fix(python): avoid removal of null characters in bindings

[domire8]

Description

The receive_bytes in python is indeed causing some troubles. The removal of the null characters at the end of the received buffer is not a good idea, because this will "destroy" clproto messages that have zeros (for example a zero joint state). Returning the buffer just like that is also not a good idea, because then clproto messages would be decoded, which will throw exceptions.

Unfortunately, I don't see a better way of handling this right now. I don't think we can make a general assumption about the messages that we receive, so its ultimately the users responsibility to correctly decode and rstrip the received message.

Review guidelines

Estimated Time of Review: 5 minutes

Checklist before merging:

• Confirm that the relevant changelog(s) are up-to-date in case of any user-facing changes

[domire8]

Also, is it okay to change the version in this PR such that I can tag it directly afterwards?

[eeberhard]

Classic message termination problems. I guess we could have anticipated it but oh well.

The tests are good but as a nitpick I would suggest a small usage example / python snippet in the README to illustrate how to handle bytes and strings when sending and receiving. Feel free to make an issue and do it another time (or a backlog item for @JasperTan97)

Also, is it okay to change the version in this PR such that I can tag it directly afterwards?

Yeah for sure, the release strategy now allows that. Separate release PRs are only needed if we decide to tag a release or write some change notes after the fact.