Repository navigation
chore: standardize CI/tooling to v2 canonical (stage 1) - #68
Merged
Merged
Conversation
- Replace ci.yml release job gate/baseline-tag guard with the canonical release.yml; add pr-policy.yml and drop the contribution job from ci.yml. - Have ci.yml fully exercise the VS Code extension (build + vsce package). - Remove the VSCE_PAT/OVSX_PAT fail-fast guard from scripts/release-vscode-extension.mjs (the release job already fails loudly if vsce/ovsx reject an empty token). - Canonicalize .oxfmtrc.json (repo-specific submodule exclusion moves to .prettierignore; routeTree.gen.ts is already covered by **/*.gen.ts). - Canonicalize .oxlintrc.json; keep unicorn/no-new-array off for now since turning it on produces 10 pre-existing errors (stage 3 will fix and flip it to canonical's "error"). - engines: >=26 -> >=22 (root + published packages hdrify/hdrify-react); no Node-26-only runtime API is used anywhere in the codebase. - test script: pnpm vitest (watch mode) -> vitest run; drop format:check and the check aggregate script (CI calls oxfmt --check directly). Closes #67
|
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
release.test.mjs asserted that publish() throws a clear error when VSCE_PAT/OVSX_PAT are unset -- that guard was removed from release-vscode-extension.mjs in this same PR. Without it the test called the real publish() with no PAT, which shelled out to a live `vsce publish` against the Marketplace (failing on auth, not on the assertion). Drop the test along with the now-unused publishExtension import.
This was referenced Sep 25, 2026
bhouston
added a commit
that referenced
this pull request
Sep 25, 2026
Closes #69 ## Summary - `typescript` was already `^7.0.2` on main (no-op for this PR). - `vitest` / `@vitest/coverage-v8`: `^4.1.10` -> `^5.0.1`. No config API changes or source changes were needed. ## Validation - `pnpm build`, `pnpm tsc`, `pnpm lint`, `pnpm exec oxfmt --check .` pass - `pnpm release:check` passes (12/12) - `pnpm size`: hdrify 42.15 kB / 50 kB, hdrify-react 17.19 kB / 22 kB (unchanged) - `pnpm test --coverage`: running locally (heavy EXR-decode suite; watching PR CI on dedicated runners for the authoritative result, per the same note as #68)
This was referenced Sep 25, 2026
bhouston
added a commit
that referenced
this pull request
Sep 25, 2026
Closes #71 ## Summary Deferred from stage 1 (#68): `unicorn/no-new-array` was kept `"off"` because turning it on surfaced 10 pre-existing errors in `pizHuffman.ts` and `b44Codec.test.ts`. - Replaced `new Array(n)` with `Array.from({ length: n })` at all 8 call sites in `pizHuffman.ts` and 2 in `b44Codec.test.ts`. - Added an explicit `Array.from<number>({ length: n })` type argument at the two untyped call sites (`hufTableBuffer`, `freq` in `hufUncompress`) — `new Array(n)` without an annotation inferred `any[]`, so removing it surfaced real `tsc` errors (arithmetic on possibly-`unknown`/`null` values) that needed the explicit type to keep compiling; behavior is unchanged since every one of these arrays is fully populated by index before being read elsewhere (the dense-vs-sparse difference between `Array.from` and `new Array` doesn't matter for direct index access). - `.oxlintrc.json`: flips `unicorn/no-new-array` to `"error"`, now byte-identical to canonical alongside the other two tightened rules (already enabled since stage 1). ## Validation - `pnpm build`, `pnpm tsc`, `pnpm lint`, `pnpm exec oxfmt --check .` pass - `pnpm exec vitest run packages/hdrify/src/exr/b44Codec.test.ts` (the touched test file): 8/8 pass - `pnpm test --coverage` (full suite): running; watching PR CI on dedicated runners for the authoritative result (local EXR-decode tests intermittently time out on this machine under load, as noted in #68/#70) - `node /path/to/standards/check-drift.mjs <repo>`: clean, no exceptions remaining
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
2 participants
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.
Closes #67
Summary
contributionjob that lived inside ci.yml).pnpm --filter hdrify-vscode-extension run package, which builds +vsce package --no-dependencies), in addition to the existing dist/pack check,release:check, and size budget.**/*.gen.ts).unicorn/no-new-array, kept"off"for now: turning it on surfaces 10 pre-existing errors inpizHuffman.ts/b44Codec.test.ts; stage 3 will fix those and flip it to canonical's"error".no-explicit-anyandno-unused-private-class-membersare already clean and go straight to canonical.engines.node:>=26->>=22(root, packages/hdrify, packages/hdrify-react). Grepped for Node-26-only runtime APIs; found none, so no true floor above 22.testscript:pnpm vitest(watch mode) ->vitest run. Removedformat:checkand thecheckaggregate script (CI callsoxfmt --checkdirectly).Validation
pnpm install --frozen-lockfilepnpm build,pnpm tsc,pnpm lint,pnpm exec oxfmt --check .all passpnpm test --coverage: 747 passed / 6 timed out locally on heavy EXR-decode tests under coverage instrumentation (pre-existing 30s timeout tuned for CI; unrelated to this diff — no test or source files touched). Watching PR CI (dedicated runners) to confirm.node /path/to/standards/check-drift.mjs <repo>: clean except the intentionally-deferredunicorn/no-new-arrayexception noted above (stage 3 will close it).