def test_extend(self):
tv1 = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1)
tv2 = TokenVector.make_token_vector(
["E", "F", "G", "H"],
sep_size=1)
tvUnion = TokenVector.make_token_vector(
["A", "B", "C", "D", "E", "F", "G", "H"],
sep_size=1)
tv1.extend(tv2, sep_size=1)
self.assertListEqual(tv1, tvUnion)
def test_get_index_length(self):
vector = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1,
sentence_idx=0)
self.assertEquals(vector.get_index_length(), 7)
def test_get_sentences(self):
sentence_one_vector = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1,
sentence_idx=0)
sentence_two_vector = TokenVector.make_token_vector(
["E", "F", "G", "H"],
relindex=9,
sep_size=1,
sentence_idx=1)
union_vector = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1,
sentence_idx=0)
union_vector.extend(sentence_two_vector)
sentences = union_vector.get_sentences()
self.assertListEqual(sentences[0], sentence_one_vector)
self.assertListEqual(sentences[1], sentence_two_vector)
def test_garbage_collection(self):
tv = TokenVector.make_token_vector(
["A", "", "B", "C", "D", "", "E", ""],
sep_size=1,
sentence_idx=0)
clean_tv = TokenVector([])
clean_tv.append(Token('A', sindex=0, eindex=0, sentence_idx=0))
clean_tv.append(Token('B', sindex=3, eindex=3, sentence_idx=0))
clean_tv.append(Token('C', sindex=5, eindex=5, sentence_idx=0))
clean_tv.append(Token('D', sindex=7, eindex=7, sentence_idx=0))
clean_tv.append(Token('E', sindex=10, eindex=10, sentence_idx=0))
tv.collect_garbage()
self.assertSequenceEqual(tv, clean_tv)
def test_make_token_vector(self):
made_tv = TokenVector.make_token_vector(
["A", "", "B", "C", "D", "", "E", ""],
sep_size=1,
sentence_idx=0)
real_tv = TokenVector([])
real_tv.append(Token('A', sindex=0, eindex=0, sentence_idx=0))
real_tv.append(Token('', sindex=2, eindex=2, sentence_idx=0))
real_tv.append(Token('B', sindex=3, eindex=3, sentence_idx=0))
real_tv.append(Token('C', sindex=5, eindex=5, sentence_idx=0))
real_tv.append(Token('D', sindex=7, eindex=7, sentence_idx=0))
real_tv.append(Token('', sindex=9, eindex=9, sentence_idx=0))
real_tv.append(Token('E', sindex=10, eindex=10, sentence_idx=0))
real_tv.append(Token('', sindex=12, eindex=12, sentence_idx=0))
self.assertSequenceEqual(made_tv, real_tv)
def setUp(self):
self.tv = TokenVector([Token("foo"),
Token("abc defg ijk", sindex=4),
Token("bla", sindex=17)])
class TestTokenVector(TestCase):
def setUp(self):
self.tv = TokenVector([Token("foo"),
Token("abc defg ijk", sindex=4),
Token("bla", sindex=17)])
def test_constructor_end_indices(self):
self.assertEqual(self.tv[0].eindex, 2)
self.assertEqual(self.tv[1].eindex, 15)
self.assertEqual(self.tv[2].eindex, 19)
def test_replace_tokens(self):
print self.tv
new_tv = TokenVector([Token("abc"), Token("defg"), Token("ijk")])
self.tv.replace_tokens_with_list({1: new_tv})
print self.tv
self.assertEqual(len(self.tv), 5, "size kaputt")
for index, token in enumerate(new_tv):
self.assertEqual(self.tv[index + 1], new_tv[index])
def test_update_indices(self):
self.tv.update_indices(5)
self.assertEqual(self.tv[0].eindex, 7)
self.assertEqual(self.tv[1].eindex, 20)
self.assertEqual(self.tv[2].eindex, 24)
self.assertEqual(self.tv[0].sindex, 5)
self.assertEqual(self.tv[1].sindex, 9)
self.assertEqual(self.tv[2].sindex, 22)
def test_get_index_length(self):
vector = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1,
sentence_idx=0)
self.assertEquals(vector.get_index_length(), 7)
def test_extend(self):
tv1 = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1)
tv2 = TokenVector.make_token_vector(
["E", "F", "G", "H"],
sep_size=1)
tvUnion = TokenVector.make_token_vector(
["A", "B", "C", "D", "E", "F", "G", "H"],
sep_size=1)
tv1.extend(tv2, sep_size=1)
self.assertListEqual(tv1, tvUnion)
def test_get_sentences(self):
sentence_one_vector = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1,
sentence_idx=0)
sentence_two_vector = TokenVector.make_token_vector(
["E", "F", "G", "H"],
relindex=9,
sep_size=1,
sentence_idx=1)
union_vector = TokenVector.make_token_vector(
["A", "B", "C", "D"],
sep_size=1,
sentence_idx=0)
union_vector.extend(sentence_two_vector)
sentences = union_vector.get_sentences()
self.assertListEqual(sentences[0], sentence_one_vector)
self.assertListEqual(sentences[1], sentence_two_vector)
def test_make_token_vector(self):
made_tv = TokenVector.make_token_vector(
["A", "", "B", "C", "D", "", "E", ""],
sep_size=1,
sentence_idx=0)
real_tv = TokenVector([])
real_tv.append(Token('A', sindex=0, eindex=0, sentence_idx=0))
real_tv.append(Token('', sindex=2, eindex=2, sentence_idx=0))
real_tv.append(Token('B', sindex=3, eindex=3, sentence_idx=0))
real_tv.append(Token('C', sindex=5, eindex=5, sentence_idx=0))
real_tv.append(Token('D', sindex=7, eindex=7, sentence_idx=0))
real_tv.append(Token('', sindex=9, eindex=9, sentence_idx=0))
real_tv.append(Token('E', sindex=10, eindex=10, sentence_idx=0))
real_tv.append(Token('', sindex=12, eindex=12, sentence_idx=0))
self.assertSequenceEqual(made_tv, real_tv)
def test_garbage_collection(self):
tv = TokenVector.make_token_vector(
["A", "", "B", "C", "D", "", "E", ""],
sep_size=1,
sentence_idx=0)
clean_tv = TokenVector([])
clean_tv.append(Token('A', sindex=0, eindex=0, sentence_idx=0))
clean_tv.append(Token('B', sindex=3, eindex=3, sentence_idx=0))
clean_tv.append(Token('C', sindex=5, eindex=5, sentence_idx=0))
clean_tv.append(Token('D', sindex=7, eindex=7, sentence_idx=0))
clean_tv.append(Token('E', sindex=10, eindex=10, sentence_idx=0))
tv.collect_garbage()
self.assertSequenceEqual(tv, clean_tv)
def _get_tokens(self): # tokenizer dummie
return TokenVector([Token("abc", sindex=0),
Token("defg", sindex=4),
Token("ijk", sindex=9)])