def setUp(self):
nl_filename = os.path.join(os.path.dirname(__file__), 'data/europarl/nl/ep-00-12-15.xml')
self.nl_extractor = EuroparlExtractor('nl', ['en'])
self.nl_tree = etree.parse(nl_filename)
self.nl_alignmenttrees = self.nl_extractor.parse_alignment_trees(nl_filename)
self.nl_translationtrees = self.nl_extractor.parse_translation_trees(nl_filename)
en_filename = os.path.join(os.path.dirname(__file__), 'data/europarl/en/ep-00-12-15.xml')
self.en_extractor = EuroparlExtractor('en', ['nl'])
self.en_tree = etree.parse(en_filename)
self.en_alignmenttrees = self.en_extractor.parse_alignment_trees(en_filename)
self.en_translationtrees = self.en_extractor.parse_translation_trees(en_filename)
def test_is_lexically_bound(self):
extractor = EuroparlExtractor('en', ['de', 'es', 'fr', 'nl'])
lemma_attr = extractor.config.get('all', 'lemma_attr')
mock_aux_verb = {lemma_attr: 'zijn'}
mock_perfect = {lemma_attr: 'botsen'}
self.assertTrue(extractor.is_lexically_bound('nl', mock_aux_verb, mock_perfect))
mock_aux_verb = {lemma_attr: 'zijn'}
mock_perfect = {lemma_attr: 'hebben'}
self.assertFalse(extractor.is_lexically_bound('nl', mock_aux_verb, mock_perfect))
mock_aux_verb = {lemma_attr: u'être'}
mock_perfect = {lemma_attr: 'regarder'}
self.assertFalse(extractor.is_lexically_bound('fr', mock_aux_verb, mock_perfect))
mock_aux_verb = {lemma_attr: u'être'}
mock_perfect = {lemma_attr: 'revenir'}
self.assertTrue(extractor.is_lexically_bound('fr', mock_aux_verb, mock_perfect))
mock_aux_verb = {lemma_attr: 'have'}
mock_perfect = {lemma_attr: 'collided'}
self.assertTrue(extractor.is_lexically_bound('en', mock_aux_verb, mock_perfect))
class TestEuroparlExtractor(unittest.TestCase):
def setUp(self):
nl_filename = os.path.join(os.path.dirname(__file__), 'data/europarl/nl/ep-00-12-15.xml')
self.nl_extractor = EuroparlExtractor('nl', ['en'])
self.nl_tree = etree.parse(nl_filename)
self.nl_alignmenttrees = self.nl_extractor.parse_alignment_trees(nl_filename)
self.nl_translationtrees = self.nl_extractor.parse_translation_trees(nl_filename)
en_filename = os.path.join(os.path.dirname(__file__), 'data/europarl/en/ep-00-12-15.xml')
self.en_extractor = EuroparlExtractor('en', ['nl'])
self.en_tree = etree.parse(en_filename)
self.en_alignmenttrees = self.en_extractor.parse_alignment_trees(en_filename)
self.en_translationtrees = self.en_extractor.parse_translation_trees(en_filename)
def test_init(self):
self.assertEqual(self.nl_extractor.config.get('nl', 'perfect_tag'), 'verbpapa')
self.assertIn('dunken', self.nl_extractor.aux_be_list['nl'])
def test_get_translated_lines(self):
from_lines, to_lines, align = self.nl_extractor.get_translated_lines(self.nl_alignmenttrees, 'nl', 'en', '17')
self.assertEqual(from_lines, ['17'])
self.assertEqual(to_lines, ['11'])
self.assertEqual(align, '1 => 1')
from_lines, to_lines, align = self.nl_extractor.get_translated_lines(self.nl_alignmenttrees, 'nl', 'en', '18')
self.assertEqual(from_lines, ['18', '19'])
self.assertEqual(to_lines, ['12'])
self.assertEqual(align, '2 => 1')
from_lines, to_lines, align = self.nl_extractor.get_translated_lines(self.nl_alignmenttrees, 'nl', 'en', '57')
self.assertEqual(from_lines, ['57'])
self.assertEqual(to_lines, ['46', '47'])
self.assertEqual(align, '1 => 2')
from_lines, to_lines, align = self.nl_extractor.get_translated_lines(self.nl_alignmenttrees, 'nl', 'en', '9')
self.assertEqual(from_lines, ['9'])
self.assertEqual(to_lines, [])
self.assertEqual(align, '')
from_lines, to_lines, align = self.en_extractor.get_translated_lines(self.en_alignmenttrees, 'en', 'nl', '19')
self.assertEqual(from_lines, ['19'])
self.assertEqual(to_lines, ['27'])
self.assertEqual(align, '1 => 1')
from_lines, to_lines, align = self.en_extractor.get_translated_lines(self.en_alignmenttrees, 'en', 'nl', '8')
self.assertEqual(from_lines, ['8'])
self.assertEqual(to_lines, ['13', '14'])
self.assertEqual(align, '1 => 2')
from_lines, to_lines, align = self.en_extractor.get_translated_lines(self.en_alignmenttrees, 'en', 'nl', '234')
self.assertEqual(from_lines, ['234', '235'])
self.assertEqual(to_lines, ['290'])
self.assertEqual(align, '2 => 1')
def test_get_line_by_number(self):
xml_sentence, _, pp = self.nl_extractor.get_line_by_number(self.nl_tree, 'nl', '4')
self.assertEqual(etree.fromstring(xml_sentence).get('id'), '4')
self.assertEqual(pp.get_sentence_id(), '4')
self.assertEqual(pp.verbs(), ['is', 'aangebroken'])
self.assertEqual(pp.verb_ids(), 'w4.9 w4.19')
self.assertEqual(pp.words_between(), 9)
xml_sentence, _, pp = self.nl_extractor.get_line_by_number(self.nl_tree, 'nl', '15')
self.assertEqual(etree.fromstring(xml_sentence).get('id'), '15')
self.assertEqual(pp.verbs(), ['heeft', 'bemoeid'])
self.assertEqual(pp.words_between(), 0)
xml_sentence, _, pp = self.nl_extractor.get_line_by_number(self.nl_translationtrees['en'], 'en', '6')
self.assertEqual(etree.fromstring(xml_sentence).get('id'), '6')
self.assertEqual(pp.verbs(), ['has', 'said'])
self.assertEqual(pp.words_between(), 1)
xml_sentence, _, pp = self.en_extractor.get_line_by_number(self.en_tree, 'en', '89')
self.assertEqual(etree.fromstring(xml_sentence).get('id'), '89')
self.assertEqual(pp.verbs(), ['has', 'been', 'mentioned'])
self.assertEqual(pp.words_between(), 1)
def test_list_filenames(self):
files = self.nl_extractor.list_filenames(os.path.join(os.path.dirname(__file__), 'data/europarl/nl'))
self.assertEqual([os.path.basename(f) for f in files], ['ep-00-12-15.xml'])
