def test_empty_set_format_error():
"""
Test that outputting an empty collection for the values of a column errors.
"""
token_line = '33 cintre cintre NOUN _ Gender=Fem|Number=Sing ' \
'30 nmod 2:nsubj|4:root SpaceAfter=No'
token = Token(token_line)
token.feats['Gender'].pop()
with pytest.raises(FormatError):
token.conll()
def test_all_empty_deps_component_error():
"""
Test that an error is thrown when all components of a dep value are None.
"""
token_line = '33 cintre cintre NOUN _ Gender=Fem|Number=Sing ' \
'30 nmod 2:nsubj|4:root SpaceAfter=No'
token = Token(token_line)
cur_list = [None] + list(token.deps['2'][1:])
token.deps['2'] = cur_list
with pytest.raises(FormatError):
token.conll()
def test_to_string():
"""
Test if a token's string representation is accurate.
"""
token_line = '26 surmonté surmonter VERB _ ' \
'Gender=Masc|Number=Sing|Tense=Past|VerbForm=Part 22 acl _ _'
token = Token(token_line)
assert token.conll() == token_line
def test_feats_keep_case_insensitive_order():
"""
Test that the features are kept sorted via case insensitive attributes.
"""
token_line = '10 gave give VERB _ gender=Fem|Number=Sing ' \
'0 root _ SpaceAfter=No'
token = Token(token_line)
conll = token.conll()
assert conll == token_line
def test_deps_parsing():
"""
Test that the deps field is properly parsed.
"""
token_line = '33 cintre cintre NOUN _ Gender=Masc|Number=Sing ' \
'30 nmod 2:nsubj|4:nmod SpaceAfter=No'
token = Token(token_line)
assert token.deps['2'] == ('nsubj', None, None, None)
assert token.deps['4'] == ('nmod', None, None, None)
assert token.conll() == token_line
def test_feats_induce_case_insensitive_order():
"""
Test that case insensitive sorting of feature attributes is induced.
"""
token_line = '10 gave give VERB _ Number=Sing|gender=Fem ' \
'0 root _ SpaceAfter=No'
token = Token(token_line)
conll = token.conll()
formatted_line = '10 gave give VERB _ gender=Fem|Number=Sing ' \
'0 root _ SpaceAfter=No'
assert conll == formatted_line
def test_deps_sort_order_decimal():
"""
Test that enhanced dependencies are sorted properly for ranges.
"""
token_line = '10 gave give VERB _ Number=Sing|Gender=Fem ' \
'0 root 10.2:nsubj|2:nmod|10.1:nsubj SpaceAfter=No'
token = Token(token_line)
conll = token.conll()
formatted_line = '10 gave give VERB _ Gender=Fem|Number=Sing ' \
'0 root 2:nmod|10.1:nsubj|10.2:nsubj SpaceAfter=No'
assert conll == formatted_line
def test_deps_sort_order_double_digits():
"""
Test that enhanced dependencies are sorted via numeric index and not string.
"""
token_line = '10 gave give VERB _ Number=Sing|Gender=Fem ' \
'0 root 10:nsubj|2:nmod SpaceAfter=No'
token = Token(token_line)
conll = token.conll()
formatted_line = '10 gave give VERB _ Gender=Fem|Number=Sing ' \
'0 root 2:nmod|10:nsubj SpaceAfter=No'
assert conll == formatted_line
def test_deps_sort_order():
"""
Test that the enhanced dependencies order is properly sorted.
"""
token_line = '10 gave give VERB _ Number=Sing|Gender=Fem ' \
'0 root 4:nsubj|2:nmod SpaceAfter=No'
token = Token(token_line)
conll = token.conll()
formatted_line = '10 gave give VERB _ Gender=Fem|Number=Sing ' \
'0 root 2:nmod|4:nsubj SpaceAfter=No'
assert conll == formatted_line
def test_remove_feature_to_string():
"""
Test a token's string representation after removing a feature completely.
"""
token_line = '33 cintre cintre NOUN _ Gender=Masc|Number=Sing ' \
'30 nmod _ SpaceAfter=No'
token = Token(token_line)
del token.feats['Gender']
new_token_line = '33 cintre cintre NOUN _ ' \
'Number=Sing 30 nmod _ SpaceAfter=No'
assert token.conll() == new_token_line
def test_modify_dict_field_to_string():
"""
Test a token's string representation after adding a feature.
"""
token_line = '33 cintre cintre NOUN _ Gender=Masc|Number=Sing ' \
'30 nmod _ SpaceAfter=No'
token = Token(token_line)
token.feats['Gender'].add('Fem')
new_token_line = '33 cintre cintre NOUN _ ' \
'Gender=Fem,Masc|Number=Sing 30 nmod _ SpaceAfter=No'
assert token.conll() == new_token_line
def test_modify_unit_field_to_string():
"""
Test a token's string representation after changing one of it's fields.
"""
token_line = '33 cintre cintre NOUN _ Gender=Masc|Number=Sing ' \
'30 nmod _ SpaceAfter=No'
token = Token(token_line)
token.lemma = 'pain'
new_token_line = '33 cintre pain NOUN _ ' \
'Gender=Masc|Number=Sing 30 nmod _ SpaceAfter=No'
assert token.conll() == new_token_line
def test_del_values():
"""
Test that values and features can be deleted from different token columns.
"""
token_line = '33 cintre cintre NOUN _ Gender=Fem|Number=Sing ' \
'30 nmod 2:nsubj|4:root SpaceAfter=No'
token = Token(token_line)
del token.feats['Gender']
del token.misc['SpaceAfter']
expected = '33 cintre cintre NOUN _ Number=Sing ' \
'30 nmod 2:nsubj|4:root _'
assert expected == token.conll()
def test_misc_parsing_output():
"""
Test that the misc field is properly output in CoNLL-U format.
"""
token_line = '33 cintre cintre NOUN _ Gender=Fem|Number=Sing ' \
'30 nmod 2:nsubj|4:root SpaceAfter=No'
token = Token(token_line)
token.misc['Independent'] = None
token.misc['SpaceAfter'].add('Yes')
token.misc['OtherTest'] = set()
token.misc['OtherTest'].add('X')
token.misc['OtherTest'].add('Z')
token.misc['OtherTest'].add('Y')
expected_output = '33 cintre cintre NOUN _ Gender=Fem|Number=Sing ' \
'30 nmod 2:nsubj|4:root Independent|OtherTest=X,Y,Z|SpaceAfter=No,Yes'
assert expected_output == token.conll()
