Setup the "last reschedule time" counter at the first pass,
to avoid useless immediate reschedule.
Update the counter when running "when blocked" reschedules.
This seems to avoid the problem with the bugdoc.
Change-Id: Ib5958a1a2b048f5ec654c69ee9e977e8a26de6f5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/146215
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
...rather than on the deprecated OComponentHelper.
The two classes SfxDialogLibrary and SfxScriptLibrary, both deriving from
SfxLibrary, had been found to implement their respective queryInterface in a way
that is incompatible with the XAggregation protocol inherited via
OComponentHelper. It looks like no code actually made use of the XAggregation
offered by these Sfx*Library classes, so the easiest fix for those
queryInterface implementations appears to switch from OComponentHelper to
WeakComponentImplHelper (thereby dropping XAggregation, and thus rendering the
existing queryInterface implementations OK).
Ideally, SfxLibrary would derive from WeakComponentImplHelper<XInitialization,
XStorageBasedLibraryContainer, XLibraryContainerPassword, ...> covering all the
UNO interface classes from which it currently derives manually. But changing
that manual implementation across SfxLibrary and its SfxDialogLibrary and
SfxScriptLibrary derived classes looks tricky, so merely introduce an "empty"
WeakComponentImplHelper<> for now and keep all the manual stuff, and leave
proper clean up for later.
Change-Id: I12dc5bad2c017b8d76ce28ac189e95cf2e3810e9
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/145792
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
(See the upcoming commit introducing that loplugin:unocast on why such
dynamic_casts from UNO types are dangerous.)
Change-Id: I962b2a11a6407a543fb8b4ad972128d59f62a8a8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/144142
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
...that hits at least when building with Clang and --with-latest-c++ against
recent libc++ or MSVC standard library (where C++20 and esp. C++23 made more and
more class template member functions constexpr). My understanding is that there
is some leeway at what point a compiler should instantiate such function
specializations, and Clang decides to instantiate constexpr ones early (cf.
<https://github.com/llvm/llvm-project/commit/242ad89a15d5466d166d47978bfff983d40ab511>
"C++11 half of r147023: In C++11, additionally eagerly instantiate:" and its "Do
not defer instantiations of constexpr functions" comment, and the discussion at
<https://discourse.llvm.org/t/point-of-instantiation-of-constexpr-function-template/65129>).
> In file included from basic/source/comp/codegen.cxx:21:
> In file included from include/basic/sberrors.hxx:23:
> In file included from include/vcl/errcode.hxx:23:
> In file included from include/rtl/ustring.hxx:34:
> In file included from ~/llvm/inst/bin/../include/c++/v1/ostream:168:
> ~/llvm/inst/bin/../include/c++/v1/__memory/unique_ptr.h:47:19: error: invalid application of 'sizeof' to an incomplete type 'SbiExprList'
> static_assert(sizeof(_Tp) >= 0, "cannot delete an incomplete type");
> ^~~~~~~~~~~
> ~/llvm/inst/bin/../include/c++/v1/__memory/unique_ptr.h:281:7: note: in instantiation of member function 'std::default_delete<SbiExprList>::operator()' requested here
> __ptr_.second()(__tmp);
> ^
> ~/llvm/inst/bin/../include/c++/v1/__memory/unique_ptr.h:247:75: note: in instantiation of member function 'std::unique_ptr<SbiExprList>::reset' requested here
> _LIBCPP_INLINE_VISIBILITY _LIBCPP_CONSTEXPR_SINCE_CXX23 ~unique_ptr() { reset(); }
> ^
> basic/source/inc/expr.hxx:118:66: note: in instantiation of member function 'std::unique_ptr<SbiExprList>::~unique_ptr' requested here
> SbiExprNode( const SbiSymDef&, SbxDataType, SbiExprListPtr = nullptr );
> ^
> basic/source/inc/expr.hxx:30:7: note: forward declaration of 'SbiExprList'
> class SbiExprList;
> ^
Change-Id: I3bde85164ae6b829b7c24d9645fd412ed3fd815d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/141810
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
where at restoring the SaveIndex the underlying aLine string had
been replaced, so the index is for a different string. Save and
restore the aLine too. Not entirely sure this is what we want,
but it does seem to be what this restoring code expects.
