And while at it, merge the change from e2bd5afd726abd5df438b6b821416bd7cf496e4d
"Make the C++/UNO bridge work to some extent on macOS on arm64" into the
existing code in a better way, reviving the assertion that openMax will be non-
negative when entering the loop.
Change-Id: I4b4e173a79ae6a0bbbf07df87d46761e86f460c0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100841
Reviewed-by: Tor Lillqvist <tml@collabora.com>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Tested-by: Jenkins
Otherwise the transparency gets converted into b&w mask here before
applying the gradient. Required for tdf#134052.
Change-Id: Ife00f75024f7d776fdb11e16fcafbd237f07eb04
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100830
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
which was dropped by
commit bba8e5aa6d1968e5279b3fe368c0f81145513d04
Date: Thu Mar 1 21:00:24 2018 +0100
tdf#116070: Use a valid PaM when confirming the dialog.
when fixing
commit 34527cec54ca634b831cfa5ac1098c4046818ac3
Date: Tue Jan 16 18:54:19 2018 +0100
lokdialog: Convert the Format -> Paragraph... dialog to async exec.
so, go back to using pPaM instead of rWrtSh.GetCursor but make a copy
of pPaM to avoid the crash of tdf#116070
Change-Id: I0aaaf5173c6be47dd306b77eacbb07d0e1c4eaf1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100670
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Regression after commit a0dcc8f78061d22d925f31ef8d03ca751ad6dc19.
The commit had added extensions to the filter names sent to file
dialog, but kept internal list of filter names aFilterEntryList
without extensions. Then dialog returned the chosen filter's name
in the form it was sent there, i.e. with extension; and that name
couldn't be found in aFilterEntryList.
So let's keep aFilterEntryList synchronized with the UI names in
the dialog.
Change-Id: Ic21d19d1a8d0b7f64ccfb517a7d0ee8ab36fc6a8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100821
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Change-Id: I52c6fe32e7737e1e0160cfd585d8616a035320c0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100657
Tested-by: Jenkins
Reviewed-by: Muhammet Kara <muhammet.kara@collabora.com>
Change-Id: I41b0887d0bc6bf4618bd1cf9afd06f4b1ddbabb1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100778
Tested-by: Jenkins
Reviewed-by: Muhammet Kara <muhammet.kara@collabora.com>
MAXROW_JUMBO/MAXCOL_JUMBO should be max - 1, similarly to
MAXROW/MAXCOL.
Since f6064b13586aa8681907b69e4f43643251f9b803.
Change-Id: I3d300e916b9584a2231f39bec5a044bd82c81752
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100777
Tested-by: Jenkins
Reviewed-by: Aron Budea <aron.budea@collabora.com>
bDesctructorCalled doesn't exist in the code anymore.
I guess we can delete the comment.
Change-Id: I551efe2298422e5139f1dd07a6b3bf4728763026
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100774
Tested-by: Jenkins
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
This abstracts the just refactored vcl::TrueTypeFont class, so the
Qt5 backend can provide it's own QRawFont based font table access.
Change-Id: Ic71bc95bc8fe39bc7a32086d4adc78cfa00d15be
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100718
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
This hides all members, which are needed to create a new TTF font
by calling vcl::CreateTTFromTTGlyphs, and adds an abstraction
to access individual font tables. This is needed, because Qt5
just allows access to the raw font based on font tables.
Change-Id: I74e34cf1aa2529a06ec5ec53aa2be751e58982ad
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100717
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
..presumably since cf596c43315bb96b5e7256a82256f1ccb8c9c4d0 "tdf#133163 DOCX:
export formula cells", see e.g. CppunitTest_sw_rtfexport3 failing at
<https://ci.libreoffice.org/job/lo_ubsan/1728/> with
> warn:sal.osl:27291:27291:sal/osl/unx/module.cxx:162: dlopen(/home/tdf/lode/jenkins/workspace/lo_ubsan/instdir/program/libmswordlo.so, 257): /home/tdf/lode/jenkins/workspace/lo_ubsan/instdir/program/libmswordlo.so: undefined symbol: _ZTI17SwTableBoxFormula
Change-Id: I01f8753ff82e41e84ad5d9806e3582ccd671065e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100762
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Based on observation (no documentation found),
MS Word 2013 ignores the actual value of LayoutInCell
and always considers it to be true. The UI shows it,
but it is grayed out when CompatibilityMode == 15.
The code's default value of bLayoutInCell is also true,
so compatMode only needs to be checked if the property
is actually set in the document.
I probably could have left off the IsInTable() check,
but decided to just be as safe as possible.
I assume it doesn't apply, but LayoutInCell sets
a IsFollowingTextFlow property - which might have
meaning outside of a table in SW - I don't know.
Change-Id: Ief073ea96c6c61efb3bcbbc11d114d739b04c3af
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100651
Tested-by: Jenkins
Reviewed-by: László Németh <nemeth@numbertext.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
Regression after 56d5531373cf0588e02c254ea8ffa00221c8342b
The missing "lx03249" caused failure in ImplImage::getSizePixel when
exiting from presentation mode in Impress:
assert(maBitmapChecksum == 0);
when res/lx03249.png couldn't be loaded; the resource is referenced
as BMP_OO_IMPRESS_DOC_LC in svtools/inc/bitmaps.hlst.
Neccessary icons added.
Change-Id: Iab240a4587e1aa1ff44b3478a029756db4b4da43
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100734
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Tested-by: Jenkins
