> w32-util.c(176,1): error: conflicting types for 'utf8_to_wchar'
> utf8_to_wchar (const char *string)
> ^
> workdir/UnpackedTarball/libgpg-error/src\gpg-error.h(1109,10): note: previous declaration is here
> wchar_t *utf8_to_wchar (const char *string, size_t length, size_t *retlen);
> ^
with clang-cl on Windows, while in a case like this where there is only one
definition, the mismatching declaration merely gets warned about by MSVC with
"warning C4029: declared formal parameter list different from definition". (And
on non-Windows that w32-util.c apparently doesn't get compiled at all.)
Change-Id: I76cfc3ec086325c527c04dbe0e8341cb9b775c50
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100091
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
("the source reference is already a subtype of the destination reference")
Change-Id: Ifba02593d77057eaa356ec7c3c76de64ba65cf09
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100089
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
* Update helpcontent2 from branch 'master'
to 92d4f32170e6d3a1b65857bb83e22c4bf7f565bd
- Fix some "D'oh! You found a bug" here and there
Change-Id: I0b3b4eec07378dcc3a110bfb7e2205220f30c579
Reviewed-on: https://gerrit.libreoffice.org/c/help/+/100132
Tested-by: Jenkins
Reviewed-by: Olivier Hallot <olivier.hallot@libreoffice.org>
...like
> canvas/source/directx/dx_impltools.cxx(137,31): error: rather use make_shared than constructing from 'Gdiplus::Graphics *' [loplugin:makeshared]
> GraphicsSharedPtr pRet(Gdiplus::Graphics::FromImage(rBitmap.get()));
> ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
on Windows, where those functions like FromImage are provided by Windows (so we
can't help it that they are returning pointers)
Change-Id: Iae5c4b20d64cc6b38ff66409519fbd25f6e509cd
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100095
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
("the source reference is already a subtype of the destination reference")
Change-Id: I1183eaa2015bd9d6567001bb14f7809dffcae7ba
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100088
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
and keep the original around for use from ::Reset and overwrite
the original SwTableRep on commit
Change-Id: I18e078e577382c4c13d6f5e576cc1eeec15a7906
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100072
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
* Update translations from branch 'master'
to 1c00f748202eb64ffa0be5ce38e57a47f6e5798a
- update translations for master
and force-fix errors using pocheck
Change-Id: I7e90d467017ce8e49d81e914d60d944e2e69e0f1
... and SwDoc::CopyMasterFooter(); this is the same as commit
9f7ee38ace but for 2 other functions that
do the same thing.
Change-Id: Id7ed449a004ee3c9452d4603bf8632e2720651ed
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100077
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@cib.de>
This changes the order of children, which matters when they have no
explicit ZOrder. With this, the text shapes on the arrow shape are on
top of the arrow, not behind it.
The trick is that nominally chOrder can be "t"(op) or "b"(ottom),
defaulting to bottom, but there is a difference between an explicit "b"
and not setting it. When not setting it, the layout node is expected to
inherit it from its parent layout node, recursively.
This would probably make sense for other algorithms as well, but set it
only for the linear algorithm for now, as that's where we have a bug
document showing the PowerPoint behavior.
Change-Id: I433f92c620149ef5590aebc8cbf43110e1d2fb85
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100047
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
As discussed at <https://gerrit.libreoffice.org/c/core/+/99724> "fix shutdown
crash in basic" the code was added in error, and at least clang-cl with latest
MSVC standard library and C++20 mode gives a helpful
> basic/source/sbx/sbxbase.cxx(53,5): error: ignoring return value of function declared with 'nodiscard' attribute [-Werror,-Wunused-result]
> std::move(m_Factories);
> ^~~~~~~~~ ~~~~~~~~~~~
now.
Change-Id: I052efe51d4415838b50de06bb308692fa937b7b2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100076
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
To be honest, I don't know why this fixes the leak, but it's generally
good practice anyway to use unique_ptr.
