def __init__(self, path=None, lang=None, result_type=Parse, units=None,
probability_estimator_cls=auto, char_substitutes=auto):
# save arguments for pickling/unpickling
self._path = path
self._lang = lang
if path is None and lang is None:
lang = 'ru'
path = self.choose_dictionary_path(path, lang)
with self._lock:
self.dictionary = opencorpora_dict.Dictionary(path)
self.lang = self.choose_language(self.dictionary, lang)
self.prob_estimator = self._get_prob_estimator(
probability_estimator_cls, self.dictionary, path
)
if result_type is not None:
# create a subclass with the same name,
# but with _morph attribute bound to self
res_type = type(
result_type.__name__,
(result_type,),
{'_morph': self, '_dict': self.dictionary}
)
self._result_type = res_type
else:
self._result_type = None
self._result_type_orig = result_type
self._init_char_substitutes(char_substitutes)
self._init_units(units)
def __init__(self,
path=None,
result_type=Parse,
units=None,
probability_estimator_cls=SingleTagProbabilityEstimator):
path = self.choose_dictionary_path(path)
with threading.RLock():
self.dictionary = opencorpora_dict.Dictionary(path)
if probability_estimator_cls is None:
probability_estimator_cls = DummySingleTagProbabilityEstimator
self.prob_estimator = probability_estimator_cls(path)
if result_type is not None:
# create a subclass with the same name,
# but with _morph attribute bound to self
res_type = type(result_type.__name__, (result_type, ), {
'_morph': self,
'_dict': self.dictionary
})
self._result_type = res_type
else:
self._result_type = None
self._result_type_orig = result_type
self._init_units(units)
def __init__(self, path=None, lang='ru', result_type=Parse, units=None,
probability_estimator_cls=auto, char_substitutes=auto):
self._path = path
self._lang = lang
self.lang = None if path is None else lang
path = self.choose_dictionary_path(path, lang)
with threading.RLock():
self.dictionary = opencorpora_dict.Dictionary(path)
if self.lang is None:
# When path is passed explicitly, set language
# to dict language.
self.lang = self.dictionary.lang
elif self.dictionary.lang != self.lang:
warnings.warn("Dictionary language (%r) doesn't match "
"analyzer language (%r)." % (self.dictionary.lang, self.lang))
self.prob_estimator = self._get_prob_estimator(
probability_estimator_cls, self.dictionary, path
)
if result_type is not None:
# create a subclass with the same name,
# but with _morph attribute bound to self
res_type = type(
result_type.__name__,
(result_type,),
{'_morph': self, '_dict': self.dictionary}
)
self._result_type = res_type
else:
self._result_type = None
self._result_type_orig = result_type
self._init_char_substitutes(char_substitutes)
self._init_units(units)
def __init__(self, path=None, result_type=Parse, units=None):
path = self.choose_dictionary_path(path)
self.dictionary = opencorpora_dict.Dictionary(path)
if result_type is not None:
# create a subclass with the same name,
# but with _morph attribute bound to self
res_type = type(result_type.__name__, (result_type, ), {
'_morph': self,
'_dict': self.dictionary
})
self._result_type = res_type
else:
self._result_type = None
self._result_type_orig = result_type
# initialize units
if units is None:
units = self.DEFAULT_UNITS
self._unit_classes = units
self._units = [cls(self) for cls in units]
