Fix compilation of fields named 'bytes'

[SpencerSharkey]

If you have a bytes field named "bytes", a recursion error occurs when importing the compiled stubs.

[Gobot1234]

Why is bytes a special case, should it not happen for all builtin types?

[nat-n]

@SpencerSharkey thanks for doing this.

The solution looks reasonable, however as @Gobot1234 it may be that you've scratched the surface of a larger issue here.

That said it's not immediately obvious to me exactly why this error is happening. I don't have a lot of time to look into this now, but if you could post an (abridged) stack trace of the recursion error then that would help :)

Also it would be good to have a regression test for this, would you be willing to add an example to the test suite following the pattern described here https://github.com/danielgtaylor/python-betterproto/tree/master/tests ?

[Gobot1234]

I can't reproduce this

syntax = "proto3";

message Test {
    bytes bytes = 1;  // or string bytes = 1;
}

both compile fine for me. Can you show an example where this doesn't work?

[giantryansaul]

Commented on the related issue #210, I'm seeing a recursive issue from a definition we have that uses bytes. I tested out this PR locally and it fixed my runtime issue.

[nat-n]

@SpencerSharkey Thanks again for your contribution.

I'm closing in favour of #226 because I made some tweaks to extend the fix and enable a test that has already been written to cover it, but didn't have write access to push to the fork (there's a checkbox to allow this when creating the PR).

I hope to have it in a release soon.