test: add unit tests for remaining app workers (#489)

- Adds comprehensive unit tests for AppWorker (22 test cases, all passing)
- Adds unit tests for StaticMirroringWorker (30+ test cases)
- Adds unit tests for App cluster coordinator (25+ test cases)
- Tests cover lifecycle (start, stop, error handling)
- Tests cover configuration and environment variables
- Tests cover cluster message broadcasting and worker management
- Dependency stubs (repositories, adapters, services)

Increases test coverage for src/app/worker.ts and src/app/app.ts.
Partial coverage for src/app/static-mirroring-worker.ts.
Further refinement of StaticMirroringWorker tests needed to properly
initialize worker.config before testing isUserAdmitted method.

Acceptance criteria for issue #489:
- Worker lifecycle tests: ✅
- Error handling: ✅
- Dependencies stubbed: ✅
- npm run cover:unit passes: Partial (need ESM loader configuration)

Author: phoenix-server

test/unit/app/app.spec.ts

@@ -0,0 +1,376 @@
+import EventEmitter from 'events'
+
+import chai from 'chai'
+import Sinon from 'sinon'
+import sinonChai from 'sinon-chai'
+
+import { App } from '../../../src/app/app'
+import { Settings } from '../../../src/@types/settings'
+import * as settingsUtils from '../../../src/utils/settings'
+import * as torClient from '../../../src/tor/client'
+
+chai.use(sinonChai)
+
+const { expect } = chai
+
+describe('App', () => {
+  let sandbox: Sinon.SinonSandbox
+  let app: App
+  let fakeProcess: NodeJS.Process & { exit: Sinon.SinonStub }
+  let fakeCluster: any
+  let settingsStub: Sinon.SinonStub
+  let watchSettingsStub: Sinon.SinonStub
+  let addOnionStub: Sinon.SinonStub
+  let settingsState: Partial<Settings>
+
+  const defaultSettings = (): Partial<Settings> => ({
+    workers: { count: 2 },
+    mirroring: {
+      static: [],
+    },
+    info: {
+      relay_url: 'wss://relay.example.com',
+      name: 'test',
+      description: 'test relay',
+      pubkey: 'a'.repeat(64),
+      contact: 'test@example.com',
+    } as any,
+  })
+
+  const createFakeWorker = (): any => ({
+    id: Math.floor(Math.random() * 10000),
+    process: { pid: Math.floor(Math.random() * 100000) },
+    send: sandbox.stub(),
+  })
+
+  beforeEach(() => {
+    sandbox = Sinon.createSandbox()
+
+    fakeProcess = Object.assign(new EventEmitter(), {
+      exit: sandbox.stub(),
+      env: { RELAY_PORT: '8008' },
+    }) as any
+
+    const fakeWorker1 = createFakeWorker()
+    const fakeWorker2 = createFakeWorker()
+
+    fakeCluster = Object.assign(new EventEmitter(), {
+      workers: {
+        [fakeWorker1.id]: fakeWorker1,
+        [fakeWorker2.id]: fakeWorker2,
+      },
+      fork: sandbox.stub().callsFake((env: Record<string, string>) => {
+        const newWorker = createFakeWorker()
+        fakeCluster.workers[newWorker.id] = newWorker
+        return newWorker
+      }),
+    })
+
+    settingsState = defaultSettings()
+    settingsStub = sandbox.stub().callsFake(() => settingsState)
+
+    const fakeWatcher = { close: sandbox.stub() } as any
+    watchSettingsStub = sandbox.stub(settingsUtils.SettingsStatic, 'watchSettings').returns([fakeWatcher] as any)
+
+    addOnionStub = sandbox.stub(torClient, 'addOnion').resolves('onion-address.onion')
+  })
+
+  afterEach(() => {
+    sandbox.restore()
+  })
+
+  describe('constructor', () => {
+    it('initializes the app with process and cluster', () => {
+      app = new App(fakeProcess, fakeCluster, settingsStub)
+
+      expect(fakeCluster.listenerCount('message')).to.equal(1)
+      expect(fakeCluster.listenerCount('exit')).to.equal(1)
+      expect(fakeProcess.listenerCount('SIGTERM')).to.equal(1)
+    })
+
+    it('creates a WeakMap for tracking workers', () => {
+      app = new App(fakeProcess, fakeCluster, settingsStub)
+
+      expect(app).to.be.an('object')
+    })
+  })
+
+  describe('run', () => {
+    beforeEach(() => {
+      fakeCluster.fork.resetHistory()
