Change-Id: I2b4cdb3809de0e9c69253d2cd05714e15fdc8913

Change-Id: I5b3d8e0e20938d8e66045ea6282f37afe29474e2

Change-Id: I7a49ed58cb20c365308db703c67e43c6687f35e4

Change-Id: I562b1f7cd1e5e7f7fc03b8f54bd6117255703eb5

Change-Id: I8fc80274e7eb33c81062f8514f63e6d0682d1b8f

Change-Id: I212c6b990da655ad7947154ca9e62981b049489e

Change-Id: Ifc44041290d44c5e4dcbd9a0882b99fe0f8f62a5

Change-Id: I96a1090ea8617644748afdea2786949fb0506b72

Change-Id: Ieddd13abc6bb23904db10f303dfc85cd87cfaa03

Change-Id: If2e8e7eddd5e75a211940416b4370c8f7768849e

...at least the Linux-deb-x86_64@56-lhm-ubuntu-trusty tinderbox complains that

> include/LibreOfficeKit/LibreOfficeKitInit.h:223:77: error: ISO C++ forbids casting between pointer-to-function and pointer-to-object [-Werror=pedantic]
> pSym2 = (LokHookFunction2 *) lok_dlsym(dlhandle, "libreofficekit_hook_2");

Change-Id: Ibd75a44fcc7da35ea32bde3fd2a300a16493789f

This way it's more clear that sw::FrameClient has nothing to do with
sw::Frame, also later renaming SwFrm to SwFrame won't cause confusion
about why do we have both sw::Frame and SwFrame.

Change-Id: I87c4d199d418c77fdbb600fc388c8b9f026cb0fa

The map mode is in general disabled during tiled rendering, so mouse
positions can be sent in using twips, but here we have to temporarily
enable it, otherwise the width will be returned always in pixels.

With this, the gray background of the sidebar has the proper width, not
e.g. fifteenth of the expected value (using default zoom).

Change-Id: I4380ee0ba6bcda97cf71735161dbdc826e7a2532

Change-Id: Ie682cfe97b8c48bc66575b89875ca617dbb8ab4e

Change-Id: If15ca89f2b7c54441f076eb233c4e1414dac25c0

Unconditional call to lok_doc_view_reset_view() resulted in hiding the
cursor, but then nothing enabled it. As a result, the cursor got hidden
after the doc size changed.

Change-Id: I06e7c9293e7a4cdbc73421a82430816ab0001f90

Change-Id: I7e266a0e29d86247d01df38da00899a8a1700719

Change-Id: I9ff07fcd5160cf218dc5add2db362aa329544708

Change-Id: I83702435e9f8e0e73d6a3ecee1e6a7a30dda52d9
Reviewed-on: https://gerrit.libreoffice.org/19886Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

Change-Id: Icbd163bac65cbb9e4ad59697b87f6c979ecb41cc
Reviewed-on: https://gerrit.libreoffice.org/19718Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Project: help  c3b9e2711f3670e3a9f5fdf7abb7e4dca8ea1f2b

tdf#95698 correction a parent tag for paragraphs

ful_func.xhp -- changed the parent tag from 'variable' to 'section'
	for paragraphs

func_im* -- canged corresponding links form embedvar to embed

Change-Id: If150119308bd5ce93a74dee83f1280810a8c5fee
Reviewed-on: https://gerrit.libreoffice.org/19859Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

According to RFC 4918 the value used for lock timeout is unsigned
32 bit, see:
<http://tools.ietf.org/html/rfc4918#section-10.7>
for info.

This patch fix the way the timeout element of lock response
payload is parsed in Windows.

Change-Id: I335a1cb884c3ef1c2362b00981a2784d9232b23e
Reviewed-on: https://gerrit.libreoffice.org/19867Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: Ifb85cac1289af94102947b9064a8c12ca65ad43b
Reviewed-on: https://gerrit.libreoffice.org/19404Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: I10e532dcbe98384fde9c553ccd2e151abf53251f

by using std::assert rather thatn DBG_ASSERT.

Change-Id: I8fca2b82f8d86a843c024556a0a29c7848b1e602
Reviewed-on: https://gerrit.libreoffice.org/19294Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

SwAccessibleParagraph::GetWordBoundary is called when you navigate
word by word normally done by pressing Ctrl+Right/Left arrow. This works
well as long as the word is an actual word. Problem is that what we
land on is not always an actual word the cursor will for example stop
at . (periods) and at the moment that won't be exposed to accessibility
tools such as NVDA screen reader. Orca on Linux seem to have some type
of workaround but it dosen't really work. I'll try to find out if the
workaround is actually in Orca or in the way LibreOffice exposes things
to atk.

