test: inherit the real @object-ui/react export surface in 25 hand-listed mocks - #6847
Merged
Merged
Conversation
Each of these files mocked `@object-ui/react` with a factory that hand-lists the
exports it returns, so the mock's export surface is frozen at whatever the author
typed that day. That makes every one of them a latent COLLECTION failure for the
next export any widely-imported module reads at module scope: the file dies before
a single test runs, so the suite goes red with zero failed assertions and reads like
flake or infrastructure.
Converted to the `importOriginal` form already used by the other 81 sites, so the
mock is a superset of the real module and only the overrides each test means to
control are replaced. The spread is inserted FIRST, so every existing override still
wins.
No assertion was edited or deleted — every removed line in this diff is a
`vi.mock('@object-ui/react', ...)` header line.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CRJge11jso9TpXRWFt1Z49
This was referenced Aug 30, 2026
Contributor
✅ Console Performance Budget
The eager closure is every chunk the entry reaches through static imports — what the browser fetches and parses before the app renders. The entry chunk on its own is a small fraction of it. 📦 Bundle Size Report
Size Limits
|
os-sam
marked this pull request as ready for review
August 30, 2026 03:36
This was referenced Aug 30, 2026
os-sam
pushed a commit
that referenced
this pull request
Aug 30, 2026
…al surface A `vi.mock` factory that hand-lists the exports it returns freezes the mock's export surface at whatever was typed that day. The next export any module in the file's import graph reads AT MODULE SCOPE then kills the file during COLLECTION: `Test Files 3 failed | 546 passed` with `Tests 6694 passed` -- zero failed assertions, because the tests in those files never ran. It reads as flake, and the bill lands on whoever added the export. PR #6847 swept 25 such sites. Nothing stopped the 26th, and this installs the ratchet that does. The recogniser is SEMANTIC, never a grep for `importOriginal`. That spelling is wrong in both directions and both were measured on this tree: it called eleven already-correct files broken (nine zero-parameter `vi.importActual` factories in plugin-dashboard, one parameter named `importActual`, one named `orig`), and it missed a genuinely frozen one that contains the token nowhere. The criterion is what the code DOES -- obtain the real module under any binding name, and spread it into the returned object. Obtaining without spreading is still frozen. Narrow by triage's ruling (objectui#6849, R+34): only the workspace specifiers in COVERED_SPECIFIERS are judged, and widening one in needs a sweep first. Local whole-module replacement, third-party packages and unswept workspace packages are counted in the census and never judged -- out of scope by construction, with no per-file exception list anywhere. Also converts the 26th, which this gate's first run found: plugin-view's ObjectView.contractEnvelope-6726.test.tsx was byte-identical at PR #6847's own commit and the sweep passed over it. Fixes #6849 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013hfmP9hoMd3dJwTh85J4yB
akarma-synetal
pushed a commit
to akarma-synetal/objectui
that referenced
this pull request
Sep 1, 2026
…al export surface (objectstack-ai#6894) * test(tooling): ratchet the vi.mock factories that must inherit the real surface A `vi.mock` factory that hand-lists the exports it returns freezes the mock's export surface at whatever was typed that day. The next export any module in the file's import graph reads AT MODULE SCOPE then kills the file during COLLECTION: `Test Files 3 failed | 546 passed` with `Tests 6694 passed` -- zero failed assertions, because the tests in those files never ran. It reads as flake, and the bill lands on whoever added the export. PR objectstack-ai#6847 swept 25 such sites. Nothing stopped the 26th, and this installs the ratchet that does. The recogniser is SEMANTIC, never a grep for `importOriginal`. That spelling is wrong in both directions and both were measured on this tree: it called eleven already-correct files broken (nine zero-parameter `vi.importActual` factories in plugin-dashboard, one parameter named `importActual`, one named `orig`), and it missed a genuinely frozen one that contains the token nowhere. The criterion is what the code DOES -- obtain the real module under any binding name, and spread it into the returned object. Obtaining without spreading is still frozen. Narrow by triage's ruling (objectui#6849, R+34): only the workspace specifiers in COVERED_SPECIFIERS are judged, and widening one in needs a sweep first. Local whole-module replacement, third-party packages and unswept workspace packages are counted in the census and never judged -- out of scope by construction, with no per-file exception list anywhere. Also converts the 26th, which this gate's first run found: plugin-view's ObjectView.contractEnvelope-6726.test.tsx was byte-identical at PR objectstack-ai#6847's own commit and the sweep passed over it. Fixes objectstack-ai#6849 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013hfmP9hoMd3dJwTh85J4yB * chore(tooling): declare the changeset and fix the suite's inferred reason type Empty-frontmatter changeset: the only released-package source file in this diff is a plugin-view test converted to the inheriting mock form, so nothing published changes. `check-changeset-presence` guards `/src/**` of released packages, which is what pulls that one file into its scope. Also narrows the census figure in the suite's comment to this tree, and marks `reason` optional where the scan's inferred union makes it so (TS2345 under `tsconfig.scripts.json`). Part of objectstack-ai#6849 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013hfmP9hoMd3dJwTh85J4yB * docs(tooling): point the guard's cross-references at the filed findings The header and suite carried placeholder numbers for two findings that had not been filed yet. They are objectstack-ai#6891 (the shared comment-masker reads a JSX closing tag as a regex literal, which is why this guard neutralises them locally) and objectstack-ai#6892 (the 298 frozen factories on other workspace specifiers, i.e. the per-specifier sweep worklist that widening COVERED_SPECIFIERS depends on). A header naming a wrong issue number is worse than one naming none: it reads as a citation. Part of objectstack-ai#6849 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013hfmP9hoMd3dJwTh85J4yB --------- Co-authored-by: Claude <noreply@anthropic.com>
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
3 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.
