Version 7.1.4.2, tag libreoffice-7.1.4.2

bump product version to 7.1.4.2

Change-Id: Idba2aaaa443c5d43bb52c6a6ab9c4aa307c0f5df

properly draw only parts of FillGradientPrimitive2D (tdf#139000)

The size of the whole gradient is getDefinitionRange(), while
getOutputRange() gives the area of it to draw.

Change-Id: I85dd7fe51bcc9c332a6fb0e6748d5ac89266b910
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115923
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
(cherry picked from commit 9bf9fea0afadef0912d38f0d08b2bf2959569100)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115850
(cherry picked from commit 887f1dbbda9dd371f52111c2f5a3bbbc67fc5615)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115960
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
Tested-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>

update credits

Change-Id: I46f39d507cc96cbcd0d59b4edfcbb109efefdb69
(cherry picked from commit 0d42ca4f39160bdc28a5855261c227c63161dd8a)
(cherry picked from commit 5f5c4036e5469893c3eec7a532a1df84bf78da8d)

Update git submodules

* Update translations from branch 'libreoffice-7-1-4'
  to aa878d129a9f44b37b3c8f220cdc0f6ddee18d7d
  - update translations for 7.1.4 rc2

    and force-fix errors using pocheck

    Change-Id: Id46165455a260dfa20c5481c1e19a8b2a9426f5a
    (cherry picked from commit ee29ff1930cf1907812365209b69326a17400969)

tdf#142546 broken formulas in .FODT files

regression from
    commit 81fe31d2c8977791f9b90ab1da4fbb1a778f87fb
    Date:   Tue Nov 3 18:25:10 2020 +0200
    use fastparser in XMLEmbeddedObjectImportContext

Change-Id: Ib5bae81001ae97c574c078c979484f016c7d4323
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116664
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 06a983b213b8fb71fdac1004868d7b4fdd6a9833)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116470
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
(cherry picked from commit 5f70c1f1ebd67db3f98ec132baf40e629de7a1c1)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116471
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Michael Stahl <michael.stahl@allotropia.de>

tdf#136292: Fix crash on invalid JSON from the extensions API

Change-Id: I91ed555f93173dee10e87025b231560693be50b5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116390
Tested-by: Jenkins
Reviewed-by: Muhammet Kara <muhammet.kara@collabora.com>
(cherry picked from commit e98b0891d6e0f1d6be4e4a4761bd1ab16bfab85c)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116168
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
(cherry picked from commit 9a9089d0f7c205ed6428e55fb86062c715483a4b)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116171
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

tdf#138302 partial revert tdf#108496: DOCX: redesign of override in lists

LO 6.4.5 commit cf13fe3e6f6a40f6db064d65d4514d13a23a31f0
was only concerned about DOCX, but just elinated shared code.
So restore the shared code and just don't execute it for DOCX.

It doesn't look like RTF was working before this,
but this does fix DOC format.
I looked for existing unit tests in rtf and doc,
but none were found that hit this code.

Change-Id: Id624f73181384f38e1ef9f27575e0fb82eea19c7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116349
Tested-by: Jenkins
Reviewed-by: Vasily Melenchuk <vasily.melenchuk@cib.de>
(cherry picked from commit f49e5902a3737c02fabf5bec23243ccb126426c4)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116170
Reviewed-by: Justin Luth <justin_luth@sil.org>
(cherry picked from commit 5107bc515fba130bf4a488324f97971789a92f37)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116453
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Miklos Vajna <vmiklos@collabora.com>

tdf#142150 nothing already selected on MouseUp without previous MouseDown

if mouse down was on tooltip (possible on macOS)

Change-Id: Icb4328fa6977356c3bac5f37d42e99ae8d4e4b98
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116158
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
(cherry picked from commit 2a0d5643234cd0989f91589695694c809d82e344)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116159
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

tdf#117957 WMF Replace ellipses with rects to fix tests on arm64

The issue is caused by floating numbers,
On x86_64 it is:

  338.499728160392-445.269903703769

on arm it is:

  338.499728160392-445.26990370377

As in test files, we have two ellipses (which are made from small lines),
which points rounds differently on different architectures

Change-Id: I44dd5c7b3b13f3b1680d5a3a8348d6f467cb2f7c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116270
Tested-by: Jenkins
Tested-by: René Engelhard <rene@debian.org>
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116314
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit a1168e2c58daf2efde944970472c67c46c2e8c03)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116161

tdf#142479 on 'return' select the entry the cursor is in before calling Select

Change-Id: Ic11f8a12c93d1999dd6fbe555b9823660eb703d0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115968
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
(cherry picked from commit eb5b345172516b31ec43e226c242e536e93a3413)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115971
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Adapt to recent LO download.lst changes

Change-Id: Iac481bb016c69bf444cbf4c91f3d0ff79714783c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115882
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit d401b26b647dcabcb52243fe9a965df12001ca7b)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115841
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Tested-by: Stephan Bergmann <sbergman@redhat.com>

tdf#142294 drop placeholder on-demand node on directly populating node

Initially there's a <placeholder> entry to indicate the load-on-demand
state.  It needs to be there for the expander indicator to be
drawn/shown when there are no "real" entries yet.

Typically the user clicks on the expander, the expansion remove the
<placeholder> entry and calls the provided expander-callback which has
the change to populate the node.

Here RequestingChildrenHdl is the expander-callback (set via
connect_expanding) and populates the node like it should.

But then there are two places where it is called directly outside the
expander-callback so it merrily appends entries into the node, but
because it's called outside the expander-callback the <placeholder>
entry was never removed which is then a problem because of the resulting
inconsistency.

Change-Id: Ie7d034e3f41832591438127c82eba85deb610952
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115952
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
(cherry picked from commit c4767868f1ee024efdd672cb21de9e3258c56694)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115959
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

bump product version to 7.1.4.1.0+

Change-Id: If62d284aa7bdbf805af58d07a2b27c174c3866bd

Branch libreoffice-7-1-4

This is 'libreoffice-7-1-4' - the stable branch for the 7.1.4 release.
Only very safe changes, reviewed by three people are allowed.

If you want to commit more complicated fix for the next 7.1.x release,
please use the 'libreoffice-7-1' branch.

If you want to build something cool, unstable, and risky, use master.

Update git submodules

* Update translations from branch 'libreoffice-7-1'
  to 9b312fbdba1d75c6d086f4b3ee182e86de9b7797
  - update translations for 7.1.4 rc1

    and force-fix errors using pocheck

    Change-Id: I5cbaffa0b102e88fc4a075332f18b72e32cf5855

Delete duplicated item in Korean autocorr list

Delete duplicated item in Korean autocurr list (DocumentList.xml)

Change-Id: I3cfaf27abb859dbcfc849082932006d20d0b22ad
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115799
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
(cherry picked from commit be622ccf9b60290d0577314ac9d76fbcf1cce636)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115832

tdf#138209 ODF export: work around forms problem in LO < 7.0

LO without commit 519d96fd8b83ef4c61576d87b58f97b7e6e6e3c6 makes a mess
when storing form documents it has loaded from ODF 1.3 documents: the
XML parts are stored as ODF 1.2, but the storage (and therefore manifest
entry) keeps version 1.3.

To avoid this, store form documents as ODF 1.2 extended by default.

