and put the init/destruct in constructor/destructor instead
of setup(), we want this to run once, not for every test method.
Change-Id: I7d5fcdd2974677f1509048c16fb40d03dff289c7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100004
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
The dialog is called from a modal dialog which means a modeless dialog
does not make sense. Some of our internal dialog tracking code is confused
by that case which confused the UI tests.
Change-Id: I705877e8a751c55bfbd00ddeaf18ab86c95321c3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99944
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
This patch adds a dedicated hrc file for RID's
of all the properties that can be displayed in
the Inspector
Change-Id: I258ca060508e28b7b66e96393fcef4fd104bb781
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99983
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
63049e98a6 "Reliably set up controls of hyperlink
dialog in constructor" fixed an issue that could cause Python UITest code to
issue a bogus SELECT request that would fire an assert in
rtl_uString_newFromSubString. Even if that issue is fixed now, it is probably a
good idea to validate the requested range here.
(An alternative would be to validate it in the underlying shell's SelectText,
but that function is also called from internal code (which presumably already
ensures that it is passing valid arguments), so it is probably better to stay
with that function's narrow interface.)
It would probably be nice if this function reported failure (by throwing a
css::lang::IllegalArgumentException, say) instead of silently (modulo SAL_WARN)
clamping the range, but it is called from Scheduler::ProcessTaskScheduling
(vcl/source/app/scheduler.cxx) in a try/catch block that prohibits all
exceptions.
Change-Id: I5b7b4255861766a81a81501e391b1ff4e09b7db6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99933
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
* Update helpcontent2 from branch 'master'
to 8b7b468cfcb1591972ee2e47a295ee4cf86a46e8
- Fix tag contents
Found you finally...
Change-Id: I71e11b570e31ad1cc470dfea3e54935e000d80f0
Reviewed-on: https://gerrit.libreoffice.org/c/help/+/100011
Tested-by: Jenkins
Reviewed-by: Olivier Hallot <olivier.hallot@libreoffice.org>
For whatever reason, CppunitTest_dbaccess_hsqldb_test had recently started to
modify binary dbaccess/qa/unit/data/hsqldb_empty.odb in-place.
Change-Id: I07b5185ddf4b809ef1e1f977c5ab1c34e2e7f18a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100002
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Constraints are OK to request more, but it seems PowerPoint doesn't
allow leaving the parent, which simplifies the layout as well.
Change-Id: Id67a8740f1eff506e4beae0c797ad50e0218dfe6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99993
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
* Update helpcontent2 from branch 'master'
to dc36eea48b4a2120911d0f99322ff4a35ea39eb6
- Update version 7.0 New features video
Change-Id: I35fff323f6ccb6a035d87abcd8b08189276054e6
Reviewed-on: https://gerrit.libreoffice.org/c/help/+/100000
Tested-by: Jenkins
Reviewed-by: Olivier Hallot <olivier.hallot@libreoffice.org>
instead of declaring them as a base-type, and then static-casting
everwhere
Change-Id: I53b901c5353bb39ca9a0357aa442f50d2f475e7d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99995
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
to make it "clearer" this is part of the memory cleanup on destruction
process
Change-Id: I789bdfa0323cf759ed4db2e2a657e73e581d9eed
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99990
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
caused by a ref-counting cycle.
to reproduce the leak do
make CppunitTest_editeng_core \
VALGRIND='memcheck --leak-check=full \
--suppressions=$$BUILDDIR/solenv/sanitizers/valgrind-suppressions' \
CPPUNIT_TEST_NAME="testBoldItalicCopyPaste"
Change-Id: I94dca092a75db8ddfaae4a210e1158fa066ed609
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99989
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
this doesn't fix the leak, just reduces the number of cycles involved
here
Change-Id: I31eda167864dc8b5cea2448f598e01f3a807481a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99988
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
With this, finally the arrow shape has the correct horizontal position
and width, even if the markup is as complex as the PowerPoint UI
generates it (the previous version was a more minimal version).
