def __init__(self, is_title, text, sentence_opins, objects_list, sentence_index):
assert(isinstance(is_title, bool))
assert(isinstance(sentence_opins, list))
assert(isinstance(objects_list, list))
assert(isinstance(sentence_index, int))
super(RuAttitudesSentence, self).__init__(text=split_by_whitespaces(text))
self.__is_title = is_title
self.__sentence_opins = sentence_opins
self.__objects = objects_list
self.__sentence_index = sentence_index
self.__owner = None
return __process_indices_list(value)
parse_value = {
const.ID: lambda value: value,
const.S_IND: lambda value: int(value),
const.T_IND: lambda value: int(value),
network_input_const.FrameVariantIndices: lambda value:
__process_indices_list(value) if isinstance(value, str) else empty_list,
network_input_const.FrameConnotations: lambda value:
__process_indices_list(value) if isinstance(value, str) else empty_list,
network_input_const.SynonymObject: lambda value: __process_indices_list(value),
network_input_const.SynonymSubject: lambda value: __process_indices_list(value),
network_input_const.Entities: lambda value: __process_indices_list(value),
network_input_const.PosTags: lambda value: __process_int_values_list(value),
"text_a": lambda value: filter_whitespaces([term for term in split_by_whitespaces(value)])
}
class ParsedSampleRow(object):
"""
Provides a parsed information for a sample row.
TODO. Use this class as API
"""
def __init__(self, row):
assert(isinstance(row, pd.Series))
self.__uint_label = None
self.__params = {}
def __iter_processed_part(self, part):
for word in split_by_whitespaces(part):
for term in self.__process_word(word):
yield term
def __calculate_terms_in_line(line):
assert (isinstance(line, str))
return len(split_by_whitespaces(line))
def apply(self, pipeline_ctx):
assert (isinstance(pipeline_ctx, PipelineContext))
return pipeline_ctx.update(param="src",
value=split_by_whitespaces(
pipeline_ctx.provide("src")))