dify/web/service/debug.spec.ts

import type { AppModeEnum } from '@/types/app'
// service/base is the dependency we're mocking in this test; the
// no-restricted-imports rule targets production imports, not test
// instrumentation — mirrors sibling service specs (annotation.spec.ts etc.).
// oxlint-disable-next-line no-restricted-imports
import { get, post, ssePost } from './base'
import { consoleClient, streamWorkflowGeneration } from './client'
import {
  fetchConversationMessages,
  fetchPromptTemplate,
  fetchSuggestedQuestions,
  fetchTextGenerationMessage,
  generateBasicAppFirstTimeRule,
  generateRule,
  sendCompletionMessage,
  stopChatMessageResponding,
} from './debug'
import {
  fetchWorkflowInstructionSuggestions,
  generateWorkflow,
  generateWorkflowStream,
} from './workflow-generator'

// Stub the shared `post` wrapper so tests verify only what `generateWorkflow`
// composes on top of it — URL, body, and the typed response surface.
vi.mock('./base', () => ({
  post: vi.fn(),
  get: vi.fn(),
  ssePost: vi.fn(),
}))

vi.mock('./client', () => ({
  streamWorkflowGeneration: vi.fn(),
  consoleClient: {
    workflowGenerate: {
      post: vi.fn(),
      suggestions: {
        post: vi.fn(),
      },
    },
  },
}))

describe('debug service — generateWorkflow', () => {
  beforeEach(() => {
    vi.clearAllMocks()
  })

  // The new endpoint lives at /workflow-generate; the controller mirrors
  // /rule-generate so the body must flow through unchanged.
  it('should call the generated workflow contract with the body verbatim', () => {
    const body = {
      mode: 'workflow' as const,
      instruction: 'Summarize a URL',
      ideal_output: 'A 3-sentence summary.',
      model_config: {
        provider: 'openai',
        name: 'gpt-4o',
        mode: 'chat' as const,
        completion_params: {},
      },
    }

    generateWorkflow(body)

    expect(consoleClient.workflowGenerate.post).toHaveBeenCalledWith(
      { body },
      { signal: expect.any(AbortSignal) },
    )
  })

  // The optional fields must still POST cleanly — `ideal_output` defaulting
  // server-side requires the helper to forward the body as-is, not augment it.
  it('should pass the body through even when ideal_output is omitted', () => {
    const body = {
      mode: 'advanced-chat' as const,
      instruction: 'Friendly support bot',
      model_config: { provider: 'openai', name: 'gpt-4o', mode: 'chat' as const },
    }

    generateWorkflow(body)

    expect(consoleClient.workflowGenerate.post).toHaveBeenCalledWith(
      { body },
      { signal: expect.any(AbortSignal) },
    )
  })

  it('should expose the controller whose signal is passed to the generated client', () => {
    const body = {
      mode: 'workflow' as const,
      instruction: 'Long-running generation',
      model_config: { provider: 'openai', name: 'gpt-4o', mode: 'chat' as const },
    }
    const getAbortController = vi.fn()

    generateWorkflow(body, { getAbortController })

    const controller = getAbortController.mock.calls[0]![0]
    expect(consoleClient.workflowGenerate.post).toHaveBeenCalledWith(
      { body },
      { signal: controller.signal },
    )
  })

  it('should create an internal abort signal when no callback is provided', () => {
    const body = {
      mode: 'workflow' as const,
      instruction: 'Plain call',
      model_config: { provider: 'openai', name: 'gpt-4o', mode: 'chat' as const },
    }

    generateWorkflow(body)

    expect(consoleClient.workflowGenerate.post).toHaveBeenCalledWith(
      { body },
      { signal: expect.any(AbortSignal) },
    )
  })

  describe('other endpoints', () => {
    it('stopChatMessageResponding', async () => {
      await stopChatMessageResponding('app-1', 'task-1')
      expect(post).toHaveBeenCalledWith('apps/app-1/chat-messages/task-1/stop')
    })

