def test_lemma_graminfo(self):
info = morph_ru.get_graminfo(u'СУСЛИКАМИ')
self.assertEqual(len(info), 1)
info = info[0]
gram_form = GramForm(info['info'])
self.assertEqual(gram_form.form, set([u'мр', u'тв', u'мн']))
self.assertEqual(info['norm'], u'СУСЛИК')
def assertHasInfo(self,
word,
norm=None,
cls=None,
method=None,
scan=False,
form=None,
standard=False,
has_info=True):
def is_correct(frm):
correct = True
if norm:
correct = frm['norm'] == norm
if method:
correct = correct and (method in frm['method'])
if cls:
correct = correct and (frm['class'] == cls)
if form:
gram_filter = GramForm(form)
gram_form = GramForm(frm['info'])
correct = correct and gram_form.match(gram_filter)
return correct
if scan:
forms = get_graminfo_scan(morph_ru, word, standard=standard)
else:
forms = morph_ru.get_graminfo(word, standard=standard)
self.assertEqual(any([is_correct(frm) for frm in forms]), has_info)
def test_lemma_graminfo(self):
info = morph_ru.get_graminfo(u'СУСЛИКАМИ')
self.assertEqual(len(info), 1)
info = info[0]
gram_form = GramForm(info['info'])
self.assertEqual(gram_form.form, set([u'мр', u'тв', u'мн']))
self.assertEqual(info['norm'], u'СУСЛИК')
def main():
# print _gram_info_match('[imper,1p,pl]', '1p,pl,imper')
# exit()
all = 0
failed1 = 0
failed2 = 0
failed_words = {}
for test_data in BASIC_TESTS:
collection = parse_test_data(test_data)
for word in collection:
expected_results = collection[word]
actual_results = morph_ru.get_graminfo(word, standard=True)
actual_results_aot = morph_ru.get_graminfo(word)
# проверяем, что все стандартные варианты разбора учтены
for expected_form in expected_results:
match = any([forms_match(expected_form, actual_form) for actual_form in actual_results])
if not match:
failed1 += 1
print (' ----', word, expected_form['class'], expected_form['norm'], expected_form['info'])
for actual_form in actual_results:
print (" \____ %s %s %s %s" % (word, actual_form['class'], actual_form['norm'], actual_form['info']))
all += 1
# проверяем, нет ли у нас лишних (неправильных) вариантов разбора
for actual_form, actual_form_aot in zip(actual_results, actual_results_aot):
# для каждого варианта разбора смотрим, соответствует ли ему
# хотя бы 1 разбор в стандарте
match = any([forms_match(expected_form, actual_form) for expected_form in expected_results])
if not match:
print (" ++++ %s %s %s %s" % (word, actual_form['class'], actual_form['norm'], actual_form['info']))
failed2 += 1
# else:
# for expected_form in expected_results:
# if forms_match(expected_form, actual_form):
# print(_format_row(word, expected_form, actual_form_aot))
print (' total: %d, failed: %d, extra: %d' % (all, failed1, failed2))
def assertHasInfo(self, word, norm=None, cls=None, method=None, scan=False, form=None, standard=False, has_info=True):
def is_correct(frm):
correct = True
if norm:
correct = frm['norm'] == norm
if method:
correct = correct and (method in frm['method'])
if cls:
correct = correct and (frm['class'] == cls)
if form:
gram_filter = GramForm(form)
gram_form = GramForm(frm['info'])
correct = correct and gram_form.match(gram_filter)
return correct
if scan:
forms = get_graminfo_scan(morph_ru, word, standard=standard)
else:
forms = morph_ru.get_graminfo(word, standard=standard)
self.assertEqual(any([is_correct(frm) for frm in forms]), has_info)
if __name__ == '__main__':
# print _gram_info_match(u'[imper,1p,pl]', u'1p,pl,imper')
# exit()
all = 0
failed1 = 0
failed2 = 0
failed_words = {}
for test_data in BASIC_TESTS:
collection = parse_test_data(test_data)
for word in collection:
expected_results = collection[word]
actual_results = morph_ru.get_graminfo(word, True)
# проверяем, что все стандартные варианты разбора учтены
for expected_form in expected_results:
match = any([forms_match(expected_form, actual_form) for actual_form in actual_results])
if not match:
failed1 += 1
print ' ----', word, expected_form['class'], expected_form['norm'], expected_form['info']
for actual_form in actual_results:
print u" \____ %s %s %s %s" % (word, actual_form['class'], actual_form['norm'], actual_form['info'])
all += 1
# проверяем, нет ли у нас лишних (неправильных) вариантов разбора
for actual_form in actual_results:
# для каждого варианта разбора смотрим, соответствует ли ему
# хотя бы 1 разбор в стандарте