def _get_hit_phrases()

in distant_supervision/synthetic_data_creator.py [0:0]

• 8 lines of code
• 4 McCabe index (conditional complexity)

    def _get_hit_phrases(self, es_hit):
        entities = [tuple(e) for e in json.loads(es_hit['_source']['entities'])]
        noun_chunks = [tuple(e) for e in json.loads(es_hit['_source']['noun_chunks'])]

        if self.phrase_mode is PhraseMode.NER_ONLY:
            phrases = self.text_preprocessor.get_phrases(entities=entities, noun_chunks=[])
        else:
            phrases = self.text_preprocessor.get_phrases(entities=entities, noun_chunks=noun_chunks)
        return set(phrases)