Gap capture: an optional call's short-circuit record can pair with another evaluation (four cases)

未关闭
#162 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

维护者通常 1 天内回复

还没有人认领这个 Issue。

评估

难度
5/5
预计耗时
一周以上
新手友好度
25/100
Issue 类型
缺陷
描述清晰度
描述清楚
活跃度
活跃
技术栈
javascript
领域
testing-qa

调研方向

Start with pr147-review.md at “Delta round 2”, D1, then read the gap oracle’s Limits and the simulations in /tmp/rev147-scripts/q2.mjs and q3.mjs. The issue describes four record pairing failures and states that fixing them may require a statement level rewrite or another design. Done means choosing and implementing a design that prevents those failures; no tests or source files are named.

由索引模型根据 Issue 内容生成。

描述

#147 makes an optional call f?.() pass its short-circuit on to the next ?. link of its chain:

  • the link's receiver opens a record (open);
  • the call's empty argument spread marks the newest open record as run (ran), and it runs only when the call does;
  • the link reads the newest record and closes it (chain).

The records of one call site form a stack, so a record pairs with its evaluation only where evaluations of the same expression nest and each one ends. The oracles repo has no coverage map, so this issue records the four cases that break the pairing. Each needs an owner. No history of the gap oracle generates any of them, and the gap oracle's Limits state them.

# Case Record left behind Consequence for the next ?. link
1 An evaluation of the same expression nested in the receiver, which throws after its call ran, and the throw is caught there run, left open With the outer method absent: a false infection, and a false throw under the link's mutant (P3)
2 The same nesting in the method lookup run, left open The same: a false throw (P3)
3 An evaluation nested in the call's arguments or the method's body, which throws before its own call runs, and the throw is caught there not run, left open With the outer call having run and returned nothing: a false skip, so the mutant is judged without a run as not-distinguished. This is the class of #109 (P5)
4 Two async evaluations of the same expression interleaved at an await ((await r(a))?.f?.(await g())?.v) the evaluations close each other's records a false skip or a false throw

The review that found them is pr147-review.md, "Delta round 2", D1. The simulations are in /tmp/rev147-scripts/q2.mjs and q3.mjs.

A pairing per evaluation would need a binding that an expression rewrite cannot create without changing this, await or arguments. For example, an arrow wrapper breaks await and yield. A fix would need a statement-level rewrite or another design.

The numeric path for nested ?. links (shortCircuited) is not affected. Nothing runs between a link that short-circuits and its consumer.

Filed under v21.2 P11 and G36, as a stated limit with an owner.

This issue was written by Isaac.

主要语言
JavaScript
星标
5
派生
1
平均合并
11 小时 18 分钟
30 天内合并 PR
66

环境准备

这个项目没有提供开发容器、Dockerfile 或贡献指南,环境需要你自己搭建:先看它的 README,通用步骤见我们的新手贡献指南。

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

TanStack/oracles 的其他 Issue

查看 TanStack/oracles 的全部 Issue

相似的 Issue

更多 JavaScript Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。