support intermediate interface

[Code-Hex]

#169 is inactive. I created a new one

required also: #178

[coveralls]

Coverage increased (+0.002%) to 92.396% when pulling daaed17 on Code-Hex:intermediate-interfaces into 2776c98 on vektah:master.

[Code-Hex]

@StevenACoffman Could you handle this PR?
I will create a new PR to support this feature for gqlgen after merged and released this one

[StevenACoffman]

Have you already figured out what is necessary to change in gqlgen to support this? We had to reverta PR here recently because it broke gqlgen and I'd like to avoid repeating that.

What is different in your change here from the changes in #169 : https://github.com/vektah/gqlparser/pull/169/files

@wilhelmeek FYI

[Code-Hex]

@StevenACoffman I understand the case (maybe this one?). I believe the broken is caused by updated introspection.
this change is not included updating introspection. so this works fine

What is different in your change here from the changes in

These are the same PR. but another one is draft. If #169 is open, I'm OK to merge it and close this one (but I want this change as soon as possible 🙏 )

[Code-Hex]

In addition, If I were to mention this issue, We would have to modify the InputFields() to ignore deprecated directives for code generation as in this code

https://github.com/99designs/gqlgen/blob/5ad012e3d7be1127706b9c8a3da0378df3a98ec1/graphql/introspection/type.go#L73-L75

- func (t *Type) InputFields() []InputValue {
+ func (t *Type) InputFields(includeDeprecated bool) []InputValue {
	if t.def == nil || t.def.Kind != ast.InputObject {
		return []InputValue{}
	}

	res := []InputValue{}
	for _, f := range t.def.Fields {
+		if !includeDeprecated && f.Directives.ForName("deprecated") != nil {
+			continue
+		}
		res = append(res, InputValue{
			Name:         f.Name,
			Description:  f.Description,
			Type:         WrapTypeFromType(t.schema, f.Type),
			DefaultValue: defaultValue(f.DefaultValue),
		})
	}
	return res
}

[StevenACoffman]

@Code-Hex I rebased and merged #169 since it preserves the original author for attribution. However, when I then rebased #178 the tests continued to fail. I wasn't able to push the rebased version back up to @vvakame fork, so I created #181 to demonstrate the continued failure.

Unfortunately, I'm not going to be able to spend more time looking into this at present, so if anyone else can sort it out, I would welcome a PR that passed the tests.

[Code-Hex]

@StevenACoffman Thank you for merged!

Sure. Because this is necessary a part of process.
If I can take to solve the problem I will send PR.

However, when I then rebased #178 the tests continued to fail

Thank you

[Code-Hex]

@StevenACoffman Could you release for merged one?

You said

We had to reverta PR here recently because it broke gqlgen and I'd like to avoid repeating that.

If so, We have to do "small change, small release". Thanks.