Age | Commit message (Collapse) | Author |
|
Change-Id: I1c4cff09a7c549e9a14886c1460a7e2c2c74f757
|
|
Change-Id: I5655dd4c07769d3c9832a3e7246e750cd2f1908d
Reviewed-on: https://gerrit.libreoffice.org/32806
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Found thanks to Buovjaga
Change-Id: I8859cbec896c4019a5137d801f753ad87287a369
Reviewed-on: https://gerrit.libreoffice.org/32809
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
Tested-by: Julien Nabet <serval2412@yahoo.fr>
|
|
Found thanks to Buovjaga
Change-Id: I73e36f50a1fd393859ba775ab084a488047864a3
Reviewed-on: https://gerrit.libreoffice.org/32810
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
|
|
Change-Id: I33b5a2343eb7701a478c8d039d818e679471d4c0
|
|
Change-Id: I425692130bd7ef6b8a9ebce3592c1976fa44dafc
|
|
SwFrame::RemoveDrawObj() calls SwAnchoredObject::ChgAnchorFrame(0)
so the SwAnchoredDrawObj has no anchor frame later when
SwAccessibleMap::InvalidateCursorPosition() asks for the parent.
(regression from 76c549eb01dcb7b5bf28a271ce00e386f3d388ba)
Change-Id: Id55cb5fc41a4e37e863498265d1565e1621d508e
|
|
Because that's what Word does to show only part of the shape in the
bugdoc.
Change-Id: Ic5cb84cace9237671d71eda0c64e9dadfe244cb9
Reviewed-on: https://gerrit.libreoffice.org/32782
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
|
|
I don't see an easy way to implement this via UNO, so use the internal
API. As to the internal API usage,
SdrEditView::MirrorMarkedObjVertical() (for UI) and
SvxMSDffManager::ImportShape() (for WW8 import) are example client code.
Change-Id: I9bf27788db32fd35d6b56e0f1a240c4b7abc5604
|
|
Change-Id: I75b077cbc2d100964d4157220561d53c1b235199
|
|
This fixes the specific test document here, but likely there
need to be other qualifiers such as "if the table is the first
node in the section". Likely this commit will be reverted.
We will use Milos' document tester to hunt for the side-effects
of this patch - which should give us a good set of test documents.
Change-Id: I7d3a2168556c82bc4758ad665da319d55df4ba2b
Reviewed-on: https://gerrit.libreoffice.org/32759
Reviewed-by: Justin Luth <justin_luth@sil.org>
Tested-by: Justin Luth <justin_luth@sil.org>
|
|
- do not iterate repeatedly over objects already handled
- filter out AS_PAGE and AS_CHAR before looping
- use circular buffer as we are only ever removing, never adding formats
to handle, so we can get away with one alloc
Change-Id: I0244bda6db4704891111f6c6a8630a72bede68e6
Reviewed-on: https://gerrit.libreoffice.org/32747
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Use vcl::PDFExtOutDevData::SetScreenStream() for embedded media to make
sure that vnd.sun.star.Package: URLs don't end up in PDF out literally.
Acrobat Reader obviously doesn't understand that protocol.
Change-Id: I384891b3ef2dcea25bbf591bd210ccf899d30a61
|
|
- checking the type of Format hopefully allows skipping contact checks
Change-Id: I11690e4ec1e94b809b0ec13f2d9b4e3a22ded213
Reviewed-on: https://gerrit.libreoffice.org/32729
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Change-Id: I78f428e9ddc50df60f7e9652015d49f3cf7b3286
Reviewed-on: https://gerrit.libreoffice.org/32746
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Change-Id: Iaf8011f1bd6a10b324653899b7f51458f7d0b68a
|
|
The hidden cells in a vertical merge retained their
numbered list properties.
It is possible in LO's UI to create a table similar to this
by numbering the cells first, and then merging them.
So this behaviour also needs to be prevented in
a future patch.
