chore(#454): re-verify callback-external guard + fix stale null-bundled comment

[apotema]

Two low-risk cleanups from #454 (extraction tech-debt, follow-up to the completed epic #386).

1. Callback-external guard — investigated, kept (correct as-is) + regression test

src/codegen/lifecycle/render.zig ExternalCallbackBackendUnsupported is not dead code. It was already narrowed in #416: the guard fires only when use_callback_lifecycle and isExternal() and !callback_dispatch_handled, where callback_dispatch_handled = is_bgfx_android or cfg.backend == .sokol. External sokol-desktop (today's production state) keeps cfg.backend == .sokol via the enum-as-shorthand, so it's handled and the guard does NOT fire.

The guard still protects two genuinely-unsupported live paths: (1) a 3rd-party external callback backend sitting at the .raylib enum default, and (2) any external non-sokol backend on wasm. Added a regression test asserting external sokol-desktop generates the sokol callback main rather than erroring — fails loudly if anyone drops cfg.backend == .sokol from callback_dispatch_handled.

2. Stale comment (src/cache_cmd.zig)

"no-op when null is still bundled" was wrong post-extraction (null → labelle-null, no builtinProvider arm returns null). Rewrote to state the real no-op is the isPluginCached guard. Comment-only; logic unchanged.

zig build test + zig build both exit 0.

Ref #454, #386.

https://claude.ai/code/session_017pW3ifKf9wgxNg4viy6okw

Summary by CodeRabbit

• Bug Fixes

  • Improved handling for external desktop backends using a callback run loop, so valid setups are no longer incorrectly rejected.
  • Refined cache-related behavior to better match the current backend flow and avoid unnecessary fetches when the plugin is already cached.

• Tests

  • Added a regression test covering an external sokol desktop backend with callback-based execution.

[gemini-code-assist]

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 1752e26f-623c-40d7-a4f3-65b5edb44bf8

📥 Commits

Reviewing files that changed from the base of the PR and between 2eb7fbd and 5687433.

📒 Files selected for processing (2)

• src/cache_cmd.zig
• test/main_zig_tests.zig

---

📝 Walkthrough

Walkthrough

This PR clarifies a comment in ensureCache regarding .null backend handling in src/cache_cmd.zig, explaining that caching idempotency now relies on isPluginCached rather than a null-return check. It also adds a regression test verifying an external sokol desktop backend with callback run-loop is not rejected by the external-callback guard.

Changes

Null-backend caching comment and callback guard test

Layer / File(s)	Summary
Cache guard comment update src/cache_cmd.zig	Replaces the outdated comment describing null-return no-op behavior with an explanation that effectiveBackendPackage() resolves the external backend and idempotency comes from the isPluginCached guard.
External sokol callback regression test test/main_zig_tests.zig	Adds a test generating main.zig for an external sokol desktop backend and asserts the callback export frame is present, confirming the external-callback guard does not falsely reject this case.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~5 minutes

Possibly related issues

• Backend extraction remainder: surface-loss adapter slot cleanup and optional fixture slimming #454: Updates the same cache_cmd.zig comment and adds the same sokol callback regression test described in that issue.

Possibly related PRs

• labelle-toolkit/labelle-assembler#409: Targets the same external callback backend rejection behavior exercised by the new regression test.
• labelle-toolkit/labelle-assembler#419: Also modifies src/cache_cmd.zig around effectiveBackendPackage() usage for external backend handling.
• labelle-toolkit/labelle-assembler#431: Also concerns ensureCache's .backend = .null external-backend handling logic.

Poem

A comment tidied, a test set free,
Sokol callbacks now run happily.
No guard shall block the frame() call,
This little rabbit approves it all. 🐰
Hop along, cache clean and bright!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately reflects both the callback-guard regression recheck and the stale cache comment fix.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch chore/454-extraction-cleanup

---

_{Comment @coderabbitai help to get the list of available commands.}