def read_yaml_config(config_file):
return yaml.load(open(config_file))
if __name__ == "__main__":
args = setup_argparse()
nlp = spacy.load('en_vectors_web_lg')
nlp.add_pipe(nlp.create_pipe('sentencizer'))
print("Reading config...")
config = read_yaml_config(args.config_file)
print("Reading corpus...")
corpus = Corpus.from_config(config, nlp)
print("Reading summarizer...")
summarizer = Summarizer.from_config(config, nlp)
print("Creating orderer...")
information_orderer = setup_information_orderer()
num_topics = len(corpus.topics)
for i, topic in enumerate(corpus.topics, 1):
candidates = summarizer.summarize(topic)
summary = information_orderer.order_all(candidates)
print("Summarized {0}/{1} topics".format(i, num_topics))
with open('{0}{1}'.format(args.output_dir, make_filename(topic.id(), config.get(Summarizer.WORD_LIMIT_KEY))), 'w') as outfile:
