def fill_data(self, num_of_docs):
num = 0
for f in os.listdir(self.path_to_corpus):
if num == num_of_docs:
return
print str(num), str(len(self.data.keys()))
num += 1
corpus = Corpus()
corpus.fill(self.path_to_corpus + f)
for sentence in corpus.sentences:
for wordform in sentence.wordforms:
wf_noun = wordform
if wf_noun.upostag != 'NOUN' or wf_noun.deprel != 'ROOT':
continue
wf_verb = sentence.wordforms[int(wf_noun.head)]
if wf_verb.upostag != 'VERB' or wf_verb.lemma.lower(
) == 'be':
continue
main_word = wf_verb.lemma.lower()
context = wf_noun.lemma.lower()
# needed_pos = sentence.wordforms[int(wordform.head)].upostag == 'VERB' and wordform.upostag == 'ADP'
# # conditions = wordform.xpostag not in self.pos_to_remove and wordform.lemma.lower()!='not'
# if not needed_pos:
# continue
# main_word = sentence.wordforms[int(wordform.head)].lemma.lower()
# context = wordform.lemma.lower()
if main_word in self.data.keys():
if context in self.data[main_word].keys():
self.data[main_word][context] += 1
else:
self.data[main_word][context] = 1
else:
w = {context: 1}
self.data[main_word] = w
