When adding a signature, first we export it to a temp. storage, then
read it back, show the verification to the user, and then later we do or
do not write the temp. storage back to the original one.
This means the signature gets exported two times, and MSO only considers
the final result valid. So when caching signatures (to avoid a real
export based on our data model), don't cache the one we just added to
the temp. storage, but do a real export second time as well.
With this, MSO considers our appended signature (next to an existing
one) valid, too.
Change-Id: I4d615298463e037ea4e654ff5c3addcef8b0a094
We append a new signature to a document by re-exporting the existing
ones, then writing the new signature. Given that existing signatures
aren't canonicalized before hashing, write them back as-is.
With this, our own signature verification is happy about the export
result, containing an existing and a newly created signature.
Change-Id: I0ff57a2266c6070a945f0c45ca5793406678be60
The previous solution had problems. InfoPlist.strings were empty,
because localization has not been working since the CFBundleTypeIconFile
entries were removed from Info.plist. Icon file basename was the key
in documents.ulf. So we packaged 0 bytes long files. The second problem
was that we used LibreOffice language codes, and OS X language codes are
different in some cases. This caused problems such as French strings on
English UI (e.g. Open/Save dialogs), because the system did not
recognize en-US.lproj and en-GB.lproj, and fell back to the next one: fr.
Conflicts:
Makefile.in
(cherry picked from commit 91902ef1411943f65da296fefd15fff9170d9c0c)
Change-Id: I9c502cdf737b497ca2ceef8f3c535ccfea2f6134
So no need for the related entitlements.
Change-Id: I54ba7c0586ee77f30096b50755a9a85bbb7965a5
(cherry picked from commit a1379c6d2555c77b1b096165ac2aab42808e90a7)
(cherry picked from commit d78c567c78e9d0f4e122caab527da23f69dec6d3)
Creating the pipe fails when sandboxed. This caused us to not start
the OfficeIPCThread, and that then meant that the file open requests
coming in through VCL_NSApplication's application:openFile: method in
vclnsapp.mm were not processed properly.
The OS takes care of not starting multiple LO apps simultaneously
anyway, so we don't really need any pipe, I hope.
Conflicts:
desktop/source/app/officeipcthread.cxx
Change-Id: Ia920520ce2928787313f83199028f9c9942f61f3
(cherry picked from commit e2f4c9eae3f53aaba968d7567f912df6fa911326)
I doubt end-users will miss gengal.bin, regview, or regmerge.
We need the gengal program at build time. But don't bother shipping
it on OS X, at least not in the sandboxed (App Store) case.
Change-Id: Id73bef1ba71d126c2d2962fe846e9c31963d6c24
Add SyncFocusState(). Before we needed
pRefInputEdit or bRefInputMode.
Change-Id: Ib1d8c4d26f45824fd6652ef29d56ddf8d276319d
Reviewed-on: https://gerrit.libreoffice.org/22823
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>
<https://msdn.microsoft.com/en-us/library/windows/desktop/ms647466%28v=vs.85%29.aspx>
"About Strsafe.h":
When you include Strsafe.h in your file, the older functions replaced by the
Strsafe.h functions will be deprecated. Attempts to use these older functions
will result in a compiler error telling you to use the newer functions. If you
want to override this behavior, include the following statement before
including Strsafe.h.
#define STRSAFE_NO_DEPRECATE
Appears to no (longer?) be relevant here, as reported by clang-cl.
Change-Id: I94a576b71a2a34edd6009059fc5047ba59ec0f12
Where wglChoosePixelFormatARB is defined as a macro expanding to
__wglewChoosePixelFormatARB (as WGLEW_GET_FUN is just expanding to its
argument), and __wglewChoosePixelFormatARB is declared in global scope in
workdir/UnpackedTarball/glew/include/GL/wglew.h.
itself) in workdir/UnpackedTarball/glew/include\GL/wglew.h
Change-Id: I0c4d09e9112c2233d25a262ea1f2b35bdf49645c
According to
<https://msdn.microsoft.com/en-us/library/windows/desktop/aa383681%28v=vs.85%29.aspx>
("Enabling STRICT"), it is used in combination with including windows.h, so
maybe this #define STRICT (which was there ever since
a934115b3ea90d7883681a6ca8922a4b859be36f "Initial revision," w/o also including
windows.h) was always unnecessary (as reported by clang-cl).
Change-Id: I02cf2584fe2fd52e9513791b18fdf6e7982e89b3
"VC_EXTRALEAN can only be used in MFC projects."
(<https://support.microsoft.com/en-us/kb/166474> "How to use VC_EXTRALEAN and
WIN32_LEAN_AND_MEAN to enhance the build process in Visual C++")
So presumably indeed not needed here (as reported by clang-cl).
Change-Id: Id98c6e9e862e716cf5e238623c6ac4ca0603fd63
For MSVC/clang-cl: "__STDC__: Indicates conformance with the ANSI/ISO C99
tandard. Defined as the integer literal constant 1 only if the /Za compiler
option is given and you are not compiling C++ code; otherwise is undefined."
(<https://msdn.microsoft.com/en-us/library/b0084kay.aspx>)
Presumably fine to just replace #if with #ifdef here (as is used in all the
other cases surrounding it), and not care whether it actually expands to true.
Change-Id: Ie58c76d7e72321b06a11a370fd28221e85fcf871
