Make option disabledness look for ancestor optgroups #11721
Conversation
Now that option elements can be descendants of optgroups instead of just children, the disabledness algorithm should look for ancestors intead of just a parent optgroup. Fixes whatwg#11707
|
I had a peak at Chromium's code for this and |
source
Outdated
| attribute is present.</p> | ||
| data-x="attr-option-disabled">disabled</code> attribute is present or if it is a | ||
| <span>descendant</span> of an <code>optgroup</code> element whose <code | ||
| data-x="attr-optgroup-disabled">disabled</code> attribute is present.</p> |
There was a problem hiding this comment.
Should it also be disabled when we have <optgroup disabled><optgroup><option>?
There was a problem hiding this comment.
Nested optgroups are something that we decided not to support, so in this case I don't think it should be disabled. I changed the text to look for the nearest ancestor optgroup instead.
There was a problem hiding this comment.
Please also add a test for this case.
There was a problem hiding this comment.
I guess I also have some other reservations. Is <select disabled><option> not disabled? What about <optgroup disabled><select><option>? (Not sure what the right answers are here, but it should be clear and we need test coverage, even if they are edge cases.)
There was a problem hiding this comment.
<select disabled><option> is not disabled, and <optgroup disabled><select><option> is also not disabled with the latest commit which aborts if it finds a select element before an optgroup. I will add tests.
Thanks, I am fixing this in my patch which adds the WPT |
|
Should this stop going through ancestors if a Also see handling of |
Yes, I pushed a commit to abort if a select element is found. |
Now that option elements can be descendants of optgroups instead of just children, the disabledness algorithm should look for ancestors intead of just a parent optgroup.
Fixes #11707
(See WHATWG Working Mode: Changes for more details.)
/form-elements.html ( diff )