Unfortunately a bunch of ODF export code accesses the global
SvtSaveOptions variable; with this version override, only
SvXMLExport::getSaneDefaultVersion() must be used.

Change-Id: I5fa8e286f5103c578ed0d93da07a8a6cbe2f0ddd
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115357
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 16de54a5c47fbc4691ee099c1f7bb559a8fe11ac)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115390
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>

vcl PDF export: fix re-exporting PDF images with arbitrary page-level rotation

Building on top of commit bd520b177637d4b7d9d93733103cff17a3c91b0a (vcl
PDF export: fix re-exporting PDF images with page-level rotation,
2019-11-06), this was already working for 90 degrees, now generalize
this to work with 180 degrees as well.

(cherry picked from commit d7d43cf460d66354a40ffa3b34c0f9efcd42d0be)

Conflicts:
vcl/qa/cppunit/pdfexport/pdfexport.cxx

Change-Id: I5a5d51662399814d5554d7c2cb86a6c9a2974012
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115523
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#137810 sw: fix SwXNumberingRules setting CharStyleName

During ODF import, due to removal of the pDocShell parameter, this hits

            else
                rCharStyleName = sCharFormatName;

while setting the "CharStyleName" property and later
GetNumberingRuleByIndex() prefers m_sNewCharStyleNames over the
format set in the SwCharFormat??

Also, "BulletFontName" has a similar problem; otoh "HeadingStyleName"
only makes sense on chapter numbering.

The m_pDoc and m_pDocShell members are such a WTF.

(regression from ae0e4a6ba9be2fa99ac2be8e20157806e36209b2)

Change-Id: I9d4d4cd7aeb7e6e29221d53facaff213fd4e35a5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115495
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit c5e5467f6a13aba68b4706a4d7feb130e824bcc6)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115291
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#141463 avoid skew in shape group in ooxml import ..

and implement special resize handling for rotation angles larger 45deg.
This solves tdf#93952 and tdf#141953 too.

Change-Id: I798f6d2cea29c4a5285f530e9cf7bb10e7f6c41d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115296
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115715

tdf#142165 restore a SwShellTableCursor if the orig selection described that

so that we are operating on the same selection that existed when the
dialog was originally launched

we need to avoid the assert of

commit 6db71f70a3b200d4074f6cda8ce445e9861d3296
    tdf#140977 drop possible table-cursor before setting the new one

and support the multi-selection of

commit e08b446e46f56e15af58fdd4396afba1a316f9e5
    tdf#140257 duplicate entire PaM ring when making copy

and support not scrolling to a different location on changing a table
page break style of

commit 9c61732677d038e32b73fc9fb883aced14c0febf
    tdf#135916 just set the target table as selection

and keep making it possible to remove aa page break on a table of

commit 81f91196b98af38e29af451b86c26a893a109abc
    tdf#135636 the selection at dialog-launch time is lost by dialog-apply time

all of which is necessitated by

commit c3a085d22742f88e91ff92f319a26d6e8d1d9a98
    lokdialog: Convert the Table -> Properties... to async exec.

Change-Id: Ide5f953e63da46a8750e85d623310af0062b048b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115522
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#142303 connect adjustments to spinbuttons

missing since

commit 87e21946fa9a81c403a146d03579691537c1352c
Date:   Sat Aug 17 20:41:26 2013 -0300

    Convert text animations tab page to widget UI

Change-Id: I8eba96030a1f5060ba60166d462130eb2cef1298
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115695
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#117957 WMF Add support for selecting colors from palette

Change-Id: I8f995dab566d9fae79d87fe13741b8ea9658b408
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/112998
Tested-by: Jenkins
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit e923f752a3adfe1a941dcbc2fdffc626a569d59e)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115520
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

vcl PDF tokenizer: fix EOF position when \r is not followed by \n

Otherwise this would break partial tokenize when we only read a trailer
in the middle of the file: m_aEOFs.back() is one byte larger than
rStream.Tell(), so we reader past the end of the trailer, resulting in a
tokenize failure.

What's special about the bugdoc:

- it has 2 xrefs, the first is incomplete, and refers to a second which
is later in the file
- the object length is as indirect object, triggering an xref lookup
- the first EOF is followed by a \r, but then not with a \n

This results in reading past the end of the first trailer and then
triggering a lookup failure.

FWIW, pdfium does the same in
<https://pdfium.googlesource.com/pdfium/+/59d107323f6727bbd5f8a4d0843081790638a1dd/core/fpdfapi/parser/cpdf_syntax_parser.cpp#446>,
we're on in sync with it.

Change-Id: Ia556a25e333b5e4f1418d92a98d74358862120e2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115466
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
(cherry picked from commit 6b1d5bafdc722d07d3dc4980764275a6caa707ba)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115516
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>

tdf#141571 Base crashed on creation of user function

Revert "reportdesign/core: create instances with uno constructors"
This reverts commit 79c1d1234bdac0d41109930529f42d39c70b0f7b.

I have no idea what is causing this. Something is stomping
on memory belonging to something else, but why?

Change-Id: Ieb758e5788f07e3f477063274ccd0721e3bf7726
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115567
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 9ca8a4e229d270fe7ec16f3aa7b5cb2e1f8a522f)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115515
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#142014 Fix displaying strokes when line width is 0

Change-Id: I80e05ff2f24f5da2f5c124c0ee1b302a1c8226ea
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115570
Tested-by: Jenkins
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit 699295ca7cab3a4f4e801a14496f202c05d18899)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115514
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#112603 tdf#142014 tdf#142139 WMF/EMF Fix line width

Previosly line width was always 1, and changing width do not affect
line.

Change-Id: I462096b915e053fa089e85860f124466b650558a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115497
Tested-by: Jenkins
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit b5ece3fbc7f878846298fd9196e5a30ba50e0dc2)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115512
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

fix some VirtualDevice leaks

Change-Id: I645543fa27ca8b75b36ed7798f0af7b81fd09f49
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115488
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>

tdf#141307: adapt null-check

- if we check pNd for nullptr, we need to expect it to be nullptr
- in that case we shouldnt use the pointer before the null check

Change-Id: Ia9d11abe32fac10f36fdbfcefeadd51891f1834e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115264
Tested-by: Jenkins
Reviewed-by: Bjoern Michaelsen <bjoern.michaelsen@libreoffice.org>
(cherry picked from commit b0ae5aaaafd236663f7e81af06a567ade8a2a745)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115287
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

Resolves: tdf#142186 Accept 123.45 fractional input on weird formats like 0"."

... or 0"."0 where the literal "." is also the decimal separator
but can only occur in the integer part.

Change-Id: I95093fdddf7759346f2869ee322222de3d130e55
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115338
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Jenkins
(cherry picked from commit 3561978410579c5222889eb7dce68f917b550334)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115277
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

Revert "tdf#141015: Show Mail Merge toolbar when DBData has data source"

This reverts commit 346153f08c907739ca310934ad45814379971027.
It seems to create more problems than expected, and there's no
way to remove DB association other than editing document XML.

Change-Id: I93171f17c01e231abbcfa9a816fdadce83c983ec
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115283
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit 2009af1c0fb4fead653dd7688347c15d74eaa1b0)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115286

tdf#139241: partially revert 059f07f9f33460c809a93e0fda1165f5c6f6d805

And instead normalize the start and end values before creating
rectangle.

