refactor(signals): shared read predicates — readerSeesCommitted, visibleOverride, one hasActiveOverride

[ryansolid]

Step 1 of DESIGN-CONSOLIDATION move 3b (one implementation per rule). Zero semantic change — the gate is the point of the PR, not the code.

What moved

• readerSeesCommitted(el, c, owner, noCommitted) (core.ts): the full committed-vs-staged arm — untracked / lane-reads-committed (Existing and newly revealed latest readers display different values #3460) / nothing staged / children-forbidden (A32) / stale-of-foreign (heldFromStale) / HELD-truth mask (2.0.0-rc until doesnt entangle the update resolving it with the action update #3164) — that read()'s slow tail used to inline. Documented as the rule it enforces; the store's backing selection and the verdict arms adopt it in step 2.
• hasActiveOverride: existed twice (lanes.ts, store.ts). One definition in core.ts now; lanes.ts and store.ts re-export.
• visibleOverride (hasActiveOverride && !unflushedOverride, A28 (5)): was the store's helper, restated by hand at latestRead, computePendingState ×2 and the store's reader selection. One definition in core.ts; five call sites.

Hot paths untouched: readNodeFast, read()'s fast block, setSignal, recompute are byte-identical. read()'s override arm keeps its shape (the AUTHORITATIVE_OBSERVED mark still fires for an unflushed override) — only the inline test became hasActiveOverride(el).

Gate

check	result
signals suite	197 files / 2746 pass
visibility oracles (signal, store)	report byte-identical to next
posture matrix (621 cells, 738-line report)	byte-identical to next
fuzzer #3446, seed 3289 × 1000	994 pass / 0 fail / 6 policy — same as next
solid-js	623 pass
web	same 12 pre-existing server-diagnostics failures as origin/next in this worktree (component-location naming, unrelated)
minified (treeshake fixture)	core +13 B, +createStore −38 B, isPending/latest +34 B, full bundle −72 B
brotli, pure-signals fixtures	−4 / −29 / +26 / −5 B

Three size-limit caps ratcheted (+9 / +44 / +21 B brotli on esbuild app bundles) with notes: the ten scenarios moved −19…+45 B in both directions, compressor layout, not code.

Local wall-clock microbench A/B was order-dependent (−7% one order, +4% the other) — CodSpeed on this PR is the perf verdict; structurally only read()'s slow tail gained a call (small, monomorphic, inlined).

Next (step 2)

serve(el, reader): the store's pendingBackingVisible and serveDataKey adopt readerSeesCommitted; then readerLive(reporter, verdict); then the verdict-deferred list.

Made with Cursor

[changeset-bot]

🦋 Changeset detected

Latest commit: 449ef6e

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 11 packages

Name	Type
@solidjs/signals	Patch
test-integration	Patch
@solidjs/web	Patch
@solidjs/babel-plugin	Patch
@solidjs/compiler	Patch
@solidjs/diagnostics	Patch
@solidjs/element	Patch
@solidjs/h	Patch
@solidjs/html	Patch
solid-js	Patch
@solidjs/universal	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[coveralls]

Coverage Report for CI Build 35255614381

Warning

No base build found for commit 590ff3a on next.
Coverage changes can't be calculated without a base build.
If a base build is processing, this comment will update automatically when it completes.

Coverage: 71.304%

Details

• Patch coverage: No coverable lines changed in this PR.

Uncovered Changes

No uncovered changes found.

Coverage Regressions

Requires a base build to compare against. How to fix this →

---

Coverage Stats

Relevant Lines:	1035
Covered Lines:	783
Line Coverage:	75.65%
Relevant Branches:	798
Covered Branches:	524
Branch Coverage:	65.66%
Branches in Coverage %:	Yes
Coverage Strength:	15.06 hits per line

---

💛 - Coveralls

[codspeed]

Merging this PR will not alter performance

✅ 172 untouched benchmarks
⏩ 3 skipped benchmarks¹

---

_{Comparing refactor/shared-read-predicates (449ef6e) with next (590ff3a)}

Footnotes

1. 3 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