ImplImageTree was used outside of VCL which is not consistent with
the name and the header also contains a lot of implementation
detail. This separates the implementation to ImplImageTree and
the public interface and singleton to ImageTree only.
Change-Id: I3a26444f0f6971a6b1d83472e9cef19c93192d3e
Reviewed-on: https://gerrit.libreoffice.org/32134
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Tested-by: Tomaž Vajngerl <quikee@gmail.com>
no point in having a template and a virtual base class when it's only
used for one type
Change-Id: Idb1a1a551064cc10896eff33652038eb5be0297e
Reviewed-on: https://gerrit.libreoffice.org/32041
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
...deriving from VclReferenceBase. Complicated by the fact that the argument
type may be incomplete at the time of template instantiation. So this approach
may be less precise than the change to loplugin:vclwidgets from
cbf5b21f2a65bbb342295200f6ad93a00f90733e "Catch some misuses of VclPtr
construction" when the argument type becomes complete later in the comilation
unit. However, this approach would also catch the two misuses in UnoControls
found by cbf5b21f2a65bbb342295200f6ad93a00f90733e, so go with this approach for
now and revert the change to loplugin:vclwdigets.
Change-Id: I7888f23d2b9e2db81ae2ce4bf4c8277912317685
Reviewed-on: https://gerrit.libreoffice.org/31966
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
sal_Bool and sal_uInt8 are typedefs for the same underlying type, so any use of
ORowSetValue with sal_Bool instead of bool, apparently intending to treat the
value as a boolean, actually treated it as a TINYINT. (See e.g. recent
7b0c57b2faec875c790051d233d1e9abaed2a3bc "some compilers don't like implicit
bool-to-ORowSetValue conversion".)
Now that there's no way to create a sal_uInt8 ORowSetValue, getUInt8 and the
m_uInt8 union member can probably go away, too.
Change-Id: Ia27554f76e7e9edce6410284b578064573e54fd3
Reviewed-on: https://gerrit.libreoffice.org/31909
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
To avoid this:
sc/source/core/tool/formulalogger.cxx:55:26: error: bad static variable causes crash on shutdown [loplugin:badstatics]
static FormulaLogger aLogger;
~~~~~~~~~~~~~~~~~~~~~^~~~~~~
sc/inc/formulalogger.hxx:42:31: note: ... due to this member of 'FormulaLogger' [loplugin:badstatics]
const ScFormulaCellGroup* mpLastGroup = nullptr;
~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~
sc/inc/formulacell.hxx:66:20: note: ... due to this member of 'ScFormulaCellGroup' [loplugin:badstatics]
ScFormulaCell *mpTopCell;
~~~~~~~~~~~~~~~^~~~~~~~~
sc/inc/formulacell.hxx:114:21: note: ... due to this member of 'ScFormulaCell' [loplugin:badstatics]
ScDocument* pDocument;
~~~~~~~~~~~~~~~~^~~~~~~~~
sc/inc/document.hxx:312:27: note: ... due to this member of 'ScDocument' [loplugin:badstatics]
VclPtr<SfxPrinter> pPrinter;
~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~
Change-Id: I533e45f655ca928a801188aa48ee818d89a962ac
As per sberg' suggestion
Limit it to == and !=, because some people like the flow of inequalities
like: "0 < a && a < 42"
The changes to sal/ were made using the rewriter.
The rewriter still has one bug, in pipe.cxx, it managed to pick up
some random piece of macro. No idea why.
Change-Id: I01305f9c5396a4b6c7421d6e92f1b4b529388e82
Reviewed-on: https://gerrit.libreoffice.org/30962
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
Inspired by a recent bug report where we were assigning the result
of VclPtr<T>::Create to a raw pointer.
As a consequence, we also need to change various methods that were
returning newly created Window subclasses via raw pointer, to
instead return those via VclPtr
Change-Id: I8118e0195a5b2b4780e646cfb0e151692e54ae2b
Reviewed-on: https://gerrit.libreoffice.org/31318
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
...to catch new places where defined'ness of OSL_BIG/LITENDIAN would be checked
without osl/endian.h being included; cf.
e2f08f9def0869460ad38a1c2adb450778290f6e "connectivity, sc: add missing #include
<osl/endian.h>" and 2b14fb3a4f92b928f0a5fc536c6a5f4a6e51a9b8 "cppcanvas, oox:
add missing #include <osl/endian.h>".
Change-Id: I167c8916a01391b7dacad7325153ccf35d3ba9dc
(like in vcl/inc/osx/a11ywrapper.h's AquaA11yWrapper), at least on recent Clang
trunk, from within the call to cast() in
cast<RecordDecl>(getDeclContext())
as the decl context apparently is something other than a RecordDecl.
Change-Id: I238bae44d6db0f04bf8f90b0032489e3b4822eee
...other options to avoid such irrelevant warnings can be to move code to an
include file and/or to define a dummy main() accessing otherwise unreferenced
entities.
Change-Id: Ifd44e376b35ef68496f3aba6a3c046d684824000
look for final classes, and make sure they don't have protected members
Change-Id: I1fa810659bba02b61a5160dbfd8e24185ec9abf4
Reviewed-on: https://gerrit.libreoffice.org/30895
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
...to check that loplugin produces warnings/errors as expected:
* Clang has a -verify switch that makes it easy to write test input .cxx files
that list in comments all the warnings/errors that are expected, and let Clang
check those expectations instead of generating object code. See
include/clang/Frontend/VerifyDiagnosticConsumer.h in the Clang source tree for
documentation.
* Introduce a CompilerTest gbuild class that uses the existing LinkTarget class
as much as possible. Checking the input files is implicitly phony, as the
compilation step doesn't generate any object files, and the link step does
nothing because there is no gb_LinkTarget_set_targettype for CompilerTest.
The setup at least works for Clang on Linux (will need adaptions for Clang on
Windows; compilers other than Clang are not relevant for now given this is
used to check compilerplugins).
* Definition of gb_CFLAGS_WERROR in solenv/gbuild/platform/com_GCC_defs.mk needs
to be lazy ('=' vs. ':=') so that CompilerTest can override it: The Clang
-verify mode wants the input files to specify whether the loplugin diagnostics
are warnings or errros, so they consistently need to be errors independent of
--enable-werror configuration.
* A first (example) test is in compilerplugins/clang/test/salbool.cxx. The
corresponding gbuild CompilerTest instance is in
solenv/CompilerTest_compilerplugins_clang.mk for now. The reason for that odd
split across compilerplugins/ and solenv/ is that there is no
compilerplugins/Modules_compilerplugins.mk file, so this setup is the easiest
hack for now (to be cleaned up). (Another area that could be improved is that
all test files need to be listed explicitly in the CompilerTest_*.mk file,
instead of, say, using all .c/.cxx/.m/.mm files in a specified directory.)
* The test is run somewhat late during a top-level 'make', after loplugin has
already been used in compilation. But it can be run manually (e.g., 'make
solenv') when making changes to loplugin during development.
Change-Id: I01e12fb84887d264ac03ef2484807458c2075af4