Change-Id: Ib9921f1a537ca88a70cedf989d2f696180ad68e7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115284
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit 05f8f239d49bd66145e736b4e2e28ad073ef6f2f)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115285
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#141982 tdf#142139 Add rotation and line width support to ROUNDRECT

With this commit the ROUNDRECT is able to change line width and
transformation (including rotation) is supported.

Change-Id: Ic303a74adf0fd0dd452353f250a13140603d492e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115429
Tested-by: Jenkins
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit f11ed681df15728abe6a0b6b7b1612f190aa1707)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115282
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

vcl: fix build with vlc backend enabled

avmedia/source/vlc/vlcplayer.cxx tries to access the aWindow member
directly, which leads to a compilation failure, since it's private.

Change-Id: I08ba00f23652ced7184aa24fd1d3ad1cef441a66
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115367
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>

tdf#55058 tdf#141982 EMF: Add rotation support for INTERSECTCLIPRECT record

With this commit the rotation support was added
for INTERSECTCLIPRECT.

Before that change rotation was not applied to these CLIP rectangles.

Change-Id: I3da66790e0aeeaaeeb28d2fc30780cba8dbda390
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115102
Tested-by: Jenkins
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit 1ef26ffe39618a745d3367310565e7eeb184a4c2)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115207
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#43175 - Adjust chart references on every copied sheet

Change-Id: I92fcfeecb9aa10bf8974de953399d780ffe02176
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115058
Tested-by: Jenkins
Tested-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit d55bf8aab008d84fac673f3cd907f5199347ce77)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115210
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

gbuild: work around GDB 10 bug with DWARF5 in split debug info

GCC 11 defaults to -gdwarf-5 and GDB can only read it if
-gsplit-dwarf isn't used.

Dwarf Error: wrong unit_type in compilation unit header (is DW_UT_split_compile (0x05), should be DW_UT_type (0x02)) [in module /workdir/CObject/desktop/source/app/main.dwo]

https://sourceware.org/bugzilla/show_bug.cgi?id=27354
https://bugzilla.redhat.com/show_bug.cgi?id=1956475

Change-Id: Ie2ac7193a29a8f257cf6f1d711f9fa6941df48ad
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115054
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit dc8c8a4aa20ddd3139a25d5c052cac9bae944cb8)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115192
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

tdf#138518 sw: layout: unbreak fdo80206-1.doc

The 7 flys on the para on page 3 create ~15 extra pages with one
paragraph each.

Argh! One of the bPageHasFlysAnchoredBelowThis checks was inverted.
How dumb of me.

(regression from c799de145f7e289f31e3669646e5bd12814e6c5e)

Still doesn't look good but now it looks same as in 7.0.

Change-Id: Ib10c46f48746362d8d679c147ddc8b85157be508
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115242
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Jenkins
(cherry picked from commit 59d96acec8c4d9e472daa3e2c287b3a754e01817)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115206
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

set also dashing cap style for directx canvas (tdf#136957)

Change-Id: I7022ffba519faad5a333d7ef6eabf1ba5f0ae3bd
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115232
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
(cherry picked from commit ec73a21bccf4a055ae9dc575dbad3d67c91f481e)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115202
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#138518 sw: layout: fix assert on ooo43913-1.doc

Assertion `rAnchoredObjPage.GetPhyPageNum() == _nFromPageNum' failed.

Because not only the fly's anchor frame moves forward in
FormatAnchorFrameForCheckMoveFwd(), but also the fly itself,
apparently because it's in a table:

0  SwAnchoredObject::SetPageFrame(SwPageFrame*) (this=0x5a1d3d8, _pNewPageFrame=0x8cfbeb0) at sw/source/core/layout/anchoredobject.cxx:162
1  SwPageFrame::MoveFly(SwFlyFrame*, SwPageFrame*) (this=0x8cbd8c0, pToMove=0x5a1d280, pDest=0x8cfbeb0) at sw/source/core/layout/flylay.cxx:985
2  lcl_ArrangeLowers(SwLayoutFrame*, tools::Long, bool) (pLay=0x8cf80c0, lYStart=179488, bInva=false) at sw/source/core/layout/tabfrm.cxx:5000
3  SwCellFrame::Format(OutputDevice*, SwBorderAttrs const*) (this=0x8cf80c0, pAttrs=0x8ce78c0) at sw/source/core/layout/tabfrm.cxx:5359
4  SwLayoutFrame::MakeAll(OutputDevice*) (this=0x8cf80c0) at sw/source/core/layout/calcmove.cxx:1036
5  SwFrame::PrepareMake(OutputDevice*) (this=0x8cf80c0, pRenderContext=0x5b7fcf0) at sw/source/core/layout/calcmove.cxx:375
6  SwFrame::Calc(OutputDevice*) const (this=0x8cf80c0, pRenderContext=0x5b7fcf0) at sw/source/core/layout/trvlfrm.cxx:1792
7  SwFrame::MakePos() (this=0x8cebdb0) at sw/source/core/layout/calcmove.cxx:627
8  SwTextFrame::MakePos() (this=0x8cebdb0) at sw/source/core/text/frmform.cxx:340
9  SwContentFrame::MakeAll(OutputDevice*) (this=0x8cebdb0) at sw/source/core/layout/calcmove.cxx:1412
10 SwFrame::PrepareMake(OutputDevice*) (this=0x8cebdb0, pRenderContext=0x5b7fcf0) at sw/source/core/layout/calcmove.cxx:286
11 SwFrame::Calc(OutputDevice*) const (this=0x8cebdb0, pRenderContext=0x5b7fcf0) at sw/source/core/layout/trvlfrm.cxx:1792
12 SwTextFrame::CalcFollow(o3tl::strong_int<int, Tag_TextFrameIndex>) (this=0x5ae2c60, nTextOfst=...) at sw/source/core/text/frmform.cxx:279
13 SwTextFrame::AdjustFollow_(SwTextFormatter&, o3tl::strong_int<int, Tag_TextFrameIndex>, o3tl::strong_int<int, Tag_TextFrameIndex>, unsigned char) (this=0x5ae2c60, rLine=..., nOffset=..., nEnd=..., nMode=1 '\001') at sw/source/core/text/frmform.cxx:611
14 SwTextFrame::FormatAdjust(SwTextFormatter&, WidowsAndOrphans&, o3tl::strong_int<int, Tag_TextFrameIndex>, bool) (this=0x5ae2c60, rLine=..., rFrameBreak=..., nStrLen=..., bDummy=false) at sw/source/core/text/frmform.cxx:1166
15 SwTextFrame::Format_(SwTextFormatter&, SwTextFormatInfo&, bool) (this=0x5ae2c60, rLine=..., rInf=..., bAdjust=false) at sw/source/core/text/frmform.cxx:1613
16 SwTextFrame::Format_(OutputDevice*, SwParaPortion*) (this=0x5ae2c60, pRenderContext=0x5b7fcf0, pPara=0x8d07490) at sw/source/core/text/frmform.cxx:1720
17 SwTextFrame::Format(OutputDevice*, SwBorderAttrs const*) (this=0x5ae2c60, pRenderContext=0x5b7fcf0) at sw/source/core/text/frmform.cxx:1910
18 SwContentFrame::MakeAll(OutputDevice*) (this=0x5ae2c60) at sw/source/core/layout/calcmove.cxx:1525
19 SwFrame::PrepareMake(OutputDevice*) (this=0x5ae2f80, pRenderContext=0x5b7fcf0) at sw/source/core/layout/calcmove.cxx:321
20 SwFrame::Calc(OutputDevice*) const (this=0x5ae2f80, pRenderContext=0x5b7fcf0) at sw/source/core/layout/trvlfrm.cxx:1792
21 SwObjectFormatterTextFrame::FormatAnchorFrameAndItsPrevs(SwTextFrame&) (_rAnchorTextFrame=...) at sw/source/core/layout/objectformattertxtfrm.cxx:905
22 SwObjectFormatterTextFrame::FormatAnchorFrameForCheckMoveFwd() (this=0x8ce5720) at sw/source/core/layout/objectformattertxtfrm.cxx:919
23 SwObjectFormatterTextFrame::DoFormatObjs() (this=0x8ce5720) at sw/source/core/layout/objectformattertxtfrm.cxx:368
24 SwObjectFormatter::FormatObjsAtFrame(SwFrame&, SwPageFrame const&, SwLayAction*) (_rAnchorFrame=..., _rPageFrame=..., _pLayAction=0x0) at sw/source/core/layout/objectformatter.cxx:160
25 SwContentFrame::CalcLowers(SwLayoutFrame&, SwLayoutFrame const&, long, bool) (rLay=..., rDontLeave=..., nBottom=192048, bSkipRowSpanCells=true) at sw/source/core/layout/tabfrm.cxx:1534
26 lcl_RecalcRow(SwRowFrame&, tools::Long) (rRow=..., nBottom=192048) at sw/source/core/layout/tabfrm.cxx:1653
27 SwTabFrame::MakeAll(OutputDevice*) (this=0x8cf5f80, pRenderContext=0x5b7fcf0) at sw/source/core/layout/tabfrm.cxx:2425

