def store_parse(self, analysed_article, data):
analysis_sentences = {
sentence.id:
AnalysisSentence.objects.create(analysed_article=analysed_article,
sentence=sentence).id
for sentence in sbd.get_or_create_sentences(
analysed_article.article)
}
result = interpret_output(analysis_sentences, data)
wordcreator.store_analysis(analysed_article, *result)
def add_sentence(asent):
rdf = get_rdf(asent.sentence.sentence)
bytes = rdf.read()
fn = "/tmp/sent_{asent.id}.rdf".format(**locals())
open(fn, 'w').write(bytes)
log.info("Parsing sent {asent.id} from {fn}".format(**locals()))
rdf = open(fn)
sentences, words = parse_rdf(rdf)
tokens, triples = create_values(asent.id, words)
wordcreator.store_analysis(asent.analysed_article, tokens, triples)
def add_sentence(asent):
rdf = get_rdf(asent.sentence.sentence)
bytes = rdf.read()
fn = "/tmp/sent_{asent.id}.rdf".format(**locals())
open(fn, 'w').write(bytes)
log.info("Parsing sent {asent.id} from {fn}".format(**locals()))
rdf = open(fn)
sentences, words = parse_rdf(rdf)
tokens, triples = create_values(asent.id, words)
wordcreator.store_analysis(asent.analysed_article, tokens, triples)
def store_parse(self, analysed_article, data):
if data.startswith("CoreNLP failed"):
raise Exception(data)
root = ElementTree.fromstring(data)
# if the analysis sentences already exist, check there are no tokens and line the analysis_sentence up.
# otherwise, create new ones
sentences = list(sbd.get_or_create_sentences(analysed_article.article))
if AnalysisSentence.objects.filter(analysed_article=analysed_article).exists():
if Token.objects.filter(sentence__analysed_article=analysed_article).exists():
raise Exception("Article already has tokens!")
analysis_sentences = [AnalysisSentence.objects.get(analysed_article=analysed_article, sentence=sentence).id
for sentence in sentences]
else:
analysis_sentences = [AnalysisSentence.objects.create(analysed_article=analysed_article, sentence=sentence).id
for sentence in sentences]
result = interpret_xml(analysis_sentences, root)
import pickle; pickle.dump(result[1], open("/tmp/triples", "w"))
wordcreator.store_analysis(analysed_article, *result)
def store_parse(self, analysed_article, data):
analysis_sentences = {sentence.id : AnalysisSentence.objects.create(analysed_article=analysed_article, sentence=sentence).id
for sentence in sbd.get_or_create_sentences(analysed_article.article)}
result = interpret_output(analysis_sentences, data)
wordcreator.store_analysis(analysed_article, *result)