Due to license problems (see mail dev list) disabled swarmsolver

Change-Id: I39889ddab2563a857c3bf0982244f889a402097f

Change-Id: If9a0b5c185e828b5a1cf50fd42879aa3fa0f8635
Reviewed-on: https://gerrit.libreoffice.org/44910Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>

Change-Id: Id5aa07f87603879fe7a21dc96cad207f1b168286
Reviewed-on: https://gerrit.libreoffice.org/44850Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Kohei Yoshida <libreoffice@kohei.us>

Change-Id: Ia3e0748d0b4ba11ec84a85e823472cdac9bf987a

For transformed FlyFrames with no Border and no Padding it
would be nice to immediately start using AutoContour, added
first implementation to do so.
Added several conditions for AutoContour, need to
work on reaction on changes to these. Corrected
needed transform adaptions to Contour(s)

Change-Id: Ia3d7845fd5d50c8a413d592ae07ce2041ccc91b9

When rotated, the contour has to be adapted to that state
and also needs to be calculated in untransformed state,
added needed code. Also need to add a ClipRegion set at
the target OutputDevice (e.g. from Contour) to the prepared
GraphicPrimitive. It was sometimes used due to a VCL_Based
PrimitiveRenderer being used, but that is just a coincidence.

Change-Id: I1888ecd0468243bf2f41233b0cb3f99a50ac9805

To correct interactive Crop in transformed states, I had to
rework some stuff involved that anyways was in a non optimal
state. Added functionality to translate the object to make
Crop seem to work more seamlessly. Some mapping was needed
to make the Cropped, transformed object to be in the correct
relative position to the uncropped, untransfomed one.
Restructured TransformableSwFrame to directly re-create
last non-transformed SwFrame(s) from the SwFrameAreaDefinition
from the current Transformations, offering the untransformed
SwRect(s) now for usage.
Identified and corrected error when FlyFrame was translated
using keyboard, the accessing method needed to adapt position
in the transformed case.
Started to look at Contour stuff, adapted a set contour to be
correctly used by adapting it as needed in the transformed case.

Change-Id: I0d5f14958bcd6f826b9abd53f1f47b7d0bc5a0e2

Have now created working interactive Rotation, Scale, Move
and Crop. Using as much of the old code as possible since
it does corrections based on BorderWidth/s stuff and similar.
To do so, need to back-transform and execute in unrotated/
untransformed state as much as possible. Had some problems
with SwVirtFlyDrawObj::NbcMove usages and hacks there, had
to restructure some of these.

Change-Id: Ibedf6d834f96b2f8e9c1ace2a0c984aa26cedc74

No longer derive a SwFrame that needs Transformation
support from TransformableSwFrame, but keep it as a
member (in a std::unique_ptr). Adapted all usages,
also added reset of local SwRect(s) in local SwFrameAreaDefinition
during layouting. That makes PrintPreview work correctly.

Change-Id: I2e7720e67b57d4d29c0feb065878e246192a8a4c

In lcl_MoveAllLowers SwFrame(s) were directly modified,
not even by calling any member method what makes it hard to
react on changes to geometric definition(s). Added logic
to Transform a SwFrame, currently basic by using e.g.
SwFrame::Ltransform_translate, may be later unified to
general Transformation usage with a homogen Matrix

Change-Id: I7582fbd1472e12e481adacedda2e65cc0c282eac

Isolated tooling stuff for SwFrames that support Transformations
to TransformableSwFrame, adapted all usages. Developed functionality
to reset temporarily the already adapted SwFrame of the outer frame
to non-rotated to mate the layouting work. Made the outer frame
being layouted always first.