It looks like the _nFromPageNum is always from the
SwObjectFormatter::mrPageFrame anyway because that's a precondition of
the mpPgNumAndTypeOfAnchors->Collect() being called, so just rely on
that to get the correct page.

(regression from c799de145f7e289f31e3669646e5bd12814e6c5e)

Change-Id: Ibdffb2121cffbc04320d17a29ab2e160dec4250b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115188
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 533a998e540b0f04068c876dde0e74adc3f79c93)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115201
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

tdf#141540 fix docx import of group or line with rotation

... and fix case wrap 'Square' and 'in Line' with them.
Non-uniform scaling of a rotated shape might produce skew. Such had
happened, when setting group or line to the size contained in
GraphicImport. Avoid it.
Writer has special rules for shape position and marging in case of wrap
'Square' and 'in Line', depending on rotation angle. The patch adds
the needed margins.
The patch changes some unit tests where we now get slightly different
values.
The patch fixes the wrong skew in sample document of tdf#73022.

Change-Id: Ic743790c3fc8b8b10a4324d9e0184ad945cdceb6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114193
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit 2a70cfb09c4d89154d229b6a95cf076e8bd76798)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115195
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#55058 tdf#141982 EMF: Add rotation support for ARC, ARCTO, CHORD, PIE

Change-Id: I5d9b76f0ddd2b7f12604f472986dd95976a8b04d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115185
Tested-by: Jenkins
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit 6bf2239a189423d087b2536dd7054b21df58ddc4)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115198
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#126226, tdf#126735 sw Next Change: cycle through tracked changes

Next/Previous Change commands don't stop at the start or
end of the document, but continue on the end or start
of the document to cycle through all tracked changes,
like Find Next/Previous and other office suites do.

Change-Id: I5578d6b98b81ca1f8f222ba78e7d3c08339eca89
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/113716
Tested-by: Jenkins
Reviewed-by: László Németh <nemeth@numbertext.org>
Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115183

Resolves: tdf#138855 crash on use of Outliner freed by GetFocus

probably since...

commit 3a662445a0e3666637e5eb140b106a1347f050eb
Date:   Sun Jun 23 17:54:49 2019 -0800

    tdf#126061 Make text edit outliner view show cursor

    Change-Id: Ic4e8b5984d5b21c3b789eba943d35d4cad1ba867
    Reviewed-on: https://gerrit.libreoffice.org/74676

Change-Id: Ibd6cf9e6f539e651d6e98f2af4a2eba55f081b58
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115196
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#142004 tdf#141982 EMF Import: Add rotation and path support for EMR_ELLIPSE

Previous implementation of EMR_ELLIPSE, doesn't support rotation
and EMR_ELLIPSE was not work with EMR_BEGINPATH, EMR_ENDPATH and EMR_ABORTPATH

The EMR_BEGINPATH opens path bracket construction.

Once path bracket construction is open, an application can begin specifying records to
define the points that lie in the path. Path bracket construction MUST be closed by an
EMR_ABORTPATH or EMR_ENDPATH record.

With this patch all these issue was resolved for EMR_ELLIPSE

Change-Id: I6d352e0ff0326dd788d43272bf1330fa6c777df4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115101
Tested-by: Jenkins
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
(cherry picked from commit 761fdaf26dc9ed7cd0d25a7630576e7800813e2f)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115194
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

update credits

Change-Id: Ifa86e93a66d54d5ad8f288ba111cbe4ca1695415
(cherry picked from commit b6516acdb96a7417f0d58749723750d83547f98a)

vcl pdfium render: handle widget annotations for form fields

Note that we render the bitmaps without FPDF_ANNOT, so comments are not
rendered into the bitmaps, rather we create them on top of the bitmaps
in Draw, explicitly.

FPDF_FFLDraw() draws content which is already an annotation, but not yet
interactive content; so this just fixes "missing text", as far as the
user is concerned.

Verified that e.g. vcl/qa/cppunit/data/PangramAcrobatAnnotations.pdf
indeed still doesn't render comments into bitmaps after this.

(cherry picked from commit 92cba30d5ce45e4f4a9516a80c9fe9915add6905)

Conflicts:
include/vcl/filter/PDFiumLibrary.hxx
vcl/source/pdf/PDFiumLibrary.cxx

Change-Id: I2b74d585729305cc1d3a9fefa258d4d76d1bd038
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115143
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>

tdf#142080 sw: layout: fix infinite loop in CalcContent()

On page 9, SwSectionFrame::Format() calls CalcContent() and that formats
all its content frames, then on SwTextFrame 91
SwObjectFormatter::FormatObj() fails becuase it moved forward.

With commit c799de145f7e289f31e3669646e5bd12814e6c5e this now sets the
o_rbPageHasFlysAnchoredBelowThis to true, which prevents a call to
SwLayouter::InsertMovedFwdFrame(), and the flys anchored in next frames
aren't moved off the page at this time.

Then the loop starts over at the beginning of the SwSectionFrame, and
frame 91 will be formatted again because the loop tries to format the
first frame on the next page to see if it will move back; now the
MoveBwd() isn't prevented any more so the result is the same failure
in SwObjectFormatter::FormatObj().

Fix this by ignoring the bRestartLayoutProcess in case the current frame
was originally on the next page and didn't move back (or, as is the case
here, moved back and then forward again); it should just be formatted
again on the next page.  Once that happens, it will eventually be
entered into SwLayouter::InsertMovedFwdFrame() too.

