We don't do unit testing when cross-compiling anyway. (And even if we
did, in some magic fashion, I don't think we would want
debugging/testing functionality like this dumping thing to be dragged
in into a production build of an app. It's not a huge amount of code,
but still.)
Change-Id: I12bc552b10452263572074aae865138ff30fbfc8
Avoid linking in various functions and potentially largish data
objects if the corresponding locale is deliberatedly skipped by using
the --with-locales configure option.
Change-Id: I8ec8c2b6176972d048a3c0e180ac0e6e805f2378
and make a real copy of the TargetFont
the 0,0 font size appears to mean "default" size so it can
take the desktop env font size or some such, with a hard-coded
value then the various "bold" entries in the UI are huge
This reverts commit 121aac3767.
Change-Id: Iec1d3867403808370d89d6b5ad2fec60ffdcd881
The color in the toolbar buttons doesn't change when we use the Floating
Font/Background/Highlight Color Toolbar.
This problem seems happen sometimes,not always.
This patch modifies that SvxColorExtToolBoxControl::StateChanged()
in /svx/source/tbxctrls/tbcontrol.cxx handles SFX_ITEM_DEFAULT message
always.
Change-Id: If7410297db768c161287b75396186fbec3605f63
Reviewed-on: https://gerrit.libreoffice.org/9107
Reviewed-by: Kohei Yoshida <libreoffice@kohei.us>
Tested-by: Kohei Yoshida <libreoffice@kohei.us>
1.The Font/Background/Highlight Color panels identify the actual color
by a small frame.
(Regression from the commit 35aa48d80b
See also fdo#32376 - Set default color to the current one in toolbar
popups)
2.Fix this in the Floating Font/Background/Highlight Color Toolbar,too.
Change-Id: I5c2db8ab8923ab146cbf921dcb90894765f8420c
Reviewed-on: https://gerrit.libreoffice.org/9103
Reviewed-by: Kohei Yoshida <libreoffice@kohei.us>
Tested-by: Kohei Yoshida <libreoffice@kohei.us>
ImpObjStack uses it's own home-grown stack and stack functions. There
is a function that unwinds the stack, but really it would be better if
we used std::set. In fact, this is better, because the name ImpObjStack
is really not terribly descriptive. I've replaced it with a stack of
OutDevState objects.
Change-Id: I87bdd4340ad77b7ffd9ff176fa5a9ffeac8b8666
In a dbgutil build LO aborts on an assertion failure the moment
KFileDialog is open. It's definitely on okay to release SolarMutex just
like that, since the Qt event loop is integrated with LO's, it'll call
back to LO code without the mutex held.
This reverts commit 13a34f4c63.
Conflicts:
vcl/unx/kde4/KDE4FilePicker.cxx
vcl/unx/kde4/KDEXLib.cxx
Change-Id: Idfa27fbb4728b529c37c25f710ea208fdaa4455c
... not just if there is no error; popping a value from the stack also
pops an error value, and the error is pushed again instead of the value
to be pushed, see call of IfErrorPushError() in PushDouble().
Change-Id: I92f486fa226c90870990f608cd43d4a1379a2361
At least with my test documents, this is the only component we now need
from the chartcontroller library.
Change-Id: Ia2c203413a585f4d9207bb22fa5f548eda3fb9a8
It might be that having fprintfs leftover in an app submitted to the
App Store is a no-go, but will fix that then if/when it happens.
Change-Id: Ic671fcce567c655a06fd4ed139100dc9b6f435be
