[Breaking][CDTOOL-1187] Add support for the 'mode' parameter when enabling DDoS Protection

[rcaril]

Change summary

This PR adds support for passing the mode parameter when calling the Enable DDOS Protection endpoint. This was previously missing and only available through Update.

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

New Feature Submissions:

• Does your submission pass tests?

Changes to Core Features:

• Have you written new tests for your core changes, as applicable?
• Have you successfully run tests with your changes locally?

[philippschulte]

This looks functionally correct, and the added fixtures/tests for explicit mode are good.

My only concern is backward compatibility: changing Enable(...) to require an EnableInput argument is a breaking API change for existing consumers!

[kpfleming]

Given the very low number of users of go-fastly, I think a breaking change would be fine, and preferable to having a separate function. For the future, we should probably standardize on all 'action' functions accepting an input structure even if it is empty, so we can add fields to that structure in the future without a breaking change.

[philippschulte]

Looks good to me now. Preserving Enable(...) and adding EnableWithInput(...) avoids the breaking API change while still exposing mode. The added tests for explicit mode and verification are great. Thank you!

[rcaril]

@kpfleming would you prefer if I reverted the secondary function changes at this point and moved forward with the breaking change for the sake of less complexity downstream?

[kpfleming]

I'm a little concerned about setting a precedent that we'll be asked to continue in the future, that's it. If we can handle the breaking change in our own tools, I suspect that's a better path because it avoids any indication that we'll added secondary/tertiary/etc. functions in the future to avoid breaking changes.