for the same reason as...
commit fb2c146752
Author: Eike Rathke <erack@redhat.com>
Date: Mon Jul 25 13:32:17 2016 +0200
use div() instead of /0 to content ubsan
While we do handle double floating point division by 0 at least on
IEEE754 systems, ubsan builds mock about.
Change-Id: Id2e34686b053707c83ca535689b693d80b606533
Makes use of a dedicated interaction wrapper to server
ssl certificates correctly.
See comphelper::SimpleFileAccessInteraction for details.
Change-Id: I5728225065ec03f9b793940de316ccd97ee7114d
Reviewed-on: https://gerrit.libreoffice.org/27596
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
This is a follow-up to commit 9d91d371e9
(svx lok: draw text edits in all views, 2016-07-26). Two corner-cases
are now handled:
1) When the SfxViewShell is created after begin text edit and
2) When the other draw view is already created, but at the time begin
text edit happens, the other draw view shows a different page.
And the opposite of these: switching away from a page were we observe a
text edit done in an other view or destroying a view that observes a
text edit.
When the complete view goes away, then SdrObjEditView::HideSdrPage() is
not called, so also try to destroy the outliner view of the text edit
from SdrObjEditView::DeleteWindowFromPaintView(). The GetSfxViewShell()
call in SdrObjEditView::ShowSdrPage() is important, because we let the
other draw view create the outliner view, but the outliner view should
invoke our view shell, not the view shell of the other draw view.
Also improve the SdTiledRenderingTest::testCursorViews() testcase, so
that it asserts it managed to begin text edit and use a test document
that still has a single slide and shape, but the shape is not
auto-sized; otherwise invalidations happen even if outliner views are
not created in all draw views, so the test would pass even without the
fixes.
Change-Id: I2c3bb27826c6887115366db818599fc8adabc5a5
Reviewed-on: https://gerrit.libreoffice.org/27583
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
... so empty cells with notes are included as last "data" position if they are
below or right of the last "real" data, which GetLastDataPos() doesn't.
Change-Id: I9bb7464033736e7e7fa24e635ef1a3d39626002b
Any string can be used as delimiter between integer and fraction.
It is now saved/loaded to/from ODF, as it was from XLS.
Change-Id: Ie6364d1cdefc020ea615c18099118135c619f96b
Reviewed-on: https://gerrit.libreoffice.org/27262
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
frames anchor to the following paragraph. Don't remove
an empty paragraph if it follows a frame or else the frame
will jump to the next page.
This gets a bit complicated because headers/footers contain
paragraphs that aren't really "following" the frame paragraph,
and so wouldn't be used as anchor paragraphs.
There may be similar sub-paragraphs for comments etc, but
exceptions for those can be added when proof documents are found.
Change-Id: I46988b40abe65e23a5c407dde38a951937978005
Reviewed-on: https://gerrit.libreoffice.org/27455
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
change since...
commit c23ab5eaed
Author: Steve Yin <steve_y@apache.org>
Date: Tue Nov 26 14:25:22 2013 +0000
Integrate branch of IAccessible2
keep the accessible description, but drop thea accessible name which
is reportedly an undesirable change.
Change-Id: I8f9c53ef967c4e566fde1829832aa035efce4f23
even when non-printing chars is off.
I think this began as an intermediate step for both page and
column breaks to the header/footer overlay we now have for
page breaks. Leaving this just for column breaks. It may
have been set to always on with page-breaks in mind.
Change-Id: I1617683e2bb8c1f9229e9c2313e7817e184fa5ea
There is really no need for end-users to have access to these
settings. We don't want end-users to shoot themselves in the foot. If
power users really think they need to modify these settings, they can
poke the LibreOffice registry.
Change-Id: I567506f2d2adf3c3b1817ca82be808af4e9698f7
Its intended use case has never been particularly clear. (I wrote it,
so I know.) We now test OpenCL usability silently at each from-scratch
start.
Change-Id: I4e04ae5876153f61089ece6a75ebfdecf4cfcc09
We haven't displayed any list of OpenCL devices to select from in this
dialog for a long time, so remove the associated code.
Change-Id: Ife0c624a8bfb829bd1e62e0a000c7681d696f40b
Currently when a text edit is started, then in the LOK case if there is
an other view that shows the same page, then both draw views will have
an outliner view showing the text edit. This means that in case a view
shell is created after starting the text edit, that won't have an
outliner view for the text edit.
Before fixing this, calls to SfxViewShell::Current() has to be avoided
when we're in the process of setting up a new LOK view. In case of
Impress, this is a double initialization, and by the time
SdrObjEditView::ImpMakeOutlinerView() is called, we're already in the
process of setting up the second SfxViewShell (as part of
SdXImpressDocument::initializeForTiledRendering()), but
SfxViewShell::Current() still points to the old view shell. Which means
that the outliner view would refer to a view shell that's deleted soon,
and we crash as soon as it tries to invoke a LOK callback.
Fix this by adding a virtual member function to SdrObjEditView, and
override it in sd, so in case applications want to provide a more
precise way of giving the view shell owning a draw view, then they can.
Change-Id: Ie0005f73237d4ff9cf576bf16fa5b46280f13759
Reviewed-on: https://gerrit.libreoffice.org/27561
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
remove meBeff and GetBiff().
adapt uses.
remove enum BiffType, because only use is gone.
keep BiffType symbols as comment for documetation.
Change-Id: I7019a21b80fd5b5f0b92458f83dd9e34bc704731
Reviewed-on: https://gerrit.libreoffice.org/27253
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
From reading the code of vcl::Window::Update, this already calls
Invalidate and Flush in case of top-level widgets and also handles
child windows.
And there is no need to invalidate the progress bar text, if we
just update the progress value.
Change-Id: Ie296c6ffe5db5a927ad69104ce8ffe3d0d29034a
Reviewed-on: https://gerrit.libreoffice.org/27285
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