This happens to fix another problem with this bugdoc too: the first
column of the section on page 9 is empty.  This also happens in LO 6.4
but not LO 6.1.

An alternative would be to move the flys anchored below off the page as
is done in SwLayAction::FormatContent() now but sections can be in flys
or footnotes or headers so perhaps it should be done only at the
top-level.

(regression from c799de145f7e289f31e3669646e5bd12814e6c5e)

Change-Id: I0965aebb4e3cec687f4e70f8d5e3aa8a55da3393
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115144
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit ed12269c42f75f553bb8a8770923406f7824e473)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115087
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

tdf#123476: also use filter by extension when its service is the same

... as passed explicitly in the media descriptor, for 0-byte files.
This makes proper handling for such files when passing module name
in command line, like --writer for DOCX.

Change-Id: If8fd51e65dcf8a67b2653026f5fc1d5c964074af
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114924
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit dff586735b6618d9b011823594a33287d8f7f223)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114925
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

tdf#141945 sw: layout: check master frame when moving fly forward

The problem is that in the finished layout the fly frames are
positioned on the first page but are in SwPageFrame::m_pSortedObjs
of the second page.

Don't use FindPageFrameOfAnchor() because that looks up the follow-frame
that contains the anchor position.  This was unintentional; the idea
was to get flys anchored in subsequent paragraphs out of the way.

This situation where it's on a follow-frame of the same paragraph is
more complicated and less obvious, so don't try to solve it now.

(regression from eb85de8e6b61fb3fcb6c03ae0145f7fe5478bccf)

Change-Id: Ic6809a43c467955368807ccc82a7873c29fbfc2e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115100
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 30512746c182b52f37f9a818d4e206c98e715cb7)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115080
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

tdf#55058 tdf#141982 EMF: Add rotation and path support for RECTANGLE record

Previous implementation of EMR_RECTANGLE, doesn't support rotation
and EMR_RECTANGLE was not work with EMR_BEGINPATH, EMR_ENDPATH and EMR_ABORTPATH

The EMR_BEGINPATH opens path bracket construction.

Once path bracket construction is open, an application can begin specifying records to
define the points that lie in the path. Path bracket construction MUST be closed by an
EMR_ABORTPATH or EMR_ENDPATH record.

With this patch all these issue was resolved for EMR_RECTANGLE
Change-Id: Ic51442df8905e47c92eed811cc776762c9752af2

Change-Id: I111f183e509f03c0b276a038680f61156b37b235
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115065
Tested-by: Jenkins
Tested-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl>
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
(cherry picked from commit 24e71494d7d1a68b2cb5f5d34083ab02009e0982)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115079

tdf#140296: Added Edit Fields(context menu) inside of tables

Change-Id: I22b6a179c1575171315d9eb40cd4895c35d77b6f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114986
Tested-by: Jenkins
Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org>
(cherry picked from commit 1cf47d932580a58a7855fe29ac39d17cdd79ae01)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115091
Reviewed-by: Ayhan Yalçınsoy <ayhanyalcinsoy@pisilinux.org>
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

scale stroked line properly in cairocanvas

- Line width should be scaled by the scaled vector, not just its X
component (see VclMetafileProcessor2D::getTransformedLineWidth()).
- Lenths of dashes should not be scaled by the scaled line width,
but by the scaling.
Testcase document sd/qa/unit/data/pptx/tdf134053_dashdot.pptx .

Change-Id: I4116f82e91620f5612f5e4e187468508f683b93e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115147
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
(cherry picked from commit 3a1d150eb99b5875e6e86117029a72e4c87b0547)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115089
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

transform by vector length, not just X

I missed this one in b71d9a6d15cfb8a50afdea5ac064f40d84c561f8.

Change-Id: Ibbfd6ed1e064030a6a2818df94d7d61dd2b92caa
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115088
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

fix dashed line info conversion for metafile (tdf#136957)

My previous change had two problems:
- It didn't handle correctly the case when something repeated,
  such as dash-dot-dot.
- The rounding when setting lengths was a left-over from my first
  attempt when LineInfo used integers and not floats.

Change-Id: I914241590b1ddec22df04c05dfe65e76e921ee52
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114940
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
(cherry picked from commit 565824df07913f47851804daed9efa28a4a95e9d)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115039
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

do not apply line dashing in drawinglayer (tdf#136957)

basegfx::utils::applyLineDashing() is not as good as the actual
VCL backend dashing, and there are some rounding errors because of
all the canvas transformation matrices or whatever, which leads
to the drawing problem. So use LineInfo to carry the dashing
information.
As a part of this change, also make LineInfo use doubles instead
of ints. The use of transformation matrices means that the values
may be fractional and less than one.

Change-Id: Ia5ac7d266cab344b7137052c81fbd96c1ce28003
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114710
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
(cherry picked from commit b71d9a6d15cfb8a50afdea5ac064f40d84c561f8)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115038
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#141963 put focus in vcl parent during Function key processing

Change-Id: Ib63567c9ce208cb7dd522a1bce9d99fa44a72554
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114884
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

Resolves: tdf#142049 expand widget to use unused extra space

also includes...

Related: tdf#142049 restore some elements default visibility in the ui

these defaulted to not visible before...

commit ec548bce309d965a25e807ae79432114926e06f3
Author: Caolán McNamara <caolanm@redhat.com>
Date:   Mon Oct 15 11:14:02 2018 +0100

    weld ChartTypeDialog and SchLayoutTabPage

    Change-Id: I101c56b855d0bdc43559ca4561d298fa4ad92572
    Reviewed-on: https://gerrit.libreoffice.org/61797

Change-Id: I7ff6aa4feeccd790b6176cb60b6adb9b21db0f63
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115136
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

rhbz#1957034 crash in embedded videos

Related tdf#138888: fix assertion on avmedia/MediaControlBase

Assertion when trying to insert a video in Impress, see bt:
https://bugs.documentfoundation.org/attachment.cgi?id=168161

According to https://en.cppreference.com/w/cpp/algorithm/clamp
"behavior is undefined if the value of lo is greater than hi"

Regression from:
commit 36b3b357fe2a882db6a5f5a006239e16200fb847
Author: Noel <noelgrandin@gmail.com>
Date:   Thu Nov 12 16:10:56 2020 +0200

    replace std::min(std::max()) with std::clamp

Change-Id: Idf33eb6c662b60174678f990f98aefc72edec2a6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/107722
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit f4e496af66bc6a779f600158cc42b986d654f143)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115078
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Caolán McNamara <caolanm@redhat.com>

rhbz#1956977 crash on switch from active comment to slide sorter

Change-Id: I2b5a5cb8b0631971b775a70d96f2cbf0c118c090
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115076
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Jenkins

tdf#139329: Shift widgets down and make hatch&bitmap list visible

regression from commit cbe86ee37018dc4bf04783ecf70fef2863e61ad7
which inserted additional gradient colour picker but didn't
adjust the positions of widgets below accordingly so in effect
there were 2 widgets at position y=4

Sad that neither glade nor our own code detects this condition and
e.g. asserts or warns this is the case

Change-Id: I4c635eeb8b740765542631a5f0319bfe1a5d94d0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115056
Tested-by: Jenkins
Reviewed-by: Katarina Behrens <bubli@bubli.org>
(cherry picked from commit c46a231b987b92bd866b415d1a0ba0d3a26a2457)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115074
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#138782 DOCX import: fix frame positions of old docs

by limiting AddFrameOffsets compatibility option
for docs created by MSO 2010 or older.

