[raunak2007]

Fixes #424.

genqlient generated a selected field's Go type purely from its schema nullability, ignoring any @skip/@include on the selection. Those directives can make a non-null field absent from the response, so the generated type couldn't represent that absence: a skipped field decoded to the zero value with no error.

This treats a field with @skip/@include as nullable (as if id: ID rather than id: ID!) and lets the existing optional-handling decide the representation, per the discussion on #424:

• optional: value (default): unchanged, still the zero value.
• optional: pointer / optional: generic: the field becomes a pointer / Option and can represent an absent value.

Technically a breaking change for non-null fields with @skip/@include under optional: pointer or optional: generic. That seems right, and such users are probably rare and may already set @genqlient(pointer: true) by hand.

Tests: SkipInclude.graphql covers the default config (output unchanged), and it's added to the OptionalPointer case, where id now generates as *string instead of string.

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

[csilvers]

A good catch! -- the rationale makes sense to me. And the fix is nice and self-contained. But I'll let @benjaminjkraft give the final OK, especially in regards to how we want to handle the fact this is technically a breaking change.