`stubgen` produces `UnicodeEncodeError` and doesn't allow setting an encoding

[tadeu]

Bug Report

Can't use stubgen in Windows if the source contains special characters not belonging to the current "code page".

To Reproduce

1. In Windows, create a file example.py with UTF-8 encoding and the following contents:

def get_π():
    return 3

2. Execute stubgen -m example -o .

Expected Behavior

Should execute normally:

Processed 1 modules
Generated .\example.pyi

Actual Behavior

Error:

Traceback (most recent call last):
  File "<env>\Scripts\stubgen-script.py", line 10, in <module>
    sys.exit(main())
  File "mypy\stubgen.py", line 1655, in main
  File "mypy\stubgen.py", line 1533, in generate_stubs
  File "mypy\stubgen.py", line 1489, in generate_stub_from_ast
  File "<env>\lib\encodings\cp1252.py", line 19, in encode
    return codecs.charmap_encode(input,self.errors,encoding_table)[0]
UnicodeEncodeError: 'charmap' codec can't encode character '\u03c0' in position 8: character maps to <undefined>

Your Environment

• Mypy version used: 0.910

Notes

The problem seems to be this open here, without an encoding argument. Perhaps the encoding could be passed via command line argument, or via an environment variable?

[sterliakov]

Doesn't reproduce on current master, stubgen handles that case now and generates the correct stub file (as of 1.15.0)