WebdriverIO中是否有ScenarioContext？如何跨测试场景存取值？

Great question! WebdriverIO doesn’t come with a built-in ScenarioContext like the one you have in Java, but it’s straightforward to build an equivalent (or even better) solution using JavaScript’s features. Let’s break this down step by step.

Implementing a Custom ScenarioContext #

First, we’ll create a reusable module that mimics the HashMap-based context from your Java code. We’ll use a Map (JavaScript’s equivalent of HashMap) and export a singleton instance to ensure data is shared across all test files.

1. Create a scenarioContext.js file:

class ScenarioContext {
  constructor() {
    this.scenarioContext = new Map();
  }

  // Set a value for a given key
  setContext(key, value) {
    this.scenarioContext.set(key.toString(), value);
  }

  // Retrieve a value by key
  getContext(key) {
    return this.scenarioContext.get(key.toString());
  }

  // Check if a key exists in the context
  isContains(key) {
    return this.scenarioContext.has(key.toString());
  }
}

// Export a singleton instance so all imports use the same context
module.exports = new ScenarioContext();

2. (Optional) Define a Context "enum" to avoid magic strings (just like your Java Context enum):

// contextKeys.js
const Context = {
  USER_ID: 'USER_ID',
  AUTH_TOKEN: 'AUTH_TOKEN',
  CURRENT_PAGE: 'CURRENT_PAGE'
};

module.exports = Context;

Using the Context in Your Tests #

Now you can import this context into any test file to set and retrieve data across scenarios.

Example 1: Setting data in one test #

// login.spec.js
const scenarioContext = require('./scenarioContext');
const Context = require('./contextKeys');

describe('Login Flow', () => {
  it('should log in and store user ID', async () => {
    // Perform login steps...
    const userId = await $('[data-testid="user-id"]').getText();
    
    // Store the user ID in the context
    scenarioContext.setContext(Context.USER_ID, userId);
    
    // Verify the key exists
    expect(scenarioContext.isContains(Context.USER_ID)).toBe(true);
  });
});

Example 2: Reusing data in another test #

// userProfile.spec.js
const scenarioContext = require('./scenarioContext');
const Context = require('./contextKeys');

describe('User Profile', () => {
  it('should retrieve stored user ID and validate profile', async () => {
    // Get the user ID from the context
    const userId = scenarioContext.getContext(Context.USER_ID);
    
    // Use the ID to navigate to the profile page
    await browser.url(`/profile/${userId}`);
    
    // Validate the profile displays the correct ID
    const displayedId = await $('[data-testid="profile-user-id"]').getText();
    expect(displayedId).toBe(userId);
  });
});

Handling Parallel Test Execution #

If you run tests in parallel, the singleton ScenarioContext will cause data leaks between test threads (since Node.js modules are process-wide). For this case, use WebdriverIO’s built-in browser.sharedStore (available in v7+), which is isolated to each browser instance:

// Setting data with sharedStore
await browser.sharedStore.set(Context.USER_ID, userId);

// Retrieving data
const userId = await browser.sharedStore.get(Context.USER_ID);

// Checking if a key exists
const hasUserId = await browser.sharedStore.has(Context.USER_ID);

This ensures each test run gets its own isolated context, preventing cross-test contamination.

内容的提问来源于stack exchange，提问作者Mitali S