Gracefully handle encoding errors when writing to stdout

[brianschubert]

Fixes #12692

Sets the encoding error handler for stdout to "backslashreplace". This prevents mypy from crashing if an error message has a character that can't be represented by the current I/O encoding.

No change is made to stderr since its default is already "backslashreplace".

Before

$ PYTHONIOENCODING=ascii mypy -c "x=γ"
Traceback (most recent call last):
    ...
UnicodeEncodeError: 'ascii' codec can't encode character '\u03b3' in position 50: ordinal not in range(128)

After:

$ PYTHONIOENCODING=ascii mypy -c "x=γ"
<string>:1: error: Name "\u03b3" is not defined  [name-defined]
Found 1 error in 1 file (checked 1 source file)

Externally setting the error handler to something other than "strict" still works. For example:

$ PYTHONIOENCODING=ascii:namereplace mypy -c "x=γ"
<string>:1: error: Name "\N{GREEK SMALL LETTER GAMMA}" is not defined  [name-defined]
Found 1 error in 1 file (checked 1 source file)

[github-actions]

According to mypy_primer, this change doesn't affect type check results on a corpus of open source code. ✅