Change-Id: Ibe5c5899d489b0449282998a3ba284f689ca9a36
Reviewed-on: https://gerrit.libreoffice.org/19424Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: I9822c930bb7e133306a3e90fd80f29648877d5f9
Reviewed-on: https://gerrit.libreoffice.org/19596Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: I26e448de315b56f36cfeb74a66c0db6a9148d9b0

Change-Id: I56336b8163de48e424526f5f426c2ad350292627

Change-Id: Id2359f6ff4bddb2afbc0b346e17cd858f00179e3

Since we now expose the bullet character to accessible text we should
not also expose the attributes, that will make NVDA read the bullet
twice. We can also remove some special cases for the fact that the bullet
earlier was not apart of the accessible text.

Change-Id: I5a8b84f92c32da7ffabdd5136655d88a2d149e75
Reviewed-on: https://gerrit.libreoffice.org/19106Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Project: help  886917bcac6656b66f78c84c57cc308299992008

tdf#92825 Rearrange the view menu in writer, calc, and impress

Change-Id: I707e1daa263acb5a09a6a90e745a619f9dd67388
Reviewed-on: https://gerrit.libreoffice.org/19721Reviewed-by: Lera Goncharuk <lera.goncharuk@gmail.com>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

This removes the restriction for returned coodrinates from ScViewData::GetScrPos().
The reasoning is the following:
1. In the existing implementation, the limit enforcement is flawed. The routine is
capable to return negative coordinates, but only positives are checked. Also, the
routine code itself uses value greater than 32767 (specifically, 65535 in case of
col/row exceed MAXCOL/MAXROW), that later gets truncated.
2. In case of tiled rendering, it's already possible to see the numbers greater
than 32767.
3. Given that the code works for long time and produces big negatives, as well as
big positives in case of tiled rendering, it's highly probable that existing code
is prepared for large returned values.

Another option could be additionally check for !bAllowNeg before emposing the limit

Change-Id: Ie78cfe5907253e6648fed27dfac78b292922f5b4
Reviewed-on: https://gerrit.libreoffice.org/19816Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: Ibb9114d34f2695aa85858e456518e36dbcbe0af1

As many users asked, now there is a checkbox inside Tools->Math
called "Auto close brackets, parentheses and braces". This option is
enabled by default, but can be turned off now.

Change-Id: I6f96201c0720fb62fc2dce99222f97194e930fbe
Signed-off-by: Marcos Paulo de Souza <marcos.souza.org@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/19750Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: Iee9987aeafed126e4324d8d487289469577fafa2

Moved the page resizing from SwLayoutFrm::Format to
SwPageFrm::MakeAll and calculated the new page height
more accurately based on the content (stolen from
the browser-mode height calculation, to be refactored).
This fixes issues with repagination with certain contents.

Removed unnecessary ToggleLayoutMode calls and
prevented clicking between pages in single-view
mode to jump the cursor.

Change-Id: I5662dd26efff66d6e95b8fd4dcf8564588adbbb7
Reviewed-on: https://gerrit.libreoffice.org/19699Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

This also makes SalColormap::GetColor() return an RGB color
if m_hColormap is valid.

Change-Id: I154dc43a2be54f7da473263f9f2a29db0cf2fd4b
Reviewed-on: https://gerrit.libreoffice.org/19836Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: I977d1fbedb1b784cb95273184ec5e5a7fa156eef

Exemple:

<svg width="100%" height="100%" viewBox="0 0 800 400"
     xmlns="http://www.w3.org/2000/svg">
<defs>
    <linearGradient id="grad1" x1="0%" y1="0%" x2="100%" y2="0%">
      <stop offset="0%" style="stop-color:rgb(255,255,0);stop-opacity:1" />
      <stop offset="100%" style="stop-color:rgb(255,0,0);stop-opacity:1" />
    </linearGradient>
  </defs>
  <ellipse cx="200" cy="70" rx="85" ry="55" fill="url(#grad1)" />
</svg>

Change-Id: I908e3d235b3e5ba0b7e4fe3666b6334e3c409d03
Reviewed-on: https://gerrit.libreoffice.org/19866Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>