How does the "magic lines(s)" in python work, when specifying encoding in python file?

Question

At the start of a python file (first line) sometimes I read

# -*- coding: utf-8 -*-

and sometimes I read

# encoding: utf-8

Both lines seem to do the same thing: specifying utf8 as encoding for all the text put in the file.

I have to questions:

1. Why does this even work? I thought the interpreter ignores everything after a # because it invokes a comment.
2. What is the difference between the two lines above? Does the interpreter just ignore the -*-?

Answer 1

The two forms are equivalent. The -*- version is a special kind of comment that Emacs understands. See PEP 263 for more information.

If a comment like in either of these forms is one of the first two lines of a file, the interpreter will use the specified encoding to read the file.

Answer 2

It works because the implementation looks for it, there is nothing magical about it. There is no difference, all possible variants are defined by PEP 263 (the only difference is that the first one is Emacs-compatible).