In sc/qa/unit/ucalc_formula.cxx, dropping the capture-default from the
lExpectedinF lambda revealed that MSVC in C++17 mode (i.e., when building
without --with-latest-c++) requires ROW_RANGE (a local const int variable from
the enclosing TestFormula::testTdf97369) to be captured, even though all uses of
that variable within the lambda body are constant expressions. That is still
true at least for the latest Visual Studio 2019 version 16.11.1. (This is not
an issue for the lExpectedinH and lExpectedinI lambdas a few lines further down,
as they, in addition to using that ROW_RANGE, also use the local const double
variables SHIFT1 and SHIFT2, whose uses are not constant expressions, so
they are implicitly captured and loplugin:unusedcapturedefault does not suggest
dropping those lambdas' capture-defaults in the first place.)
Change-Id: Iee7efb485187cbe8eba6a2d470afca4993eb1816
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120693
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Use the new write functionality. This change also sets SvmWriter::Write
second argument as const, as it's value does not change
Change-Id: I1620756542a0bc54b5f8fc7ce3f57d0243a2ee55
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120408
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
The change allowed to simplify many places where previously this API was
used, to avoid inefficient calculations (e.g., moving rectangle keeping
its size, and then immediately changing the size).
Change-Id: Ica2dc594d91cae83e2c2740c1f4fb23f44998916
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120461
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
since it uses o3tl::cow_wrapper, so it is really just a wrapper
around a pointer, no point in allocating it on the heap
Remove assert in SdrText::SetOutlinerParaObject, which was
bogus anyhow, because it was comparing pointers, not deep equality.
And since we are now being more efficient and avoiding
copying of the internal data in OutlinerParaObject, we hit
this assert.
Change-Id: I6dbfaab5ee2ca05b2001baf63110041e469df9c5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120510
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
This changes its value from 56 to the proper 57.
See 9be17c52b7, which had
changed the previous 5-mm value to the incorrectly rounded
1-mm value; and cefce1a247,
which had unified the other places mentioning i19922.
Change-Id: I7186320b4c1a2a65a6f728ed8bafcbf0eb45c4cb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120441
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
If a LogicToLogic uses fixed units, we can use o3tl::convert
instead. We can also do the same for all other cases where
LogicToLogic is used, but that needs additional investigation to
determine if it is safe to do so.
Note:
MapUnit::Pixel is converted to o3tl::Length::pt because it assumed
72 PPI for a logical pixel, which corresponds with the conversion
rate of a point (72 PPI).
Today, 96 PPI is standard, which is also used for o3tl::Length:px.
Change-Id: I29126df38bfcfda74b5d83d4cb880a378aecd18b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120230
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
is set each time here
This prevents a crash when an object is in text edit mode and a
different object is selected in the navigator to jump to.
Change-Id: I7c0eaa7770f0e3f23e6a62780f18e0a973019c62
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120170
Tested-by: Jenkins
Reviewed-by: Jim Raykowski <raykowj@gmail.com>
As effect1 with command (e.g. the one from misc/MEDIACALL category)
is created, 'mnCommand' member var is set.
But when it is replaced by command-less effect2 via
CustomAnnimationEffect::replaceNode, this var is never reset (simply
because the relevant if-else branch is never hit) so it looks as if
the command was still there. This causes UI glitch in sidebar:
setting incorrect category of the next selected effect
To fix that, we reset 'mnCommand' in ::setNode function, just as
it is done with resetting audio associated w/ the effect
Change-Id: I86016dd1acadfb31f460bde749bbe15164f1eb81
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119492
Tested-by: Jenkins
Reviewed-by: Katarina Behrens <bubli@bubli.org>
since it is just a wrapper over SvtCJKOptions and SvtCTLOptions.
Later I will replace those two with the equivalent officecfg calls
Change-Id: I61c5667a05f75d42643175f2c28c29d7a590b15c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119516
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
...as seen when running under UBSan `instdir/program/soffice --headless
--convert-to epub` of caolan/SIGSEGV-270412-142321-SIGSEGV-150412-121455-91.pptx
from the crash-testing corpus
Change-Id: I7ce0151689f90b8cba7cd86d053f327fdccc82d4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119487
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
It's been there since the beginning, but doesn't help us much:
e4239e0414 (Initial checkin of remote control., 2012-07-09)
Change-Id: I57425001cd0b9d0d035916405bd6d94329964b4e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119466
Tested-by: Jenkins
Reviewed-by: Andrzej Hunt <andrzej@ahunt.org>
This follows the pattern in osl::Socket which also cleans up on
destruction if necessary.
By closing on destruction we can avoid leaking the underlying socket or
filedescriptor in the numerous scenarios where we throw away the socket
without bothering to call close(). This isn't a big deal in normal
usage - we don't use many sockets in the first place - but you can
quickly run out of filedescriptors when running sufficiently complex
tests.
We also need to make BufferedStreamSocket final to avoid triggering
loplugin:fragiledestructor - BufferedStreamSocket probably should
have been final anyway, so there's no reason not to do so.
Change-Id: I90c271df4b598a6c2b326fde13543e6b27d7a110
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119093
Tested-by: Jenkins
Reviewed-by: Andrzej Hunt <andrzej@ahunt.org>
Before commit 653b53287c, editengine
objects had a pair IsVertical/SetVertical, which queried and set
a boolean flag (and SetVertical also had a second argument to set
another flag).
The mentioned commit had introduced an inconsistency, changing
SetVertical to only set a single flag, but at the same time making
IsVertical to return a synthesized result from two values: vertical
and rotation. Additionally, GetDirectVertical was introduced to
complement SetVertical.
In many places, the use of synthetic IsVertical looks suspicious,
especially where it is used in combinations with SetVertical. But
here I don't change existing logic, and only rename the methods,
so that in case someone sees an actual problem, it would be easier
to spot the method mismatch.
The end result is that now we have a proper getter/setter pair
GetVertical/SetVertical, and also IsEffectivelyVertical, named to
reflect that it calculates its return value.
Change-Id: I38e2b7c5bd7af0787dd7a1c48e1385138dac80b1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119315
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Versions before 166671f4aa do not
import the datetime field as expected if there's no text in the
datetime field.
To compensate that, added placeholder text "Date" inside them.
Fixes regression from: 25ae0de974
tdf#59323: pptx export: add support for slide footers
Change-Id: I8e35102bc0bb9a2e540cfdd59fb0dc2a0b23f185
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118008
Tested-by: Jenkins
Reviewed-by: Jan Holesovsky <kendy@collabora.com>
Stephan Bergmann