Change-Id: I0c3ae50aada3fa03e624b1d4212b2776b7559036
Reviewed-on: https://gerrit.libreoffice.org/32512
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
Change-Id: If996c17dbfee2302a1d0f58428b088a953abcc9f
|
|
Fixes size and position issues under gtk3/wayland
Change-Id: I357a5935e2b44fd3d5effde8cf7d3499e26384db
|
|
Change-Id: I66033eeff8ebb636d5f55d61102a0f30f3353f85
|
|
... by fixing the size of SwFootnoteNum so it can be forward declared.
Change-Id: I1a9d32a1ec7e7201de1910407d32020b75fb7342
|
|
way back in 2001 commit a7afc46b3766e4000ae5a9b8a914ede36bbd2991
removed the option to select per-chapter footnote-numbering for
footnotes placed at the end of the document (or perhaps chapter
since the variable is called FTNPOS_CHAPTER, but the UI refers to
it as document). So only "Per Document" remained. However the
function GetNumbering was never updated to reflect that fact that
2 options were removed from the list now, instead of one.
So, the UI reported the choice as "Per Page", the only dropdown choice
available was "Per Document", and the actual implemented choice
was "Per Chapter". How's that for a mess?
Change-Id: Ib8dc3d07c0ef62d39afdd2a1c78c14527649c6df
Reviewed-on: https://gerrit.libreoffice.org/32700
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
Tested-by: Justin Luth <justin_luth@sil.org>
Reviewed-by: Michael Stahl <mstahl@redhat.com>
|
|
The "DocShell must be cleared..." comment was a leftover from
2613afee051fee15ac78b06f611da493989f893c "INTEGRATION: CWS sw008", pertained to
code that had been moved into implFlushDocument. The extra code block and
accompanying "Thus, an own block..." comment are just nonsense.
Change-Id: I7e60b488c609796d86bedf518218b06c612e494d
|
|
Liberals of Number format code (NFC) were no longer recognized
by dmapper after it had been removed in
d30a8ec448bcd08c6a52a37d6ae41a4b71c235da.
Replace them with enumeration.
Change-Id: I3ed1e09d9f3987fce95bd708b9c0b0568d3f15f9
Reviewed-on: https://gerrit.libreoffice.org/32527
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mark Hung <marklh9@gmail.com>
|
|
SwXAutoTextEntry::GetBodyText calls SwGlossaries::EditGroupDoc to create a
Sw[Web]GlosDocShell (which it then stores in xDocSh), and there must be a call
to that shell's DoClose to clean up its resources. However, if a reference to
the SwXAutoTextEntry is still held during XDesktop::terminate (as can happen in
JunitTest_sw_unoapi_1, where such a reference is held from Java, thus
arbitrarily delaying destruction of the SwXAutoTextEntry until the JVM does GC),
SwXAutoTextEntry::Notify's PrepareCloseDoc case will release xDocSh without
calling DoClose, so an eventual ~SwXAutoTextEntry will not call DoClose either.
(And calling DoClose from withing SwXAutoTextEntry::Notify would cause a crash,
as it would destroy the SwGlosDocShell recursively from within its
SfxBroadcaster::Broadcast call.)
So introduce a disposal protocol for XAutoTextEntry and call it at least in the
place that caused the resource leak in JunitTest_sw_unoapi_1. For lack of a
better inerface, reuse css.lang.XComponent for that protocol, even though this
scenario doesn't involve breaking cyclic references among UNO objects caused by
listener patterns.
Change-Id: Id6501b84e562950e40b83e0b1afd1b3184807856
|
|
hideMark in Word does not force the minimum possible row size,
but simply ignores any cell marker styles/sizings. So an empty row
then takes on the "at least" size defined for the row.
Change-Id: Id24a52f2a360a5040c2eed5c17757c0d32d03520
Reviewed-on: https://gerrit.libreoffice.org/32380
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
Be explicit about the page number, this way a video on the second page
doesn't end up as an annotation for the first page. (In the Impress case
each slide is exported separately, so there this wasn't a problem.)
