def perform(self, input): message = Serializer.parse(input, Message) if message.type == 'command': if message.body == 'DIE': self.logger.info("Received command message DIE") self.stop() else: self.logger.error('Unknown command message: %s', message.body) self.deliver(message) question = Question(message.body) question.tokens = word_tokenize(question.body) for snippet in question.snippets: snippet.sentences = self.tokenize(snippet.text) message.body = question self.deliver(message)
def perform(self, input): """""" """ Kaichen Chen 0304 """ print "Splitter {}: received, start to perform.".format(time()) message = Serializer.parse(input, Message) if message.type == 'command': if message.body == 'DIE': self.logger.info("Received command message DIE") self.stop() else: self.logger.error('Unknown command message: %s', message.body) self.deliver(message) question = Question(message.body) question.tokens = word_tokenize(question.body) for snippet in question.snippets: snippet.sentences = self.tokenize(snippet.text) message.body = question self.deliver(message)
}, { u'offsetInBeginSection': 1276, u'offsetInEndSection': 1608, u'text': u'The increased odds ratios with African Americans was retained in post-menopausal women, while the protective odds ratios for pregnancy, smoking and oral contraceptives (OCs) became stronger in pre-menopausal women. The pattern by duration and timing of use does not suggest an etiologic role for OCs or hormone replacement therapy. ', u'beginSection': u'abstract', u'document': u'http://www.ncbi.nlm.nih.gov/pubmed/16570277', u'endSection': u'abstract' }, { u'offsetInBeginSection': 1044, u'offsetInEndSection': 1221, u'text': u'The use of hormone replacement therapy in symptomatic postmenopausal women either with previously treated disease or with dormant tumors is discussed, but remains controversial.', u'beginSection': u'abstract', u'document': u'http://www.ncbi.nlm.nih.gov/pubmed/15006250', u'endSection': u'abstract' }] } question = Question(data) question.tokens = word_tokenize(question.body) for snippet in question.snippets: snippet.sentences = tokenize(snippet.text) sci = CoreMMR() sentences = sci.getRankedList(question) print sentences print 'Q. ' + question.body for i, s in enumerate(sentences): print "{} {}".format(i, s)