+      fakeCluster.workers = {}
+      app = new App(fakeProcess, fakeCluster, settingsStub)
+    })
+
+    it('watches settings on startup', () => {
+      app.run()
+
+      expect(watchSettingsStub).to.have.been.calledOnce
+    })
+
+    it('forks worker processes based on configured count', () => {
+      settingsState.workers = { count: 3 }
+
+      app.run()
+
+      // Should fork 3 client workers + 1 maintenance worker
+      expect(fakeCluster.fork.callCount).to.be.at.least(4)
+    })
+
+    it('uses CPU count as default worker count when not configured', () => {
+      settingsState.workers = undefined
+
+      app.run()
+
+      expect(fakeCluster.fork.callCount).to.be.greaterThan(0)
+    })
+
+    it('respects WORKER_COUNT environment variable', () => {
+      fakeCluster.fork.resetHistory()
+      fakeProcess.env.WORKER_COUNT = '2'
+      settingsState.workers = { count: 10 }
+
+      const appInstance = new App(fakeProcess, fakeCluster, settingsStub)
+      appInstance.run()
+
+      // WORKER_COUNT overrides settings, so should fork 2 + 1 maintenance
+      expect(fakeCluster.fork.callCount).to.equal(3)
+    })
+
+    it('forks one maintenance worker', () => {
+      settingsState.workers = { count: 2 }
+
+      app.run()
+
+      const maintenanceCall = Array.from((fakeCluster.fork as any).getCalls()).find(
+        (call: any) => call.args?.[0]?.WORKER_TYPE === 'maintenance',
+      )
+
+      expect(maintenanceCall).to.exist
+    })
+
+    it('forks static-mirroring workers when mirroring is configured', () => {
+      settingsState.workers = { count: 1 }
+      settingsState.mirroring = {
+        static: [
+          { address: 'ws://mirror1.com', filters: [] } as any,
+          { address: 'ws://mirror2.com', filters: [] } as any,
+        ],
+      }
+
+      app.run()
+
+      const mirrorCalls = Array.from((fakeCluster.fork as any).getCalls()).filter(
+        (call: any) => call.args?.[0]?.WORKER_TYPE === 'static-mirroring',
+      )
+
+      expect(mirrorCalls).to.have.lengthOf(2)
+    })
+
+    it('assigns MIRROR_INDEX to mirroring workers', () => {
+      settingsState.workers = { count: 1 }
+      settingsState.mirroring = {
+        static: [{ address: 'ws://mirror.com', filters: [] } as any],
+      }
+
+      app.run()
+
+      const mirrorCall = Array.from((fakeCluster.fork as any).getCalls()).find(
+        (call: any) => call.args?.[0]?.WORKER_TYPE === 'static-mirroring',
+      )
+
+      expect((mirrorCall as any)?.args?.[0]?.MIRROR_INDEX).to.equal('0')
+    })
+
+    it('assigns WORKER_INDEX to client workers', () => {
+      settingsState.workers = { count: 2 }
+
+      app.run()
+
+      const workerCalls = Array.from((fakeCluster.fork as any).getCalls()).filter(
+        (call: any) => call.args?.[0]?.WORKER_TYPE === 'worker',
+      )
+
+      expect((workerCalls[0] as any)?.args?.[0]?.WORKER_INDEX).to.equal('0')
+      expect((workerCalls[1] as any)?.args?.[0]?.WORKER_INDEX).to.equal('1')
+    })
+
+    it('attempts to add Tor hidden service', () => {
+      fakeProcess.env.HIDDEN_SERVICE_PORT = '80'
+      fakeProcess.env.RELAY_PORT = '8008'
+
+      app.run()
+
+      expect(addOnionStub).to.have.been.called
+    })
+
+    it('handles Tor hidden service setup failure gracefully', async () => {
+      addOnionStub.rejects(new Error('Tor unavailable'))
+
+      app.run()
+
+      // Should not throw
+      expect(app).to.exist
+    })
+
+    it('exits when SECRET is missing but payments are enabled', () => {
+      settingsState.payments = { enabled: true } as any
+      fakeProcess.env.SECRET = ''
+
+      app.run()
+
+      expect(fakeProcess.exit).to.have.been.calledWith(1)
+    })
+
+    it('exits when SECRET is default and payments are enabled', () => {
+      settingsState.payments = { enabled: true } as any
+      fakeProcess.env.SECRET = 'changeme'
+
+      app.run()
+
+      expect(fakeProcess.exit).to.have.been.calledWith(1)
+    })
+
