daemon inputs: reads from a test's worker threads and child processes are no input yet
メンテナーはふだん 1 日以内に返信
まだ誰も着手していません。
評価
- 難易度
- 5/5
- 見積もり時間
- 1週間以上
- 初心者へのやさしさ
- 25/100
- issue の種類
- 機能追加
- 明瞭さ
- 明確に書かれている
- 活発さ
- 活発
- 技術スタック
- javascript, node.js
調査の方向性
Start with packages/oracles/oracle-gaps/inputs.mjs (watchReads) and how capture-setup.mjs records reads; compare that with the CLI path in check.mjs. The issue gives an oracle-first sequence: add childRead and workerRead gates and histories, then run the named Vitest command and shard 4. Done means child and worker reads reach the copy's reads, the regression history passes against a cold run, and the mutant check fails only for that history; the non-Node child behavior still needs a maintainer ruling.
索引モデルが issue の本文から書いたものです。
説明
Family: reads from a test's worker threads and child processes. The recorder runs in the Vitest worker only (capture-setup.mjs). A new Worker(...) or child_process.spawn/fork/exec started by a test reads files and variables in another thread or process, and none of those reads is an input. Today a test that spawns a child does make env:* and every variable an input, because spawning lists the environment. The child's own fs reads are missing.
Gate kinds childRead and workerRead: file b runs spawnSync(process.execPath, ['-e', "process.stdout.write(require('fs').readFileSync(<gate path>, 'utf8'))"]) and parses its stdout. The workerRead kind does the same through new Worker(..., { eval: true }) and receiveMessageOnPort. The edit changes only tests/gate.json.
Fix (larger): inject the recorder into the child and the worker, and carry their reads back into the copy's reads.
- Children: wrap
child_processspawn calls in the worker so that a Node child gets--import <session>/inputs-child.mjsthroughNODE_OPTIONS. That module callswatchReadsand appends{ key, hash }rows to a file named by anORACLE_GAPS_CHILD_READSvariable. The copy merges those rows before its closing row. A non-Node child (a shell, a binary) cannot be watched. Such a run has to be treated as reading everything, which is not reusable: it is stored under a key that never matches, so the daemon recomputes it, the conservative choice. This needs a maintainer ruling on whether that cost is acceptable. - Workers: Node
WorkertakesexecArgvandenv. Do the same with aMessageChannel, or with the same rows file. - Cost to measure:
NODE_OPTIONSreaches grandchildren too. The rows file needs the torn-row care from #83.
Authority: maintainer decision M 2026-11-16 "Every read is a daemon input" (the user ruled it with no exceptions), v21.2 R46, G42, I8. The watch family (watch, watchFile) stays no input (C-19).
Context: #133 records the six #55 fs calls on any path, node_modules included, plus environment variables, plugin inputs and directory listings (readdir/opendir, dir:/dirs: keys). The recorder is packages/oracles/oracle-gaps/inputs.mjs (watchReads), shared by the worker (capture-setup.mjs) and the CLI process (check.mjs). Keys are hashed by inputHash and looked up through servedHash. This family is still unbuilt, so I8 does not hold for it.
Oracle-first plan (the pattern from #133):
- Add a value to
inputRead(drawn, appended; update the canary'sinputReadline) and one constructed history appended at the end, with steps[inputGate, recheck]. It entersedit inputGate changed the cold result,an edit to the <kind> gate input changed the cold resultandan unchanged check through the daemon computed nothing. Add it toscripts/pr-shards.mjsshard 4 (and a Windows shard) and toscripts/e2e-shards.mjs. - RED: run
pnpm exec vitest --run --config oracle-gaps/vitest.config.mjs gap-report-oracle -t "fixed seed 20261105: <name>"on the oracle-only commit. Expectincremental equals cold: "after inputGate, the daemon's report differs from the cold run's", with the daemon reusingkilledwhere the cold run sayssurvived. - Fix in
inputs.mjs(see below), then show GREEN on shard 4. - Mutant: remove the new wrapper or key from
watchReads. Only this history should fail.
- 主要言語
- JavaScript
- スター
- 5
- フォーク
- 1
- 平均マージ
- 8時間 31分
- マージ済み PR(30日)
- 60
環境構築
このプロジェクトには開発コンテナ、Dockerfile、コントリビューションガイドがありません。まず README を読み、一般的な手順ははじめてのコントリビューションガイドを参照してください。
はじめの一歩
- issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
- 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
- リポジトリをフォークし、ブランチを切って変更します。
- issue 番号を参照したプルリクエストを送ります。
TanStack/oracles のほかの issue
-
難易度 2/5 1〜3時間 初心者へのやさしさ 62/100
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 72/100
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 64/100
メンテナーはふだん 1 日以内に返信
-
難易度 5/5 1週間以上 初心者へのやさしさ 35/100
メンテナーはふだん 1 日以内に返信
-
難易度 4/5 3〜5日 初心者へのやさしさ 46/100
メンテナーはふだん 1 日以内に返信
TanStack/oracles の issue をすべて見る
似ている issue
-
Edit: Telemante [YT]オープンchannels:edit check:passed
難易度 2/5 1〜3時間 初心者へのやさしさ 65/100
メンテナーはふだん 1 日以内に返信
-
fix
難易度 2/5 1〜3時間 初心者へのやさしさ 68/100
idean3885/claude-ops-agent#1083 ·
メンテナーはふだん 1 日以内に返信
-
入群码过期オープン
難易度 1/5 1時間未満 初心者へのやさしさ 84/100
aliyun/ai-agent-handbook#29 · コメント 1 件 ·
-
docs documentation need-triage SaaS
難易度 2/5 1〜3時間 初心者へのやさしさ 72/100
メンテナーはふだん 1 日以内に返信
-
bug
難易度 2/5 1〜3時間 初心者へのやさしさ 76/100
apache/cloudstack#14719 ·
メンテナーはふだん 1 日以内に返信