Add all-validator support

[leventeliu]

PGV changed its interface (again 😹) and a new interface ValidateAll() error was added to implement a partial function of the legacy interface Validate(bool) error.

I think it's a good idea to keep this middleware up-to-date and allow users to choose this new method ValidateAll over Validate. Also, backward compatibility is considered in these changes. Please review.

[google-cla]

Thanks for your pull request. It looks like this may be your first contribution to a Google open source project (if not, look below for help). Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

📝 Please visit https://cla.developers.google.com/ to sign.

Once you've signed (or fixed any issues), please reply here with @googlebot I signed it! and we'll verify it.

---

What to do if you already signed the CLA

Individual signers

• It's possible we don't have your GitHub username or you're using a different email address on your commit. Check your existing CLA data and verify that your email is set on your git commits.

Corporate signers

• Your company has a Point of Contact who decides which employees are authorized to participate. Ask your POC to be added to the group of authorized contributors. If you don't know who your Point of Contact is, direct the Google project maintainer to go/cla#troubleshoot (Public version).
• The email used to register you as an authorized contributor must be the email used for the Git commit. Check your existing CLA data and verify that your email is set on your git commits.
• The email used to register you as an authorized contributor must also be attached to your GitHub account.

ℹ️ Googlers: Go here for more info.

[leventeliu]

@googlebot I signed it!

[johanbrandhorst]

Why is this signature change needed? This is a potentially breaking change

[leventeliu]

Hi, I don't see how this can break old code using zero arguments, and this is a common way to add (one) argument for a function without breaking things. Any better idea?

[johanbrandhorst]

Yeah this is a common assumption, but this is actually a breaking change if someone has assigned this function to a variable at some point:

var interceptor func() grpc.UnaryClientInterceptor = validator.UnaryClientInterceptor

I think it's going to be very rarely cause an action breakage but it is technically breaking. Why do we need to add this parameter?

[leventeliu]

Ok, as I said in the PR description, I changed this signature to allow users to choose the new method ValidateAll over Validate.

As I know this new method uses a multi-error to return all validation failures. IMHO this would be much more friendly while designing user-oriented and non-performance sensitive APIs -- to tell the user which fields you provided are invalid, at one time.

And as you can see in the previous implementation, while using the old interface Validate(all bool) error, this middleware directly calls Validate(false). In this way, it somehow hides part of PGV's functionality, which doesn't seem to be a good practice to me too.

So, here comes the point we should weigh the pros and cons of changing this signature. As you mentioned such kind of breaking is a real rare case, should we change it here? Or maybe we should create a new type of interceptor if I would like to use ValidateAll?

[johanbrandhorst]

I see now that this is actually targeting the master branch - I think we can just move this PR to the v2 branch and then we can change the signature without worrying about breaking users as it is still in development. What do you think?

[leventeliu]

Surely, I cherry-picked it into a new branch and opened a new PR #443. Now I will close this, thanks.