Skip to content

test: organize recipe sync cases without changing discovery - #1569

Closed
risu729 wants to merge 1 commit into
mainfrom
refactor/service-recipes-tests-20261112
Closed

risu729 wants to merge 1 commit into
mainfrom
refactor/service-recipes-tests-20261112

Conversation

@risu729

@risu729 risu729 commented Nov 12, 2026 •

Copy link
Copy Markdown
Owner

The recipe-sync test entry exceeded the line limit, combining sync recovery, capacity, disconnect/layout lookup and cached lookup. Keep the sole discovered entry, all seven original suite wrappers, helper/fixture initialization and remaining suite bodies in their original scopes; move four contiguous sync groups and cached-lookup cases into five passive synchronous registrars outside production coverage.

The entry is 831 lines, leaves 271–418 lines and context 15 lines. Erased fixture types refer to the original helpers through typeof, avoiding duplicated declarations. Callbacks, assertions, tables, comments, raw templates, references and order remain unchanged. Remove only this entry's line-limit exemption and document ownership. Continues the approved refactoring work.

Validation at final head:

  • Fresh independent source review found no issues. Independently expanded AST, raw tokens/comments, imported binding ownership, all 62 forwarded references and erased fixture types match the original.
  • Original 109/109 cases pass in 3.62 s; the actual service task passes 769 cases across 20 files with coverage in 17.47 s. All 109 expanded IDs, full titles, order and outcomes plus seven suite IDs/nesting match exactly.
  • All 20 discovered service entries remain unchanged. Actual coverage contains 258 production files and no helper modules; runner/discovery/coverage configuration is unchanged.
  • Extension typecheck, scoped static checks, docs validation, full CI and CodeQL pass, including all six packaged browser shards.
  • Every Verify task matches the actual tested-parent baseline: 9,186 Vitest cases across 274 files plus 73 tool tests; components retain 1,814 cases across 108 files. The tested merge includes 14 newer diagnostics/lint/docs paths with no owning service source/test difference.
  • Single-run Verify was 275 → 209 s, runtime 211.72 → 162.08 s, service task 50.65 → 39.90 s and recipe entry 34.692 → 25.782 s. Components were slower, 139 → 172 s (step 117 → 145 s), including unchanged options/orders/DOM tasks. These observations do not establish a causal speedup or zero regression.

No real-account validation was performed. Leave open for owner review; no merge or auto-merge.

@risu729 risu729 left a comment

Copy link
Copy Markdown
Owner Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fresh independent source review: no actionable findings at the reviewed head against the original main base.

Independently expanded all five synchronous registrars without the author's manifest. Complete AST and all 20,787 raw tokens, including 65 ordered comments and 15 templates, exactly match after removing imports and the added erased fixture alias. Preserved 96 declarations / 109 expanded cases, seven describes, no hooks and 405 expect calls. All 62 injected fields preserve original initialized const/function bindings and fixture/cache closures; no reassigned capture, receiver change or module-load registration. Import owners/value-type roles match, fixture types use exact typeof queries, and discovery/coverage remain unchanged. Only the target max-lines exemption is removed; owning docs accurately describe the split.

Runtime verification and CI are separate evidence. The reviewer made no edits, runtime/browser tests or real-account actions.

risu729 commented Nov 12, 2026

Copy link
Copy Markdown
Owner Author

Merged as #1577. That PR contains these commits, merged @pateat/contracts imports, the context alias dropped, a docs section of its own, and a merge of main.

The collected test IDs and their order match main (109 tests).


Generated by Claude Code

@risu729 risu729 closed this Nov 12, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant