[cssom-1] A more direct method than `caretPositionFromPoint()` about csswg-drafts HOT 2 OPEN

Hi Keith,

I think what you want is the selection bounds regardless of where is current selection. For contenteditable div, you can get current selection bounds using document.getSelection().getRangeAt(0).getBoundingClientRect(). For textarea/input element, there is no way to get selection bounds today.

Instead of adding currentCaretPosition() API on Element, I would like to propose the following:

partial interface HTMLTextAreaElement {
  [NewObject] DOMRect getSelectionBoundingClientRect();
}

partial interface HTMLInputElement {
  [NewObject] DOMRect getSelectionBoundingClientRect();
}

getSelectionBoundingClientRect() would return the bounding rect of current selection in textarea/input element. The bounding rect is the caret rect if the selection is collapsed. If there is no active selection in textarea/input, it would return empty rect.

The new proposed API aligns with existing selection APIs on textarea/input such as select(), selectionStart, selectionEnd, etc.

from csswg-drafts.

That sounds like a great idea! We can comfortably use an API such as that, provided it returns a bounding rect for the caret when empty.

from csswg-drafts.