Support for files that specify their encoding

I ran into the same issue many had here:
https://github.com/bndr/pipreqs/issues/214

Python source code does have a standard way of specifying the encoding of a file. I suggest pipreqs uses this if available to detect the encoding of the file.

PEP-263: Defining Python Source Code Encodings
https://peps.python.org/pep-0263/

Start by reviewing pipreqs' source-file reading path and the referenced PEP-263 encoding rules, then compare the behavior described in issue #214. Done means files declaring a supported encoding can be processed without the reported decoding failure.

Written by the indexing model from the issue text.