Support for newCDPSession()

[fedono]

Feature Description

could support for context.newCDPSession() to enable developers to send commands via the Chrome DevTools Protocol
or is there any way to create newCDPSession by current method?

Suggested Solution (optional)

No response

Already existing or connected issues / PRs (optional)

No response

[inancgumus]

Yes, that's correct. However, we don't currently expose the newCDPSession method. Can you explain your use case for sending custom commands?

[fedono]

Yes, that's correct. However, we don't currently expose the newCDPSession method. Can you explain your use case for sending custom commands?

@inancgumus we have some scenes, eg

1. use Page.screencastFrame for recording videos when run tests
2. set Log.enable to recording the page logs so that we can check page errors
3. set Network.enable and use some methods like Network.dataReceived to recording whole network traffics

[inancgumus]

From what I see, you need three extra features, but (I believe) you want to use them without waiting for us to implement them. However, adding direct access to CDPSession might have security implications we cannot foresee for now.

1. use Page.screencastFrame for recording videos when run tests

We're open to contributions for this extra feature if you're interested 🙇

2. set Log.enable to recording the page logs so that we can check page errors

We automatically enable Log.enable. You can see these logs by passing K6_BROWSER_DEBUG=debug environment variable to the k6 command. Search for browser_source in the logs using grep or any other favorite tool.

3. set Network.enable and use some methods like Network.dataReceived to recording whole network traffics

We automatically enable Network.enable. For you to process the requests and responses, the next version will come with #4281 and #4296. It looks like it can be helpful for your usecase.

[fedono]

really thanks! I can't wait to see the next version.