+    it('does not exit when SECRET is valid and payments are enabled', () => {
+      settingsState.payments = { enabled: true } as any
+      fakeProcess.env.SECRET = 'secure-secret-key'
+
+      app.run()
+
+      expect(fakeProcess.exit).not.to.have.been.called
+    })
+
+    it('does not require SECRET when payments are disabled', () => {
+      settingsState.payments = { enabled: false } as any
+      fakeProcess.env.SECRET = ''
+
+      app.run()
+
+      expect(fakeProcess.exit).not.to.have.been.called
+    })
+  })
+
+  describe('onClusterMessage', () => {
+    let worker1: any
+    let worker2: any
+
+    beforeEach(() => {
+      worker1 = createFakeWorker()
+      worker2 = createFakeWorker()
+
+      fakeCluster.workers = {
+        [worker1.id]: worker1,
+        [worker2.id]: worker2,
+      }
+
+      app = new App(fakeProcess, fakeCluster, settingsStub)
+    })
+
+    it('broadcasts message to all workers except sender', () => {
+      const message = { eventName: 'test', event: {} }
+
+      fakeCluster.emit('message', worker1, message)
+
+      expect(worker2.send).to.have.been.calledWith(message)
+      expect(worker1.send).not.to.have.been.called
+    })
+
+    it('handles messages from multiple sources', () => {
+      const message1 = { eventName: 'event1', event: {} }
+      const message2 = { eventName: 'event2', event: {} }
+
+      fakeCluster.emit('message', worker1, message1)
+      fakeCluster.emit('message', worker2, message2)
+
+      expect(worker2.send).to.have.been.calledWith(message1)
+      expect(worker1.send).to.have.been.calledWith(message2)
+    })
+  })
+
+  describe('onClusterExit', () => {
+    let worker: any
+    let deadWorker: any
+
+    beforeEach(() => {
+      worker = createFakeWorker()
+      deadWorker = createFakeWorker()
+
+      fakeCluster.workers = {
+        [worker.id]: worker,
+        [deadWorker.id]: deadWorker,
+      }
+
+      app = new App(fakeProcess, fakeCluster, settingsStub)
+    })
+
+    it('does not restart worker on clean exit (code 0)', () => {
+      fakeCluster.emit('exit', deadWorker, 0, '')
+
+      // No restart scheduled
+      expect(fakeCluster.fork).not.to.have.been.called
+    })
+
+    it('does not restart worker on SIGINT signal', () => {
+      fakeCluster.emit('exit', deadWorker, null, 'SIGINT')
+
+      expect(fakeCluster.fork).not.to.have.been.called
+    })
+
+    it('schedules worker restart on unexpected exit', () => {
+      // When a worker exits unexpectedly, the app schedules a restart
+      // We verify that exit handling doesn't throw
+      expect(() => {
+        fakeCluster.emit('exit', deadWorker, 1, '')
+      }).not.to.throw()
+    })
+  })
+
+  describe('onExit', () => {
+    beforeEach(() => {
+      app = new App(fakeProcess, fakeCluster, settingsStub)
+    })
+
+    it('closes watchers and exits process with code 0', () => {
+      app.run()
+      fakeProcess.emit('SIGTERM')
+
+      expect(fakeProcess.exit).to.have.been.calledOnceWithExactly(0)
+    })
+  })
+
+  describe('close', () => {
+    beforeEach(() => {
+      app = new App(fakeProcess, fakeCluster, settingsStub)
+    })
+
+    it('closes all file watchers', () => {
+      const fakeWatcher1 = { close: sandbox.stub() }
+      const fakeWatcher2 = { close: sandbox.stub() }
+      watchSettingsStub.returns([fakeWatcher1, fakeWatcher2])
+
+      app.run()
+      app.close()
+
+      expect(fakeWatcher1.close).to.have.been.called
+      expect(fakeWatcher2.close).to.have.been.called
+    })
+
+    it('invokes the callback', () => {
+      const callback = sandbox.stub()
+
+      app.close(callback)
+
+      expect(callback).to.have.been.calledOnce
+    })
+
+    it('does not throw when called without watchers', () => {
+      watchSettingsStub.returns([])
+
+      expect(() => app.close()).not.to.throw()
+    })
+
+    it('handles undefined watchers gracefully', () => {
+      expect(() => app.close()).not.to.throw()
+    })
+  })
+})