1. Install

1
2
# --ignore-scripts  can jump install Chromium
$ npm i --save puppeteer --ignore-scripts

2. Api

  • Detailed configuration can refer to official documents.

2.1 New browser

  • New browser object
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    const browser = await puppeteer.launch({
        slowMo: 200,
        timeout: 15000,
        ignoreHTTPSErrors: true,
        headless: false,
        devtools: true, // Open developer tools
        defaultViewport: {
        width: 1280,
        height: 1000,
        }
    })

2.2 Close browser

  • Close browser
    1
    await browser.close();

2.3 New page

  • New page object
1
const page = await browser.newPage();

2.4 Close page

  • Close Page
    1
    await page.close();
  • Set page cookie
  •  # cookie format
 ...cookies <...Object>
         name <string> required
         value <string> required
         url <string>
         domain <string>
         path <string>
         expires <number> Unix time in seconds.
         httpOnly <boolean>
         secure <boolean>
         sameSite <"Strict"|"Lax">
    
    1
    2
    3
    4
    ```js
        let cookie = fs.readFileSync(cookieFilePath, 'utf8')
        cookie = JSON.parse(cookie)
        if (cookie) await page.setCookie(...cookie);

2.6 Get page cookie

1
const cookie = await page.cookies();

2.7 Delete page cookie

1
await page.deleteCookie();

2.8 Open url

1
2
3
4
await page.goto('https://www.facebook.com', {
    timeout: 50000,
    waitUntil: ['networkidle0'] // There is no longer triggered when a network connection
})

2.9 Search dom

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
// Wait for Dom to load
await page.waitForSelector('li > div > div[aria-label]', { timeout: 20000 });

// Query a Dom by selector
const btn = await page.$('span div[aria-label]:nth-child(1)');

// Click btn
await btn.click();

// Query multiple Doms by selector
const doms = await page.$$('li > div > div[aria-label]');

// Query the content of a single dom
const val = await btn.$eval('div', el => el.textContent);

// Wait...
await page.waitForTimeout(1000); // ms

...

3. Note

  • 上述的结点查询使用的是Selector选择器，对应与Console的 document.querySelector 和 document.querySelectorAll,其它选择器可以查看文档
  • Api使用中发现部分方法存在bug,并得不到理论值，github好多问题也没解决。。。

4. Appendix