Stugben: Handling files with file encoding comment #3726

I was trying to generate stubs for sre_compile and it fails with:

$ python stubgen.py --py2 sre_compile
.../2.7/lib/python2.7/sre_compile.py:0: error: encoding declaration in Unicode string

This is probably happening because ast27_parse_impl converts source from bytes to string and tries to parse it without caring about source code encoding comment.

PEP 263 says:

If a Unicode string with a coding declaration is passed to compile() , a SyntaxError will be raised.

One quick fix would be to remove such comment from source before it is passed to mypy.parse.parse() in stubgen.py.

I am wondering if this is worth fixing? If yes, is the above the best way to do it.