While playing with the ast module, I noticed the below:
import ast
>>> mu = 'µ'
>>> mu.encode()
b'\xc2\xb5'
>>> root = ast.parse(mu, "<string>", mode="eval")
>>> root.body.id.encode()
b'\xce\xbc'
It seems like ast.parse converts 'µ' which is encoded as UTF-8 Latin-1 Supplement (b'\xc2\xb5') to UTF-8 Greek and Coptic (b'\xce\xbc').
When looking at the cpython code of ast.parse(), I couldn't find any reference to such a conversion.
Can someone help find the reason for the above conversion?
