Index discovered test files once during shard selection validation (#24532)

Verified selection plans previously validated each selected filename with a scan of all discovered files. One per-call Set now handles membership checks. Exact source/discovery checks, nonempty selection, fallback to all tests, shard balancing and manifests remain intact.

config/scripts/ci-unit-sequencer.mjs

@@ -16,13 +16,14 @@ export default class TimingSequencer extends BaseSequencer {
           'utf8'
         )
       )
+      const discovered = new Set(plan.files)
       if (
         plan.version !== 1 ||
         !plan.sourceSha ||
         plan.sourceSha !== process.env.ORCA_SHARD_SOURCE_SHA ||
         JSON.stringify([...plan.files].sort()) !== JSON.stringify(specs.map(key).sort()) ||
         !Array.isArray(plan.executionFiles) ||
-        plan.executionFiles.some((file) => !plan.files.includes(file))
+        plan.executionFiles.some((file) => !discovered.has(file))
       ) {
         throw new Error('Selection provenance or discovery differs')
       }

config/scripts/ci-unit-sequencer.test.mjs

@@ -6,39 +6,86 @@ import TimingSequencer from './ci-unit-sequencer.mjs'
 
 let root
 afterEach(() => {
+  vi.restoreAllMocks()
   vi.unstubAllEnvs()
   if (root) {
     rmSync(root, { recursive: true, force: true })
   }
 })
 
-it.each(['valid', 'stale', 'missing-file', 'missing-artifact'])(
-  'preserves complete shard coverage with %s planning evidence',
-  async (kind) => {
-    root = mkdtempSync(join(tmpdir(), 'unit-sequencer-'))
-    const files = ['src/a.test.ts', 'src/b.test.ts', 'src/c.test.ts', 'src/d.test.ts']
-    const plan = {
-      version: 1,
-      sourceSha: kind === 'stale' ? 'old' : 'current',
-      files: kind === 'missing-file' ? files.slice(1) : files,
-      executionFiles: files.slice(0, 2)
-    }
-    const planPath = join(root, 'selection.json')
-    if (kind !== 'missing-artifact') {
-      writeFileSync(planPath, JSON.stringify(plan))
-    }
-    vi.stubEnv('ORCA_UNIT_SELECTION_PLAN', planPath)
-    vi.stubEnv('ORCA_SHARD_SOURCE_SHA', 'current')
-    vi.stubEnv('ORCA_SHARD_MANIFEST', join(root, 'assignment.json'))
-    const assigned = []
-    for (const index of [1, 2]) {
-      const sequencer = new TimingSequencer({ config: { root, shard: { index, count: 2 } } })
-      const specs = files.map((file) => ({ moduleId: join(root, file) }))
-      assigned.push(...(await sequencer.shard(specs)).map((spec) => spec.moduleId))
-    }
-    expect(assigned.sort()).toEqual(
-      (kind === 'valid' ? files.slice(0, 2) : files).map((file) => join(root, file)).sort()
-    )
-    expect(new Set(assigned).size).toBe(assigned.length)
+it.each([
+  'valid',
+  'stale',
+  'missing-file',
+  'missing-artifact',
+  'outside-selection',
+  'empty-selection'
+])('preserves complete shard coverage with %s planning evidence', async (kind) => {
+  root = mkdtempSync(join(tmpdir(), 'unit-sequencer-'))
+  const files = ['src/a.test.ts', 'src/b.test.ts', 'src/c.test.ts', 'src/d.test.ts']
+  const plan = {
+    version: 1,
+    sourceSha: kind === 'stale' ? 'old' : 'current',
+    files: kind === 'missing-file' ? files.slice(1) : files,
+    executionFiles:
+      kind === 'outside-selection'
+        ? ['src/unknown.test.ts']
+        : kind === 'empty-selection'
+          ? []
+          : files.slice(0, 2)
   }
-)
+  const planPath = join(root, 'selection.json')
+  if (kind !== 'missing-artifact') {
+    writeFileSync(planPath, JSON.stringify(plan))
+  }
+  vi.stubEnv('ORCA_UNIT_SELECTION_PLAN', planPath)
+  vi.stubEnv('ORCA_SHARD_SOURCE_SHA', 'current')
+  vi.stubEnv('ORCA_SHARD_MANIFEST', join(root, 'assignment.json'))
+  const assigned = []
+  for (const index of [1, 2]) {
+    const sequencer = new TimingSequencer({ config: { root, shard: { index, count: 2 } } })
+    const specs = files.map((file) => ({ moduleId: join(root, file) }))
+    assigned.push(...(await sequencer.shard(specs)).map((spec) => spec.moduleId))
+  }
+  expect(assigned.sort()).toEqual(
+    (kind === 'valid' ? files.slice(0, 2) : files).map((file) => join(root, file)).sort()
+  )
+  expect(new Set(assigned).size).toBe(assigned.length)
+})
+
+it('validates a large selection without scanning the discovered array for each file', async () => {
+  root = mkdtempSync(join(tmpdir(), 'unit-sequencer-scale-'))
+  const files = Array.from({ length: 1600 }, (_, index) => `src/scale-${index}.test.ts`)
+  const executionFiles = files.slice(800)
+  const planPath = join(root, 'selection.json')
+  writeFileSync(
+    planPath,
+    JSON.stringify({ version: 1, sourceSha: 'current', files, executionFiles })
+  )
+  vi.stubEnv('ORCA_UNIT_SELECTION_PLAN', planPath)
+  vi.stubEnv('ORCA_SHARD_SOURCE_SHA', 'current')
+  vi.stubEnv('ORCA_SHARD_MANIFEST', join(root, 'assignment.json'))
+  const sequencer = new TimingSequencer({ config: { root, shard: { index: 1, count: 1 } } })
+  const specs = files.map((file) => ({ moduleId: join(root, file) }))
+  const includes = Array.prototype.includes
+  let discoveredArrayScans = 0
+  const scan = vi
+    .spyOn(Array.prototype, 'includes')
+    .mockImplementation(function (value, fromIndex) {
+      if (
+        this.length === files.length &&
+        this[0] === files[0] &&
+        typeof value === 'string' &&
+        value.startsWith('src/scale-')
+      ) {
+        discoveredArrayScans += 1
+      }
+      return includes.call(this, value, fromIndex)
+    })
+  const selected = await sequencer.shard(specs)
+  scan.mockRestore()
+  expect(selected.map((spec) => spec.moduleId).sort()).toEqual(
+    executionFiles.map((file) => join(root, file)).sort()
+  )
+  expect(discoveredArrayScans).toBe(0)
+})