Change-Id: I25bd02fad72d9b0729b9504fe04416340d0e070a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/136286
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
I started from a copy/paste of FormatNumber.
Then I deduplicated the code (it saved about 99% of it).
Change-Id: Ibcb9ffbf8cebf45d5ffac4713e3d220b8499ba11
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133133
Tested-by: Jenkins
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
1. Make BasicCollection::implGetIndexForName take OUString again,
after commit f7de7de118 made it take
std::u16string_view. All call sites pass OUStrings, and commit
ef32c3b4f9 made it create OUString
from the argument.
2. Have SbxVariable cache a case-insensitive variant of the name.
It is currently only used in the Collection implementation, but
may be used in other places; Tthe names are case-insensitive in
Basic, and VBA allows non-ASCII characters in names, so this
caching might be useful elsewhere.
3. Skip non-ASCII characters when calculating name hash, to allow
non-ASCII-containing strings still have some hash variance, when
at least some of the first 6 characters are ASCII.
Change-Id: If90ccea2c4b44c34967e6b764b6fab45b2976c40
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132493
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
If you want to keep the string to floating point conversion semantics
in BASIC as they were in LibreOffice 6, you can either set the
LIBREOFFICE6FLOATINGPOINTMODE environment variable to some non-empty
value, or set the org.openoffice.Office.Scripting/Basic/Compatibility/
UseLibreOffice6FloatingPointConversion flag in the Expert
Configuration. (Or use a registry modification file with that effect.)
Adapt the relevant unit test accordingly. Thanks to Mike Kaganski for
showing how to access the setting in the LibreOffice registry from
Basic.
Change-Id: I13d6d5d834e1bb81ef8df489db2b1da79f01dfc5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125756
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
So it's not a good idea to change the patch without testing it
again. I'm still hoping for a better usable MediaItem then the
previous dummy. The ODF stores a preview image and it should
allow to export the media.
The scripting fix feels like a workaround. somehow the final
class specifier triggers something that makes the linker fail
with missing destructors for all std::unique_ptr objects
derivated from SbxFactory. Implementing these as virtual
overrides didn't help, neither as {} or = default. So this
removes them in the !"#if HAVE_FEATURE_SCRIPTING" case.
Change-Id: I9f6751f44e0247ed31c27be0ee816239795e9372
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126330
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
If a name is ending with an underscore at the end of the line, it won't
be recognized correctly and it will be wrongly interpreted as the
respective name without the underscore at the end. Instead of changing
the macro source, use a flag in order to correclty identify the line end.
Change-Id: I6f933d7382b510bffed3e2692d32c232b7084624
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/123833
Tested-by: Jenkins
Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de>
In order to support the correct data type of numeric constants,
booleans, and default values for strings, the strategy to transport the
data type character to the runtime has been changed.
The type character will be added after the literal and the string
termination symbol in order to keep compatibility. This allows to
retrieve the correct type in StepLOADNC and in StepPARAM.
Any legacy written images are still possible to process, since if there
is a suffix type character where the data type character was directly
written after the numeric constant, it will be processed in StepLOADNC.
Without this fix, an optional parameter of type Variant, would still
include the suffixe type character, and will not be converted to the
correct type in StepPARAM.
Change-Id: I86c8192c6dc28457053fa7b23a073420e45407b2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119945
Tested-by: Jenkins
Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de>
This makes sure that results of comparison of literals have proper boolean type,
and that this type is properly stored to/read from the string pool.
This introduces a new non-standard "type" character used in the pool: 'b'. It is
not a proper type character used in Basic (unlike '%'/'&'/'!'/'@'), but we have
to use this trick locally, because we need to pass the type that has no own type
character.
The change should be backward-compatible: older versions reading the value should
just ignore the 'b', and read the value as double.
Change-Id: Ibd4a70b366331342346eb171c8ed3c7026279596
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/117655
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
"As New" variables use lazy instantiation, but that should not re-apply each
time one re-enters the module. If the object is still there, don't replace it.
Change-Id: Ic568a56b93db9bc9139d434d97a4a4deb05840df
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/112018
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Mike Kaganski