This reverts the previous attempt of commit
250ad9311a ("tdf#115227 svtools:
suppress UNO notifications for color selectors"), in favor of
a different fix. The advantage of this new approach is that
is doesn't affect events other than activate/deativate, and that
it covers more cases: Toolbar popups even if not based on
svtools::ToolbarPopup, or if closed by clicking outside them,
font name/size and paragraph style toolbar controls, notebookbar,
context and main menus (for the non gtk3 native case, the gtk3
native menus left for a future investigation). For now, keep this
logic inside toolkit, but ideally it should be in vcl (after
reviewing vcl internal listeners).
One side effect that I noticed after this change, is that there is
a deactivation/activation pair which suppressed when e.g. opening a
new document from the start center. But I'm not sure if that's a
problem, given that the focus state of the top window as a whole
wasn't changed, only its contents, and there are other APIs to track
activation of document components. This happens because the source
and the target windows of those events are the same, and we need to
suppress this case to fix the font name/size style controls, and also
the color picker after 27473d1c0f
("tdf#114935 Move the focus back to the document"). Otherwise we'll
need to find another way to move the focus to the document w/o
triggering listeners.
Another case that will see a change in behavior, is document event
listeners in dbaccess, as the Focus/UnFocus events there are based
on top window activation/deactivation. However, I think it's a good
change, as currently just opening of a toolbar popup or main/context
menus there triggers those document events, making them useless.
I would like to also mention here, that in fact those top window
activation events never really worked as tdf#115227 expects them, as
the superfluous events for at least the font name/size, style and
color toolbar buttons existed already in OOo. The behavior of the
color buttons changed in LO for a few years, but regressed again in
the work on a real focus grabbing for floating windows starting with
commit dd46727b99 ("Resolves; tdf#87120
no keyboard navigation inside floating windows"). That work also
introduced superfluous events when using menus.
And a note about the change in menubarwindow.cxx: When a menubar
popup is closed it's deleted using the lazydelete mechanism, which
reparents the popup window, so it doesn't appear anymore in the
hierarchy of the document window. Moreover, I suspect that at some
point the lazydelete thing will be replaced by a VclPtr mechanism,
which might break this even further, so the event won't appear as
coming from the popup window (which might be already disposed at
this stage). So instead, temporarily set the menubar window as the
current focus window, so the activation will appear as if it was
coming from the menu bar window itself.
Change-Id: I292232adfcbd1a31d66ce394cd2f1bf42a013ecb
Reviewed-on: https://gerrit.libreoffice.org/48746
Reviewed-by: Maxim Monastirsky <momonasmon@gmail.com>
Tested-by: Jenkins <ci@libreoffice.org>
This is nominally a regression from commit
43bc303148 (unify color selectors,
2016-10-28), but that just changed the Writer color picker to behave the
same way as the Impress one.
The Impress one started to emit these events with
daeed90f45 (re-base on ALv2 code.
Includes:, 2012-11-15), to fix i#118707, improving accessibility.
That means either the focus changes and then accessibility is happy or
the focus does not change and then the UNO API client only gets events
when the user actually switches to an other window.
Fix the problem with suppressing UNO-level notifications for the
problematic events, by moving the existing VclListenerLock to a public
header and using it at two more places in svtools.
This should address not just color selectors, but in general other
toolbar menus / popup windows.
Change-Id: If427708c5b9fe4fa49cb8f00ec04b32cb28eb391
Reviewed-on: https://gerrit.libreoffice.org/48570
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
look for possibly bogus implicit conversions to bool when passing
(normally pointer) args to bool params.
this plugin comes in the wake of a couple of bugs caused by refactoring,
where some of the call sites were not currently updated.
Of the changes, the following are real bugs:
desktop/../dp_persmap.cxx
StartInputFieldDlg
in sw/../fldmgr.cxx
which occurred as a result of
commit 39d719a80d
Date: Tue May 3 11:39:37 2016 +0200
tdf#99529 sw: don't pop up input field dialog before inserting field
CSerializationURLEncoded::encode_and_append in
forms/../serialization_urlencoded.cxx
XclExpCFImpl::XclExpCFImpl
in sc/../xecontent.cxx
I have no idea how to properly fix this, just made a guess.
SwDocTest::test64kPageDescs
in sw/qa/core/uwriter.cxx
which looks like a simple copy/paste error.
Change-Id: I795ebd5ef485a1d36863dc27fe13832989f5a441
Reviewed-on: https://gerrit.libreoffice.org/48291
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
auto-rewrite with <https://gerrit.libreoffice.org/#/c/47798/> "Enable
loplugin:cstylecast for some more cases" plus
solenv/clang-format/reformat-formatted-files
Change-Id: Iaee0fee0cbc6e5311342b83d847cb89bd1056973
Insert constructor everywhere, except a couple places that apparently
want to compare GetMapUnit().
Change-Id: I1910deb60562e5e949203435e827057f70a3f988
Translates leftovers found using a custom regex.
Additionally:
- A few spelling fixes
Change-Id: I3772e1b914acc487d80ab14efb815cb178ca3dcb
Reviewed-on: https://gerrit.libreoffice.org/41831
Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>
had to change the structure of the plugin considerably, was too messy to
structure it to do the calculations on a per-function basis
Change-Id: I4edee7735f726101105c607368124a08dba21086
Reviewed-on: https://gerrit.libreoffice.org/40516
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
I can only imagine the weird bugs that must have periodically resulted
when we had a hash value collision.
In the process, fix hasElements() to return a useful value
Change-Id: I1d9a052e73332b4b2bbc9c1fd8142c13eb22f1be
Reviewed-on: https://gerrit.libreoffice.org/40226
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
std::binary_function is deprecated since C++11 and removed in C++17
90% done with regexp magic.
removed obsolete <functional> includes.
The std::binary_function base class was used by deprecated
std::bind2nd, this was solved in individual commits.
The members first_argument_type and second_argument_type were used
in chart2/source/controller/dialogs/DataBrowserModel.cxx:
DataBrowserModel::implColumnLess and are inlined in this commit.
Change-Id: I60ded60a8d4afd59e15ac15a58e18d2498c9be5a
Reviewed-on: https://gerrit.libreoffice.org/39659
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
std::unary_function is deprecated since C++11 and removed in C++17
90% done with regexp magic.
removed obsolete <functional> includes.
The std::unary_function base class was used in 3 places:
* chart2/source/tools/DataSeriesHelper.cxx: lcl_MatchesRole
is used in a std::not1 function helper who uses the members
return_type and argument_type.
- replace deprecated std::not1 with a lambda
* chart2/source/tools/ModifyListenerHelper.cxx:
lcl_weakReferenceToSame used the argument_type member in the
operator() parameter.
- inline the parameter type.
* xmloff/source/chart/SchXMLExport.cxx: lcl_SequenceToMapElement
used result_type and argument_type in operator().
- inline the types
Also fix compile error with gcc about finding std::for_each.
Change-Id: I073673beb01410c3108e7d0346d9e7d6b9ad2e2f
Reviewed-on: https://gerrit.libreoffice.org/39358
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Tested-by: Stephan Bergmann <sbergman@redhat.com>
Jochen Nitschke