chrome-debugging-client

chrome-debugging-client

An async/await friendly Chrome debugging client with TypeScript support, designed with automation in mind.

Features:

• Promise API for async/await (most debugger commands are meant to be sequential).
• Launches Chrome with a new temp user data folder so Chrome launches an isolated instance. (regardless if you already have Chrome open).
• Opens an ephemeral remote debugging port so you don't need to configure a port.
• A TypeScript codegen for API autocomplete and tooltips with documentation for the debugger protocol https://chromedevtools.github.io/devtools-protocol/
• Cleans up processes and connections at end of session.

Example:

import { createSession } from "chrome-debugging-client";
// import protocol domains "1-2", "tot", or "v8"
import { HeapProfiler } from "chrome-debugging-client/dist/protocol/tot";

createSession(async (session) => {
  // spawns a chrome instance with a tmp user data
  // and the debugger open to an ephemeral port
  const process = await session.spawn("canary", {
    additionalArguments: ['--headless'],
    windowSize: { width: 640, height: 320 }
  });

  // open the REST API for tabs
  const client = session.createAPIClient("localhost", process.remoteDebuggingPort);

  const tabs = await client.listTabs();
  const tab = tabs[0];
  await client.activateTab(tab.id);

  // open the debugger protocol
  // https://chromedevtools.github.io/devtools-protocol/
  const debuggerClient = await session.openDebuggingProtocol(tab.webSocketDebuggerUrl);

  // create the HeapProfiler domain with the debugger protocol client
  const heapProfiler = new HeapProfiler(debuggerClient);
  await heapProfiler.enable();

  // The domains are optional, this can also be
  // await debuggerClient.send("HeapProfiler.enable", {})

  let buffer = "";
  heapProfiler.addHeapSnapshotChunk = (evt) => {
    buffer += evt.chunk;
  });
  await heapProfiler.takeHeapSnapshot({ reportProgress: false });
  await heapProfiler.disable();

  return JSON.parse(buffer);
}).then((data) => {
  console.log(data.snapshot.meta);
}).catch((err) => {
  console.error(err);
});

customize browser executable path

By default, this tool launches Chrome Canary. It may error if it cannot find the executable. For this and other reasons, you can configure the executable path like so:

// example for macOS
let browser = await session.spawnBrowser('exact', {
 executablePath: '/Users/someone/Applications/Google Chrome Canary.app/Contents/MacOS/Google Chrome Canary'
});