-
Notifications
You must be signed in to change notification settings - Fork 760
Description
In Run-In Layout, CSS Display says
If a run-in sequence is immediately followed by a block box that does not establish a new block formatting context, it is inserted as direct children of the block box after its
::markerpseudo-element’s boxes (if any)
This is clear for most cases but not completely for some degenerated ones:
- The
::markerpseudo-element may havedisplay: contentsand thus generate no box, but its descendants (created via thecontentproperty) may do so. For consistency, I guess the run-in should be inserted after the boxes or text runs generated by these descendants. - The
::markerpseudo-element may be wrapped inside an anonymous parent box (unless the run-in munging happens first, which I'm not sure in cases other than CSS2§9.2). For consistency, I guess the run-in should be inserted after that anonymous box.
So it could say something like
If a run-in sequence is immediately followed by a block box that does not establish a new block formatting context, it is inserted as direct children of the block box after any box or text run which is or contains any box or text run generated by any node in the subtree rooted at the
::markerpseudo-element.
However, this doesn't seem much natural. That's because ::marker lives in the (pseudo-)element-tree while the run-in munging is done in the box-tree. Would it be bad to simply insert the run-in at the very beginning of the block box?