Binding to a type in the same package as generated code produces invalid output about genqlient

[costela]

This is such a weird coincidence! I just started working on a PR for this 😅

However, my PR currently deals with it from the package_bindings side, using a "magic" . package. The more general solution of matching the package path more intelligently requires a bit more refactoring - I think.

Will link it here as soon as the tests are done and we can see if it's enough.

[benjaminjkraft]

Ah, good find! I think the ideal place to update this would be in generator.ref/generator.addImportFor? Those should be able to see whatever we know about the current package and just have a special case. (But somewhere you may have to figure out -- or ask Go -- what the full package path for the current package even is.)

[costela]

Got a first attempt in #285.

Reworked my approach to make it work with bindings + package_bindings equally well.

Getting the fully qualified package is a bit fiddly and I'm not entirely sure it's "safe" (it relies on relative imports, which AFAICT have a rocky support history). But is seems to work.

PTAL