Should rm be applied to the output got too?

**Describe the feature you'd like**
The ``rm`` feature removes a string from the expected output *before* this is turned into a regex. Should be applied to the output *got* too?

If an example prints a weird symbol we can remove it. However, if it prints bytes that are invalid in the given encoding, the ``rm`` would be too late and this is one of the principal reasons to do this feature.

This should be a backward-compatible change.

**Describe alternatives you've considered (optional)**
May be we need another option like ``sanitation`` to fix any encoding problem?

**Additional context (optional)**
We need more brainstorming here.