452787 - Couldn't drag elements which contain text or it is difficult for empty elements

Status: RESOLVED FIXED

(Core :: DOM: Copy & Paste and Drag & Drop, defect)

Description

[Laurent Jouanneau]

Attachment: HTML page which shows the bug

With the new HTML5 drag and drop API, I found an issue: 

- if a div is draggable, the drag cannot start if it contains text.
- If it is empty, we should first click on it (by pushing and releasing the mouse button), and then we can begin a drag.
- If the div have the -moz-user-select:none;, all works as expected.

Look at the given html page to see this behaviors.

Comment 1

[Neil Deakin]

Attachment: allow dragging within a selection

This patch checks the draggable attribute within the mouse down for selection handling.

Comment 2

[Olli Pettay [:smaug][[e-mail]]]

Comment on attachment 336138 [details] [diff] [review]
allow dragging within a selection

> // static
> PRBool
>+nsContentUtils::ContentIsDraggable(nsIContent* aContent)
>+{
>+  nsCOMPtr<nsIDOMNSHTMLElement> htmlElement = do_QueryInterface(aContent);
>+  if (htmlElement) {
>+    PRBool draggable = PR_FALSE;
>+    htmlElement->GetDraggable(&draggable);
>+    if (draggable)
>+      return PR_TRUE;
>+  }
>+
>+  // special handling for content area image and link dragging
>+  return IsDraggableImage(aContent) || IsDraggableLink(aContent);
>+}
What if image or link has dragable="false", should this return PR_FALSE in that case?
nsHTMLImageElement and nsHTMLAnchorElement have their own implementation for ::GetDraggable.

Comment 3

[Neil Deakin]

> What if image or link has dragable="false", should this return PR_FALSE in that
> case?

Such as:

nsCOMPtr<nsIDOMNSHTMLElement> htmlElement = do_QueryInterface(aContent);
if (htmlElement) {
  PRBool draggable = PR_FALSE;
  htmlElement->GetDraggable(&draggable);
  if (draggable)
    return PR_TRUE;

  if (aContent->AttrValueIs(kNameSpaceID_None, nsGkAtoms::draggable,
                            nsGkAtoms::_false, eIgnoreCase)
    return PR_FALSE;
}

?

Comment 4

[Olli Pettay [:smaug][[e-mail]]]

Um, I wasn't CC'd to this bug - sorry didn't notice the comment.

But yes, something like that if you can't find any better way.

Comment 5

[Neil Deakin]

Attachment: updated patch