feat(launcher): add browserUrl option to puppeteer.connect

docs/api.md

@@ -441,7 +441,8 @@ puppeteer.launch().then(async browser => {
 
 #### puppeteer.connect(options)
 - `options` <[Object]>
-  - `browserWSEndpoint` <[string]> a [browser websocket endpoint](#browserwsendpoint) to connect to.
+  - `browserWSEndpoint` <?[string]> a [browser websocket endpoint](#browserwsendpoint) to connect to.
+  - `browserUrl` <?[string]> a browser url to connect to, in format `http://${host}:${port}`. Use interchangeably with `browserWSEndpoint` to let Puppeteer fetch it from [metadata endpoint](https://chromedevtools.github.io/devtools-protocol/#how-do-i-access-the-browser-target).
   - `ignoreHTTPSErrors` <[boolean]> Whether to ignore HTTPS errors during navigation. Defaults to `false`.
   - `defaultViewport` <?[Object]> Sets a consistent viewport for each page. Defaults to an 800x600 viewport. `null` disables the default viewport.
     - `width` <[number]> page width in pixels.
@@ -3533,4 +3534,4 @@ TimeoutError is emitted whenever certain operations are terminated due to timeou
 [Worker]: #class-worker "Worker"
 [Accessibility]: #class-accessibility "Accessibility"
 [AXNode]: #accessibilitysnapshotoptions "AXNode"
-[ConnectionTransport]: ../lib/WebSocketTransport.js "ConnectionTransport"
+[ConnectionTransport]: ../lib/WebSocketTransport.js "ConnectionTransport"

lib/Launcher.js

@@ -15,6 +15,8 @@
  */
 const os = require('os');
 const path = require('path');
+const http = require('http');
+const URL = require('url');
 const removeFolder = require('rimraf');
 const childProcess = require('child_process');
 const BrowserFetcher = require('./BrowserFetcher');
@@ -274,18 +276,33 @@ class Launcher {
   }
 
   /**
-   * @param {!(Launcher.BrowserOptions & {browserWSEndpoint: string, transport?: !Puppeteer.ConnectionTransport})} options
+   * @param {!(Launcher.BrowserOptions & {browserWSEndpoint?: string, browserUrl?: string, transport?: !Puppeteer.ConnectionTransport})} options
    * @return {!Promise<!Browser>}
    */
   async connect(options) {
     const {
       browserWSEndpoint,
+      browserUrl,
       ignoreHTTPSErrors = false,
       defaultViewport = {width: 800, height: 600},
-      transport = await WebSocketTransport.create(browserWSEndpoint),
+      transport,
       slowMo = 0,
     } = options;
-    const connection = new Connection(browserWSEndpoint, transport, slowMo);
+
+    let connectionUrl;
+    let connectionTransport;
+
+    if (browserWSEndpoint)
+      connectionUrl = browserWSEndpoint;
+    else if (!browserWSEndpoint && browserUrl)
+      connectionUrl = await getWSEndpoint(browserUrl);
+
+    if (transport)
+      connectionTransport = transport;
+    else
+      connectionTransport = await WebSocketTransport.create(connectionUrl);
+
+    const connection = new Connection(connectionUrl, connectionTransport, slowMo);
     const {browserContextIds} = await connection.send('Target.getBrowserContexts');
     return Browser.create(connection, browserContextIds, ignoreHTTPSErrors, defaultViewport, null, () => connection.send('Browser.close').catch(debugError));
   }
@@ -373,6 +390,48 @@ function waitForWSEndpoint(chromeProcess, timeout, preferredRevision) {
   });
 }
 
+/**
+ * @param {string} browserUrl
+ * @return {!Promise<string>}
+ */
+function getWSEndpoint(browserUrl) {
+  let resolve, reject;
+  const endpointUrl = URL.resolve(browserUrl, '/json/version');
+  const requestOptions = Object.assign(URL.parse(endpointUrl), { method: 'GET' });
+  const promise = new Promise((res, rej) => { resolve = res; reject = rej; });
+
+  function handleRequestEnd(data) {
+    try {
+      const {webSocketDebuggerUrl} = JSON.parse(data);
+      resolve(webSocketDebuggerUrl);
+    } catch (e) {
+      handleRequestError(e);
+    }
+  }
+
+  function handleRequestError(err) {
+    reject(new Error(`Failed to fetch browser webSocket url from ${endpointUrl}: ${err}`));
+  }
+
+  const request = http.request(requestOptions, res => {
+    let data = '';
+    if (res.statusCode !== 200) {
+      // consume response data to free up memory
+      res.resume();
+      handleRequestError(res.statusCode);
+      return;
+    }
+    res.setEncoding('utf8');
+    res.on('data', chunk => data += chunk);
+    res.on('end', () => handleRequestEnd(data));
+  });
+
+  request.on('error', handleRequestError);
+  request.end();
+
+  return promise;
+}
+
 /**
  * @typedef {Object} Launcher.ChromeArgOptions
  * @property {boolean=} headless

lib/Puppeteer.js

@@ -37,7 +37,7 @@ module.exports = class {
   }
 
   /**
-   * @param {!(Launcher.BrowserOptions & {browserWSEndpoint: string, transport?: !Puppeteer.ConnectionTransport})} options
+   * @param {!(Launcher.BrowserOptions & {browserWSEndpoint?: string, browserUrl?: string, transport?: !Puppeteer.ConnectionTransport})} options
    * @return {!Promise<!Puppeteer.Browser>}
    */
   connect(options) {

test/puppeteer.spec.js

@@ -344,6 +344,24 @@ module.exports.addTests = function({testRunner, expect, defaultBrowserOptions})
         expect(await restoredPage.evaluate(() => 7 * 8)).toBe(56);
         await browser.close();
       });
+      it('should be able to connect using browserUrl, with and without trailing slash', async({server}) => {
+        const originalBrowser = await puppeteer.launch(Object.assign({}, defaultBrowserOptions, {
+          args: ['--remote-debugging-port=21222']
+        }));
+        const browserUrl = 'http://127.0.0.1:21222';
+
+        const browser1 = await puppeteer.connect({browserUrl});
+        const page1 = await browser1.newPage();
+        expect(await page1.evaluate(() => 7 * 8)).toBe(56);
+        browser1.disconnect();
+
+        const browser2 = await puppeteer.connect({browserUrl: browserUrl + '/'});
+        const page2 = await browser2.newPage();
+        expect(await page2.evaluate(() => 8 * 7)).toBe(56);
+        browser2.disconnect();
+
+        originalBrowser.close();
+      });
     });
     describe('Puppeteer.executablePath', function() {
       it('should work', async({server}) => {