Windows console, Python and encodings - how does it work?

Question

Python

I noticed that subprocess.getoutput('ping ya.ru') returns krakozyabry instead of Russian text. I went to figure out what encoding the console is working in, the console says that it is cp866. Okay, in principle, if you use subprocess.check_output('ping ya.ru'), then I will get a response in bytes that can be represented as normal text using decode("cp866").

And now it seems that the problem is no more, but curiosity does not let go, I wanted to get a normal result from subprocess.getoutput. Googled a recipe in which the resulting string is first encoded in CP1251 and then decoded in CP866.

subprocess.getoutput('ping ya.ru').encode('cp1251').decode('cp866')

But now I don't understand why this recipe works. Why is this intermediate encode needed in CP1251? It seems to us that the cp866 console returned. I understand that before decode, you need to somehow translate the string from the response into bytes, but why doesn't encode('cp866').decode('cp866') work then?

Answer 1

As far as I remember, there is a difference if you output directly to the (interactive) console or redirect to a file or to PIPE (then ANSI). At least for Python itself. Here is a little...

Answer 2

The docs say that this is a legacy function, maybe you shouldn't use it anymore.
This is how it works normally:

subprocess.check_output('ping ya.ru', encoding='cp866')