GeezerVison

This commit changed decoding processing of getting content to use
ContentFile.decoded_content method instead of self-implemented one.
Because the PyGithub, which this pack depends on, have already
implemented decoding processing for content of ContentFile.

…amter

In the previous implementation uses custom decoding method of
`lib.formatters.base64_decode` when `github.create_file` and
`github.update_file` actions are called with "encoding" parameter. And
those will be failed because of its implementation.

This removes `lib.formatters.base64_decode` method because it is not
used any more except for above actions, which is wrong the way of using.

It's necessary for decode processing for the value that is get from
`github.ContentFile.ContentFile.decoded_content` because it returns
bytes typed value. This commit added its processing at
`lib.formatters.contents_to_dict` method.
(c.f. https://pygithub.readthedocs.io/en/latest/github_objects/ContentFile.html)

…amter