2017-10-09 01:02:23 +03:00
|
|
|
'''Test that the parser respects preset sentence boundaries.'''
|
2017-10-09 01:59:49 +03:00
|
|
|
from __future__ import unicode_literals
|
2017-10-09 01:02:23 +03:00
|
|
|
import pytest
|
|
|
|
from thinc.neural.optimizers import Adam
|
|
|
|
from thinc.neural.ops import NumpyOps
|
|
|
|
|
|
|
|
from ...attrs import NORM
|
|
|
|
from ...gold import GoldParse
|
|
|
|
from ...vocab import Vocab
|
|
|
|
from ...tokens import Doc
|
2017-10-26 13:38:23 +03:00
|
|
|
from ...pipeline import DependencyParser
|
2017-10-09 01:02:23 +03:00
|
|
|
|
|
|
|
@pytest.fixture
|
|
|
|
def vocab():
|
|
|
|
return Vocab(lex_attr_getters={NORM: lambda s: s})
|
|
|
|
|
|
|
|
@pytest.fixture
|
|
|
|
def parser(vocab):
|
2017-10-26 13:38:23 +03:00
|
|
|
parser = DependencyParser(vocab)
|
2017-10-09 01:02:23 +03:00
|
|
|
parser.cfg['token_vector_width'] = 4
|
|
|
|
parser.cfg['hidden_width'] = 32
|
|
|
|
#parser.add_label('right')
|
|
|
|
parser.add_label('left')
|
|
|
|
parser.begin_training([], **parser.cfg)
|
|
|
|
sgd = Adam(NumpyOps(), 0.001)
|
|
|
|
|
|
|
|
for i in range(10):
|
|
|
|
losses = {}
|
|
|
|
doc = Doc(vocab, words=['a', 'b', 'c', 'd'])
|
|
|
|
gold = GoldParse(doc, heads=[1, 1, 3, 3],
|
|
|
|
deps=['left', 'ROOT', 'left', 'ROOT'])
|
|
|
|
parser.update([doc], [gold], sgd=sgd, losses=losses)
|
|
|
|
return parser
|
|
|
|
|
|
|
|
def test_no_sentences(parser):
|
|
|
|
doc = Doc(parser.vocab, words=['a', 'b', 'c', 'd'])
|
|
|
|
doc = parser(doc)
|
2017-10-12 22:18:22 +03:00
|
|
|
assert len(list(doc.sents)) >= 1
|
2017-10-09 01:02:23 +03:00
|
|
|
|
|
|
|
|
|
|
|
def test_sents_1(parser):
|
|
|
|
doc = Doc(parser.vocab, words=['a', 'b', 'c', 'd'])
|
|
|
|
doc[2].sent_start = True
|
|
|
|
doc = parser(doc)
|
2017-10-09 01:29:37 +03:00
|
|
|
assert len(list(doc.sents)) >= 2
|
2017-10-09 01:02:23 +03:00
|
|
|
doc = Doc(parser.vocab, words=['a', 'b', 'c', 'd'])
|
|
|
|
doc[1].sent_start = False
|
|
|
|
doc[2].sent_start = True
|
|
|
|
doc[3].sent_start = False
|
|
|
|
doc = parser(doc)
|
|
|
|
assert len(list(doc.sents)) == 2
|
|
|
|
|
|
|
|
|
|
|
|
def test_sents_1_2(parser):
|
|
|
|
doc = Doc(parser.vocab, words=['a', 'b', 'c', 'd'])
|
|
|
|
doc[1].sent_start = True
|
|
|
|
doc[2].sent_start = True
|
|
|
|
doc = parser(doc)
|
2017-11-03 16:36:08 +03:00
|
|
|
assert len(list(doc.sents)) >= 3
|
2017-10-09 01:02:23 +03:00
|
|
|
|
|
|
|
|
|
|
|
def test_sents_1_3(parser):
|
|
|
|
doc = Doc(parser.vocab, words=['a', 'b', 'c', 'd'])
|
|
|
|
doc[1].sent_start = True
|
|
|
|
doc[3].sent_start = True
|
|
|
|
doc = parser(doc)
|
2017-10-11 09:38:34 +03:00
|
|
|
assert len(list(doc.sents)) >= 3
|
2017-10-09 01:02:23 +03:00
|
|
|
doc = Doc(parser.vocab, words=['a', 'b', 'c', 'd'])
|
|
|
|
doc[1].sent_start = True
|
|
|
|
doc[2].sent_start = False
|
|
|
|
doc[3].sent_start = True
|
|
|
|
doc = parser(doc)
|
|
|
|
assert len(list(doc.sents)) == 3
|