class TestSentenceDelayTransform(unittest.TestCase):
maxDiff = None
def setUp(self):
self.analyzer = FinHeuristicSentenceAnalyzer()
def test_delay(self):
transform = SentenceDelayTransform(buffer_size=2)
sentences = list(self.analyzer.analyze_text('Eka lause tää on. Tämä on toka. Kolmatta viedään. Neljäskin löytyi.'))
delayed = list(transform.transform_stream(sentences))
self.assertEqual(delayed[0:2], [Sentence(), Sentence()])
self.assertEqual(delayed[2:], sentences)
def test_delay_loop(self):
transform = SentenceDelayTransform(buffer_size=2, looping=True)
sentences = list(self.analyzer.analyze_text('Eka lause tää on. Tämä on toka. Kolmatta viedään. Neljäskin löytyi.'))
delayed = list(take(10, transform.transform_stream(sentences)))
self.assertEqual(delayed[:2], [Sentence(), Sentence()])
self.assertEqual(delayed[2:6], sentences)
self.assertEqual(delayed[6:8], sentences[-2:])
self.assertEqual(delayed[8:10], sentences[-2:])
def test_delay_loop_change_state(self):
transform = SentenceDelayTransform(buffer_size=2, looping=True)
sentences = list(self.analyzer.analyze_text('Eka lause tää on. Tämä on toka. Kolmatta viedään. Neljäskin löytyi.'))
sentences2 = list(self.analyzer.analyze_text('Viides lause menossa. Kuudes tulossa. Seitsemäs päättää.'))
delayed = list(take(3, transform.transform_stream(sentences)))
delayed2 = list(take(7, transform.transform_stream(sentences2)))
delayed = delayed + delayed2
self.assertEqual(delayed[:2], [Sentence(), Sentence()])
self.assertEqual(delayed[2:5], sentences[:3])
self.assertEqual(delayed[5:8], sentences2)
self.assertEqual(delayed[8:], sentences2[-2:])
class TestFinNounGenerator(unittest.TestCase):
def setUp(self):
# analyzer returns lemmas only with unambiguous analyses
self.analyzer = FinHeuristicSentenceAnalyzer()
self.generator = SentenceTokenGenerator()
def test_generate(self):
text = self.generator._generate('vuosi+N+Pl+Par')
self.assertEqual(text, 'vuosia')
def test_token_with_new_lemma_known(self):
token = self.analyzer.analyze_word('keppien')
new_word = self.generator.token_with_new_lemma(token, 'gouda').text
self.assertEqual(new_word, 'goudien')
def test_token_with_new_lemma(self):
token = self.analyzer.analyze_word('keppien')
new_word = self.generator.token_with_new_lemma(token, 'Truuda').text
self.assertEqual(new_word, 'Truudien')
def test_token_with_new_lemma_harder(self):
token = self.analyzer.analyze_word('kasveilla')
new_word = self.generator.token_with_new_lemma(token, 'Truuda').text
self.assertEqual(new_word, 'Truudilla')
def test_token_with_new_lemma_no_replacement(self):
token = self.analyzer.analyze_word('lakkien')
new_word = self.generator.token_with_new_lemma(token, 'Merhab').text
self.assertEqual(new_word, 'Merhabien')
def test_token_with_new_lemma_no_replacement_harder(self):
token = self.analyzer.analyze_word('hipeillä')
new_word = self.generator.token_with_new_lemma(token, 'Mozokoz').text
self.assertEqual(new_word, 'Mozokozeilla')
class TestFinAnalysisTense(unittest.TestCase):
def setUp(self):
self.analyzer = FinHeuristicSentenceAnalyzer()
def test_tense_analysis(self):
text = 'Tietäisin. Mistä tiesit? Miten olisitte voineet haluta tietää? Miten hän oli tietänyt?'
sentences = list(self.analyzer.analyze_text(text))
self.assertEqual(sentences[0].tokens[0].morphology,
'tietää+V+Act+Cond+Sg1')
self.assertEqual(sentences[1].tokens[1].morphology,
'tietää+V+Act+Ind+Prt+Sg2')
self.assertEqual(sentences[2].tokens[1].morphology,
'olla+V+Act+Cond+Pl2')
self.assertEqual(sentences[2].tokens[2].morphology,
'voida+V+Act+PrfPrc+Pl+Nom')
self.assertEqual(sentences[2].tokens[3].morphology,
'haluta+V+Act+InfA+Sg+Lat')
self.assertEqual(sentences[2].tokens[4].morphology,
'tietää+V+Act+InfA+Sg+Lat')
self.assertEqual(sentences[3].tokens[2].morphology,
'olla+V+Act+Ind+Prt+Sg3')
self.assertEqual(sentences[3].tokens[3].morphology,
'tietää+V+Act+PrfPrc+Sg+Nom')
class TestFinVerbGenerator(unittest.TestCase):
def setUp(self):
# analyzer returns lemmas only with unambiguous analyses
self.analyzer = FinHeuristicSentenceAnalyzer()
self.generator = SentenceTokenGenerator()
def test_generate(self):
text = self.generator._generate('juosta+V+Act+Ind+Prt+Pl3')
self.assertEqual(text, 'juoksivat')
def test_token_with_new_lemma_known(self):
token = self.analyzer.analyze_word('kepitimme')
new_word = self.generator.token_with_new_lemma(token, 'kimmeltää').text
self.assertEqual(new_word, 'kimmelsimme')
def test_token_with_new_lemma(self):
token = self.analyzer.analyze_word('kipitimme')
new_word = self.generator.token_with_new_lemma(token, 'huheltaa').text
self.assertEqual(new_word, 'huhelsimme')
def setUp(self):
self.analyzer = FinHeuristicSentenceAnalyzer()
def setUp(self):
# analyzer returns lemmas only with unambiguous analyses
self.analyzer = FinHeuristicSentenceAnalyzer()
self.generator = SentenceTokenGenerator()
def get_sentence(text: str) -> Sentence:
analyzer = FinHeuristicSentenceAnalyzer()
return list(analyzer.analyze_text(text))[0]
def get_example_sentence(self) -> Sentence:
analyzer = FinHeuristicSentenceAnalyzer()
return list(analyzer.analyze_text('On ilo testata.'))[0]