Likely regression from commit 355d25eac764713f4d52eac801ade6e2ff1deab0
(n#779627: added quite some compat options from the ww8
filter on writerfilter).

This patch fixes several bugs, which were
collected as duplicates by Gábor Kelemen.

Change-Id: I106e90c4bf00bb0b6a8986615cb3ad9c4828d5b3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114476
Tested-by: László Németh <nemeth@numbertext.org>
Reviewed-by: László Németh <nemeth@numbertext.org>
(cherry picked from commit 5812fb81013cc124a9b6a0b9912a34cc715fc495)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115077
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Jenkins

tdf#139500 svx: fix crash on changing table properties during active text edit

Regression from commit fdeb04f7c59cf8032fe17072ed779e70505cc6ab
(tdf#129961 svx: finish UI for table shadow as direct format,
2020-12-15), the problem was that the BegUndo() / EndUndo() pair can be
only used if we know that the text edit of a cell of a table shape is
not started or ended in-between.

The bugreport scenario was an active text edit, where setting attributes
on the shape ends the text edit:

    #9 0x7f6dbb417121 in SdrEditView::EndTextEditAllViews() const /svx/source/svdraw/svdedtv.cxx:1079:20
    #10 0x7f6dbb466798 in SdrEditView::SetAttrToMarked(SfxItemSet const&, bool) /svx/source/svdraw/svdedtv1.cxx:1095:9
    #11 0x7f6dbc34b0af in sdr::table::SvxTableController::SetAttrToSelectedShape(SfxItemSet const&) /svx/source/table/tablecontroller.cxx:2738:12

Which also means that the underlying edit engine is deleted. But then
undo/redo would still reference that edit engine:

==31830==ERROR: AddressSanitizer: heap-use-after-free on address 0x60c0001fc300 at pc 0x7f6dd73a9cb9 bp 0x7fff788db4b0 sp 0x7fff788db4a8
READ of size 8 at 0x60c0001fc300 thread T0
    #0 0x7f6dd73a9cb8 in EditUndo::GetComment() const /editeng/source/editeng/editundo.cxx:147:34

Fix the problem by not grouping in case there is an active text edit,
that's not something I considered when I added the original grouping.

(cherry picked from commit ece86ef173cbc070c76f180d02ac80c65e07fff9)

Change-Id: I4f3583e21a27f8380c35b3f4563ce496819bcb81
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114929
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>

tdf#142017 set default focus widget

Change-Id: I5befe345d9c6ec499efdc1cf8305513ee8a53690
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114931
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

Resolves: tdf#141857 update background to current theme

Change-Id: I1004bd9944f382ca66e48575b81170cf3314845f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115073
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

Related: tdf#141857 send style-updated after app settings are updated

so the libreoffice settings are updated before the widget updates,
otherwise the app settings are the old settings when the widget updates

Change-Id: Id32dd55a1e5102cc707d6ee74eddfae2247e105a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115072
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#141930 document set as unmodified if editengine didn't modify on keyevent

this appears to have always happened for cursor keys, and since

commit 0154f04e8c64df937b793c73ac58d553637f324f
Author: Caolán McNamara <caolanm@redhat.com>
Date:   Wed Feb 24 14:32:21 2021 +0000

    Resolves: rhbz#1931423 start update timer in IM events as well as keyevents

also happens on Input Engine events which don't cause changes

Change-Id: Ic5cdd9ef4f1bed1cc2b2c17fc847f196a9c6e23a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114930
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

avoid windres preprocessor quoting-messups with current cygwin

apparently fresh installs of cygwin behave differently with the windres
command's quoting and treats --preprocessor='cpp foo bar' as a single file
"cpp foo bar" to run instead of running "cpp" with the arguments "foo"
and "bar".
(-D and -I options are passed to the preprocessor automatically, so no
need to prefix those with --preprocessor-arg)

Change-Id: I711e968206f0769ff07152ebb9572e8b71c30cec
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114965
Tested-by: Jenkins
Reviewed-by: Georgy Litvinov <git@litvinovg.pro>
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
(cherry picked from commit 07f1ec0af02a9ab6c1bf6e6478092ee50495afa7)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114971
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>

tdf#90278 - Set the vba border to the default border

The default border width in calc is 0.75[pt] * 1/72[inch] * 2.54[cm/inch]
= 0.0264[cm] set in 2f527738ea4f8e93acafdd7f0ae06de1678cfdd8 which
contradicts the border setting of OOLineThin used in order to parse vba
borders. The latter was set to 1.00[pt] * 1/72[inch] * 2.54[cm/inch] =
0.0352[cm]. This mismatch leads to an exception when the corresponding
weight of a cell is parsed.

Change-Id: I945d8aecdf31e5ad76d0d04e76eaaf4668e3d8ea
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114980
Tested-by: Jenkins
Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de>
Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/115035

tdf#141556 Fix 100% CPU usage in Writer idle loop

Do not interrupt the idle layout processing unnecessarily, because if it
is continually interrupted before making enough progress then it will
keep resuming at the same page, never finishing, constantly using CPU.

This is achieved with two changes:

 - Revert "tdf#123583 use TaskStopwatch for Writer Idle loop"
   (commit 383032c50a3e3354f04200ce984a47ab9d2c5c67) which
   introduced a stopwatch timer to interrupt idle processing every 50ms.
   This reversion restores the previous behaviour where idle processing
   is interrupted only when there is an input event.

 - Filter out TIMER events so that they do not interrupt the idle loop;
   this fixes both tdf#123583 and tdf#141556

Conflicts:
sw/source/core/inc/layact.hxx
sw/source/core/layout/layact.cxx

Change-Id: Ic989631e5f32199209d64b66b72059253fc0167a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/113825
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit 0fedac18214a6025401c4c426466a5166553e8ec)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114944

allow utf-8 in xml names (liborcus) (tdf#141672)

Change-Id: Ib150d55b588a572e4352396f18de2331983b2aae
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114892
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
(cherry picked from commit 6b7c2fa65eb68be520ed4135cc245e33fa22e8bf)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114917
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

liborcus: newline problem with GCC 11 -E -fdirectives-only

ooxml_tokens.inl:3524:32: error: stray '#' in program

Change-Id: Ib7f50e1798e8c985f31a0338944ae1e9b439e98f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114975
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit aa38eede9d340cbb11a792ab1ebbe681521a4489)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114921
Reviewed-by: Caolán McNamara <caolanm@redhat.com>

sw: speed up DoIdleJobs(WORD_COUNT)

SwTextNode::CountWords() has early returns that don't clear the
WordCountDirty flag, for nodes that don't contain any plain text.

This speeds things up considerably for tdf135116-1.odt, which didn't
finish the idle jobs after 5 minutes in an optimized 7.1 build, because
the time slice tends to expire iterating nodes in word-count before it
can make progress.

With this it does ~25 WORD_COUNT pages and then ~90 AUTOCOMPLETE_WORDS
pages per idle, sending OnLayoutFinished after 13-14 seconds.

jmux just pointed me to commit 0fedac18214a6025401c4c426466a5166553e8ec
which is not in 7.1 today and may have improved the performance here.

