Fix in rule for repeated int32

[mortezaPRK]

Fix generating lookup for repeated.items.int32.in

This PR fixes the in field validation for repeated int32 fields.

The current generated code doesn't emit the lookup map e.g.

// message DoSomethingRequest {
//   repeated int32 field_with_int32_type = 4 [ (validate.rules).repeated.items.int32 = { in: [25,50] } ];
// }

// Generated code doesn't have this section
var _DoSomethingRequest_FieldWithInt32Type_InLookup = map[int32]struct{}{
	25:  {},
	50:  {},
}

I also added a Test cases for both int32 and int64

[CLAassistant]

All committers have signed the CLA.

[rodaine]

Hi @mortezaPRK thanks for the patch! Looks like the lookup was also not generated in the C++ codegen either.

[mortezaPRK]

Hey @rodaine

Unfortunately I can't run the tests locally (I tried with Docker and it doesn't work), and I have no idea how C++ code looks like.

Should I remove the test cases I added? I found this PR which is doing something similar tho.

[nicksnyder]

@mortezaPRK maybe the best path forward is to split this into 2 different PRs. Keep the fixes for Go in this PR and remove the tests you added so we can get this merged as an incremental improvement. Then open a separate PR to add the tests (which will still fail, but will be a reminder that there is a gap in C++ codegen).

[mortezaPRK]

@nicksnyder Agree. I updated the PR.

I'll create another PR for the test cases

[rodaine]

Awesome! Thanks again