2015-02-07 18:14:07 +00:00
|
|
|
from __future__ import unicode_literals
|
|
|
|
|
2015-07-13 16:39:38 +00:00
|
|
|
from spacy.tokens import Doc
|
|
|
|
|
2015-02-07 18:14:07 +00:00
|
|
|
import pytest
|
|
|
|
|
2015-07-26 14:23:41 +00:00
|
|
|
@pytest.mark.models
|
|
|
|
def test_getitem(EN):
|
2015-06-07 16:02:24 +00:00
|
|
|
tokens = EN(u'Give it back! He pleaded.')
|
2015-02-07 18:14:07 +00:00
|
|
|
assert tokens[0].orth_ == 'Give'
|
|
|
|
assert tokens[-1].orth_ == '.'
|
|
|
|
with pytest.raises(IndexError):
|
|
|
|
tokens[len(tokens)]
|
2015-07-13 16:39:38 +00:00
|
|
|
|
2015-10-06 08:59:11 +00:00
|
|
|
def to_str(span):
|
|
|
|
return '/'.join(token.orth_ for token in span)
|
|
|
|
|
2015-10-06 07:51:25 +00:00
|
|
|
span = tokens[1:1]
|
2015-10-06 08:59:11 +00:00
|
|
|
assert not to_str(span)
|
2015-10-06 07:51:25 +00:00
|
|
|
span = tokens[1:4]
|
2015-10-06 08:59:11 +00:00
|
|
|
assert to_str(span) == 'it/back/!'
|
2015-10-06 07:56:33 +00:00
|
|
|
span = tokens[1:4:1]
|
2015-10-06 08:59:11 +00:00
|
|
|
assert to_str(span) == 'it/back/!'
|
2015-10-06 07:51:25 +00:00
|
|
|
with pytest.raises(ValueError):
|
|
|
|
tokens[1:4:2]
|
|
|
|
with pytest.raises(ValueError):
|
|
|
|
tokens[1:4:-1]
|
|
|
|
|
2015-10-06 08:59:11 +00:00
|
|
|
span = tokens[-3:6]
|
|
|
|
assert to_str(span) == 'He/pleaded'
|
|
|
|
span = tokens[4:-1]
|
|
|
|
assert to_str(span) == 'He/pleaded'
|
|
|
|
span = tokens[-5:-3]
|
|
|
|
assert to_str(span) == 'back/!'
|
|
|
|
span = tokens[5:4]
|
|
|
|
assert span.start == span.end == 5 and not to_str(span)
|
|
|
|
span = tokens[4:-3]
|
|
|
|
assert span.start == span.end == 4 and not to_str(span)
|
|
|
|
|
|
|
|
span = tokens[:]
|
|
|
|
assert to_str(span) == 'Give/it/back/!/He/pleaded/.'
|
|
|
|
span = tokens[4:]
|
|
|
|
assert to_str(span) == 'He/pleaded/.'
|
|
|
|
span = tokens[:4]
|
|
|
|
assert to_str(span) == 'Give/it/back/!'
|
|
|
|
span = tokens[:-3]
|
|
|
|
assert to_str(span) == 'Give/it/back/!'
|
|
|
|
span = tokens[-3:]
|
|
|
|
assert to_str(span) == 'He/pleaded/.'
|
|
|
|
|
|
|
|
span = tokens[4:50]
|
|
|
|
assert to_str(span) == 'He/pleaded/.'
|
|
|
|
span = tokens[-50:4]
|
|
|
|
assert to_str(span) == 'Give/it/back/!'
|
|
|
|
span = tokens[-50:-40]
|
|
|
|
assert span.start == span.end == 0 and not to_str(span)
|
|
|
|
span = tokens[40:50]
|
|
|
|
assert span.start == span.end == 7 and not to_str(span)
|
|
|
|
|
2015-10-06 09:08:39 +00:00
|
|
|
span = tokens[1:4]
|
|
|
|
assert span[0].orth_ == 'it'
|
2015-10-06 09:45:49 +00:00
|
|
|
subspan = span[:]
|
|
|
|
assert to_str(subspan) == 'it/back/!'
|
|
|
|
subspan = span[:2]
|
|
|
|
assert to_str(subspan) == 'it/back'
|
|
|
|
subspan = span[1:]
|
|
|
|
assert to_str(subspan) == 'back/!'
|
|
|
|
subspan = span[:-1]
|
|
|
|
assert to_str(subspan) == 'it/back'
|
|
|
|
subspan = span[-2:]
|
|
|
|
assert to_str(subspan) == 'back/!'
|
|
|
|
subspan = span[1:2]
|
|
|
|
assert to_str(subspan) == 'back'
|
|
|
|
subspan = span[-2:-1]
|
|
|
|
assert to_str(subspan) == 'back'
|
|
|
|
subspan = span[-50:50]
|
|
|
|
assert to_str(subspan) == 'it/back/!'
|
|
|
|
subspan = span[50:-50]
|
|
|
|
assert subspan.start == subspan.end == 4 and not to_str(subspan)
|
2015-10-06 09:08:39 +00:00
|
|
|
|
2015-07-13 16:39:38 +00:00
|
|
|
|
2015-07-26 14:23:41 +00:00
|
|
|
@pytest.mark.models
|
|
|
|
def test_serialize(EN):
|
2015-07-22 23:19:11 +00:00
|
|
|
tokens = EN(u'Give it back! He pleaded.')
|
|
|
|
packed = tokens.to_bytes()
|
|
|
|
new_tokens = Doc(EN.vocab).from_bytes(packed)
|
|
|
|
assert tokens.string == new_tokens.string
|
|
|
|
assert [t.orth_ for t in tokens] == [t.orth_ for t in new_tokens]
|
|
|
|
assert [t.orth for t in tokens] == [t.orth for t in new_tokens]
|
|
|
|
|
|
|
|
|
2015-07-26 14:23:41 +00:00
|
|
|
@pytest.mark.models
|
2015-07-22 23:19:11 +00:00
|
|
|
def test_serialize_whitespace(EN):
|
2015-07-13 16:39:38 +00:00
|
|
|
tokens = EN(u' Give it back! He pleaded. ')
|
2015-07-22 23:19:11 +00:00
|
|
|
packed = tokens.to_bytes()
|
|
|
|
new_tokens = Doc(EN.vocab).from_bytes(packed)
|
2015-07-13 16:39:38 +00:00
|
|
|
assert tokens.string == new_tokens.string
|
|
|
|
assert [t.orth_ for t in tokens] == [t.orth_ for t in new_tokens]
|
|
|
|
assert [t.orth for t in tokens] == [t.orth for t in new_tokens]
|
2015-08-05 22:35:40 +00:00
|
|
|
|
|
|
|
|
|
|
|
def test_set_ents(EN):
|
|
|
|
tokens = EN.tokenizer(u'I use goggle chrone to surf the web')
|
|
|
|
assert len(tokens.ents) == 0
|
|
|
|
tokens.ents = [(EN.vocab.strings['PRODUCT'], 2, 4)]
|
|
|
|
assert len(list(tokens.ents)) == 1
|
|
|
|
assert [t.ent_iob for t in tokens] == [0, 0, 3, 1, 0, 0, 0, 0]
|
|
|
|
ent = tokens.ents[0]
|
|
|
|
assert ent.label_ == 'PRODUCT'
|
|
|
|
assert ent.start == 2
|
|
|
|
assert ent.end == 4
|