Change-Id: I5e9651da3a173649c49fc36edefb23298dfd6bc4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114936
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 4184569b963877c2a488ff05766654b9db194798)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114919
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>

tdf#141761 Enhance PrintDialog Preview for FormControls

The display quality of the Preview is pretty ugly when
FormControls are used. I made a deep-dive why this happens,
and in principle the reason is the Mteafile::Scale used
below. Since Metafile actions are integer, that floating point
scale leads to rounduing errors that make the lines painting
the FormControls disappear in the surrounding ClipRegions.
That Scale cannot be avoided since the Metafile contains it's
own SetMapMode commands which *will* be executed on ::Play,
so the ::Scale is the only possibility fr Metafile currently:
Giving a Size as parameter in ::Play will *not* work due to
the relativeMapMode that gets created will fail on
::SetMapMode actions in the Metafile - and FormControls DO
use ::SetMapMode(MapPixel).
This can only be solved better in the future using Primitives
which would allow any scale by embedding to a Transformation,
but that would be a bigger rework.
Until then, use this little 'trick' to improve qulatity.
It uses the fact to empirically having tested that the quality
gets really bad for FormControls starting by a scale factor
smaller than 0.2 - that makes the ClipRegion overlap start.
So - for now - try not to go below that.

Change-Id: I540de602634c6afa697b5659d69c34159c22075c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114704
Reviewed-by: Armin Le Grand <Armin.Le.Grand@me.com>
Tested-by: Armin Le Grand <Armin.Le.Grand@me.com>
(cherry picked from commit 4722ad2cf3f2b91c217e3548f811f2972f2aa60c)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114715
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>

sw: layout: let nested table move forward

The problem is that a nested table at the bottom of the page doesn't fit
into its cell, but it is split and its first row still remains on the
page where it doesn't fit.

The outer table 3 tries to split with bTryToSplit=true.

In SwTabFrame::MakeAll(), first a split of the nested table 435 with
bTryToSplit=true is attempted; this fails and is partially undone -
the follow flow line is removed, but the last 2 rows remain in follow.

The bTryToSplit=false is skipped because there's only one row now.

Then MoveFwd() is tried, but it fails because the top-level table
doesn't have a previous frame.

Then bTryToSplit=false is tried, and succeeds, again leaving row 436
on page 1 and other 2 in follow.

Now another attempt with bTryToSplit=true is made, failing again, not
effectively changing anything.

During all of this, growing of the outer cell frame 434 and row frame
433 is prevented by the outer table 3 having a follow flow line, see
SwRowFrame::GrowFrame().

The result is that cell 434 has content area height 190 but the table
435 with its single row is all valid with height 406, so it's cut off in
the rendering.

This doesn't happen for SwTextFrame inside table because that one does
the MoveFwd().

Plausibly it's the inner table's responsibility to finish with a valid
state that fits the constraints of the current page; there are some
checks in lcl_RecalcSplitLine() to check for no content frame in the
row but none for the row being too small to contain the content that
was formatted.

So the only valid results for the inner table are that it either moved
forward, or it left behind a row containing no content (such as that
produced by its own failed attempt to split with bTryToSplit=true),
which could be handled by the outer table split code - but the latter
could be insufficient in case the outer table is itself a follow, or at
least would require further changes in lcl_RecalcSplitLine().

So fix this by removing a condition in MoveFwd() that doesn't make any
sense to me - why is it relevant for an inner table during a split of
the outer table whether the outer table itself can move forward?

Change-Id: I1e01ce233383cc70b9aea72d25369b7278eb75f0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114760
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit ca19177728c66d913996a48c91a0ba47d08825d6)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114718
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>

tdf#137945: sc: Add UItest

Change-Id: I6fa935d0c77556113ae087a67c2932ddcee12a1e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114871
Tested-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

vcl: print dialog: remove broken code

This was introduced in 149807eb6100090e178505ba4a264bb38eba1e0c
< Resolves tdf#127782 - Print dialog height >
trying to fix the problem of the window size not being remembered
after reopening the dialog, however, the problem is still
reproducible in master with GEN, thus, remove this code

Change-Id: I65f327fa692cd3e2b5597d04e9e15abc952fe6d9
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114874
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114890
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>

tdf#141854 Resize called during dispose

Change-Id: Ie5eaafa95704f9567b8bb9881865571a39bb2f6b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114708
Tested-by: Jenkins
Tested-by: Caolán McNamara <caolanm@redhat.com>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
(cherry picked from commit aea139c440d7c7a3683fcbbc48799bc79ba590e7)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114860
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>

apparmor: Fix "oopslash" typo

Change-Id: I8b355bceadeebebc9e6fedbf92a35aa44a356e09
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114819
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
(cherry picked from commit f616d96bd8ce8986e4cc204953db0467e6060b5c)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114866
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#141970 Revert "tdf#129606: Round the mean of the two subtractions"

This reverts commit d5ebe7c3089ab9f4d3fe0707169fc1ce024cdb70, and
brings number of inexact values in testing back to lower numbers
(see test document in the bug).

Change-Id: I5cfb34b7260fc6e4866585af6a2a50e79696eea2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114861
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit aa096331ba468f19951e43e2550105d4dcf50053)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114865
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

Related: tdf#130326 skip calling Refresh if its already just called

Change-Id: I80d3ae7f10b56dfe53a4ab5101ae771b47848092
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114877
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Jenkins

Related: tdf#130326 drop ScContentTree::ObjectFresh

this was introduced as part of the large IA2 integration

commit b41332475783c31136673fb44cf4c411bb0148f8
Author: Steve Yin <steve_y@apache.org>
Date:   Mon Dec 2 15:54:29 2013 +0000

    Integrate branch of IAccessible2

Both calc and writer gained the feature that "space" in the navigator
for drawing objects selects the highlighted object, extending the
selection of another one is already selected. (notably impress was
left unchanged)

In calc, but not writer, an ObjectFresh was added which fires
on every SfxHintId::ScKillEditView which is broadcast on every
exit of the cursor from a cell. Its purpose seems to be primarily to
re-highlight the last selected-by-keyboard entry on a content refresh.

As far as I can tell this doesn't seem necessary and we already refresh
content on SfxHintId::ScDrawChanged events so we shouldn't need to
additionally refresh on ScKillEditView and refreshing on every
ScKillEditView is very slow on navigating through a calc document with a
lot of drawing objects.

Change-Id: I64b2840f8510d474314d108e657fc3367f8ab6c0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114876
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Jenkins

tdf#135997: make sure that the two lists are same length

This fixes the strange assumption that when searching the two
lists (character names and font names) independently, the two
found positions will necessarily correspond to each other.

Instead, the positions of the match must be the same, which is
implemented now. Also the input from configuration is sanitized.

Change-Id: I920de7414387e181e11183b8a22776a72b6be419
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114722
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit b6ab5914d8b2fc7041430796890f086f8e3e6369)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114852
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>

tdf#116384 only drag hyperlink if user's not currently setting the selection

which is what writer does for normal text.

Under gtk its common on selecting by dragging the mouse over a hyperlink
for one or two letters of the hyperlink to get selected before that
selection then gets dragged as dnd.

