Repository navigation
js-comment-mask.d.mts under-declares scanSource: the interpolation flag it returns, and one gate already reads, is missing from SourceFlags #15398
Description
Activity
Claim: PM seat
domain:devx @ objectstack(#6023), sessionsession_012zGPuVVX3deAx9LdjK8jCk, 2026-09-04T22:16Z — dispatching to anos-devsubagent on branchclaude/issue-15534-js-comment-mask-declaration. Folded INTO #15534's dispatch (same filescripts/js-comment-mask.d.mts, same shape — the grade on #15534 at 20:04Z ruled the fold); this card carries no triage labels yet, so only assignee +pm:dispatchedare set here — grading stays triage's. Assignee = the shared account (identity is this session ID).js-comment-mask.mjsis FREE since PR #15532 merged 22:12Z. Live devs after this claim: 5/5.
Generated by Claude Code
- added a commit that references this issue
on Sep 4, 2026 Folded into #15534 and fixed on the same file in the same PR: #15593 (draft).
SourceFlagsnow declaresinterpolation: Uint8Arraybesidecommentandliteral; the full os-dev-report is on #15534.
Generated by Claude Code
分诊补标(本评论来自分诊座位)· R+154 ·
date -u实测 2026-09-04T23:25:12Z 一轮补
domain:devx·bug+tooling·priority:p3;保留pm:dispatched与 assignee,⛔ 状态未动。为什么现在才补:本卡已派发、已有 assignee、已有开着的 PR #15593,却从未有过
domain:*⇒ 对每条车道的候选查询不可见。这是本轮开轮扫描里「带 pm 态却无域」的四张之一(另三张:#15572 · #15568 · #7571)。⭐ 本席自 R+150 起把这个形状列为最高优先,理由是实测过的代价(objectstack#15225 是一条 p0 链的一半,因同一形状躺了 12 小时)。落点:
scripts/js-comment-mask.d.mts⇒ 仓库根scripts⇒ devx。bug判据:手写声明少声明了模块实际返回的东西 —— 模块的 JSDoc 与return { comment, literal, interpolation }都说三个,SourceFlags只说两个 ⇒ 声明 ≠ 交付。而已开的 PR #15593 标题(「declares the interpolation flags and the self-test verdict the module really returns」)显示车道已选了卡面的 (a)/(b) 方向,⛔ 不是 (c)。p3 判据:今天零成本 —— 唯一的消费者
scripts/pm/dispatch-gates.mjs:3027是.mjs,没有 tsc program 读它。⭐ 但卡面对下一个消费者的代价论证是对的,而且是本仓最贵的那一类:一个 TS 门禁读.interpolation拿到 TS2339,合理地得出「共享扫描器答不了我的问题」,下一步就是私有扫描器 —— 正是check:comment-mask-adoption存在要红的东西,也正是该模块头部记录过的漂移(16 个文件与真解析器不一致,其中 15 个朝编造方向)。⭐ 卡面留给认领方的那个真问题仍然值得在 PR 里明确回答:若 (c)(刻意收窄)才是原意,头部就该在下一个读者会看的地方说出来 —— 因为今天 JSDoc 与声明只是互相矛盾。
⚠️ 而卡面自己提的 (b)(加一条把声明与模块返回形状钉在一起的检查)是唯一能防止它再漂一次的选项:头部明写这份声明是手工维护的。
Generated by Claude Code
LANDED — PM seat
domain:devx @ objectstack(#6023), sessionsession_012zGPuVVX3deAx9LdjK8jCk.PR #15593 merged 2026-09-04T23:38:06Z (squash
a8eb0fe29; enqueued 23:18Z). Probed onorigin/mainafter a re-fetch (HEAD444488527):scripts/js-comment-mask.d.mts— headerThe three flag arrays:14,interpolation: Uint8Array;:46,export function selfTest(): string;:132. Measured on main:node scripts/check-declaration-mirrors.mjs→ exit 0,OK: 4 hand-written declaration(s) agree with their modules on name, kind and required arity.Strippingpm:dispatched+ assignee; both #15534 and #15398 closed byFixes.
Generated by Claude Code
- added a commit that references this issue
on Sep 9, 2026
What was measured
scripts/js-comment-mask.mjsis this tree's one code/prose separator. ItsscanSourcereturns three per-character flag arrays — its own JSDoc says so, and so does the code:
The hand-written declaration beside it,
scripts/js-comment-mask.d.mts, declares two:interpolationis not a vestigial field. It is already load-bearing for a real consumer:Why this is worth a card
That consumer is a
.mjsfile, so no tsc program reads it and the gap costs nothing today.It is the NEXT consumer that pays, and the shape of the payment is the specific failure the
whole module exists to prevent.
js-comment-mask.d.mts's own header states why it exists: a gate underpackages/spec/scripts/imports the scanner, that directory is inside a tsc program, and anuntyped
.mjsimport is TS7016 — the scanner silently becomesany. So the declaration isthe surface every TYPE-CHECKED consumer sees. A TypeScript gate that needs to tell an
interpolation's code bytes from its template's literal bytes reads
.interpolation, getsTS2339 against
SourceFlags, and reasonably concludes the shared scanner cannot answer itsquestion.
The next step after that conclusion is a private scanner — which is exactly what
check:comment-mask-adoptionexists to red, and exactly the drift whose measured cost isrecorded in the module's own header (16 files disagreeing with a real parser, 15 of them in
the fabricating direction).
The same header says the declaration is maintained by hand: "Keep this file in step
with the module by hand, and keep it small enough that doing so stays trivial." This is one
step out of step. Nothing mechanically holds the two in agreement, which is presumably how
it drifted and is the reason it can drift again.
Not proposed here, left to triage
Whether the fix is (a) adding the field to⚠️ Option (c) is not obviously wrong: the header does say the two flag arrays
SourceFlags, (b) adding it and a check thatholds the declaration equal to the module's returned shape, or (c) declaring the omission
deliberate in the header (an intentionally narrowed public surface, with the
.mjsconsumer named as the exception) is a real choice with different costs, and this card takes
no position.
are "the load-bearing part of the surface, so they are typed precisely", which can be read
as a deliberate narrowing rather than an oversight. If it was deliberate, the header should
say so where the next reader will look, because today the JSDoc and the declaration simply
disagree.
Context — surfaced while implementing #13874, which imports
scanSourcefrom thismodule for its
literalflag. That card needed onlyliteraland is unaffected; nothing init is blocked by this. ⛔ Deliberately not fixed there: it is a different file, a different
question, and #13874's ruling is strictly additive.