 'b', 'c', 'd']
5893 ['START', 'relevant text ', 'END']
5895 (See also :class:`DelimitedList`.)
5896 """
5898 def __init__(self, expr: Union[ParserElement, str], savelist: bool = False):
5899 if expr is ...:
5900 expr = _PendingSkip(NoMatch())
5901 super().__init__(expr)
5903 def __add__(self, other) -> "ParserElement":
5904 if isinstance(self.expr, _PendingSkip):
5905 return Suppress(SkipTo(other)) + other
5907 return super().__add__(other)
5909 def __sub__(self, other) -> "ParserElement":
5910 if isinstance(self.expr, _PendingSkip):
5911 return Suppress(SkipTo(other)) - other
5913 return super().__sub__(other)
5915 def postParse(self, instring, loc, tokenlist):
5916 return []
5918 def suppress(self) -> ParserElement:
5919 return self
5922def trace_parse_action(f: ParseAction) -> ParseAction:
5923 """Decorator for debugging parse actions.
5925 When the parse action is called, this decorator will print
5926 ``">> entering method-name(line:<current_source_line>, <parse_location>, <matched_tokens>)"``.
5927 When the parse action completes, the decorator will print
5928 ``"<<"`` followed by the returned value, or any exception that the parse action raised.
5930 Example::
5932 wd = Word(alphas)
5934 @trace_parse_action
5935 def remove_duplicate_chars(tokens):
5936 return ''.join(sorted(set(''.join(tokens))))
5938 wds = wd[1, ...].set_parse_action(remove_duplicate_chars)
5939 print(wds.parse_string("slkdjs sld sldd sdlf sdljf"))
5941 prints::
5943 >>entering remove_duplicate_chars(line: 'slkdjs sld sldd sdlf sdljf', 0, (['slkdjs', 'sld', 'sldd', 'sdlf', 'sdljf'], {}))
5944 <<leaving remove_duplicate_chars (ret: 'dfjkls')
5945 ['dfjkls']
5946 """
5947 f = _trim_arity(f)
5949 def z(*paArgs):
5950 thisFunc = f.__name__
5951 s, l, t = paArgs[-3:]
5952 if len(paArgs) > 3:
5953 thisFunc = f"{type(paArgs[0]).__name__}.{thisFunc}"
5954 sys.stderr.write(f">>entering {thisFunc}(line: {line(l, s)!r}, {l}, {t!r})\n")
5955 try:
5956 ret = f(*paArgs)
5957 except Exception as exc:
5958 sys.stderr.write(
5959 f"<<leaving {thisFunc} (exception: {type(exc).__name__}: {exc})\n"
5960 )
5961 raise
5962 sys.stderr.write(f"<<leaving {thisFunc} (ret: {ret!r})\n")
5963 return ret
5965 z.__name__ = f.__name__
5966 return z
5969# convenience constants for positional expressions
5970empty = Empty().set_name("empty")
5971line_start = LineStart().set_name("line_start")
5972line_end = LineEnd().set_name("line_end")
5973string_start = StringStart().set_name("string_start")
5974string_end = StringEnd().set_name("string_end")
5976_escapedPunc = Regex(r"\\[\\[\]\/\-\*\.\$\+\^\?()~ ]").set_parse_action(
5977 lambda s, l, t: t[0][1]
5978)
5979_escapedHexChar = Regex(r"\\0?[xX][0-9a-fA-F]+").set_parse_action(
5980 lambda s, l, t: chr(int(t[0].lstrip(r"\0x"), 16))
5981)
5982_escapedOctChar = Regex(r"\\0[0-7]+").set_parse_action(
5983 lambda s, l, t: chr(int(t[0][1:], 8))
5984)
5985_singleChar = (
5986 _escapedPunc | _escapedHexChar | _escapedOctChar | CharsNotIn(r"\]", exact=1)
5987)
5988_charRange = Group(_singleChar + Suppress("-") + _singleChar)
5989_reBracketExpr = (
5990 Literal("[")
5991 + Opt("^").set_results_name("negate")
5992 + Group(OneOrMore(_charRange | _singleChar)).set_results_name("body")
5993 + Literal("]")
5994)
5997def srange(s: str) -> str:
5998 r"""Helper to easily define string ranges for use in :class:`Word`
5999 construction. Borrows syntax from regexp ``'[]'`` string range
6000 definitions::
6002 srange("[0-9]") -> "0123456789"
6003 srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz"
6004 srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_"
6006 The input string must be enclosed in []'s, and the returned string
6007 is the expanded character set joined into a single string. The
6008 values enclosed in the []'s may be:
6010 - a single character
6011 - an escaped character with a leading backslash (such as ``\-``
6012 or ``\]``)
6013 - an escaped hex character with a leading ``'\x'``
6014 (``\x21``, which is a ``'!'`` character) (``\0x##``
6015 is also supported for backwards compatibility)
6016 - an escaped octal character with a leading ``'\0'``
6017 (``\041``, which is a ``'!'`` character)
6018 - a range of any of the above, separated by a dash (``'a-z'``,
6019 etc.)
6020 - any combination of the above (``'aeiouy'``,
6021 ``'a-zA-Z0-9_$'``, etc.)
