This was a feature requested by mmeeks, as a result of
tdf#92611.
It validates that things that extend XInterface are not
directly heap/stack-allocated, but have their lifecycle managed
via css::uno::Reference or rtl::Reference.
Change-Id: I28e3b8b236f6a4a56d0a6d6f26ad54e44b36e692
Reviewed-on: https://gerrit.libreoffice.org/16924
Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>
... at the same position.
Since commit 0d57434180 these will be
exported by the ODF filter as duplicate attributes.
Change-Id: I8befe55f61c59ab968409fa03359540c300f9198
...mostly done with a rewriting Clang plugin, with just some manual tweaking
necessary to fix poor macro usage.
Change-Id: I71fa20213e86be10de332ece0aa273239df7b61a
The GetPrevPara() method apparently has to return 0 when there is no
previous paragraph.
(regression from ac85b6cff1)
Change-Id: I09a3e1d3a3adb33562e4e03c0755447047cbd433
Thre is still some 0xffff limit left and possibly some
less than gracefully handled overflow/error cases
Change-Id: I00957ee3a30b02f73918ea49d7353056263dc638
Reviewed-on: https://gerrit.libreoffice.org/7787
Reviewed-by: Norbert Thiebaud <nthiebaud@gmail.com>
Tested-by: Norbert Thiebaud <nthiebaud@gmail.com>
Also:
- simplify the return type of the GetPrevPara method, there is no
need return an OUString by pointer
- simply the ppPara parameter of the ChgAutoCorrWord method,
passing a pointer to a pointer to a value type is unnecessary.
Change-Id: I4a8c44fdab1cf68af88c34003827d1c20704f839
The horrible SvxAutoCorrect::AutoCorrect() inserts the character into
the document half-way through, and then _expects_ this inserted
character to show up in its rTxt parameter. This broke in Writer when
SwTxtNode::m_Text was converted to OUString, because now a
temporary String copy is created. Work around this disaster area for
now.
(regression from 0295c8a34e)
Change-Id: I1cb11a20cb0c2577036176e605426105631f3311
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
avoids the problems of dangling uno singletons invalidated after the first
dispose and the chain of other singletons that don't expect to need to
re-initialize, etc.
reenable editeng cppunit test
inherit i18npool cppunit test from unotest base
drop LibreOfficeProtector, do "throwable" work in setUp/tearDown not
in ctors/dtors
Without manually releasing the EditDLL singleton instance, it gets
deleted *after* the cppunit does its cleanup, which de-initializes VCL.
The problem is, when the EditDLL instance is destroyed, its member
GlobalEditData instance deletes the OutputDevice instance that it owns,
which in turn accesses font caches in VCL. But by the time we reach
that point, VCL is already de-initialized, hence the problem.
