fractional second allowed in timestamp. But not in time. Don't ask why.

Change-Id: I5415889ab4cf4835b8b0db7363b7d9eb0a332393

Change-Id: I0aba2344afde94226cfbac5b3221e827e6774b3a

std::ostringstream::width is not sticky

Change-Id: I32d77bec68506b7691a4f86dadb24e62fdc13d42

Change-Id: I72a27a05c9e3e49b17817c2566fbd7f02b7835c5

Change-Id: Icbe5ac8e20b5a05fdba4ed7b8a287c3d9e94d2da

Change-Id: I3c0282d731082c32558a527110762dffad01b44b
Reviewed-on: https://gerrit.libreoffice.org/4558Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>

With JDK upgrade to 1.7.x focus gain/lost events were dropped and thus not
discoverable by AT tools.

The problem has nothing to do with different packaging of JAB (2.0.3 is
included with JDK 1.7u6 and higher).

The reason for the event droping was a new check in
DefaultKeyboardFocusManager.java that was introduced in JDK 1.7.x:

@@ -294,18 +271,10 @@ public class DefaultKeyboardFocusManager
        extends KeyboardFocusManager {

-      if (!(newFocusedWindow.isFocusableWindow()
-               && newFocusedWindow.isVisible()
-               && newFocusedWindow.isDisplayable()))
-         {
-             // we can not accept focus on such window, so reject it.
-             restoreFocus(we);
-             break;
-         }

With this check in place it is assumed that Component.isDisplayable() must
return true, for event to be dispatched. But particularly
`Dialog.isDisplayable()` didn't and thus all tab key events (focus traversal)
were dropped.

Many thanks to Jamie and Stuart for helping out with reproducing and tracking that down!

Change-Id: I4876773199ca06a6e73213401692e8275a1f3b22

Change-Id: I7a7e13104e7256da9a82e16e14ddfa04e35e7f8a
Reviewed-on: https://gerrit.libreoffice.org/4525Tested-by: Eike Rathke <erack@redhat.com>
Reviewed-by: Eike Rathke <erack@redhat.com>

Raptor already sets up all 4 error handlers in xmlSAXHandler so why it
would need the global ones in addition to that is a mystery anyway.
Messing with libxml2's globals can only cause trouble.

Change-Id: I2935efe5c4cd75d48cc4ecdeaa8437e91b48349e

Project: help  f92fe458ddf0b7a79a11f426eff8f271639c6ada

at the paragraph

- some minor reformatting and cleanup of older comments

(cherry picked from commit e4578fcc)

Conflicts:
	sw/source/core/text/txttab.cxx
	sw/source/filter/ww8/ww8par.cxx

Change-Id: I345a20c78f26be60de3eb01ac6f5303de46c2fc1

Change-Id: Id1028f87a6dded7d2c27fe8d793342574d06e3f9
Reviewed-on: https://gerrit.libreoffice.org/4528Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

Change-Id: Id90ef9f43578d952fd8d7872d4fd2fd622d07dbd
Reviewed-on: https://gerrit.libreoffice.org/4529Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

Change-Id: Icc70cc02d1b3b154fc1cab0419d89b437e2a366f
Reviewed-on: https://gerrit.libreoffice.org/4530Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

see origin f1ecd5ff, f9d08dd5c9af8f8128371540e
& follow up 0a686a9b

Change-Id: I1a10de6984f7ed78efdf6f32bfca14b433b24405

Change-Id: I0ca0edf4fc8f19adb1859eb4a679fee6480ea295

There's a document which has coordsize="6094,0" path="m444,424r6094,e" ,
where the 0 is somewhat questionable (and the spec doesn't seem to say
anything about it), but given it's a horizontal line, vertical coordinate
space size shouldn't matter, so treat it as 1, making the space visible
and avoiding division by 0 in our code.

Change-Id: I53725cd73a699b8e6d3c96e8fc39c276d082bc0c

a) make sure backcolor is skipped when libreoffice backcolor
is default. Libreoffice default is 0 ( but if we write that out mso
reads it as black )
b) don't take the label default backcolor in exchange for the
libreoffice default. Libreoffice default backcolor for label is white,
mso's default is rather different ( and never seems to be used )
c) re-enable compObj export ( seemed this was disabled unintentionally
except for buttons )

Change-Id: Ie68267809cee2e92bd06e88e698f17a10eeeb6ad

Reported by: Du Jing
Patch by: Jianyuan Li
Review by: sunying

There were multiple problems here:

- \f had a special handler for the listlevel destination, \af had not
- in case of multiple \f or \af, the first one is used for the bullet
  font
- in case only \af is used for the bullet font, its value should be used
  for \f as well

Change-Id: I6631504c1aa9f2e0792a3469a5fdce5b7bd49518

change for all text

- group intrinsic actions of language change for all text into one Undo action
- secure <SwRootFrm::CalcRects(..)> - catch NULL pointer
- correct <CursorGuard> - really restore the cursor

(cherry picked from commit 4207db47)

Conflicts:
	svl/inc/svl/undo.hxx
	svl/source/undo/undo.cxx
	sw/inc/IDocumentUndoRedo.hxx
	sw/source/core/edit/edundo.cxx
	sw/source/core/inc/UndoManager.hxx
	sw/source/core/layout/trvlfrm.cxx
	sw/source/core/undo/docundo.cxx

Change-Id: I972988bbc21a519e0956ff196aa93a46287b9a2e

... always so don't need to check for it.

