Viewport X of the click (from the mouse event).
Viewport Y of the click.
The row's root element; the position is computed over the text of its
.textual-note-body descendant (see RowTextKind in this module for what each piece of that
text counts as).
The position (see FootnoteCaretPosition), or 'end' when the click cannot be
mapped (no browser support, click outside the row's text).
Map a mouse click on a read-only footnote row to a caret position in the footnote's text, so an editor swapped into the row can place its caret where the user clicked (PT9-parity caret-where-you-clicked). Uses the browser caret APIs; positions land only at valid caret boundaries, so graphemes are never split.
A click on a marker lands inside that marker, which the note editor renders as editable text; a click on the note's own marker or caller (the row's header cell) lands right after the caller, the first place in the note the user can type.