def setUpClass(cls):
"""set up wikitext examples with expected output for each parser step.
examples -- dictionary of word: text, where text is a dictionary
text['text'] -- full wikitext
text['lang'] -- only the language section
text['pron'] -- only the pronunciation section of the language section
text['ipa'] -- only the extracted ipa
"""
# examples = {'word': {'text': ..., 'lang': ...}, 'word2': {...}}
cls.examples = wp.json_load('test/wikitext_examples.json')
def test_extract_ipa_meets_threshold(self):
pronunciation = wp.json_load('test/pron.json')
self.assertGreater(len(pronunciation), 45000)
# dictionaries from words to lists of ipa
ipa = {}
ipa_lenient = {}
for word, pron_section in pronunciation.items():
pron_info = wp.Wikitext(pron_section).extract_pronunciation()
if 'ipa' in pron_info:
ipa[word] = pron_info['ipa']
ipa_lenient_results = wp.Wikitext(pron_section).extract_ipa_lenient()
if ipa_lenient_results:
ipa_lenient[word] = ipa_lenient_results
ipa_diff = {k: pronunciation[k] for k in ipa_lenient.keys()
if k not in ipa or len(ipa_lenient[k]) > len(ipa[k])}
self.assertGreater(len(ipa_lenient), 32000)
self.assertGreater(len(ipa), 32000)
self.assertGreater(len(ipa_diff), 500)