def cql_search(request):
from pynlpl.formats import fql, cql
# парсинг входящих параметров
params = json.loads(request.body.decode('utf-8'))
# обновление фолиа-документа по актуальным данным
doc = folia.Document(id='doc')
text = folia.Text(doc, id='doc.text')
sentences = Sentence.objects.all()
# поиск слов в документе
for s in sentences:
sen = text.append(folia.Sentence(doc, id=doc.id + '.s.' + str(s.id)))
words = Word.objects.filter(Sentence_id=s.id)
for w in words:
sen.append(
folia.Word(doc,
id=doc.id + '.s.' + str(s.id) + '.w.' + str(w.id),
text=w.value))
doc.append(text)
query = fql.Query(cql.cql2fql(params['title']))
texts = query(doc)
arr = []
for t in texts:
arr.append(t[0].parent.id.split('s.')[1])
sens = Sentence.objects.filter(id__in=arr)
# вывод результатов
return render(request, 'cabinet/cql_results.html', {
'texts': texts,
'sens': sens
})
def splitcorrection(self, word, newwords, **kwargs):
str_newwords = ''
for nw in newwords:
str_newwords += str(nw) + ' '
self.errout("Splitting " + str(word) + " into " + str_newwords)
sentence = word.sentence()
newwords = [
folia.Word(self.doc, generate_id_in=sentence, text=w)
for w in newwords
]
kwargs['suggest'] = True
kwargs['datetime'] = datetime.datetime.now()
word.split(*newwords, **kwargs)
def mergecorrection(self, newword, originalwords, **kwargs):
str_originalwords = ''
for ow in originalwords:
str_originalwords += str(ow) + ' '
self.errout("Merging " + str_originalwords + " into " + newword)
sentence = originalwords[0].sentence()
if not sentence:
raise Exception("Expected sentence for " +
str(repr(originalwords[0])) + ", got " +
str(repr(sentence)))
newword = folia.Word(self.doc, generate_id_in=sentence, text=newword)
kwargs['suggest'] = True
kwargs['datetime'] = datetime.datetime.now()
sentence.mergewords(newword, *originalwords, **kwargs)
if pos:
words[i].append( folia.PosAnnotation(foliadoc, cls=pos) )
else:
print >>sys.stderr,"WARNING: Out of sync after calling Frog! ", i, word
else:
#pass untokenised sentence
try:
sentext = s.text()
except folia.NoSuchText:
continue
response = frogclient.process(sentext)
for i, (word, lemma, morph, pos) in enumerate(response):
if legacy: legacyout(i,word,lemma,morph,pos)
if word:
w = folia.Word(foliadoc, text=word, generate_id_in=s)
if lemma:
w.append( folia.LemmaAnnotation(foliadoc, cls=lemma) )
if pos:
w.append( folia.PosAnnotation(foliadoc, cls=pos) )
s.append(w)
if not found_s:
#pass paragraph
try:
partext = p.text()
except folia.NoSuchText:
continue
s = folia.Sentence(foliadoc, generate_id_in=p)
response = frogclient.process(partext)