Change-Id: I59f237c582053067e890180a1ae40471e5f46dea
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99894
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
Recheck after 7-0 branchoff
Also drop the now unused file include/vcl/field.hxx
Found with bin/find-unneeded-includes
Only removal proposals are dealt with here.
Change-Id: I9e54c82f50d1e02a0f99858939cac999fc66f7de
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99261
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
This check was originally introduced by me for a few toolbar
controllers (save, underline, currency), as an attempt to reuse
the same controllers also for non-split buttons (see tdf#98745
and tdf#101672). But since then it was copy-pasted to other
places where it probably makes no sense.
Change-Id: I7af2a3f5864ced88c91f330e87f32046d0c21383
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99960
Tested-by: Jenkins
Reviewed-by: Maxim Monastirsky <momonasmon@gmail.com>
The vcl implementation expects menu item attributes to be
set on the sub menu they appear in, not on the parent menu.
Change-Id: I7d6553126ed1dd5c1434529342b776aee01409a3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99954
Tested-by: Jenkins
Reviewed-by: Maxim Monastirsky <momonasmon@gmail.com>
by #include <unotest/bootstrapfixturebase.hxx> so that CPPUNIT_TEST_FIXTURE
is defined
Change-Id: Iacbfb2f513df6e7eb7ca04f7900cd4e6b3fd69c7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99963
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Seems code never really bothered not to touch default style
info, when inserting from file. Original commit is:
Author: Sascha Ballach <sab@openoffice.org>
Date: Wed Feb 28 08:24:41 2001 +0000
import of default styles added
Change-Id: Ibb639a585bedabdcc5987900ecca1e04f4bb593a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99015
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
If we removed something that is a link to a real file from our list of images
to pack, but that real image is not in the list of images to pack, add it in
instead so the real image does get packed
Change-Id: I71bcbdf872a59194a1d94f287dda8fc27e4a6464
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99961
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Unfortunately we need this on Android for the tablets, otherwise the
sidebar lacks the icons.
Also it is not easily possible to use the isLOKMobilePhone() check here,
because that is per-view in general, and handled in sfx in particular,
making it hard to use here in vcl.
This reverts commit 9a38b194ee.
Change-Id: I2b599e884ad4d00b7c246743c180a5324c9a143d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99300
Tested-by: Jenkins
Reviewed-by: Andras Timar <andras.timar@collabora.com>
This only handles lists with PositionAndSpaceMode::LABEL_ALIGNMENT.
TODO: handle PositionAndSpaceMode::LABEL_WIDTH_AND_POSITION.
TODO: handle NumberingType::BITMAP.
It doesn't align number according to Adjust, since that requires to
use a tab stop at the number position with appropriate alignment, and
it's unclear how would that interact with LabelFollowedBy modes other
than LISTTAB.
When first tab stop position is greater than ParaLeftMargin, Writer
uses left margin as implicit tab stop position, thus in this case
list appearance is different from processed text appearance.
In case of justified paragraphs, space after number (which does not
participate in justification when part of numbering) becomes part
of justification after conversion to text.
Change-Id: I88ad6617f8a09307ecad9d28edee92a59c68a4d4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99939
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Typically in these OutputDevice methods, the record-to-metafile case the
MetaFile is already written to before the test against mbOutputClipped to
determine that output to the current device would result in no visual output
(due to the output being outside the bounds of the device).
In this case the metafile is written after the test, so we must continue past
mbOutputClipped if recording to a metafile. It's typical to record with a
device of nominal size and play back later against something of a totally
different size.
Change-Id: Id1249b10f919165582f7de65b4cdc4d48074abaa
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99900
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Change-Id: I00f6fea1ee93bf2598d0cfde73fb2de17f0eb379
--- and not # of columns - 1.
Life is already too confusing to add that complication to it.
Not quite NFC. There is one place where a column
count of 1 would have set column separator/space.
I didn't think that would be necessary.
Change-Id: I87a7bfb5e746e8b7e4c57ddf40b0740d0ef35aba
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99930
Tested-by: Jenkins
Reviewed-by: Justin Luth <justin_luth@sil.org>
