Documentation
¶
Overview ¶
Package plutotest wires a Pluto pack execution into ordinary go test, and gates a test on the derived qualification disposition. It runs no trial loop of its own: each runnable table plan expands to exactly one eval.Suite executed once via eval.Run. Execution itself lives in pkg/run — this package is a thin go-test wrapper around run.Execute that adds t.Fatalf-on-error and t.Logf-on-skip test ergonomics.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func RequireDisposition ¶
func RequireDisposition(t *testing.T, card qual.Scorecard, p profile.Profile, allowed ...profile.Disposition)
RequireDisposition evaluates card against p (profile.Evaluate) and t.Fatalf's when the derived disposition is not one of allowed, printing the disposition and every requirement outcome. An empty allowed list is itself a t.Fatal: a gate that accepts no disposition is a configuration bug, not a legitimate all-reject gate.
func Run ¶
Run executes every pack in spec against spec.Target under spec.Manifest and returns the resulting Scorecard, by delegating to run.Execute (pkg/run): qual.Plan expands the manifest into per-table plans, a runnable plan is executed with eval.Run using spec.Target for every table, and a non-runnable plan (missing capability) becomes a Skipped TableResult instead of being executed. Run logs each skipped table's missing capabilities via t.Logf, and t.Fatalf's on any error run.Execute returns — a broken pack or manifest, or eval.Run's own preflight failure, is a bug in the test setup, not a verdict to record. Pluto implements no trial loop of its own — Trials passes straight through to eval.RunConfig.
Types ¶
type RunSpec ¶
type RunSpec struct {
Manifest qual.Manifest
Packs []qual.Pack
Target eval.Target
Trials int // passed straight to eval.RunConfig.Trials
}
RunSpec is one full offline-or-live Pluto execution: a manifest, the packs to plan against it, the target to observe, and how many trials to run each scenario.
Source Files