Fixes #6768
25 test files mocked
@object-ui/reactwith a factory that hand-lists the exports it returns. That freezes each mock's export surface at whatever the author typed that day, making every one of them a latent collection failure for the next export any widely-imported module reads at module scope. They are converted to theimportOriginalform already used by every other site, so the mock is a superset of the real module and only the overrides each test means to control are replaced.Verified at
9ddab598a.Re-derived enumeration (against
origin/mainat26896c689, not the card's numbers)The card counted 36 sites "without
importOriginal". That is a name-based count, and it is wrong in the safe direction: 11 of those 36 already inherit the real surface by another spelling. The real remediation set is 25.A semantic classifier was used instead — a mock inherits iff its factory both obtains the real module (its
importOriginalcallback parameter, whatever it is named, orvi.importActualof this specifier) and spreads it:vi.mockof@object-ui/reactControl: the 81 come back from the same query as a non-zero count, so the matcher is not simply failing to match. Their spellings: 70 spread an
importOriginalparam, 9 spreadvi.importActual, 1 names the paramimportActual, 1 names itorig.The 11 the card would have had us touch, which need nothing:
plugin-dashboardalready doconst actual = await vi.importActual('@object-ui/react')and spread it (ObjectDataTable.*,lookupRelationalMeta-6694).vi.importActualisimportOriginalbound to the path; the export surface is equally complete.EnvironmentListToolbar.test.tsxnames the paramimportActual.PageView.test.tsxnames itorig.They are left alone deliberately. Note for the next audit: a grep for the literal
importOriginalwill mis-flag all 11.Contention
Enumerated before editing. All 30 open PRs were listed and every changed file fetched (266 files, plus 633 on the release PR #5400, paginated to exhaustion). Zero of the 25 targets is occupied by an open PR, so the sweep is complete rather than partial, with nothing deferred.
Behaviour preservation
Spreading the real module means the mock now returns exports it previously omitted, so this was checked rather than assumed:
import * as ns from '@object-ui/react'. Absent a namespace object, there is noObject.keysorintest over the module for the added exports to change.undefined, so any export the graph reads is already hand-listed in a passing file. EverytoBeUndefined/not.toHaveProperty/Object.keysassertion in these files was read: all of them are about schema objects, column objects and query params, none about the mocked module.SchemaRendererContextseveral files create) shadow the real export exactly as before.Did any existing test move?
None. Mechanically, not by assertion: every removed line in the whole diff is a
vi.mockheader.Lint agrees the added lines are inert: of 204 eslint messages on these files, 0 land on a line this PR added (all are pre-existing
no-explicit-any).Ablation — the point of the card, not just the edit
A new export was added to
@object-ui/reactand a real, widely-imported module (packages/components/src/hooks/related-count-store.ts, which already readssubscribeDataChangesfrom it at module scope) was made to read it at module scope — the exact shape of the next PR that would have broken these files. Both legs ran on the same mutated tree; the only variable is the mock form.The mutation was confirmed on disk before either leg (1 occurrence injected into the index, 2 into the store, and the old bare import confirmed gone), because an edit tool's exit code proves nothing about a zero-hit anchor.
Leg A — converted file (this PR):
Leg B — same probe, that one file reverted to its pre-conversion frozen mock:
Tests no testsis the whole argument: the file died at collection, so its 32 tests never ran. That is a red suite with zero failed assertions — the signature that reads like flake and is neither.No rebuild is involved on either leg, and that is not an omission: the root vitest config aliases
@object-ui/reacttopackages/react/src, so the probe is resolved from source anddistplays no part. Restore was proved by an emptygit diff HEAD, and the probe confirmed gone from both mutated files.Verification
Run from the repo root (a package-scoped
vitestre-roots and reports another package's files):pnpm exec vitest runover the 25 converted files plus the 9 already-inheriting ones as an untouched control — 34 files passed, 319 tests passed, at9ddab598a.tsc -p packages/PKG/tsconfig.test.json --noEmitfor all six affected packages — all exit 0, zeroerror TS. Coverage proved rather than assumed: the package build tsconfigs exclude test files, so--listFileswas used to confirm all 25 edited files are inputs of the chained test program (0 missing). This mattered — the concise spelling has 17 precedents in the tree, but the block-body spelling this PR introduces had zero, so it was not covered by any existing green.check:vi-mock-specifiers—OK (3955 tracked source file(s), 2245 test-named; 490 carry a mock; 751 relative specifier(s) resolved, ... 0 non-static, ... 3 via the import() form).check:control-bytes—OK (scanned 5648 tracked text file(s); skipped 85 binary).check-changeset-presence—25 source file(s) of 6 released package(s) changed, and this change declares 1 changeset(s) ... Every one of them has an EMPTY frontmatter — declared as releasing nothing, which is the explicit exemption and a complete answer to this gate.check-changeset-no-major,check-changeset-fixed,check-changeset-overwrite— exit 0.Declared narrowing: repo-wide
pnpm lintwas not run; eslint was run on the 25 changed files only (25/25 linted by eslint's own config, count read from--format json, 0 errors). This narrowing excludes nothing:eslint.config.jsenables no type-aware linting (noproject/projectService), so a test-file diff cannot move the verdict of any file it does not touch. CI runs the full farm regardless.Changeset
Empty frontmatter — the gate's own printed verdict calls that "the explicit exemption and a complete answer". Nothing published changes: no source file is touched.
Generated by Claude Code
Generated by Claude Code