def temp():
art_files = [glob.glob(art_dirs_pre + art_dir + "*.txt") for art_dir in art_dirs]
with open(art_files[1][134], "r") as f:
art_text = f.read()
art_sents = nltk.sent_tokenize(art_text)
sent_ = art_sents[7]
sent_doc = source.nlp(unicode(sent_))
_questRewr.is_sentence(sent_doc)
good = wn.synset("good.a.01")
good.antonyms()
def temp():
art_files = [glob.glob(art_dirs_pre + art_dir + '*.txt') for art_dir in art_dirs]
with open(art_files[1][134], 'r') as f:
art_text = f.read()
art_sents = nltk.sent_tokenize(art_text)
sent_ = art_sents[7]
sent_doc = source.nlp(unicode(sent_))
_questRewr.is_sentence(sent_doc)
good = wn.synset('good.a.01')
good.antonyms()
def select_sents(art_text, nlp, min_sent_n=1, max_sent_n=5, max_sent_l=30, rng=np.random.RandomState(156), max_iter=20):
sent_offs = 3
sents = [unicode(s) for s in nltk.sent_tokenize(art_text)]
sents_docs = [nlp(s) for s in sents]
sents_n = len(sents)
all_sents = False
i_ = 0
try:
while not all_sents and i_ < max_iter:
sent_n = rng.randint(min_sent_n, max_sent_n + 1)
sent_i0 = rng.randint(sent_offs, sents_n - sent_offs - sent_n)
sent_in = sent_i0 + sent_n - 1
if len(nltk.word_tokenize(sents[sent_in])) < max_sent_l:
all_sents = np.all([_questRewr.is_compl_sentence(s) for s in sents_docs[slice(sent_i0, sent_in)]])
i_ += 1
except ValueError:
return None
if all_sents:
return sents[slice(sent_i0, sent_in)]
else:
return None
def select_sents(art_text, nlp, min_sent_n=1, max_sent_n=5,
max_sent_l=30, rng=np.random.RandomState(156), max_iter=20):
sent_offs = 3
sents = [unicode(s) for s in nltk.sent_tokenize(art_text)]
sents_docs = [nlp(s) for s in sents]
sents_n = len(sents)
all_sents = False
i_ = 0
try:
while not all_sents and i_ < max_iter:
sent_n = rng.randint(min_sent_n, max_sent_n+1)
sent_i0 = rng.randint(sent_offs, sents_n - sent_offs - sent_n)
sent_in = sent_i0 + sent_n - 1
if len(nltk.word_tokenize(sents[sent_in])) < max_sent_l:
all_sents = np.all([_questRewr.is_compl_sentence(s)
for s in sents_docs[slice(sent_i0, sent_in)]])
i_+=1
except ValueError:
return None
if all_sents:
return sents[slice(sent_i0, sent_in)]
else:
return None
