test: key differential oracle cache by environment - #51
Open
patlux wants to merge 2 commits into
Open
Conversation
Contributor
|
@patlux is attempting to deploy a commit to the Vercel Labs Team on Vercel. A member of the Team first needs to authorize it. |
vinikjkkj
pushed a commit
to vinikjkkj/scriptc
that referenced
this pull request
Aug 19, 2026
… `in` rules are untouched Main moved a fourth time (`a9c1c756`, block/toprimitive), and this one is not a formality: it closed both bundle `SC1040`s — the two sites this block priced and declined to build — by ANSWERING rather than fencing. No conflict, in either file, which is weaker evidence than a resolved one. Checked instead. `order-parity.json`, by value and on the RAW TEXT because `JSON.parse` collapses a duplicated key silently: main 1642, mine 1642, auto-merged **1644** — 1644 raw key occurrences, **1644 distinct, 0 duplicated**, my two added, main's 4301/4302 present, **0 of main's 1642 rewritten**, 0 dropped. Accounting green. Symbol counts unchanged across the auto-merge (`OBJECT_PROTOTYPE_NAMES` 2, `EMITTER_INSTANCE_STATE_NAMES` 2, `classHasKeyHelper` 2, `classInMemberNames` 4, the leaf guard 1, `%cls.haskey` 2, `a.prim` 7), zero conflict markers. This merge needed more than a symbol count. `block/toprimitive` changed a SHARED runtime path (`scr_json.c`, `scr_runtime.h`) reachable from ~7000 zapo call sites, and the one line of this block that could ride it is `L.ensureString` over a narrowed primitive arm in the `in` classifier. So the fixtures were re-run rather than trusted: **six programs byte-exact against Node v25.9.0 on BOTH backends** — mine (4271, 4272), toprimitive's own (4301, 4302), and the two earlier neighbours' (4282, 4291). Its correction to this block's handover is right and is recorded in the report: asking whether `this.high` can hold an object was the wrong question, because all 33 base traps are runtime fences that never fire and they all still count, so a fenced object arm would simply have been a 34th trap. That is this block's own "-1, not -2" one level up — a fence that never runs is still a trap.
techfreaque
added a commit
to techfreaque/scriptc
that referenced
this pull request
Aug 22, 2026
techfreaque
added a commit
to techfreaque/scriptc
that referenced
this pull request
Aug 22, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Prevent stale Node differential-oracle results when unchanged program bytes observe a different inherited environment.
Both variables already affect output in existing corpus programs, so an environment allowlist can reuse the wrong cached verdict.
NODE_ENVandSCRIPTC_NEVERand bump the key namespace.Focused helper tests and a
NODE_ENVdifferential run, lint, workspace build, andgit diff --checkpass. A shared-cache probe produced distinct Oracle records for both output variants of each corpus program. The full GitHub Actions matrix is pending maintainer approval for this fork PR.