Change-Id: Ia60a971b8eaa28859d0f2a5243eabf68fe949846

When Layouting the update of the Areas is needed,
but during layout needs to be stable on the unrotated
values. Hard to find the right spot to do this, decided
to use the inner/outer frame, layout both, and trigger
the updates when the inner frame is layouted to non-
rotated Areas for inner and outer frame. Still not sure
if this will be correct, but is close to working well.
Some repaints/updates missing

Change-Id: I3d7dcbf624f3f32392e5e98420cb348d11d7d322

The layout element SwFrame now has methods to get linear
transformations for the geometry-defining FrameAreaDefinitions.
These return by default the SwRect (without being relative
for the FramePrintArea to make things easier), but are
replaced by the now two SwFrame derivates that support
rotation. The layout will now use the BoundRects of the
transformations, thus supporting a future that will allow
all kinds of free transformations (including mirror, rotation
and shear)

Change-Id: I18b85f5ddc2970a1b8c137d30b0c2ee49bb8df72

First steps to get a rotated FlyFrame and content
that only uses layouted sizes/positions and does not
change the model data (except rotation). This works
with persistence, after reload the rotation can be
resetted with going back to the original FrameSize.
Lot of stuff not yet working, experimental state.

Change-Id: Ie29d501fe2e618a1cb4457d600ce97575ed372d0

if neither LPSOLVE or COINMP is defined swarmsolve terminates with
an error "No LIBFILENAME set", if either one is a demand for
swarmsolve it should make an error message, if not the mk should
be modified.

For now enabling coinmp for iOS builds.

Change-Id: I0c0611c119f7930f50c90ff52d382196135f2c1f

The latest Xcode upgrade caused dublicate XBUILDCONFIGURATIONS,
giving a strange behaivour when changing parameters using Xcode.

Change-Id: I3e4d43af5540ef505665cf832090c767774bd951

Change-Id: Ie7dfdf4c2ec1d5629a5bd69623074f306831a9d2
Reviewed-on: https://gerrit.libreoffice.org/44899Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
Tested-by: Jenkins <ci@libreoffice.org>

Change-Id: If925aef0109a023d5760a7f83ab570a1f0312cb6
Reviewed-on: https://gerrit.libreoffice.org/44903Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>

For sheets with low data density (many empty cells) created charts with
checked "First row as label" or "First column as label" can get too much
extra rows/columns from data range (if they contain no data).

New approach is to treat empty cells as belonging to data, not to headers.

Change-Id: Ib5fd47636b80e0b8807f75d9eb582eed0ef68627
Reviewed-on: https://gerrit.libreoffice.org/44755Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
Tested-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>

E.g., when a unit test is debugged using
make CppunitTest_foo CPPUNITTRACE=TRUE
then devenv.exe is launched with /DEBUGEXE switch, and it's impossible
to add a .natvis file to the debug session (unless one adds the file
to path where VS looks for visualizers). A natvis added to PDBs will be
used in this case.

See https://docs.microsoft.com/en-us/cpp/build/reference/natvis-add-natvis-to-pdb

Change-Id: Icc31eb7c4fd063810e4074eaff774e8146b86877
Reviewed-on: https://gerrit.libreoffice.org/44595Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>

OTools::ThrowException uses SQLGetDiagRec to get error message. It's not
the caller's responsibility to define the encoding of resulting string.
Also, using the default setting of RTL_TEXTENCODING_MS_1252 turns
returned localized strings into garbage.

To allow correct conversion into Unicode string, let's use current
thread encoding, and drop the input encoding parameter.

Change-Id: Idb02e3ebb4fc407ce5e658fb2137ea2aa3bc127d
Reviewed-on: https://gerrit.libreoffice.org/44877Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>

+ use for range loops

Change-Id: I676009baa7b7aeb8c26951a8786da22431b0bc81
Reviewed-on: https://gerrit.libreoffice.org/44896Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>

+ use for range loop

Change-Id: I7bd95df8a75979bf342e3a8920fab016593e1b18
Reviewed-on: https://gerrit.libreoffice.org/44891Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>

+ use for range loops

Change-Id: If0fcba6e06538913031c50ec878b18db3547e06c
Reviewed-on: https://gerrit.libreoffice.org/44894Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>

+ use for range loops

Change-Id: Ied18e378b73826c5a47957cad6cf86a4e19a9230
Reviewed-on: https://gerrit.libreoffice.org/44892Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>

Change-Id: I4da2a768b6b55869c3a3d6f8a8d50dc018709acd
Reviewed-on: https://gerrit.libreoffice.org/44865Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>

can't find an easy way of removing the duplication, even though
that BeginLocking() stuff looks ripe for simplification.

