It seems we need mirroring.

Change-Id: Id9ad7041eb09ec773e4174cb68d34a65a2c473fe

Change-Id: I53f6a97820da12e89d567e1bc10226fa27cafd84

Change-Id: I03580779ba69a71e51e9e4e1349fe2f5d77956d5

Change-Id: I71e7e1fbab6ec2ecc525f7d637199f6f7a69f8ea

Change-Id: I67237267be5ac4071ecdb480cacf51b4a2fb4f1f

Change-Id: I25167b28397db185cdf9fe5844feecb5a16cc13e

Change-Id: Idaa9a057bb17af2642706b144544df92139a1fce

30bb1edd originally incorrectly
changed some different 16bit numbers to 32bit, then
02abb559 reverted those 16bit->32bit
changes but accidentally changed the 32bit nAryLen to 16bit but
that was 32bit before 30bb1edd

Change-Id: I36863b62418d7a3b5ccce51aef86c63c523f9598

Change-Id: I875faea495a375b998c439a0c987b2241ca47ea8

Change-Id: I01501e99787118e7a7d8259dc65ae20a2932dea7

Change-Id: I095820c3abcb8e2b47e564096f050309b22cf901

Change-Id: I968c5cd5929e2587588f0e0547352acdc19e456a

Change-Id: I11bdec73ea12e189cca41f5f275616a138cf5793

Change-Id: Ia167c532afc8ba900629b3f056021deed45185ea

Change-Id: I475b8c9bdcea120ac7ba24862000a16d9f4127c0

Change-Id: Icf3a58257c8c8e90d7e9d38f167a9b9f80d4bf32

Setting 3 separators instead of 2 does not work.

(regression from 26315384)

Change-Id: Id0b4649404b8a49a7f89a237c46106bce4e146ff

Change-Id: Ifd6acd85a9bf33742cb62327a3b8a4320f35d0f3

Change-Id: I454e29c2cdf0857d329d84000d79973489cebb41

Otherwise we never hit cache directly, only after expensive call to
ImplFindByFont.

Change-Id: If15b368feeba94c8fff8ee7cbe049fc4a2069768

Change-Id: I35e295347a046376289f5d4fd5468860d0b8f0ae

Found when investigating fdo#79228 which was caused by the ScPatternAttr
change due to the number format being applied.

Change-Id: Ieada557f5e194401a9ce255f26d71d36e0704f35

Change-Id: I99deaf80f559e28de106b639663b262544cdf60c

Because the same dialog is used when creating a new pivot table, where
the Apply and Close semantic feels a bit weird.  Also, some features of
the dialog were designed with Ok / Cancel in mind (like the destination
location).

Change-Id: I9ac889bca9f967cb97b34ab45ea563547b58d422

Change-Id: Idd97331669ff204efb568ef1c17ba7444db4abf4

Change-Id: I24f814a402d2b8af3ddf888719b78d3158c2a38f

Conflicts:
	chart2/source/view/inc/3DChartObjects.hxx
	chart2/source/view/main/3DChartObjects.cxx

Change-Id: Ied38d4457fbc04ded42fe4fb55a8ec654cefa259

Change-Id: Id75c459cfa59fe2adfc0f748dc33d0bc9e271507

The last used color is stored inside
SvxColorToolBoxControl, so we must call its Select method
to do that. (This also fixes using the last shape, See
commit ce842113ae728e8995c7c5398204a20d0707e843 for a
fix of exactly the same problem in chart2.)

For the font color, there was also another problem:
The SID used for that button is SID_ATTR_CHAR_COLOR2.
The problem is that SvxColorToolBoxControl::Select
actually calls for that SID the ".uno:CharColorExt"
command instead of ".uno:FontColor" (which activates
in writer the format paintbrush mode). Given that since
commit 584b4159 this
command also receives the last used color as argument,
the easiest solution is to register ".uno:CharColorExt"
here as well.

Change-Id: I2db25a7fb537ce10dfec7b1c2d049dd77d9f8f6b

Clicking a shape button works only for the default shape,
not for the last used one.

Steps to reproduce:
1) Click on the drop-down arrow of one of the shape buttons,
   and select a shape. Notice that the button icon updates
   with the last selected shape.
2) Click on another button.
3) Click on the first button, and try to draw something.
   The button doesn't work anymore. Only opening the sub
   toolbar and selecting the shape there will work.

The problem is in ToolboxController::execute that requires
the command to be in the listener map, but we didn't
register sub commands (like .uno:SymbolShapes.sun).
Instead of doing that, we'll try some generic solution.
(Note that calling the corresponding method from
SvxTbxCtlCustomShapes like we do in other methods here
won't work, because ChartController::queryDispatch checks
for the "_self" target, but SfxToolBoxControl::Dispatch
doesn't provide it.)

Change-Id: I5be3dfd3ee3ca9ab9d8080929173dbe5984f61a0

also coverity#706500 and coverity#706502 and coverity#706503

Change-Id: If8c76f9fea669f3397a3db65fefaabe101bf9c35

The huge size of this dialog calls for a maximize button...

Change-Id: I68706669ac4e10ea9d5cba5526e9f1b0eb3628b6

Change-Id: I34f37587284893927463e214f0d7d1c3e914c7b2

Change-Id: I847e5e3652517d0c11eb2cf3e9625ed649595bc5

Change-Id: I48d7b746d7cc49246743c62480b0d225d2519342

Change-Id: I47582b072bb939cf270a76e430a9f7908b5c1d93

Change-Id: I5d5066413508a115f6b4c63b8b3dd19268f66574

Older versions of OOo/LO would mangle negative crop values on round-trip
of WW8, which now (since the drawing-layer does not ignore them) causes
the images to be rendered invisible; probably large factors don't make
sense anyway, so ignore the crop if it's 50x the image size, which
happens to work for the bugdoc.

Change-Id: I9f36d37e3be27234554bc91e80bfe719b1ce86af

Change-Id: I7fe1350e55bea2eead2c7fedf432da594aa6bd9c

Change-Id: I594d93c0630528841a1f243e6b6e9eda67b6138d