How to use _innerLaunchWithRetries method in Playwright Internal

I played with your example for a while and I got the same errors. These are the things I found that made my example work:

It must be Linux. I know that you mentioned that you picked a Linux plan. But I found that in VS Code that part is hidden, and on the Web the default is Windows. This is important because only the Linux plan runs npm install on the server.

Make sure that you are building on the server. You can find this option in the VS Code Settings:

Make sure you set the environment variable PLAYWRIGHT_BROWSERS_PATH, before making the publish.

Well this is one way, but not sure if it will work for all possible locators!.

// Get a selector from a playwright locator
import { Locator } from "@playwright/test";
export function extractSelector(locator: Locator) {
    const selector = locator.toString();
    const parts = selector.split("@");
    if (parts.length !== 2) { throw Error("extractSelector: susupect that this is not a locator"); }
    if (parts[0] !== "Locator") { throw Error("extractSelector: did not find locator"); }
    return parts[1];
}

I found the error. It was because of some missing libraries need. I discovered this when I downgraded playwright to version 1.9 and ran the the code then this was the error msg:

(node:12876) UnhandledPromiseRejectionWarning: browserType.launch: Host system is missing dependencies!

Some of the Universal C Runtime files cannot be found on the system. You can fix
that by installing Microsoft Visual C++ Redistributable for Visual Studio from:
https://support.microsoft.com/en-us/help/2977003/the-latest-supported-visual-c-downloads

Full list of missing libraries:
    vcruntime140.dll
    msvcp140.dll
Error
    at Object.captureStackTrace (D:\Projects\snkrs-play\node_modules\playwright\lib\utils\stackTrace.js:48:19)
    at Connection.sendMessageToServer (D:\Projects\snkrs-play\node_modules\playwright\lib\client\connection.js:69:48)
    at Proxy.<anonymous> (D:\Projects\snkrs-play\node_modules\playwright\lib\client\channelOwner.js:64:61)
    at D:\Projects\snkrs-play\node_modules\playwright\lib\client\browserType.js:64:67
    at BrowserType._wrapApiCall (D:\Projects\snkrs-play\node_modules\playwright\lib\client\channelOwner.js:77:34)
    at BrowserType.launch (D:\Projects\snkrs-play\node_modules\playwright\lib\client\browserType.js:55:21)
    at D:\Projects\snkrs-play\index.js:4:35
    at Object.<anonymous> (D:\Projects\snkrs-play\index.js:7:3)
    at Module._compile (internal/modules/cjs/loader.js:1063:30)
    at Object.Module._extensions..js (internal/modules/cjs/loader.js:1092:10)
(Use `node --trace-warnings ...` to show where the warning was created)
(node:12876) UnhandledPromiseRejectionWarning: Unhandled promise rejection. This error originated either by throwing inside of an async function without a catch block, or by rejecting a promise which was not handled with .catch(). To terminate the node process on unhandled promise rejection, use the CLI flag `--unhandled-rejections=strict` (see https://nodejs.org/api/cli.html#cli_unhandled_rejections_mode). (rejection id: 1)
(node:12876) [DEP0018] DeprecationWarning: Unhandled promise rejections are deprecated. In the future, promise rejections that are not handled will terminate the Node.js process with a non-zero exit code.

A list of missing libraries was provided. After successful installments, firefox ran fine. I upgraded again to version 1.10 and firefox still works.

This question is quite close to a "need more focus" question. But let's try to give it some focus:

Does Playwright has access to the cPicker object on the page? Does it has access to the window object?

Yes, you can access both cPicker and the window object inside an evaluate call.

Should I trigger the events from the HTML file itself, and in the callbacks, print in the DOM the result, in some dummy-element, and then infer from that dummy element text that the callbacks fired?

Exactly, or you can assign values to a javascript variable:

const cPicker = new ColorPicker({
  onClickOutside(e){
  },
  onInput(color){
    window['color'] = color;
  },
  onChange(color){
    window['result'] = color;
  }
})

And then

it('Should call all callbacks with correct arguments', async() => {
    await page.goto(`http://localhost:5000/tests/visual/basic.html`, {waitUntil:'load'})

    // Wait until the next frame
    await page.evaluate(() => new Promise(requestAnimationFrame))

    // Act
   
    // Assert
    const result = await page.evaluate(() => window['color']);
    // Check the value
})

Assuming you are not running test with the --runinband flag, the simple answer is yes but it depends ????

There is a pretty comprehensive GitHub issue jest#6957 that explains certain cases of when tests are run concurrently or in parallel. But it seems to depend on a lot of edge cases where jest tries its best to determine the fastest way to run the tests given the circumstances.

To my knowledge there is no way to force jest to run in parallel.

Aside

Have you considered using playwright instead of puppeteer with jest? Playwright has their own internally built testing library called @playwright/test that is used in place of jest with a similar API. This library allows for explicitly defining test groups in a single file to run in parallel (i.e. test.describe.parallel) or serially (i.e. test.describe.serial). Or even to run all tests in parallel via a config option.

// parallel
test.describe.parallel('group', () => {
  test('runs in parallel 1', async ({ page }) => {});
  test('runs in parallel 2', async ({ page }) => {});
});

// serial
test.describe.serial('group', () => {
  test('runs first', async ({ page }) => {});
  test('runs second', async ({ page }) => {});
});