flux-pr-1128
graphql-go-tools (Go) · W2 · GPT-5.3 Codex
Tests passed. 1/1 commands passed. Strength: strong.
go test -C v2 ./... -count=1 -timeout=300sPartial score: 1/1
Trajectory
codex · partial order onlyprovider-native trajectory captured; validation and decision steps are appended with coarse ordering only
Quality
Equivalence Reasoning
behavioral
The patch adds static `__typename` resolution, but it implements fetch-skipping at the whole `objectFetchConfiguration` level (`skipFetch`). That can suppress an entire planner fetch when root `__typename` is grouped with real root fields, causing missing backend data. It also does not add the gold patch’s resolvability/filtering adjustments for root-operation `__typename` (only union got that treatment), so behavior can diverge beyond the intended optimization.
Code Review
The patch only partially addresses the requested behavior and does so through higher-risk control-flow bypasses; it likely does not satisfy the intended planner-level static root `__typename` resolution end-to-end.
The change relies on `skipFetch` to avoid backend calls, but does not implement the corresponding datasource-resolvability handling for root `__typename`. This can leave planning/filtering inconsistent and is a likely reason tests still fail.
Both root checks require every post-root path segment to be `InlineFragmentName`; valid named fragment spreads at root will fail this test and not get static `__typename` resolution.
Early returns in `configureSubscription` and `configureObjectFetch` skip normal planner wiring based on a new boolean flag. Any misclassification can suppress required fetch configuration with little local validation.
Default root-type name fallback and matching are duplicated in `configuration_visitor` and `visitor`, which increases drift risk and makes behavior harder to reason about.