6022 """
6023 _expanded = lambda p: (
6024 p
6025 if not isinstance(p, ParseResults)
6026 else "".join(chr(c) for c in range(ord(p[0]), ord(p[1]) + 1))
6027 )
6028 try:
6029 return "".join(_expanded(part) for part in _reBracketExpr.parse_string(s).body)
6030 except Exception as e:
6031 return ""
6034def token_map(func, *args) -> ParseAction:
6035 """Helper to define a parse action by mapping a function to all
6036 elements of a :class:`ParseResults` list. If any additional args are passed,
6037 they are forwarded to the given function as additional arguments
6038 after the token, as in
6039 ``hex_integer = Word(hexnums).set_parse_action(token_map(int, 16))``,
6040 which will convert the parsed data to an integer using base 16.
6042 Example (compare the last to example in :class:`ParserElement.transform_string`::
6044 hex_ints = Word(hexnums)[1, ...].set_parse_action(token_map(int, 16))
6045 hex_ints.run_tests('''
6046 00 11 22 aa FF 0a 0d 1a
6047 ''')
6049 upperword = Word(alphas).set_parse_action(token_map(str.upper))
6050 upperword[1, ...].run_tests('''
6051 my kingdom for a horse
6052 ''')
6054 wd = Word(alphas).set_parse_action(token_map(str.title))
6055 wd[1, ...].set_parse_action(' '.join).run_tests('''
6056 now is the winter of our discontent made glorious summer by this sun of york
6057 ''')
6059 prints::
6061 00 11 22 aa FF 0a 0d 1a
6062 [0, 17, 34, 170, 255, 10, 13, 26]
6064 my kingdom for a horse
6065 ['MY', 'KINGDOM', 'FOR', 'A', 'HORSE']
6067 now is the winter of our discontent made glorious summer by this sun of york
6068 ['Now Is The Winter Of Our Discontent Made Glorious Summer By This Sun Of York']
6069 """
6071 def pa(s, l, t):
6072 return [func(tokn, *args) for tokn in t]
6074 func_name = getattr(func, "__name__", getattr(func, "__class__").__name__)
6075 pa.__name__ = func_name
6077 return pa
6080def autoname_elements() -> None:
6081 """
6082 Utility to simplify mass-naming of parser elements, for
6083 generating railroad diagram with named subdiagrams.
6084 """
6085 calling_frame = sys._getframe().f_back
6086 if calling_frame is None:
6087 return
6088 calling_frame = typing.cast(types.FrameType, calling_frame)
6089 for name, var in calling_frame.f_locals.items():
6090 if isinstance(var, ParserElement) and not var.customName:
6091 var.set_name(name)
6094dbl_quoted_string = Combine(
6095 Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"'
6096).set_name("string enclosed in double quotes")
6098sgl_quoted_string = Combine(
6099 Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'"
6100).set_name("string enclosed in single quotes")
6102quoted_string = Combine(
6103 (Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"').set_name(
6104 "double quoted string"
6105 )
6106 | (Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'").set_name(
6107 "single quoted string"
6108 )
6109).set_name("quoted string using single or double quotes")
6111python_quoted_string = Combine(
6112 (Regex(r'"""(?:[^"\\]|""(?!")|"(?!"")|\\.)*', flags=re.MULTILINE) + '"""').set_name(
6113 "multiline double quoted string"
6114 )
6115 ^ (
6116 Regex(r"'''(?:[^'\\]|''(?!')|'(?!'')|\\.)*", flags=re.MULTILINE) + "'''"
6117 ).set_name("multiline single quoted string")
6118 ^ (Regex(r'"(?:[^"\n\r\\]|(?:\\")|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*') + '"').set_name(
6119 "double quoted string"
6120 )
6121 ^ (Regex(r"'(?:[^'\n\r\\]|(?:\\')|(?:\\(?:[^x]|x[0-9a-fA-F]+)))*") + "'").set_name(
6122 "single quoted string"
6123 )
6124).set_name("Python quoted string")
6126unicode_string = Combine("u" + quoted_string.copy()).set_name("unicode string literal")
6129alphas8bit = srange(r"[\0xc0-\0xd6\0xd8-\0xf6\0xf8-\0xff]")
6130punc8bit = srange(r"[\0xa1-\0xbf\0xd7\0xf7]")
6132# build list of built-in expressions, for future reference if a global default value
6133# gets updated
6134_builtin_exprs: List[ParserElement] = [
6135 v for v in vars().values() if isinstance(v, ParserElement)
6136]
6138# Compatibility synonyms
6139# fmt: off
6140sglQuotedString = sgl_quoted_string
6141dblQuotedString = dbl_quoted_string
6142quotedString = quoted_string
6143unicodeString = unicode_string
6144lineStart = line_start
6145lineEnd = line_end
6146stringStart = string_start
6147stringEnd = string_end
6148nullDebugAction = replaced_by_pep8("nullDebugAction", null_debug_action)
6149traceParseAction = replaced_by_pep8("traceParseAction", trace_parse_action)
6150conditionAsParseAction = replaced_by_pep8("conditionAsParseAction", condition_as_parse_action)
6151tokenMap = replaced_by_pep8("tokenMap", token_map)
6152# fmt: on