Hacktoberfest 2026:維護者為十月標記出來的 issue,仍然開放、適合新手。 瀏覽 Hacktoberfest issue

Restore Query DB receiving histories omitted during PR integration

未關閉
#2,188 0 則留言 0 個 reaction 已指派 0 人 在 GitHub 檢視

維護者通常 1 天內回覆

還沒有人認領這個 Issue。

評估

難度
5/5
預估耗時
一週以上
新手友好度
8/100
Issue 類型
功能
描述清晰度
需要釐清
活躍度
活躍
技術堆疊
sql, sqlite, typescript
領域
testing-qa

研究方向

Start with docs/contributing/oracle-coverage.md near the Query DB ownership section to see which histories are listed as missing, then read load-subset-lifecycle-oracle.test.ts and ownership-lifecycle.oracle.test.ts. Done means new public-history tests with wrong-design controls, scratch failures classified as invalid fixture, production defect or open contract question, and the coverage map updated.

由索引模型根據 Issue 內容生成。

描述

Coverage gap

The #2134 branch once had public receiving histories for Query Core observer handoff, paused and reentrant replacement fetches, refetch cancellation, and nested fetch order across direct writes. Some did not survive its integration with #2158 at executable merge 2089e0630.

The current merged tree has strong named and generated lifecycle coverage. It does not replay every earlier receiving path. docs/contributing/oracle-coverage.md lists the missing paths near the Query DB ownership section. A scratch merge of the older tests with the published-Core adapter reported 68 failures among 841 tests. Some fixtures assumed an eager signal data descriptor or same-stack physical abort. The current adapter uses a lazy getter and Query Core's release turn. Those fixture failures are not evidence of product bugs; the remaining public-law failures have not been classified.

Oracle owners and acceptance

  • Extend load-subset-lifecycle-oracle.test.ts with outside-observer handoff, cleanup, paused-fetch, and reentrant replacement histories. Compare the physical signal, Query cache, caller settlement, and full public rows after the correct release and application cuts.
  • Extend ownership-lifecycle.oracle.test.ts with nested fetch-start order across accepted direct writes. Add explicit refetch and snapshot histories that compare each caller's result with its own applied replacement.
  • Derive expected outcomes from the Collection sync-adapter contract and Query Collection guide. Do not copy Query DB's current fetch classifier into the reference model.
  • Calibrate at least one wrong-design control at each new public comparison. Classify old scratch failures as invalid fixture, production defect, or unresolved contract question before changing assertions.
  • Update the coverage map with the exercised histories and remaining limits. File confirmed product defects separately. #2176 already tracks non-source Query result provenance.

This is follow-up oracle work. It does not change the bounded identity, primitive SQLite upgrade, or #2158 acquisition claims already reviewed in those PRs.

主要語言
TypeScript
星號
3.9k
分支
272
平均合併
22 小時 54 分鐘
30 天內合併 PR
236

環境準備

從這裡開始

  1. 先讀完整個 Issue,再讀專案的貢獻指南。
  2. 在 Issue 下留言說明你要接手 —— 這能避免兩個人做同樣的事。
  3. Fork 儲存庫,在一個分支上完成修改。
  4. 送出 Pull Request,並在描述裡引用這個 Issue 編號。

TanStack/db 的其他 Issue

查看 TanStack/db 的全部 Issue

相似的 Issue

更多 TypeScript Issue

把新 issue 寄到你的電子郵件信箱

精選適合新手參與的 GitHub issue 摘要。