While working on this accname wpt with Daniel Clark, we stumbled upon the computation steps, section 2C, where it says:
Combobox/Listbox: If the embedded control has role combobox or listbox, return the text alternative of the chosen option.
However, in the combobox spec, it says:
Typically, the initial state of a combobox is collapsed. In the collapsed state, only the combobox element and a separate, optional popup control button are visible. A combobox is said to be expanded when both the combobox element showing its current value and its associated popup element are visible. Authors MUST set aria-expanded to true on an element with role combobox when it is expanded and false when it is collapsed.
Elements with the role combobox have an implicit aria-haspopup value of listbox. If the combobox popup element has a role other than listbox, authors MUST specify an aria-haspopup value of tree, grid, or dialog that corresponds to the role of its popup.
We think (and are trying to confirm with the ARIA editors) that it's possible to have a combobox element that doesn't have a listbox and any option.
The combobox spec also says:
Otherwise, the value of the combobox is represented by its descendant elements and can be determined using the same method used to compute the name of a button from its descendant content.
Which appears to indicate that a combobox without a listbox should have its name from its value, and its value computed from the descendants' content. Is that right? If that's the case, should we clarify the accname computation steps?
While working on this accname wpt with Daniel Clark, we stumbled upon the computation steps, section 2C, where it says:
However, in the combobox spec, it says:
We think (and are trying to confirm with the ARIA editors) that it's possible to have a combobox element that doesn't have a listbox and any option.
The combobox spec also says:
Which appears to indicate that a combobox without a listbox should have its name from its value, and its value computed from the descendants' content. Is that right? If that's the case, should we clarify the accname computation steps?