According to the dmake documentation, ICC refers to Visual Age C++ for
OS/2, which is not a supported compiler (or platform).
Change-Id: Ic9e23bc7c44de110a3a312bd007beda3b660927d
- nanosecond precision
- signed (allowed negative) year
Also: assorted improvements / bugfixes in date/time handling code.
Some factorisation of copy/pasted code.
Change-Id: I761a1b0b8731c82f19a0c37acbcf43d3c06d6cd6
Modules sal, salhelper, cppu, cppuhelper, codemaker (selectively) and odk
have kept them, in order not to break external API (the automatic using declaration
is LO-internal).
Change-Id: I588fc9e0c45b914f824f91c0376980621d730f09
.. to Reference<XComponentContext>
mostly in the dbaccess module, but it also affected some other
modules.
Change-Id: I09b3f6fe7a9b33498b11d98b5521b5aeeb8882be
use solenv/bin/add-modulelines script for the task
and remove all UTF bom from *.src and *.hrc files
svx/source/dialog/hdft.src
Change-Id: I745d4f0fe9b05436a142a03f8512970f91c41bd4
and use it to replace usage of createWithServiceFactory, and thus
replace usage of XMultiServiceFactory.
Change-Id: Id2f44066683ce17cf7a22f80916031556ef0e82a
No wonder such a generic-sounding term is defined in some system header, like
in the Android NDK's <asm/page.h>.
Change-Id: Id813afb836d358a3749a8a9fb1c9fb162f881c04
fetched data in the current row is always cached.
This avoids trying to fetch the same data several times, which is tricky in several scenarios in ODBC.
Numerous cleanups in passing.
Change-Id: I17246aa614276e141161a64c716881242c559310
The service is deprecated, but we still have a handful of in-tree
users, and converting it lets me thread XComponentContext through
a bunch of classes.
Change-Id: Iffdfe537ada6b9e4a89f9b3c8dd82ca85f4bfaba
Through that different initialization pIUnknown
got random address. The consequence was that
if (pIUnknown)
pIUnknown->Release();
ends with access viloation in
WpBase& WpBase::operator=(const WpBase& rhs)
Change-Id: I8b3c5de233d0868fea052c990cc83aed917117ae
More type-safe way, instead of void* everywhere
void* instead of sal_Int8 for raw memory / multi-use buffers
Fixed more issues than I care to count, but at least:
- updates to a DECIMAL were truncated to integer because DecimalDigits set to 0
maybe/probably also NUMERIC
- setObjectWithInfo(... DataType::LONGVARCHAR ...) was passed as AT_EXEC, but the buffer was already free()d by then -> crash or wrong data
Change-Id: I0e6791a05b96fb345bfe3f911386263e6cfedde9
This is a preparation for merge into master, feature/mork
branch is obsoslete and shouldn't be used any more. In
context of this change the usage of old mozilla based mork
driver depends on platform: all non windows platforms use
this new mork driver, windows still uses mozilla.
Because mozilla driver shipping was stripped on feature/mork
branch for all platforms, the corresponding commits were
changed prior to creating this change (the build file changes
are reverted now). So the build and shipping of different mork
driver implementation should be implemented from scratch:
mozilla on windows, new mork dirver on all other platforms.
Currently the new mork dirver is compiled (on all platforms),
but not used. It can be tested with mork_helper executable.
fdo#51004
Change-Id: Ib2413ab6856f163337aa311c4bf7b1182d6c6f63
Reviewed-on: https://gerrit.libreoffice.org/635
Reviewed-by: Miklos Vajna <vmiklos@suse.cz>
Tested-by: Miklos Vajna <vmiklos@suse.cz>