Change-Id: I4bbc2a44a9c3d4ebde766a64a6ab78f7ead9c7af

All the comments of the form //-------- are removed from /svx/source/accessibilty/AccessibleControlShape.cxx"

Change-Id: I7ae9f656939e2512ce503a5ba103d765594f20ac
Reviewed-on: https://gerrit.libreoffice.org/4543Reviewed-by: Miklos Vajna <vmiklos@suse.cz>
Tested-by: Miklos Vajna <vmiklos@suse.cz>

The option should be used also at link time (as e.g. the gcc man page says),
but cairo doesn't do so, and since -flto makes clang generate .o with LLVM
bytecode, plain ld chokes on it.

Change-Id: Ia5ce9a98afa4f88b8d0943a5971450d7009af913

Change-Id: I0031094be63f2d8efeac42d5ca560f57af5326b6

Change-Id: I3406b5481991e459ac051868e441c0c04a2f7d5a

This reverts commit e36d227e.

The description in the orig. commit: "Font size at the end of the para
when there is a new(empty) line is set using an extra charprops
definition." That would mean the extra StyleTextPropAtom record does not
have a record header, which seems pretty fishy. I suspect the real
problem is somewhere else (or the ppt is broken). But I cannot check it,
because I cannot access the bug.

Conflicts:
	filter/source/msfilter/svdfppt.cxx

Change-Id: I1d555cd2905ef8f297d1a96f56bebbea8f83ad62

Change-Id: If7e998065390d89cc22556adcd727a63f6496aad

Check if we hit the left bracket, left parenthesis or left braces, insert the right close character
and set the current position to center the characters.

Change-Id: If1ee8a00799ef0933d5dfd51c32f669a2a27b2a7
Reviewed-on: https://gerrit.libreoffice.org/4374Reviewed-by: Michael Meeks <michael.meeks@suse.com>
Tested-by: Michael Meeks <michael.meeks@suse.com>

Change-Id: I37b2a892143892740674ce9882ed61ea88eda894

The interaction of environment variables and make variables is fun.

For some reason, the
workdir/.../UnpackedTarball/nss/mozilla/nsprpub/configure script is
run twice: Once directly from nss/ExternalProject_nss.mk, once from
the sub-make run from nss/ExternalProject_nss.mk. In the first case,
the AR and RANLIB exported by the gbuild make process propagate just
fine to the configure script. In the latter case, not. So add AR and
RANLIB assignments on the sub-make command line (to override values
set in some of the nss makefiles), *and* make sure the sub-make
exports AR and RANLIB.

Change-Id: Ibd55bc8a7e001106e12b2207500e74c7bd01c73a

With Clang, LTO means compiling to LLVM intermediate representation
instead of object code, and optimisation and code generation at link
time.

As expected, linking the single executable for the iOS experimental
app takes quite some time (several hours), and in fact it eventually
fails. Hopefully better luck on OS X.

Need to use xcrun to find the right ar and ranlib, too, from the
Xcode toolchain being used.

Change-Id: Iee69397c57bf1d622882ad78c188e1734f6cbd63

** CID 1038508: Use after free (USE_AFTER_FREE)
** CID 1038509: Use after free (USE_AFTER_FREE)
** CID 1038510: Use after free (USE_AFTER_FREE)

Change-Id: I3a1422e252af415536d9303e58ff85763c231921

The proper background is now set using HIThemeBackground.

Change-Id: I71e4287273f7345a18eee4c4fc494eea10f2deb2
Reviewed-on: https://gerrit.libreoffice.org/4287Reviewed-by: Thorsten Behrens <tbehrens@suse.com>
Tested-by: Thorsten Behrens <tbehrens@suse.com>

(cherry picked from commit 0a78e6e8)

Conflicts:
	odk/cfgWin.js
	odk/configure.pl
	odk/docs/install.html
	odk/setsdkenv_unix.sh.in
	odk/setsdkenv_windows.template
	odk/settings/std.mk

Change-Id: Ie124ffbb857e796923981e13dc3116c7af341056

Change-Id: I2db39313de7df003b59a00a6d6a210f29c3a24ea

Change-Id: I384e9d499eedbe87030417952bf057b16cf549d2

RTFParserState is private to RTFDocumentImpl, exposing it is not a good
idea. One more step towards RTFTokenizer not speaking directly to
RTFDocumentImpl, but through an interface.

Change-Id: I3f626d4f2c560a20d9e24105459e95a66d0a87ca

Change-Id: I1154b29119fd3fbb30d50246e94b3906bea24220

Change-Id: I1154b29119fd3fbb30d50246e94b3906bea24220