Change-Id: I6bfce155151b79cd0a1707744e79298b45d8752b

firebird in debug mode abort() on lo shutdown due to a lifcycle management
issue where firebord try to create a directory in a temp directory that
is not there anymore, and then exception for it and
that exception in debug mode cause an abort.

Change-Id: Iabbd6ffda34a84e5b0d8e44761226083c8c1d168

Change-Id: Ie0460c8db6b1e7fcc9bab1f5e9ae1fbd00987994

Change-Id: I10d6493997b4f3bfab769340c990c01a26f3da20

Avoid heap allocating UNO object wrappers for the underlying Node
structures only to convert to Any and immediately free them agian
when we can.

Change-Id: Iae4612e9602f872f5d8cca2e516df594c9f1118c

Also re-use a single set of OUString property names in
a sequence rather than re-allocating them constantly.

Change-Id: I7f0afc52363e57ea0c63f46f1e1f63cb752bb302

Saves 13.5m cycles on a headless start, 1.3%.

Change-Id: I73f0dd8a523f84c765a217ee95aea9bcc3db1b5d

Change-Id: Ib170b3cec17402e38bf5555e21f83f44d70bb574

Change-Id: I57c81b5ec7a6541a825e42f9b68e7c5fa98f316f

Accelerates headless startup by 6.5m cycles or so, 0.6% or
so but something.

Change-Id: I9edca3d0c1c81d865e8369fee5cf193da7adb97c

Change-Id: Ie3bb69d93b97deab0830c2b45ba3428a8931a949

It seems some of the warnings are treated as errors
in Linux-with-check builds.

Change-Id: I9ed876ba634d944c022838a625164f06f100a7df

Change-Id: Icff534ac96e0a595d71bc34b6055caf8cd096ad8

Change-Id: I20c3715bcfb5212977699d661fd3acf1be28499f

Change-Id: Icefef960e1ecc0f6529e8f1b5f903c676b62e154

Change-Id: I4a8a05905a779bb8c7fad830c049d1e403d79711

Change-Id: I78fe90f24805c5bca3548cf4c005d2c44574aa5c

Change-Id: I6d39bc1a1132972eece99f1ea05cb25962805351

Change-Id: I6342a407878e24e671615b0c12a9badcfc774f37

Change-Id: I8b045cf44b3e62616c8c3e0a896524af13310dd3

Change-Id: Icf49a339fb23d6217dd0bc41c9012233717ad808

Change-Id: Ib4a3f8401e2ddd4cb76c76c968151799c86b5845

Change-Id: I5fe887cdb800a5f5cc6f6277b52a5f02c1b9bb84

We prolly need to look at iODBC and how to carry the headers
as 'external'
Note the dylib for runtime are still in 10.9

Change-Id: I6cad2b90fcb61b5787c0599e40d48a74b8ff917a

Change-Id: I3eb71120aed85a4ec2c697d149006bce063d983e

These cause problems in drawinglayer, and nobody needs a 5km clip.

Change-Id: Ic485250413194adbcd0ccf855f12e89e532d7120

Change-Id: I37438cd3f9c8e197e5a3aa3d9cbcc5bf3681d792

...after 0181a139 "odbc properly support
platform with sizeof(SQLWCHAR) = 4" introduced usage of RTL_TEXTENCODING_UCS2/4
there, which do not make sense in combination with converting between OString
and OUString.

OTools::getStringValue will need a corresponding fix, too, in the other
direction (where the OUString(sal_uInt32 const * codePoints,
sal_Int32 codPointCount) ctor will be useful).

Change-Id: Ia94cd0deec46d269b6ee43362f4849837bb011c5

DOCX export tries to write accurately both <w:br> and
<w:pageBreakBefore>, while DOC/RTF prefers just the page break before
paragraph attribute. Given that these later two exporters are not
adapted to write their equivalent of <w:br> at all, add a virtual method
to determine if writing page breaks as "page break before" is preferred
or not, that'll give the expected RTF export result.

This conditionally reverts commit
a31fbb53 (fdo#74566:DOCX: Preservation
<w:br> tag for Break to Next Page, 2014-02-06).

Change-Id: I4a5dd295b07739a570cd8cc6709561e64d031273

Change-Id: Ie9a95b99e1f6de65985b295076f5f89e5d99ec93

(which loplugin:unreffun did not catch, as it was referenced by a
friend declaration in vendorbase.hxx)

Change-Id: I8a1afd4699b56425b580c1e63557342ee7ec1cb2

Thanks to Rene Engelhard for help to make it valid.

Plus glu is not used, so remove it.

Change-Id: I5e3631e9b3d24c2b1d9044192d7c0a55c234f7c0

Change-Id: I6e007cc3a9fd222c1c1381c8b4f890966c05cbab

when a Paragraph Style is assigned.

(cherry picked from commit 6a1564284d90c7e175f7848520924f7f6a5884b7)
(cherry picked from commit 4a381b1124f60b76cabb558fa24827448a1caab0)

Conflicts:
	sw/inc/doc.hxx
	sw/inc/editsh.hxx
	sw/inc/fmtcol.hxx
	sw/inc/ndtxt.hxx
	sw/source/core/attr/format.cxx
	sw/source/core/doc/docfmt.cxx
	sw/source/core/doc/docglbl.cxx
	sw/source/core/doc/doctxm.cxx
	sw/source/core/doc/fmtcol.cxx
	sw/source/core/edit/edfcol.cxx
	sw/source/core/txtnode/ndtxt.cxx
	sw/source/core/undo/unfmco.cxx
	sw/source/core/unocore/unosett.cxx
	sw/source/core/unocore/unostyle.cxx

Change-Id: Ifa1214876df296f2e1210cc4a35f2df67f9f6bc4

Change-Id: I658974102f544185bf27f985bc920d8d54b40656

Change-Id: I6c9b85ab587f6fb9f585c9714c3a1076f627c7d3

Without this, in case TextAutoGrowHeight is set before TextBox, the
textbox of the shape won't have automatic height.

Change-Id: I2ebf77eacbe1f1d10ccf64ef833c39ee965245d3

This dialog used to have no OK / Cancel buttons, and the data would get
flushed when the dialog got dismissed.  This behavior got lost during the
dialog conversion.

Also, SaveModified() does check whether or not the data table is modified.
No need to call IsModified() before calling SaveModified().

Change-Id: I5cea88cb52127cc5c7c1be67a5a592d87df92d0c

Change-Id: I06d5a6c93817d2623fac3962b82c1319caf13276

This was triggered by connectivity calc driver, which on shutdown
was trying to 'dispose()' it's connection
which in turn would, if the shutdown of calc was already done
try to re-create an sfxApplication instance, just to delete
illusory inexistant documents...

Change-Id: Ic166367ca399d613743f379ebbfbffb5813701c0