Though this just prevents mishappens during reformatting a broken portion that
is in a half-ready state because a11y was notified and attempts to obtain a
paragraph. Solving the real cause means not being called in such state..
However, not going out of memory and also having matching X-values for the
node's text positions that don't make the EditEngine crash later is better than
nothing.
Change-Id: Ifa8d2c5d1766b0f145b8e89ad26a78528db42056
* adds to LibreOffice markup support for /italics/ and -strikeout-
* TODO update strings in the Options dialog (they only refer *bold* and
_underline_)
* TODO update documentation with new feature
Change-Id: I6fd7bbd036bf406a9e24500d316e7f28a848faab
Reviewed-on: https://gerrit.libreoffice.org/31076
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
the usage of which looks a little dodgy, because this is a bitmask, but
it's using == everywhere to check them
Change-Id: I8e57d4f943a9048cc457a376ffbdfde0cffe22dd
SW_DRAWLAYER had the same value as SC_DRAWLAYER, so I merged it into the
ScOrSwDraw enum constant
Change-Id: I5c45d378c00364d11cc960c9e48a6e3f10928724
Reviewed-on: https://gerrit.libreoffice.org/31037
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
if the assigned-to item had loaded its original content, but the assigned-from
item hadn't, then surely the assigned-to item will never load the new content.
if the assigned-to item hadn't loaded its original content, but the
assigned-from had, then why load it again
Change-Id: I68c2cf2682a517e7e630ab6cbc637b35a2b9c7aa
Reviewed-on: https://gerrit.libreoffice.org/30781
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
if SetGraphicPos was called with GPOS_NONE then
xGraphicObject, maStrLink and maStrFilter are
in a suitable state to copy to get the same
results as explicitly forced here
Change-Id: Id072590e92e0c083a3cbc443db0e85d9dbfa73f0
Reviewed-on: https://gerrit.libreoffice.org/30780
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
maybe the PurgeMedia call in sw was meant to be a PurgeGraphic
call originally
(PurgeGraphic since removed by...
commit a22ac2c218870033822120bf0b0d6cfde6ce799f
Author: Caolán McNamara <caolanm@redhat.com>
Date: Thu Jul 14 22:06:29 2011 +0100
callcatcher: remove unused methods)
PurgeMedia releasing the stream makes no difference to the only place its used
which is SvxBrushItem::GetGraphicObject which makes a new one every time
anyway.
the SvxBrushItem assignment operator doesn't change the stream
member of the pImpl which looks utterly nuts, so its a good thing
the stream is not reused
Change-Id: Ie0dee22a6640a6916908fcddbc3541ba85034217
It saves lots of extra code: no separately checking if a line exists,
and then getting the width. Closely matches the existing CalcLineSpace.
sc/source/ui/view/printfun.cxx is another place that could use this
heavily to replace their lcl_LineTotal function. Perhaps something
good for an easyHack. (Wait until LO5.4, since much of the logic
should use CalcLineSpace(,true) instead, and that function probably
will have the default bEvenIfNoLine changed to true. Compiler doesn't
like providing "true" when the default value is also "true".)
Change-Id: I298d057b2bf04959434736f6ab2666d2de4222f9
Reviewed-on: https://gerrit.libreoffice.org/30589
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
regression from commit e598ab04476a32a08f18e8f0662fafa5f78f1a4a
very aggressively forced a new frame size via compat setting
CLIPPED_PICTURES on any fly - not just images.
This only affects MS-format documents, EXCEPT that it is a document
property, so if the file every spent any part of it's life in MS-format,
it will always retain that compatibility setting. That explains
why the problem was intermittent for me - and was hard to reproduce
in a clean document, even though I'd seen it in .ODTs.
bIgnoreLine (ignore the fact that there is no visible line)
was a confusing word choice for "if there is no line,
then return a spacing size of zero". bEvenIfNoLine=false is better.
Change-Id: I50a3bdef3a67339ae517ee6319920651bc56f9be
Reviewed-on: https://gerrit.libreoffice.org/30585
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
I have checked the normal model and the editing model after UNDO, and
all seems to be well, this is purely a rendering/lack-of-invalidation
issue.
The extra invalidation I add here is restricted to the UNDO case to
prevent tripping up a LOK unit test
(SdTiledRenderingTest::testCursorViews).
I confess to not having followed the invalidation logic all the way to
see why exactly it makes the bug go away.
Change-Id: I34f7d84526462665b1ec09aba966c98cd4e8795f
Reviewed-on: https://gerrit.libreoffice.org/30225
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
which can be replaced with using declarations.
Is there a more efficient way to code the search? Seems to slow the
build down a little.
Change-Id: I08cda21fa70dce6572e1acc71bf5e6df36bb951f
Reviewed-on: https://gerrit.libreoffice.org/30157
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
Because some stuff wants to multiple-inherit from VclAbstractDialog and
OutputDevice-subclasses, and we'd prefer to keep all the lifetime
management through a single smart pointer class (VclPtr)
The change in msgbox.cxx and window.cxx is to workaround a bug in
VS2013 to do with virtual inheritance and delegating constructors.
Change-Id: I178e8983b7d20a7d2790aa283be838dca5d14773
Reviewed-on: https://gerrit.libreoffice.org/29140
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
makeAny and Any ctor return an Any
Change-Id: Iaa361bc315d785f80153acf1009bf47d109728ec
Reviewed-on: https://gerrit.libreoffice.org/29914
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Jochen Nitschke <j.nitschke+logerrit@ok.de>
What's new:
1) when an edit view is killed, the area which was used by the edit
view is invalidated for both own window and other view windows after
the edit view has been destroyed;
2) when an edit view is created or its out area is expanded, the
windows of other views are invalidated too;
3) when a vertical scroll occurs in the edit view area the windows of
other view are invalidated too;
4) same methods renaming since now we add/remove windows not edit
views.
Change-Id: Iac54f5b182c9562f08bb724f9ddde1c26cffa2e7
Reviewed-on: https://gerrit.libreoffice.org/29783
Reviewed-by: Marco Cecchetti <mrcekets@gmail.com>
Tested-by: Marco Cecchetti <mrcekets@gmail.com>
