def test_ze_custom_nlp(self):
my_nlp = spacy.load("en_core_web_sm")
self.sentence_extractor = SentenceExtractor(custom_nlp=my_nlp)
self.test_simple_split()
self.test_starts_lower()
self.test_no_space()
self.test_worst_case_scenario()
def __init__(self, etk: ETK):
ETKModule.__init__(self, etk)
self.sentence_extractor = SentenceExtractor(
name="My sentence splitter")
def setUp(self):
self.sentence_extractor = SentenceExtractor()
class TestSentenceExtractor(unittest.TestCase):
def setUp(self):
self.sentence_extractor = SentenceExtractor()
# Test1: simple split
def test_simple_split(self):
input_text = "Spacy's parser splits sentences better if the inputs " \
"are formatted with proper grammar. However, this is " \
"not always the case for text scraped from the internet."
output_expected = [
"Spacy's parser splits sentences better if the "
"inputs are formatted with proper grammar.",
"However, this is not always the case for text "
"scraped from the internet."
]
output_extractions = self.sentence_extractor.extract(input_text)
for i, extraction in enumerate(output_extractions):
self.assertEqual(extraction.value, output_expected[i])
# Test2: ...end of sentence. start without caps
def test_starts_lower(self):
input_text = "Spacy's parser splits sentences better if the inputs " \
"are formatted with proper grammar. however, this is " \
"not always the case for text scraped from the internet."
output_expected = [
"Spacy's parser splits sentences better if the "
"inputs are formatted with proper grammar.",
"however, this is not always the case for text "
"scraped from the internet."
]
output_extractions = self.sentence_extractor.extract(input_text)
for i, extraction in enumerate(output_extractions):
self.assertEqual(extraction.value, output_expected[i])
# Test3: ...end of sentence.Start without space
def test_no_space(self):
input_text = "Spacy's parser splits sentences better if the inputs " \
"are formatted with proper grammar.However, this is " \
"not always the case for text scraped from the internet."
output_expected = [
"Spacy's parser splits sentences better if the "
"inputs are formatted with proper grammar.",
"However, this is not always the case for text "
"scraped from the internet."
]
output_extractions = self.sentence_extractor.extract(input_text)
for i, extraction in enumerate(output_extractions):
self.assertEqual(extraction.value, output_expected[i])
# Test4: ...end of sentence.worst case scenario
def test_worst_case_scenario(self):
input_text = "Spacy's parser splits sentences better if the inputs " \
"are formatted with proper grammar.however, this is " \
"not always the case for text scraped from the internet."
output_expected = [
"Spacy's parser splits sentences better if the "
"inputs are formatted with proper grammar."
"however, this is not always the case for text "
"scraped from the internet."
]
output_desired = [
"Spacy's parser splits sentences better if the "
"inputs are formatted with proper grammar.",
"however, this is not always the case for text "
"scraped from the internet."
]
output_extractions = self.sentence_extractor.extract(input_text)
for i, extraction in enumerate(output_extractions):
self.assertEqual(extraction.value, output_expected[i])
with self.assertRaises(AssertionError):
for i, extraction in enumerate(output_extractions):
self.assertEqual(extraction.value, output_desired[i])
# Test5: load custom_nlp -> run 1-4
def test_ze_custom_nlp(self):
my_nlp = spacy.load("en_core_web_sm")
self.sentence_extractor = SentenceExtractor(custom_nlp=my_nlp)
self.test_simple_split()
self.test_starts_lower()
self.test_no_space()
self.test_worst_case_scenario()