node Puppteer

1. 安装 Puppteer

	npm install puppteer -S 

2. 引入并使用 Puppteer

const puppeteer = require("puppeteer");

(async () => {
  const browser = await puppeteer.launch(); // 创建无头浏览器
  const page = await browser.newPage();
  await page.goto('https://example.com'); // 打开页面地址
  await page.screenshot({ path: 'example.png' }); // 拍照
  await browser.close(); // 关闭浏览器
})();

3. 获取页面数据

const puppeteer = require("puppeteer");
class Robot {
  constructor() {
    this.username = ''
    this.password = ''
  }
  async init() {  // 初始化 Puppteer 
    const browser = await puppeteer.launch({ timeout: 10000 });
    this.browser = browser;
    this.page = await browser.newPage();
    await this.page.setViewport({
      width: 1100,
      height: 800
    });
    await this._login();
  }
  async _login() {
    await this.open(LOGIN_URL); // LOGIN_URL 需要打开的页面地址
    await this.page.evaluate(() => {
      fetch('login相关地址', {
        method: 'POST',
        mode: 'cors',
        credentials: 'include',
        headers: {
          'Content-Type': 'application/json;charset=utf-8'
        },
        body: JSON.stringify({ username: this.username, password : this.password })
      })
    })
  }
  async open(url) { // 打开页面
    await this.page.goto(url, { waitUntil: 'domcontentloaded' }); // 等待没有网络请求
  }
  async requestInterception() {  // 截获页面网络请求
    await this.page.setRequestInterception(true);
    return this.page
  }
  async waitForSelector(selector) { // 等待页面某个元素加载出现
    return await this.page.waitForSelector(selector, { timeout: 10000 })
  }
  async getHtml() { // 获取当前页面的html 字符串
    return await this.page.content()
  }
  async getContentByClass(name) { // 通过 className 获取页面上多个同名元素
    let pages = await this.page.$$eval(name, e => {
      let arr = []
      e.map((item,index) => {
        if (item.className.includes(name)) {
          let obj = {
            content: item.innerHTML,
            pageNum: index + 1
          }
          arr.push(obj)
        }
      })
      return arr
    });
    return pages
  }
  async getContentClass(name) { // 获取页面上的单个元素
    let pages = await this.page.$eval(name, e => e.innerHTML);
    return pages
  }
  async screenshot(option) { // 拍照
    return await this.page.screenshot(option)
  }
  async exportPdf(option) {  // 把当前页面生成pdf 根据 options 配置
    return await this.page.pdf(option)
  }
  async close() { // 关闭浏览器
    await this.browser.close()
  }
}

module.exports = Robot

具体 Puppteer api 请访问以下地址

Puppteer 中文文档