Documented auto-fetch of remaining fields from `@service`-returned records does not happen (RC30)

[krhoybraten-sikt]

Summary

The README ("Auto-fetching @table types from database") states that when a service
field returns a @table type, Graphitron automatically fetches all requested
fields using the primary key from the service result:

The service needs to return a record with the primary key populated, all other
fields on the record are fetched from the database. […] For listed fields,
Graphitron extracts primary keys from all service results and performs a single
batch query using WHERE pk IN (...).

On 10.0.0-RC30 this does not happen — non-PK fields resolve to null.

Reproduction

type Query {
  mineSynligeMiljoer: [Miljo]
      @service(service: {className: "…ApplikasjonsoversiktService", method: "mineSynligeMiljoer"})
}

type Miljo implements Node @table(name: "miljo") @node(typeId: "20014", keyColumns: ["MILJOKODE"]) {
  id: ID! @nodeId
  kode: String @field(name: "MILJOKODE")
  navn: String @field(name: "NAVN")
}

public List<MiljoRecord> mineSynligeMiljoer() {
    return ctx.select(MILJO.MILJOKODE).from(MILJO)
            .fetch(r -> { var rec = new MiljoRecord(); rec.setMiljokode(r.value1()); return rec; });
}

Query { mineSynligeMiljoer { kode navn } } returns
[{"kode":"demo","navn":null}, …] — navn is never fetched even though the rows
exist with non-null values (verified directly in the database). The field is
plain (no @splitQuery), non-paginated, and the return type has @table — the
documented conditions for auto-fetch.

Workaround: the service selects complete records (ctx.selectFrom(MILJO))
instead of PK-only records. Fine for small tables; defeats the documented
pattern for large ones.

Version: 10.0.0-RC30

[folkef6]

https://graphitron.sikt.no/roadmap/plans/service-record-return-pk-autofetch.html