Incorrect detection of TIS-620 instead of UTF-8

[ulyashin]

cchardet incorrectly detects TIS-620 encoding for some UTF-8 encoded cyrillic strings:

>>> cchardet.__version__
'2.1.1'

>>> cchardet.detect(u"тест".encode("utf-8"))
{'confidence': 0.9900000095367432, 'encoding': u'TIS-620'}
>>> cchardet.detect(u"тост".encode("utf-8"))
{'confidence': 0.9900000095367432, 'encoding': u'TIS-620'}

>>> chardet.__version__
'3.0.4'
>>> chardet.detect(u"тест".encode("utf-8"))
{'confidence': 0.938125, 'language': '', 'encoding': 'utf-8'}
>>> chardet.detect(u"тост".encode("utf-8"))
{'confidence': 0.938125, 'language': '', 'encoding': 'utf-8'}