    it('sendCompletionMessage', async () => {
      const callbacks = {
        onData: vi.fn(),
        onCompleted: vi.fn(),
        onError: vi.fn(),
        onMessageReplace: vi.fn(),
      }
      await sendCompletionMessage('app-1', { text: 'hello' }, callbacks)
      expect(ssePost).toHaveBeenCalledWith(
        'apps/app-1/completion-messages',
        {
          body: { text: 'hello', response_mode: 'streaming' },
        },
        callbacks,
      )
    })

    it('fetchSuggestedQuestions', async () => {
      const getAbortController = vi.fn()
      await fetchSuggestedQuestions('app-1', 'msg-1', getAbortController)
      expect(get).toHaveBeenCalledWith(
        'apps/app-1/chat-messages/msg-1/suggested-questions',
        {},
        { getAbortController },
      )
    })

    it('fetchConversationMessages', async () => {
      const getAbortController = vi.fn()
      await fetchConversationMessages('app-1', 'conv-1', getAbortController)
      expect(get).toHaveBeenCalledWith(
        'apps/app-1/chat-messages',
        { params: { conversation_id: 'conv-1' } },
        { getAbortController },
      )
    })

    it('generateBasicAppFirstTimeRule', async () => {
      await generateBasicAppFirstTimeRule({ mode: 'chat' })
      expect(post).toHaveBeenCalledWith('/rule-generate', { body: { mode: 'chat' } })
    })

    it('generateRule', async () => {
      await generateRule({ mode: 'chat' })
      expect(post).toHaveBeenCalledWith('/instruction-generate', { body: { mode: 'chat' } })
    })

    it('generateWorkflowStream', async () => {
      const body = {
        mode: 'workflow' as const,
        instruction: 'test',
        model_config: { provider: 'test', name: 'test', mode: 'chat' as const },
      }
      const callbacks = {
        onPlan: vi.fn(),
        onResult: vi.fn(),
        onError: vi.fn(),
        onCompleted: vi.fn(),
        getAbortController: vi.fn(),
      }

      vi.mocked(streamWorkflowGeneration).mockImplementation((_url, _body, options) => {
        options?.onPlan?.({ title: 'plan' })
        options?.onResult?.({ graph: { nodes: [], edges: [], viewport: { x: 0, y: 0, zoom: 1 } } })
        return Promise.resolve()
      })

      await generateWorkflowStream(body, callbacks)

      expect(streamWorkflowGeneration).toHaveBeenCalled()
      expect(callbacks.onPlan).toHaveBeenCalled()
      expect(callbacks.onResult).toHaveBeenCalled()
    })

    it('fetchWorkflowInstructionSuggestions without getAbortController', async () => {
      await fetchWorkflowInstructionSuggestions({ mode: 'workflow' })
      expect(consoleClient.workflowGenerate.suggestions.post).toHaveBeenCalledWith(
        { body: { mode: 'workflow' } },
        { signal: expect.any(AbortSignal) },
      )
    })

    it('fetchWorkflowInstructionSuggestions with getAbortController', async () => {
      const getAbortController = vi.fn()
      await fetchWorkflowInstructionSuggestions({ mode: 'workflow' }, { getAbortController })
      const controller = getAbortController.mock.calls[0]![0]
      expect(consoleClient.workflowGenerate.suggestions.post).toHaveBeenCalledWith(
        { body: { mode: 'workflow' } },
        { signal: controller.signal },
      )
    })

    it('fetchPromptTemplate', async () => {
      await fetchPromptTemplate({
        appMode: 'chat' as AppModeEnum,
        mode: 'chat',
        modelName: 'gpt-4',
        hasSetDataSet: true,
      })
      expect(get).toHaveBeenCalledWith('/app/prompt-templates', {
        params: {
          app_mode: 'chat',
          model_mode: 'chat',
          model_name: 'gpt-4',
          has_context: true,
        },
      })
    })

    it('fetchTextGenerationMessage', async () => {
      await fetchTextGenerationMessage({ appId: 'app-1', messageId: 'msg-1' })
      expect(get).toHaveBeenCalledWith('/apps/app-1/messages/msg-1')
    })
  })
})