Read live HTML content of displayed webpage

[verelpode]

Can WebView2 support something like WebView.CaptureSelectedContentToDataPackageAsync but with an additional option to choose whether it captures the entire webpage versus only the selected portion of the webpage? For troubleshooting purposes, we would find it useful to be able to capture and analyze the HTML of the entire webpage, in order to figure out what's going wrong when a malfunction is reported by users.

I do realize that I can directly download the URL by myself using Windows.Web.Http.HttpClient but that's not the same thing because that effectively downloads the "source code" instead of the current live content of the webpage. JavaScript in the webpage is capable of making many modifications to the webpage, thus there can be a big difference between what HttpClient obtains and what is actually displayed in the WebView2.

Ofcourse the original intention of CaptureSelectedContentToDataPackageAsync is to support copy & paste scenarios, and this is also a valid usage of this feature. If this feature is already supported for capturing the selected portion of the page, then it seems like very little work to create an option to make it capture the entire page.

I think some other issues are much higher priority than this issue, but this feature would still be useful, even if it's not the most important thing currently.

AB#28476423

[david-risney]

Can you use IWebView2WebView::ExecuteScript to do this? By which I mean running some script that turns whatever portion of the DOM you like into a string and returning it back?

[verelpode]

@david-risney -- Although that's a smart idea and I'll keep it in mind, I think the answer is No, it doesn't fully replace CaptureSelectedContentToDataPackageAsync because ExecuteScript returns a string and would only allow HTML to be retrieved, not images, whereas CaptureSelectedContentToDataPackageAsync also captures images.

I just remembered another reason -- a more important reason -- why it should support capturing of the entire page. If it can capture the entire page, then it could be used to implement an external print dialog, in cases where the built-in print dialog cannot be used because it provides excessive access to users. See where I mention CaptureSelectedContentToDataPackageAsync in issue #42.

A similar example is a Kodak photo kiosk, etc. Kiosks similar to that example need to print but it's unacceptable to display the standard Windows print dialog to the kiosk users. Thus an external print dialog is needed, thus the need to capture the entire page (including images), not only the selection, and not only HTML in a string.

[ukandrewc]

You could do that using https://wkhtmltopdf.org/