Mock Data

Mocking data is a significant part of web application front end development which is the key of separating front-end and back-end. We can mock Restful API of background service to avoid development block of service side development progress.

We can use umi to mock data in Ant Design Pro.

Use umi#

In umi, files in mock directory means mocking files, which export definitions of APIs. It supports realtime refreshment and ES6, and has friendly errors prompts. More details in

export default {
  // literal Object and Array
  'GET /api/users': { users: [1, 2] },

  // GET could be omitted
  '/api/users/1': { id: 1 },

  // implement simple API by express@4
  'POST /api/users/create': (req, res) => {

  // Forward to another server
  // `/assets/0.0.50/index.css` will be forword
  'GET /assets/*': '',

  // Forward to another server with glob matcher
  // `/someDir/0.0.50/index.css` will be forword to
  'GET /someDir/(.*)': '',

When you start server by umi dev, the mock service will handle network requests (such as GET /api/users) according to your rules defined in the mock directory. You can respond to client with literal data, or process the request with a function, or forward the request to another server.

For example, there is a mapping rule like this:

'GET /api/currentUser': {
  name: 'momo.zxy',
  avatar: imgMap.user,
  userid: '00000001',
  notifyCount: 12,

Now you can see the result in browser console when visits /api/currentUser:



Using Mock.js#

Mock.js is a popular mock library that can help to generate mock data. You can use any library you like along with umi to build mocking service.

import mockjs from 'mockjs';

export default {
  'GET /api/tags': mockjs.mock({
    'list|100': [{ name: '@city', 'value|1-100': 50, 'type|0-2': 1 }],

CORS support#

Define header of response like this:

'POST /api/users/create': (req, res) => {
  res.setHeader('Access-Control-Allow-Origin', '*');

Divide mock data into different files#

We can maintain our mock data in different files for different data models in the large application, and put them in the mock directory, then they will be imported automatically.

Delay mock API#

In real world the AJAX request usually responds with a network delay which should be simulated in mocking API.


You can implement the API with setTimeout.

'POST /api/forms': (req, res) => {
  setTimeout(() => {
  }, 1000);

Use roadhog-api-doc plugin#

We provide delay function in roadhog-api-doc to make this convenient.

import { delay } from 'roadhog-api-doc';

const proxy = {
  'GET /api/project/notice': getNotice,
  'GET /api/activities': getActivities,
  'GET /api/rule': getRule,
  'GET /api/tags': mockjs.mock({
    'list|100': [{ name: '@city', 'value|1-100': 50, 'type|0-2': 1 }],
  'GET /api/fake_list': getFakeList,
  'GET /api/fake_chart_data': getFakeChartData,
  'GET /api/profile/basic': getProfileBasicData,
  'GET /api/profile/advanced': getProfileAdvancedData,
  'POST /api/register': (req, res) => {
    res.send({ status: 'ok' });
  'GET /api/notices': getNotices,

// Delay one second for all interfaces
export default delay(proxy, 1000);


If you need to integrate with real service after finishing front end development via mock data, just close the mock data or forward request to the real interfaces.

$ npm run start:no-mock