class LawsParser:
def __init__(self, gesetze_df: pd.DataFrame, verordnungen_df: pd.DataFrame,
concept_df: pd.DataFrame):
self.locale = ''
parse_columns = ('Kurztitel', 'Titel', 'Abkürzung')
dependent_columns = {'Titel': 'External Reference Normalized'}
preformed_entity = {
'External Reference Type': 'Laws and Rules',
'External Reference Source': 'BaFin',
'External Reference Issuing Country': 'Germany'
}
self.gesetze_parser = DataframeEntityParser(gesetze_df, parse_columns,
dependent_columns,
preformed_entity)
self.verordnungen_parser = DataframeEntityParser(
verordnungen_df, parse_columns, dependent_columns,
preformed_entity)
parse_columns = ('b', )
dependent_columns = {
'b': 'External Reference Normalized',
'a': 'External Reference Type'
}
preformed_entity.pop('External Reference Type')
self.concept_parser = DataframeEntityParser(concept_df, parse_columns,
dependent_columns,
preformed_entity)
def parse(self, text: str, locale: str = None) -> List[LawAnnotation]:
res = []
self.locale = locale if locale else 'de'
res.extend(self.gesetze_parser.get_entity_list(text))
res.extend(self.verordnungen_parser.get_entity_list(text))
res.extend(self.concept_parser.get_entity_list(text))
res_formatted = [] # type: List[LawAnnotation]
for i in res:
coords = (i.pop('location_start'), i.pop('location_end'))
text = i.pop('source')
ant = LawAnnotation(name=text,
coords=coords,
text=text,
locale=self.locale)
# new_item.update(i)
res_formatted.append(ant)
return res_formatted
class LawsParser:
def __init__(self, gesetze_df: pd.DataFrame, verordnungen_df: pd.DataFrame,
concept_df: pd.DataFrame):
self.locale = ''
parse_columns = ('Kurztitel', 'Titel', 'Abkürzung')
dependent_columns = {'Titel': 'External Reference Normalized'}
preformed_entity = {
'External Reference Type': 'Laws and Rules',
'External Reference Source': 'BaFin',
'External Reference Issuing Country': 'Germany'
}
split_params = LineSplitParams()
split_params.line_breaks = {'.', ';', '!', '?'}
split_params.abbreviations = DeLanguageTokens.abbreviations
split_params.abbr_ignore_case = True
proc = LineProcessor(line_split_params=split_params)
self.gesetze_parser = DataframeEntityParser(
gesetze_df,
parse_columns,
result_columns=dependent_columns,
preformed_entity=preformed_entity,
line_processor=proc)
self.verordnungen_parser = DataframeEntityParser(
verordnungen_df,
parse_columns,
result_columns=dependent_columns,
preformed_entity=preformed_entity,
line_processor=proc)
parse_columns = ('b', )
dependent_columns = {
'b': 'External Reference Normalized',
'a': 'External Reference Type'
}
preformed_entity.pop('External Reference Type')
self.concept_parser = DataframeEntityParser(
concept_df,
parse_columns,
result_columns=dependent_columns,
preformed_entity=preformed_entity,
line_processor=proc)
def parse(self, text: str, locale: str = None) -> List[LawAnnotation]:
res = []
self.locale = locale if locale else 'de'
res.extend(self.gesetze_parser.get_entity_list(text))
res.extend(self.verordnungen_parser.get_entity_list(text))
res.extend(self.concept_parser.get_entity_list(text))
res_formatted = [] # type: List[LawAnnotation]
for i in res:
coords = (i.pop('location_start'), i.pop('location_end'))
text = i.pop('source')
ant = LawAnnotation(name=text,
coords=coords,
text=text,
locale=self.locale)
# new_item.update(i)
res_formatted.append(ant)
return res_formatted