Change-Id: I97b580b3668381c12b62136ea09772d91ab11eac
Reviewed-on: https://gerrit.libreoffice.org/44884Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>

Change-Id: I1e5346e2f014b412a2f8be08553c8eeff37920f9
Reviewed-on: https://gerrit.libreoffice.org/44840Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Jens Carl <j.carl43@gmx.de>

Change-Id: I899c05a4260324fa4c5648d7164821eddb1ead08
Reviewed-on: https://gerrit.libreoffice.org/44879Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>

Change-Id: Iae8fd98b678acdd3667eeb35b110b4e405c1167b
Reviewed-on: https://gerrit.libreoffice.org/44878Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>

This is a new, simple non-linear solver that uses a swarm
(population) to do global optimization. It uses two algoritms -
Particle Swarm Optimization (PSO) or Differential Evolution (DE)
to find a (non-optimal) solution.

It is experimental as not all functions are implemented and it
needs a lot more testing so that it performs well.

Change-Id: If55dad7eda17394851a9d178ad892de771eca7c9
Reviewed-on: https://gerrit.libreoffice.org/44382Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>

Change-Id: Ibb90096f85dfab277be7888ec6762624686df85b
Reviewed-on: https://gerrit.libreoffice.org/44796Reviewed-by: andreas_kainz <kainz.a@gmail.com>
Tested-by: andreas_kainz <kainz.a@gmail.com>

Change-Id: If41b1f038a3dc737066c27837c3fdaa0531de14c
Reviewed-on: https://gerrit.libreoffice.org/44827Reviewed-by: andreas_kainz <kainz.a@gmail.com>
Tested-by: andreas_kainz <kainz.a@gmail.com>

Radio buttons are adjusted when width/height is modified

Change-Id: Ia660a25314a46f2bb04b248a4fef937f10c4941d
Reviewed-on: https://gerrit.libreoffice.org/44876Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Yousuf Philips <philipz85@hotmail.com>
Tested-by: Yousuf Philips <philipz85@hotmail.com>
Reviewed-by: Heiko Tietze <tietze.heiko@googlemail.com>

Change-Id: Idc648e4673cca1b1dec8c1f009d52cfa290508ff
Reviewed-on: https://gerrit.libreoffice.org/44895Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
Tested-by: Julien Nabet <serval2412@yahoo.fr>

Textline offset was based on horizontal font metrics.
In vertical writing for CJK fonts, use descent
as its underline offset so that the line is drawn at the edge
of the bounding box to prevent overlap the CJK glyphs.

Change-Id: Iaa01bdf3cbb694dcf925975cde3f0f7ba7b97dac
Reviewed-on: https://gerrit.libreoffice.org/44720Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mark Hung <marklh9@gmail.com>

The offsets were 0 in HarfBuzz 1.3.3 but the value became
something else and made text offsets toward upper-right
after upgrading to HarfBuzz 1.4.8. Those values are offsets
relative to the horizontal origin.  But underlying renderer
expect positionos relative to the vertical origin. Translate
them back by adding the horizontal origin.

Change-Id: If58d2c082b198bfd8789efb3df2be49869a04bf9
Reviewed-on: https://gerrit.libreoffice.org/44556Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mark Hung <marklh9@gmail.com>

can't reproduce this, but throw a guess at it

Change-Id: Iff6198c463bbab4e3ef87ec62eaa13e88d016a23
Reviewed-on: https://gerrit.libreoffice.org/44890Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: I8a1a8c4b4d5f32a1f2b8a2e03225f6ac6b747232
Reviewed-on: https://gerrit.libreoffice.org/44889Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

for current monitor mode as well as all-monitors mode

Change-Id: I9de864dc0b416d18937f6ee90f432a8254fe987f
Reviewed-on: https://gerrit.libreoffice.org/44870Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>