In commit 88a874fc "convert SfxItemState constants to a proper enum"
I made some mistakes in converting bitwise logic to boolean logic.

I fixed one of those places in commit 7ad83656 "fix bitwise->logic
conversion in SfxItemState commit"

This commit fixes the other places where I converted bitwise to normal
boolean logic. I also validated that none of the existing code tries to
uses combinations of these enum values.

This commit also introduces an exception-throwing check in the one place
where the enum is explicitly cast to make sure that no combinations
sneak in.

Change-Id: I545f7d17b76c4fd999078867caec314e83ffe165
Signed-off-by: Stephan Bergmann <sbergman@redhat.com>

from ParagraphObj, by using boost::ptr_vector.

This is also a supposed fix for ParagraphObj::operator=(),
because ParagraphObj::ImplClear() now makes the vector empty
while its original code kept deleted pointers.

Change-Id: I7710eb33fc05bed55474ff00baaeb8e19cd79f3d
Reviewed-on: https://gerrit.libreoffice.org/11379Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

hmm, make CppunitTest_i18npool_test_breakiterator fails after all

This reverts commit 693f72384e081429cddb29415276cc4d07b05645.

Change-Id: I3339324ec99bbed56f81f2aa4e6cf54edcbec8e6

Change-Id: I88eeac06413fc3935cbbdb357ff8bf9acaa383a8

Change-Id: I64be7ca711dcd3ea7c4d5840a30f2f701d055e1b

Change-Id: Ic8751562e4dbfa5b6be903d87d369a750d305495

For the temporary document there is no layout so SwTxtNode::MakeFrm()
isn't called.

(regression from 6db39dbd)

Change-Id: I711d32a81519edd6840cc29f5cd9dbc57ef21982

Change-Id: Idcd7de83ff7e551b350db0bf80e2ac78baa67686
Reviewed-on: https://gerrit.libreoffice.org/11374Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

valgrind + bff

Change-Id: Ib3ed8a6e518c0686f8cbeaf021b9ca3a07005032

Change-Id: Id1c48d99f18736de1e9b0edf4b8a77becaa1c208

Change-Id: I90135b6d7780ba8a22fff725a965f0ea7e442707

Change-Id: I94fed6a9361f21457b3e631efffc0db833068aef
Reviewed-on: https://gerrit.libreoffice.org/11256Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

Change-Id: I274236939c6de4db811f8932025c619479bebbee

so that the global aLotusPatternPool map gets destroyed at a sane point in time
where calling delete on its patterns is a legal operation and not at some
random location in the dtor chain

Change-Id: I6f202c5345c6a227b5a680d70150bf32fb9450c7

Change-Id: Ie31c16318b09699e080484292d489a378e3a6dce

Change-Id: I15ead38cadbee5d092c96010c8e9a7253c2027cc

When the Xft.dpi setting is not found, and if only one screen is
available, fall back to detecting the screen DPI from the reported
resolution and monitor size.

Change-Id: I1481d94b5dc4072c1f8da4659a221dfc7971080b
Reviewed-on: https://gerrit.libreoffice.org/11377Reviewed-by: Chris Sherlock <chris.sherlock79@gmail.com>
Tested-by: Chris Sherlock <chris.sherlock79@gmail.com>

Move the validity check for the DPI (>= 50 and <= 500) into its own
(static) function, since we're going to use it repeatedly in the
fallback DPI detection code.

Change-Id: Id5d9b10be385d45549bbb57cfa14939cd51a0324
Reviewed-on: https://gerrit.libreoffice.org/11376Reviewed-by: Chris Sherlock <chris.sherlock79@gmail.com>
Tested-by: Chris Sherlock <chris.sherlock79@gmail.com>

Just like the debug documents in SwDBManager::MergeDocuments.

Change-Id: I275575853c80d4e19d6df5ba9d8dc689353d7c85
Reviewed-on: https://gerrit.libreoffice.org/10970Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>

This adds a lot of SAL_INFO messages to SwDoc::CreateCopy and
SwDoc::Append.

Change-Id: Ie924e20698bec84482fcc5496175a9e29245729e
Reviewed-on: https://gerrit.libreoffice.org/10969Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>

This drops all the specialized, workaround code from MM, introduced
to use the SwFEShell::Paste function and additionally merges
and renames SwDoc::Paste into SwDoc::Append.

There is still a little common codepath, therefore this adds
comments to always update both functions.

Change-Id: I704b3ef3257dd977dac95e16e25049ff8ade97ed
Reviewed-on: https://gerrit.libreoffice.org/10967Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>

Change-Id: I83b9c86685f72569c0df88231458183b1ae511c8

Change-Id: I6627363aea34570b5f7d118b1d64b6a2d93c19cb

Change-Id: I10e4a0aa23fe83fa0f2edd671b321fa4588c880f

Change-Id: I97d5a8dc85fe77c6bf0235425db0a7b9e42ff869

Change-Id: Ibf478abae5aa0574279e8b45c26b96ab66bfe203

We really don't want to call on GtkSalGraphics::Init() twice, the only
reason we do this is because we pass in a new screen. To get around this,
I've created a new GtkSalGraphics constructor that takes and additional
screen parameter.

Change-Id: Ie53d609c2b8f64d77a3bfb2efe0e97027c75c5d0

Change-Id: I1dadf4f905fcf06aee616cc0cfea3d77a145a3a1
Reviewed-on: https://gerrit.libreoffice.org/11371Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: I1a840346e75a3d7dd0de761660a0ea95c7ee034b
Reviewed-on: https://gerrit.libreoffice.org/11370Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: I284a1b2bcb61695b74ec51bb46cbb683abf83449
Reviewed-on: https://gerrit.libreoffice.org/11369Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: I6d3e592de64c63e66bb788b112e188ef217e7bf7
Reviewed-on: https://gerrit.libreoffice.org/11367Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: If5e77da2b42b126e2c8fd7613b6adaed635fd683
Reviewed-on: https://gerrit.libreoffice.org/11364Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: Ic8d9c97ee6bc48c8440c27c5ae4fc9bb20fa5acd
Reviewed-on: https://gerrit.libreoffice.org/11359Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: Ibbac337b481e138ff9bfaa9d350b52e368ba5f2b
Reviewed-on: https://gerrit.libreoffice.org/11355Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: If8addcc5d95f610ade404e62e4a2a775319e9ff3
Reviewed-on: https://gerrit.libreoffice.org/11354Reviewed-by: Noel Grandin <noelgrandin@gmail.com>
Tested-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: Ide5705323579aa5ea1e725685acb7421f416cf04

Change-Id: Iff4da6d6281eb9194db348ebc10fbe7718538401

To make 'if (rBackgroundFillStyle.mnThemedIdx != 0)' actually work
as intended. Avoids invalid call to pTheme->getFillStyle() which
is not supposed to return 0 here.

Change-Id: I323e2628cba64167240e8f0b945e6693b47b8e77