:hedgehog: PostHog is the leading platform for building self-driving products. Our developer tools – AI observability, analytics, session replay, flags, experiments, error tracking, logs, and more – capture all the context agents need to diagnose problems, uncover opportunities, and ship fixes. Steer it all from Slack, web, desktop, or the MCP.
Python ab-testingai-analyticsanalyticscdpdata-warehouseexperimentsfeature-flagsjavascriptproduct-analyticspythonreactsession-replaysurveystypescriptweb-analytics
import { expectLogic } from 'kea-test-utils'
import { dispatchChangeRequestCreated } from 'scenes/approvals/utils'
import { useMocks } from '~/mocks/jest'
import { initKeaTests } from '~/test/init'
import { BULK_ARCHIVE_MAX_FLAGS, flagSelectionLogic } from './flagSelectionLogic'
jest.mock('scenes/approvals/utils', () => ({
...jest.requireActual('scenes/approvals/utils'),
dispatchChangeRequestCreated: jest.fn(),
}))
describe('flagSelectionLogic bulk archive', () => {
let logic: ReturnType<typeof flagSelectionLogic.build>
let archiveRequests: { id: number; body: Record<string, any> }[]
function useArchiveMocks(archiveHandler: (id: number) => [number, any] = () => [200, {}]): void {
archiveRequests = []
useMocks({
get: {
'/api/projects/:team/feature_flags/': { count: 0, results: [] },
},
patch: {
'/api/projects/:team/feature_flags/:id/': async ({ request, params }) => {
const id = Number(params.id)
archiveRequests.push({ id, body: (await request.json()) as Record<string, any> })
return archiveHandler(id)
},
},
})
}
beforeEach(() => {
jest.clearAllMocks()
initKeaTests()
})
afterEach(() => logic?.unmount())
it('disables every selected flag as it archives it, then reloads the list', async () => {
useArchiveMocks()
logic = flagSelectionLogic()
logic.mount()
logic.actions.bulkArchiveFlags([1, 2])
await expectLogic(logic).toDispatchActions(['bulkArchiveFlagsFinished', 'loadFeatureFlags'])
// `active: false` matters: the API rejects an archived flag that is still enabled
expect(archiveRequests).toEqual([
{ id: 1, body: { archived: true, active: false } },
{ id: 2, body: { archived: true, active: false } },
])
expect(logic.values.bulkArchiveResult).toEqual({ archivedIds: [1, 2], failed: [] })
expect(logic.values.bulkArchiveProgress).toEqual({ done: 2, total: 2 })
expect(logic.values.bulkArchiveRunning).toBe(false)
})
it('keeps archiving the remaining flags when one needs approval and another fails', async () => {
useArchiveMocks((id) => {
if (id === 1) {
return [409, { change_request_id: 'cr-1' }]
}
if (id === 2) {
return [500, { detail: 'Server error' }]
}
return [200, {}]
})
logic = flagSelectionLogic()
logic.mount()
logic.actions.bulkArchiveFlags([1, 2, 3])
await expectLogic(logic).toFinishAllListeners()
expect(archiveRequests.map(({ id }) => id)).toEqual([1, 2, 3])
expect(logic.values.bulkArchiveResult?.archivedIds).toEqual([3])
expect(logic.values.bulkArchiveResult?.failed).toEqual([
{ id: 1, errorMessage: expect.any(String), approvalPending: true },
{ id: 2, errorMessage: 'Server error', approvalPending: false },
])
// Only the approval-pending flag announces a change request; the 500 failure must not
expect(dispatchChangeRequestCreated).toHaveBeenCalledTimes(1)
expect(dispatchChangeRequestCreated).toHaveBeenCalledWith({ resourceType: 'feature_flag', resourceId: 1 })
// Failed flags still advance the progress counter, and the run ends
expect(logic.values.bulkArchiveProgress).toEqual({ done: 3, total: 3 })
expect(logic.values.bulkArchiveRunning).toBe(false)
})
it.each([
['an empty selection', []],
['a selection above the cap', Array.from({ length: BULK_ARCHIVE_MAX_FLAGS + 1 }, (_, i) => i + 1)],
])('refuses to run for %s without issuing any archive requests', async (_label, ids) => {
useArchiveMocks()
logic = flagSelectionLogic()
logic.mount()
logic.actions.bulkArchiveFlags(ids)
await expectLogic(logic).toFinishAllListeners()
expect(archiveRequests).toHaveLength(0)
expect(logic.values.bulkArchiveResult).toBeNull()
expect(logic.values.bulkArchiveRunning).toBe(false)
})
})