fix(dispatch): wire IaCProvider.BootstrapStateBackend in InvokeMethod (v0.7.5)

[intel352]

Summary

• Adds missing case "IaCProvider.BootstrapStateBackend": to doModuleInstance.InvokeMethod — fixes "unknown method IaCProvider.BootstrapStateBackend" error blocking BMW deploy
• Adds invokeProviderBootstrapStateBackend helper: extracts cfg from args (defaults to empty map if absent), calls provider, nil-safe returns empty map
• Promotes fakeIaCProvider.BootstrapStateBackend from no-op stub to tracking stub (bootstrapCalled, bootstrapCfg, bootstrapResult)
• Three new dispatch tests: dispatch+result shape, nil-result safety, nil-cfg-arg safety

Root cause

v0.7.4 added DOProvider.BootstrapStateBackend (the Go interface method and implementation) but did not add the corresponding InvokeMethod switch case. The gRPC layer routes calls by string method name — without the case, every call from wfctl hits the default branch and returns the "unknown method" error.

Test plan

• GOWORK=off go test -race -short -count=1 ./internal/... — all pass
• TestDoModuleInstance_InvokeMethod_BootstrapStateBackend_Dispatches — verifies routing + bucket/region in result
• TestDoModuleInstance_InvokeMethod_BootstrapStateBackend_NilResult — no panic on nil provider result
• TestDoModuleInstance_InvokeMethod_BootstrapStateBackend_NilCfgArg — missing cfg defaults to {}

🤖 Generated with Claude Code

[Copilot]

Pull request overview

This PR fixes a missing gRPC dispatch route for IaCProvider.BootstrapStateBackend in the DigitalOcean plugin module invoker, unblocking state-backend bootstrapping calls that were previously failing with an “unknown method” error.

Changes:

• Adds "IaCProvider.BootstrapStateBackend" handling to doModuleInstance.InvokeMethod.
• Introduces invokeProviderBootstrapStateBackend to decode cfg, call the provider, and return a map result safely.
• Extends the fakeIaCProvider and adds new tests covering dispatch and nil-result / missing-cfg behavior.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
internal/module_instance.go	Wires IaCProvider.BootstrapStateBackend into InvokeMethod via a new helper.
internal/module_instance_test.go	Adds dispatch tests and upgrades fakeIaCProvider to track Bootstrap calls/inputs/outputs.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

invokeProviderBootstrapStateBackend will panic if InvokeMethod is called with a nil args map (since it indexes args["cfg"]). Since other InvokeMethod callers/tests sometimes pass nil args, this method should defensively treat nil args the same as an empty map (and ideally also surface a clearer error if cfg is present but not a map, instead of silently defaulting to {}).

Suggested change

	cfg, _ := args["cfg"].(map[string]any)
	if cfg == nil {
	cfg = map[string]any{}
	cfg := map[string]any{}
	if args != nil {
	if rawCfg, ok := args["cfg"]; ok {
	if rawCfg != nil {
	typedCfg, ok := rawCfg.(map[string]any)
	if !ok {
	return nil, fmt.Errorf("IaCProvider.BootstrapStateBackend: cfg must be a map[string]any")
	}
	cfg = typedCfg
	}
	}

[intel352]

Fixed in 5945ba2: guarded args[cfg] index behind if args != nil — nil args now treated as empty map, consistent with how other invoke* helpers handle missing optional fields.

[Copilot]

The new BootstrapStateBackend dispatch tests cover empty args (map[string]any{}), but not the nil-args case (InvokeMethod called with args == nil). Adding a regression test for nil args would catch the potential panic path and document the expected behavior when no args are supplied by the host.

[intel352]

Fixed in 5945ba2: added TestDoModuleInstance_InvokeMethod_BootstrapStateBackend_NilArgs — calls InvokeMethod with nil args, asserts no panic/error, result non-nil, and bootstrapCalled=true.