HTML selectedcontent element

[zcorpan]

Request for Mozilla Position on an Emerging Web Specification

• Specification title: HTML selectedcontent element
• Specification or proposal URL (if available): Define the <selectedcontent> element whatwg/html#10633
• Explainer URL (if available):
• Proposal author(s) (@-mention GitHub accounts): @josepharhar
• Caniuse.com URL (optional):
• Bugzilla URL (optional):
• Mozillians who can provide input (optional): @hsivonen
• WebKit standards-position:

Other information

Previously, I had looked only at changes to select parsing and had missed whatwg/html#10633 , which appears to be treated as part of the whole package. I'm worried about breaking the so far implied invariant that there is one DOM element node per non-erroneous (possibly implied) start tag. Even template, which changed the basic expectations of how markup corresponds to DOM constructs allows parsing conforming markup to a DOM tree and re-reserializing to round trip. selectedcontent causing one option start tag to result in two option element nodes breaks at least serialization round-trippability of conforming markup. I don't at this time have more concrete reasoning about what practical badness this can cause exactly, but this kind of departure from the (implied) invariants worries me, so I don't want to say "positive" about this particular aspect.

Originally posted by @hsivonen in #1086 (comment)

[mfreed7]

Previously, I had looked only at changes to select parsing and had missed whatwg/html#10633 , which appears to be treated as part of the whole package. I'm worried about breaking the so far implied invariant that there is one DOM element node per non-erroneous (possibly implied) start tag.

I believe that invariant is still true here. The <selectedcontent> element is serialized correctly, as are its contents, and will completely round-trip. Now there is also a UA behavior in these cases (behavior that happens to be at parsing time, but also happens when the user changes selected options) that clones subsequent content back into the <selectedcontent>. In addition, that’s exactly equivalent to what you can do today with script. I.e. I don’t see that as different from this snippet:

<div id=foo>Some content</div>
<script>
  foo.innerText = "something else";
</script>

That will “round trip”, but #foo will now be “something else”. The new thing here is just that <selectedcontent> can have its content replaced. That’s new behavior for HTML elements, but the round tripping behavior seems like a natural fallout.

Even template, which changed the basic expectations of how markup corresponds to DOM constructs allows parsing conforming markup to a DOM tree and re-reserializing to round trip. selectedcontent causing one option start tag to result in two option element nodes breaks at least serialization round-trippability of conforming markup. I don't at this time have more concrete reasoning about what practical badness this can cause exactly, but this kind of departure from the (implied) invariants worries me, so I don't want to say "positive" about this particular aspect.

I’m not sure what you mean here, about one option tag resulting in two option element nodes… can you please clarify?

[hsivonen]

that’s exactly equivalent to what you can do today with script

Scripts can change the shape of the DOM tree, yes, but that argument could justify all kinds of departures from expectations about the relationship of the HTML source and the tree that the parser generates from it.

I’m not sure what you mean here, about one option tag resulting in two option element nodes… can you please clarify?

AFAICT, "maybe clone an option into selectedcontent" can cause the parser to clone a parser-inserted option element and insert it into the tree. Surely that means that one <option> start tag can result in the parser inserting two option element nodes for it? (Previously, this kind of behavior has only occurred as error handling.)

[zcorpan]

@hsivonen the children of the option element are cloned: https://github.com/whatwg/html/pull/10633/files#diff-41cf6794ba4200b839c53531555f0f3998df4cbb01a4d5cb0b94e3ca5e23947dR56379-R56380

That can include elements though.