Change-Id: I5e57e6d506504933096f4c8e371b0ddc93c04cbd

Project: help  08790edcf50d4281d55e5c28c0c07f88a99241f3

Change-Id: Ice43b710b16eed79870d561e6dcf14f4a07cad9f

Change-Id: I8a4e8fc801bae7826d37b9ff29c1535416623f12

Change-Id: Iefb2ef68cf769c78f4e41c14a02afc9356e8e495

Change-Id: Ie7612e4599699977ddca1f581eafd6aca4003f5d

Change-Id: Ie3494fe4cee904d8eec65aa5cad861ab1916e8d9

Change-Id: I5c6b2e603cfada736fcca21d5124c01bfab77831

In the bugdoc, the chart legend had 6 entries in PowerPoint, but
only 5 in Impress. I played a lot with padding and offset parameters
in layout algorithm without success. Finally I introduced this 1mm
tolerance, so the fixed legend frame can expand up to 1mm, if necessary.

Change-Id: Id1ad1df67aab41d942133cea8d16fccf0f215167
Reviewed-on: https://gerrit.libreoffice.org/15149Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

mostly changed due to loplugin:constantfunction

Change-Id: Ib677c78db256f8032a99b3f02a3e363fee68ebcc

Change-Id: If08968487f62d0bc4c827bb69be25db6f9d5a0ee

Change-Id: Icc432d6d3e40be9a6e1956c2d0c1dd2250c273a9

Change-Id: I146fa693e89b00c77df33547971fb3a49c7e1b34
Reviewed-on: https://gerrit.libreoffice.org/15148Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>

Not sure if the comment in desktop/source/deployment/inc/dp_persmap.h: "should
be removed for LibreOffice 4.0" refers to just the DISABLE_BDB2PMAP parts or
the class and functionality as a whole.

Change-Id: Ie701576efc4076ab8be02a89145b5ecfcaaebc53

Change-Id: Ieead62bcbfab5a96b5350bf30fbbb712f981f10c

Change-Id: Ifc728dc0b42d713a5eaa2fc22d3e71c6aa4f196d
Reviewed-on: https://gerrit.libreoffice.org/15146Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>

Using a similar fix as Caolan's one:
http://cgit.freedesktop.org/libreoffice/core/commit/?id=8f9bc93ba3299efbd8ad181481e6e2e87df7c066

Change-Id: Ib33f9585bf1e1c2cf6dd367bfe4eb0604794decc

Change-Id: I17dfb0a818838b00d9148ee1c6328ef094eab99b

Change-Id: Ide66aecd5bd9beda298da189c3564908c5756cdd

Change-Id: I2eb14bfb9a8f634a7fc56f581f5dca7529c5e85d

Change-Id: I15ed795340c73b2d12b0e93990ede5a1e7284533
Reviewed-on: https://gerrit.libreoffice.org/15145Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

Change-Id: I22cbcb9d1f15ab18a8d70c1793c64f04f27389a2

That makes my life so much easier when I need to debug where we screw
up.

Change-Id: Ia0b54eef47a06c43e3d0ba66b909e398e02a6d0e

Change-Id: Iffb7a766f15475e3600332a6c76a15a1d192d7c8

Change-Id: I1fd67449b9182db8eb9fb0e58b092cef00283c2a

They are not shown anyway and just confuse us internally as they contain
a width but no color and produce invalid documents during export.

Change-Id: Ibe20a27b0172ece86dbb1f734df567b903465ebf

Change-Id: I96bafad0cc73b27c2a08b5205a35b10b4901afb9

Change-Id: Iab334625fbbe2a8d991a17b05538897a23fc34db

Change-Id: I635efd806bf9a6d158fb0df7a1705f42966ba669

09:41 <@dtardon> vmiklos, maybe it's a regression in libxml2. "Fix XPath '//' optimization with predicates (Nick Wellnhofer)" in NEWS looks suspicious

So tweak the XPath not to use // for now.

Change-Id: Ia1c4508d8aede5818bc258d6e4243f026da3a3cd

Change-Id: I88a48ada4f9e261e425127bc1b0d3db83f1fabba

Change-Id: I82f66218bf02898e523e4f69fec9166aa00d0b83

Change-Id: Ie195e46e1cb14b39875efd1c9689db0f13968e87

My first idea does not work that well as it introduces some nasty
problems with the lifecycle. THe new approach will only allow to use
cond format objects of the same document.

Change-Id: I276c3750ef24cdee6d63162c678ec839d4845b08

otherwise we lose our selection with multi-page tables

Change-Id: I4f52d3cfb7a6d3c2cc1dbb68bbd5ce5c0005b1c3

Change-Id: I522a236015c91c0744b3018d3d4e4d6c65ea19ab

for listboxes in WB_DROPDOWN mode which means the dropdown exists,
but when the drop down is not activated.

I think it's plausible that there was confusion around "b_IsDropDownList" and
that the meaning taken by the author was the the dropdown is active if
b_IsDropDownList is set, so renamed that to bHasDropDownList and take the old
"b_IsDropDownList" path when "b_IsDropDownList && ..IsInDropDown()"

Folded the then both equal code paths of the LISTBOX code together. No idea if
there should be an additional equivalent AccessibleEventId::SELECTION_CHANGED
for LISTBOXES like COMBOBOX instead of
AccessibleEventId::ACTIVE_DESCENDANT_CHANGED in the dropdown not active case.

Change-Id: I6db3ec21b392b7bf1b488235464202bb326f2535

Change-Id: I1d3a6a552bd535622009b43e0633a357b1a919e6

jdiggs: I believe [VALUE_CHANGED] is expected for Windows' accessibility APIs.
However for ATK/AT-SPI2, value-change events are expected only for the sorts of
things which implement AtkValue (scrollbars, progressbars, sliders,
spinbuttons, etc.) Combo boxes, lists, and the like should not be emitting this
event.

Change-Id: I9dcab7fb4004a24f88915f904eb508dfee2d4ae5

Emitting object:state-changed:indeterminate is appropriate for things like
tri-state checkboxes when the state is neither fully checked nor fully
unchecked. ... jdiggs think that this event doesn't make sense in this instance
for any platform's accessibility API

Change-Id: Id8288522a825ad7ff61524ad4f225a79c015bb8f