Add consideration for pointer false directive when optional: pointer configuration option is used

[spencermurray]

When support for the optional configuration parameter was added in #198 it was done in a way that, if optional: pointer was set, it would not consider the presence of # @genqlient(pointer: false) graphql file declarations. As a consequence, a schema of

# schema.graphql
scalar DateTime

type Query {
  user(id: ID!): User
}

type User {
  avatarUrl: String
  brandID: ID
  brandVisible: Boolean
  createdAt: DateTime
  firstName: String
  id: ID!
  lastName: String
  name: String
  primaryEmail: String
}

and a query of

# operations.graphql
query userGetByID($id: ID!) {
  user(id: $id) {
    id
   # @genqlient(pointer: false)
    name
  }
}

would result in a generated user struct where name was of type *string

type UserGetByIDResponse struct {
	User * UserGetByIDUser `json:"user"`
}

type UserGetByIDUser struct {
	Id   string  `json:"id"`
	Name *string `json:"name"` // pointer despite query directive
}

With this change the generated structs instead are

type UserGetByIDResponse struct {
	User * UserGetByIDUser `json:"user"`
}

type UserGetByIDUser struct {
	Id   string  `json:"id"`
	Name string `json:"name"` // # @genqlient(pointer: false) directive now respected
}

I do not know if there's appetite for merging this in. To me this updates the behavior to work as I expected and gives more control to users, however, I know that may not be universally agreed upon. I had this fix already made for my own use so I decided to directly open a PR instead of creating an issue first. I updated the existing OptionalPointer test with two new queries that enforce that the configuration option optional: pointer takes effect in absence of # @genqlient(pointer: false) directive and that, if present, the directive takes precedence.

Happy to discuss or make any changes to the PR as requested.

---

I have:

• Written a clear PR title and description (above)
• Signed the Khan Academy CLA
• Added tests covering my changes, if applicable
• Included a link to the issue fixed, if applicable
• Included documentation, for new features
• Added an entry to the changelog

[benjaminjkraft]

Ah, good find, thanks for fixing! I agree this is the intended behavior.

(Edit: forgot I need to wait on the person who checks you did indeed sign the CLA. Normally she's quick but may take a day or two given the holidays. Will merge once that's checked.)

[spencermurray]

Perfect -- thank you for the timely review!

[benjaminjkraft]

Sorry for the delay in merging -- this week got quite busy!