def test_source_en_contains_en(self):
"""
The source_en return actually all the whole english sentence.
:return:
"""
sources_en, sources_de, targets_de, bases_de, pos_de = generate_train_data(
'Record file and application usage',
'Die Nutzung von Dateien und Anwendungen aufzeichnen')
for source_en in sources_en:
self.assertEqual(source_en, 'Record file and application usage\n')
def test_bases_de_contains_de(self):
"""
The base_de return actually the whole german sentence.
:return:
"""
sources_en, sources_de, targets_de, bases_de, pos_de = generate_train_data(
'Record file and application usage',
'Die Nutzung von Dateien und Anwendungen aufzeichnen')
for base_de in bases_de:
self.assertEqual(
base_de,
'Die Nutzung von Dateien und Anwendungen aufzeichnen\n')
def test_empty_string_parameter(self):
"""
Empty strings should return empty lists back
:return:
"""
sources_en, sources_de, targets_de, bases_de, pos_de = generate_train_data(
'', '')
self.assertEqual(len(sources_en), 0, '"sources_en" has lenght of 0.')
self.assertEqual(len(sources_de), 0, '"sources_de" has lenght of 0.')
self.assertEqual(len(targets_de), 0, '"targets_de" has lenght of 0.')
self.assertEqual(len(bases_de), 0, '"bases_de" has lenght of 0.')
self.assertEqual(len(pos_de), 0, '"pos_de" has lenght of 0.')
def test_generation_of_train_data(self):
"""
Checks if the returning values have all the length as expected.
:return:
"""
sources_en, sources_de, targets_de, bases_de, pos_de = generate_train_data(
'Record file and application usage',
'Die Nutzung von Dateien und Anwendungen aufzeichnen')
self.assertEqual(len(sources_en), 7, '"sources_en" has lenght of 7.')
self.assertEqual(len(sources_de), 7, '"sources_de" has lenght of 7.')
self.assertEqual(len(targets_de), 7, '"targets_de" has lenght of 7.')
self.assertEqual(len(bases_de), 7, '"bases_de" has lenght of 7.')
self.assertEqual(len(pos_de), 7, '"pos_de" has lenght of 7.')
def test_unequal_token_length(self):
"""
The lines of test data produces must match the number of tokens generated from the german sentence,
even if the
:return:
"""
sources_en, sources_de, targets_de, bases_de, pos_de = generate_train_data(
'test short translation',
'Dieser Schlüssel legt fest, ob die Karten per Ziehen-und-Ablegen oder per Anklicken gelegt werden sollen.'
)
self.assertEqual(len(sources_en), 15, '"sources_en" has lenght of 15.')
self.assertEqual(len(sources_de), 15, '"sources_de" has lenght of 15.')
self.assertEqual(len(targets_de), 15, '"targets_de" has lenght of 15.')
self.assertEqual(len(bases_de), 15, '"bases_de" has lenght of 15.')
self.assertEqual(len(pos_de), 15, '"pos_de" has lenght of 15.')
def test_sources_de_contains_de(self):
"""
The sources_de contains one token as the base form
of the corresponding word of the whole sentence at each position.
:return:
"""
sources_en, sources_de, targets_de, bases_de, pos_de = generate_train_data(
'Record file and application usage',
'Die Nutzung von Dateien und Anwendungen aufzeichnen')
de_tokens = [
'der\n', 'Nutzung\n', 'von\n', 'Datei\n', 'und\n', 'Anwendung\n',
'aufzeichnen\n'
]
for i in range(0, len(sources_de)):
self.assertEqual(
sources_de[i], de_tokens[i],
'source_de entry: "{0}" matches with token "{1}" at position {2}'
.format(sources_de[i], de_tokens[i], i))
def test_targets_de_contains_de(self):
"""
The targets_de contains the actual german token inflected in their original form. at each position
:return:
"""
sources_en, sources_de, targets_de, bases_de, pos_de = generate_train_data(
'Record file and application usage',
'Die Nutzung von Dateien und Anwendungen aufzeichnen')
targets_tokens = [
'die\n', 'Nutzung\n', 'von\n', 'Dateien\n', 'und\n',
'Anwendungen\n', 'aufzeichnen\n'
]
for i in range(0, len(targets_de)):
self.assertEqual(
targets_de[i], targets_tokens[i],
'targets_de entry: "{0}" matches with token "{1}" at position {2}'
.format(targets_de[i], targets_tokens[i], i))