WPF - Ensure only single drag operation is emit on DragSourceEndedAt

[WilliamTomOBrien]

Fixes: #4801

Summary:

• This change ensures that only none, move, copy, or link can be passed the the browser host to signify a DragEnd event. Specifying multiple flags on the bitmask seems to prevent the event from firing, and therefore the result that is returned from DragDrop.DoDragDrop is compared to the last drag drop effect we have on record.
• When doing further testing after creating the issue, I noticed that the dragEnd event was emit when I cancelled the drag drop. Beyond this, when testing and manually specifying the flag, I noticed the event was emit when using the none, move, copy, or link flags by themselves in the bitmask, and seemed to correlate with the event's dataTransfer.dropEvent field. Given the API specifies that can only be one of those four values (and not a mixture of them), that seemed reasonable. Furthermore, it seemed that the DoDragDrop function used within the StartDragging function would consistently return the mask with more than one value set.

Changes:

• I have modified the implementation of the CefSharp.Wpf ChromiumWebBrowser's StartDragging method, but no significant changes were made to the structure of the code.

How Has This Been Tested?

This has been tested by going through the same process detailed in the issue here. This was done on Windows 11, with an x64 architecture, .NET Version 4.7.2, and the WPF implementation.

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Updated documentation

Checklist:

• Tested the code(if applicable)
• Commented my code
• Changed the documentation(if applicable) N/A
• New files have a license disclaimer
• The formatting is consistent with the project (project supports .editorconfig)

[AppVeyorBot]

✅ Build CefSharp 125.0.80-CI5005 completed (commit fcd1f2ff91 by @WilliamTomOBrien)

[amaitland]

Thanks for the PR 👍

We should compare this to the cefclient reference implementation.

[amaitland]

Some quick testing and this look ok 👍 It probably needs more thorough testing though.

Rather than let it sit here I'll extract the logic out into a method and allow for customisation of the behaviour.

Someone can implement their own custom logic (or revert to the previous behaviour if they wanted).

Follow up PR #4845