Make --mutation-info-file optional

[JamesLee-Jones]

Make --mutation-info-file parameter optional as writing this file can cause issues if the mutation tree is too deep.

[JamesLee-Jones]

Encapsulating whether we want to build the mutation tree in protobufs seemed like the best choice. We can't readily use optional as mutation_info isn't initialised when it is passed to NewMutateFrontendActionFactory.

[afd]

I didn't do a proper review yet, but at a glance I don't like the idea of adding a use field to the top-level protobuf object. Presumably the idea is that if this field is false / 0, the protobuf is not in use.

I think it would be better to pass an optional around, such that if this optional is empty there is no protobuf object at all, and if it's not empty there is a complete protobuf object, so that no change to the protobuf description is required.

Would that be possible or is there a good reason why the use approach is needed?

[afd]

@JamesLee-Jones as part of this PR, can you update the project README so that the instructions no longer use this option (and also remove the comments that explain that the option is mandatory but doesn't matter for purposes of the README examples)?

[afd]

@JamesLee-Jones in my latest push I updated the README so that it does use the generated .json file in the math library example, so it's only for the simple pi example that we'd want to remove all trace of it.

[afd]

I now have bandwidth to look at this again if you want to rebase it and revise it if needed.

[JamesLee-Jones]

I now have bandwidth to look at this again if you want to rebase it and revise it if needed.

I'll revise it today along with the rest of the reviewed PRs.

[afd]

Looking good - small changes requested.

[afd]

Thanks! A few final wording things - can you attend to them and then merge? I don't need to re-review.

[afd]

@JamesLee-Jones This needs a rebase, but otherwise should be good to go if you're able to attend to the minor edits.