summaryrefslogtreecommitdiff
path: root/sw/source/filter
diff options
context:
space:
mode:
authorJustin Luth <justin.luth@collabora.com>2020-07-24 08:11:54 +0300
committerJustin Luth <justin_luth@sil.org>2020-07-24 13:41:59 +0200
commitc016fe2b5918d6e53113e100b1126076b6e1a6a3 (patch)
tree376b160a107c2626cf45b32baadb46bbfe609ffe /sw/source/filter
parentb827eefd6397dc34b49a03e186146aed6cd80042 (diff)
tdf#99492 sw layout: always paint merged cell grids
So the logic goes like this: 1.) paint table background 2.) paint row X background 3.) paint row X's cell backgrounds So far so good. The cell background overrides all others. 4.) paint row Y background. Well, now it has painted over top of any merged Y grids, and thus covered cells need to paint themselves again. 5.) paint row Y's cells, NOT excluding covered cells. Dropping this "optimization" of ignoring the covered cell shouldn't impact performance too much since merged cells are not super common. One alternative option would be to determine if the row has a background fill, and only paint all the merged grids in that case. But that isn't trivial, and so the processing of determining that might as well be used to just paint the cell background again. The only other option is to paint all rows first, and then paint all the cells. But the logic of this generic function doesn't easily allow for that. Unfortunately, I couldn't see how to create a unit test since layout.xml doesn't include color at all. However, Office Interoperability Testing should catch it since there are a couple of clear examples attached to the bug report. Change-Id: Ie2c3776ab062dbb51c88134291e8901e71838b18 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99343 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@cib.de>
Diffstat (limited to 'sw/source/filter')
0 files changed, 0 insertions, 0 deletions