def test_en_bug_correct2(self):
"""测试英文纠错bug"""
spell = EnSpell()
spell.check_init()
print(spell.word_freq_dict.get('whould'))
print(spell.candidates('whould'))
a = spell.correct_word('whould')
print(a)
r = en_correct('contend proble poety adress whould niether quaties')
print(r)
assert en_correct('whould')[0] == 'would' # no error
def en_correct_t():
assert en_correct('spelling') == 'spelling' # no error
assert en_correct('speling') == 'spelling' # insert
assert en_correct('correctud') == 'corrected' # replace 1
assert en_correct('gorrectud') == 'corrected' # replace 2
assert en_correct('bycycle') == 'bicycle' # replace
assert en_correct('inconvient') == 'inconvenient' # insert 2
assert en_correct('arrainged') == 'arranged' # delete
assert en_correct('peotrry') == 'poetry' # transpose + delete
assert en_correct('word') == 'word' # know
assert en_correct('quintessential') == 'quintessential' # unknow
assert words('the test is it.') == ['the', 'test', 'is', 'it'] # segment
assert len(spell.WORDS) > 100
assert spell.WORDS['the'] > 100
assert en_probability('word') > 0
assert en_probability('quintessential') == 0
assert 0.07 < en_probability('the') < 0.08
return 'unit_test pass'
def spell_t(tests, verbose=False):
"""
run en_correct(wrong) on all (right,wrong) pairs, and report result
:param tests:
:param verbose:
:return:
"""
import time
start = time.clock()
good, unknown = 0, 0
n = len(tests)
for right, wrong in tests:
w = en_correct(wrong)
good += (w == right)
if w != right:
unknown += (right not in spell.WORDS)
if verbose:
print('en_correct({}) => {} ({}); expected {} ({})'.format(wrong, w, spell.WORDS[w], right,
spell.WORDS[right]))
dt = time.clock() - start
print('{:.0%} of {} correct ({:.0%} unknown) at {:.0f} words per second'.format(good / n, n, unknown / n, n / dt))
def spell_t(tests, verbose=False):
"""
run en_correct(wrong) on all (right,wrong) pairs, and report result
:param tests:
:param verbose:
:return:
"""
import time
start = time.clock()
good, unknown = 0, 0
n = len(tests)
for right, wrong in tests:
w = en_correct(wrong)[0]
good += (w == right)
if w != right:
unknown += (right not in spell.word_freq_dict)
if verbose:
print('en_correct({}) => {}; expected {}'.format(
wrong, w, right))
dt = time.clock() - start
print('acc: {:.0%}, total num: {}, ({:.0%} unknown), speed: {:.0f} '
'words per second'.format(good / n, n, unknown / n, n / dt))
def test_en_bug_correct1(self):
"""测试英文纠错bug"""
r = en_correct('folder payroll connectivity website')
print(r)
assert en_correct('spelling')[0] == 'spelling' # no error
def test_en_correct(self):
"""测试英文纠错"""
print(
en_correct(
'spelling speling correctud gorrectud bycycle inconvient arrainged peotrry word quintessential'
))
print(en_correct('spelling')[0])
assert en_correct('spelling')[0] == 'spelling' # no error
assert en_correct('speling')[0] == 'spelling' # insert
assert en_correct('correctud')[0] == 'corrected' # replace 1
assert en_correct('gorrectud')[0] == 'corrected' # replace 2
assert en_correct('bycycle')[0] == 'bicycle' # replace
assert en_correct('inconvient')[0] == 'inconvenient' # insert 2
assert en_correct('arrainged')[0] == 'arranged' # delete
assert en_correct('peotrry')[0] == 'poetry' # transpose + delete
assert en_correct('word')[0] == 'word' # know
assert en_correct('quintessential')[0] == 'quintessential' # unknow
assert len(spell.word_freq_dict) > 100
assert spell.word_freq_dict['the'] > 100
assert en_probability('word') > 0
assert en_probability('quintessentialkk') == 0
print(en_probability('the'))
return 'unit_test pass'
