In general, 'utf8' and 'utf-8' are interchangeable in the codecs (and in many parts of the Python library). However, 'utf8-sig' is missing ... and it happens to also be generated by lib2to3.tokenize.detect_encoding.

>>> import codecs
>>> codecs.getincrementaldecoder('utf-8-sig')()
<encodings.utf_8_sig.IncrementalDecoder object at 0x7fecbcdbbc10>
>>> codecs.getincrementaldecoder('utf8-sig')()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python3.6/codecs.py", line 987, in getincrementaldecoder
    decoder = lookup(encoding).incrementaldecoder
LookupError: unknown encoding: utf8-sig