Change-Id: Ic5aa2a0c6ab092450e9eae96616a998c08e2e723
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100045
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
> error: 'error' diagnostics seen but not expected:
> File compilerplugins/clang/test/makeshared.cxx Line 47: rather use make_shared than constructing from 'unique_ptr<int>' [loplugin:makeshared]
> File compilerplugins/clang/test/makeshared.cxx Line 49: rather use make_shared than constructing from 'unique_ptr<int>' [loplugin:makeshared]
> File compilerplugins/clang/test/makeshared.cxx Line 53: rather use make_shared than constructing from 'unique_ptr<int>' [loplugin:makeshared]
Change-Id: I5d2d1b129c9d0fee496eceb4e2cf14f5853ba00b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100074
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
...that now defines the wide-character-to-narrow-stream inserters as deleted
too, at least in C++20 mode.
Change-Id: I554f2530d5905e46343bf0d8bf12a6feb3d63075
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100073
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
"bitwise operation between different enumeration types ('CommandTypeEnum' and
'ExecuteOptionEnum')", but where <https://docs.microsoft.com/en-us/sql/ado/
reference/ado-api/execute-method-ado-connection?view=sql-server-ver15> documents
parameter Options as "A Long [...] Can be a bitmask of one or more
CommandTypeEnum or ExecuteOptionEnum values."
Change-Id: If000490127e7215863ea750a3f9f30e69916978e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100070
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
GetMaster() is for right pages. I assume this was a
copy/paste mistake which came in already at initial import.
And since left/right are alternating, they always use
the first version of the header/footer. Although the UI
doesn't allow changing these, they still COULD be different,
as is the case with the unit test.
No existing unit tests even touch this code,
so not a common situation at all.
Change-Id: I9e3720ddf34a8f7e08ce196780fbe16e8c88940b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99628
Tested-by: Justin Luth <justin_luth@sil.org>
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Reviewed-by: Justin Luth <justin_luth@sil.org>
As pstoedit documents itself, its wmf/emf driver uses a very approximate
interletter spacing, making the text look really awful. But it provides a
-nfw option that delegates the letter placement to the emf reader, and we
happen to be doing a proper job, thus getting a proper vectorized output.
This is not a concern on Windows (and the option is ignored there). The
option is available since version 3.40 (~2005). So we can just always pass
it on.
Change-Id: I8ffd3fbf046b5a80e8011651eeaf060a8f5107e2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100035
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@cib.de>
no need to use shared_ptr here, this class is not doing copy-on-write.
Change-Id: I4e921bfc789cc5989d98b5f9ab7074eb7d5ac33e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100022
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Regression from commit 28d67b7927
(tdf#107776 sw ODF shape import: make is-textbox check more strict,
2019-08-26), now that we correctly identify what shape text to import as
"textbox" (Writer TextFrame) and what to import as editeng text, there
are documents out there that try to import mailmerge fields into
editeng-based shape text. Fix missing error handling there.
Note that the error is not just silently ignored, we do insert the field
result into the shape text, existing code provides this already.
Change-Id: Ibe631ac5d94c1c7795dd00bad05fdcca0e6741a8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100028
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
need to call DoClose() to unwind the internal ref-counting
Change-Id: Ibfd13e2300f3542e5e86363facd5b41dbfafdf6e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100001
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
This reverts commit 09aa5a9be8, now that
<https://github.com/llvm/llvm-project/commit/
551092bc3dfb86f1e11a55f3bee0c8ee1be6fdd6> "Revert AST Matchers default to AsIs
mode" reverted the Clang commit that prompted this compilerplugins change.
Change-Id: I75c8b4cb2894cd67a791db460f2886a783856c73
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/100026
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
When we don't have type attribute on slide but have on
slidelayout we have to use it instead of default type.
Change-Id: Ibb874b5ee39c48641484fe1a8686f66c31695f76
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99904
Tested-by: Jenkins
Reviewed-by: Gülşah Köse <gulsah.kose@collabora.com>