View transition layered capture: determine geometry based on box model
Relanding with a build fix(spurious -blink include)
Instead of capturing the border offset, we capture the following
layout sizes to determine sizing, in addition to the border box:
-
padding box
-
content box
-
The box-sizing property
-
In layered capture mode, we use the content box size as the
reference box for positioning the snapshot, so that padding take
effect. -
The ::view-transition-image-pair pseudo-element now has
position:relative so that it gets affected by padding. It has the
content box size. -
The width/height of the group is determined by the box-sizing
property, and it receives the computed padding.
Bug: 375395117
Change-Id: Ide0c001ea17582bb854a2fe678965504e0d4c0d5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6011700
Reviewed-by: Mike West mkwst@chromium.org
Reviewed-by: Philip Rogers pdr@chromium.org
Commit-Queue: Noam Rosenthal nrosenthal@chromium.org
Cr-Commit-Position: refs/heads/main@{#1381667}