Under gen StartDragWidth is 2 by default so the drag kicks in before
enough text has been selected for writer to select anything, but under
gtk gtk-dnd-drag-threshold is used for StartDragWidth and that is 8 by
default making it easy to select one or two characters before the
gesture is recognized as a possible drag.

it seems sane to do the same as for normal text and not dnd if the user
is still using the mouse to set a selection

Change-Id: I971d429acaf8a5a656c5ce58f52632979b94e40a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114726
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>

tdf#135997: fix toolbar control

This is the same as fc7b4bfeba39a2479c9a2c9c8ef5ced5ab3c0de7,
just for the toolbar control (the older one was for special
character dialog).

The problem was present since 800ac37021e3f8859a52c5eebca261a5d3bc5a11.

Change-Id: Ib86aa7336a3078ff28b417de37fbd5bc492c85c6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114719
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit b27de500b1412bacefc7144b5222c307ca1ae0fe)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114731
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

Accept iOS SDK 14.5

We can also surely drop iOS SDKs 12.* now.

Change-Id: I57bfcf40e823551021b58a74ef0e78a5781f5b0e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114806
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

Resolves tdf#127782 - Print dialog height

* Expander status remembered
* Fix default size for scrollwindow
* Window size remembered
* Window position center on parent
* Size doesn't depend on preview anymore

Change-Id: If14376ecd190f5d73e7cfad6f5f136d287778478
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/110631
Tested-by: Jenkins
Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org>
Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114821
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

update credits

Change-Id: Idd3d14c272c1eebca7207054dccbeae4632c51cc
(cherry picked from commit 93a90a5b0558aeac847aefe06f49f68c00a26678)

Accept macOS SDK 11.3

Change-Id: I210ae30e51d796990f88340da8b29f4c7080f5d2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114702
Tested-by: Tor Lillqvist <tml@collabora.com>
Reviewed-by: Tor Lillqvist <tml@collabora.com>
(cherry picked from commit bc35e5fe5922a71749c65a9b2d0412e4a3f34128)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114721
Tested-by: Jenkins

tdf#137470 use a more proper URL for the extensions

also hide the download-count, since those numbers are completely made up
by the script

Change-Id: I15b53a8a023c34ecc6a544b88ad35800891e4327
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114740
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
(cherry picked from commit 452fcbe0792aa10042bb8cd2cfd6cd29ca754be5)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114774
Tested-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>

tdf#138604 sw textbox copy: fix missing fly frames of as-char draw frame

Regression from commit 682e0488df819c191c13a03758fad0690706e508
(tdf#134099 sw: fix textbox anchors on copy-paste and undo, 2020-06-29),
now that we don't copy draw+fly format pairs more than once, the fly frame
of a fly format was missing.

This was broken because
sw::DocumentContentOperationsManager::CopyFlyInFlyImpl() calls
CopyLayoutFormat() with bMakeFrames=true, but the frame format list
doesn't contain as-char frames.

For as-char frames CopyLayoutFormat() is called by
SwTextFlyCnt::CopyFlyFormat() instead, but with bMakeFrames=false, so
the fly frame's layout frame were never made.

Fix the problem by explicitly making frames in
DocumentLayoutManager::CopyLayoutFormat(), which also requires
delete+make in SwTextFlyCnt::SetAnchor(). [ This is now consistent with
the start of SwTextFlyCnt::SetAnchor(), which also deletes frames for
fly frame formats. Otherwise CppunitTest_sw_core_txtnode would crash in
testTextBoxCopyAnchor. ]

Change-Id: I87003ee09ca75b9fecd70a1aa5c42f762f715be8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114680
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit feeaa5d3ca0c784344549816fbaab2c8716b4105)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114611
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

Update git submodules

* Update translations from branch 'libreoffice-7-1'
  to 0a7477257a9aa9c557c090f29f9d3b52e2cfda77
  - update translations for 7.1.3 rc2

    and force-fix errors using pocheck

    Change-Id: I01bef69ec2c4e02df560e3334fac5d1226cb2c56

tdf#99991 - Don't resize toplevel widget containing the options expander

Change-Id: I930445e9047478b26142d049598fbe656d25d666
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114628
Tested-by: Jenkins
Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de>
(cherry picked from commit dc83f99c466296caaa6b236e32b39e04ae9e6d8c)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114598

AboutDialog: imgBrand: Fix alignment

The imgBrand was "centered" aligned with the text bellow which was
resulting in not only going against the dialog layout (2 sections) but
was also appearing "false" centered aligned:

- Left section: imgAbout
- Right section: text left aligned
- Right section: imgBrand was not left aligned and not centered aligned
etheir (with the dialog)

Fix: Left align imgBrand to be consistent with other elements in the
same section.

Signed-off-by: Pedro Pinto Silva <pedro.silva@collabora.com>
Change-Id: I227f67863d5b6e951c498ee7acdfadc0fac75e6c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114648
Tested-by: Jenkins
(cherry picked from commit 4e9df130a7f4084d1642a213792100187569900b)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114606
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

tdf#141887 SW don't crash on missing SwWrtShell

The code assumes that SwEditWin::Command can just be called,
while there is still an m_rView.GetWrtShell(), which is at least
not true, if you have an IM in pre-selection and close the Writer
window via Window decoration.

Change-Id: I67d5f1f917e9b2180037babb8c70c72d75c4a7ae
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114624
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
(cherry picked from commit 9a1c4a10b08ccb8d874a85a2cd984fd71f08c022)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114604
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#141914: Allow to unset modified anyway

IsEnableSetModified returning true means kind of "treat this document
as read-only". Marking object unmodified does not break this idea.

Change-Id: I3efd5b3b25e2d9318702cdbf24ac5eb61aeda3ba
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114666
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit eeb2854e085fdd9f7a38d6e952a8aedf43e90323)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114605
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#141924: SwFormatAutoFormat::GetStyleHandle may return nullptr

Change-Id: Id1b6fe687f5b328fa0743036a287bacfefe5e5fb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114685
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114687
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#141902 Fix domain in ui files (found with ui-checkdomain.sh)

Backport to 7-1 to fix a user visible problem

Change-Id: I6ec07748616c6ddbe12a48077a777c91ffb8f96f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/108046
Tested-by: Jenkins
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114682
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>

tdf#141761 Avoid vanishing FormControls

For details see task documentaion, but BG is that
a Rectangle gets not transformed in
OutputDevice::LogicToPixel and corresponding
methods of OutputDevice due to the given
Rectangle.IsEmpty() == true. This ignores that
the Rectangle has a valid position and that
we have cases where a Rectangle is used to hold
a Position without having a Size != 0 in X and Y.
Thus I added transforming the Position in these
methods in all cases - a Rectangle always has
a valid position - and take care of Sizes that
exist.
This *may* need to be done different due to old
hiccups of Rectangle itself, see explanation at
OutputDevice::ImplLogicToDevicePixel
Had to do it different now, see comment in map.cxx
for expalantions.
Also needed to adapt the ancient optical geometric
paint position correctionsfor Buttons - these can
get zero or negative due to using Zoom and MapMode
settings. A Width shoul not get zero or negative.

Adapted to LibreOffice-7-1 as needed.

Change-Id: Ide085fbe74f9c0387de3c1bf6320b7e873238029
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114304
Tested-by: Jenkins
Reviewed-by: Armin Le Grand <Armin.Le.Grand@me.com>
(cherry picked from commit 35e4a45260f128f353d25e2a2f2b800e6bd11d61)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114596