Change-Id: I83ba9cb4a3b2a6734bd88a138654e391199651c6
Reviewed-on: https://gerrit.libreoffice.org/32696
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
|
|
Change-Id: Ia3afaf2abd0f0a56ccb56d4f967117f1adc52fb2
Reviewed-on: https://gerrit.libreoffice.org/32357
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Change-Id: I2a0cf6d6b663e00cd67447089f65512667b18f3a
Reviewed-on: https://gerrit.libreoffice.org/32297
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Change-Id: I6148ffccd4284cfc72fed2c7f4538054a5aaf24c
Reviewed-on: https://gerrit.libreoffice.org/32358
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Change-Id: I487db955caff589b9f62c50b83067628bb768d50
Reviewed-on: https://gerrit.libreoffice.org/32460
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Eike Rathke <erack@redhat.com>
|
|
This is the sw-specific part only, the real work is done by the common
CreateScreen() / SetScreenURL() API added for sd earlier.
Change-Id: Ief9fd80082960ddd1f92f413eac79577621551ce
Reviewed-on: https://gerrit.libreoffice.org/32687
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
|
|
The problem seems to be the table's repeated rows value which is
setted to max table rows in some docx file.
Decreasing repeated rows value by one avoids loops but renders a bad
layout.
Table continuity is lost and lots of rows are hidden.
So I set repeated rows value directly to one in order to avoid bad
layout.
Change-Id: If4adc33febe088a866b704e87e49339710e5aaf5
Reviewed-on: https://gerrit.libreoffice.org/32349
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
See commit 1be0a3fa9ebb22b607c54b47739d4467acfed259 (n#825305:
writerfilter RTF import: override style properties like Word,
2014-06-17) for the details on style override in RTF.
Here the problem was that we added an unneeded "reset to 0" property, the
opposite situation that commit 657c6cc3acec0528209a8584b838cd6de581c437
(tdf#104228 RTF import: fix override of style left/right para margin,
2016-12-13) was fixing (there a "reset to 0" was missing).
Change-Id: I37f079b9cb4773214d2531c2e34920b3b8927211
|
|
Change-Id: Id0ac2e87344fdab4db97357158c5cc60eed70898
Reviewed-on: https://gerrit.libreoffice.org/32663
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
The hidden cells in a vertical merge retained
their numbered list properties while the
document was loaded (bad). However, on closing and
reopening, the hidden cells no longer had any
numbering (good).
This patch immediately removes the numbering,
so that the working copy matches the saved copy.
Now vertical merging works the same way as
horizontal merging already worked.
Change-Id: Ia37151031eab4fd57bf958e9cddf6dce06dafc3c
Reviewed-on: https://gerrit.libreoffice.org/32518
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
The PEP8 suggest to use join() instead of + operator for string concatenation.
Change-Id: I35325cc73845a450f5c8ce40b7594e6d57ad6f45
Reviewed-on: https://gerrit.libreoffice.org/32504
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
|
|
Change-Id: I05a4bc59e178d0e94e88540ab71fc6c46040490b
|
|
Change-Id: I9be657cd0b5ed964db8db2ca28fd2874ae5406aa
cppcheck: performance
Reviewed-on: https://gerrit.libreoffice.org/32431
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
|
|
fixes commit cbd0fbc287051f918e4adb32b3e9b58dfbf8059d
which removed the insertion code with the explanation saying
we should simply not do anything,
and that'll lead to the right behavior,
don't try to be smart and try to set TABLE_WIDTH here
but still left the flag set, falsely indicating that the
tableSizeType had been inserted.
Change-Id: I0720083c992d1d03a5fe259d5b4b177a0c8108f8
TODO: the table size still isn't correct.
Reviewed-on: https://gerrit.libreoffice.org/32385
Reviewed-by: Justin Luth <justin_luth@sil.org>
Tested-by: Justin Luth <justin_luth@sil.org>
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
In mailmrge.cxx/mailmrge.hxx manula allocated pImpl pointer is changed to
std::unique_ptr.
Change-Id: Icb973cdb449e59184a24c359c0ca9fe2da4a8ebd
Reviewed-on: https://gerrit.libreoffice.org/32464
Reviewed-by: Tamás Zolnai <tamas.zolnai@collabora.com>
Tested-by: Tamás Zolnai <tamas.zolnai@collabora.com>
|
|
Change-Id: Idb4bfc97e7d758c58118742affd9a9345c7a99f2
Reviewed-on: https://gerrit.libreoffice.org/32491
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Include ooxmlexport_setup.mk only once in Module_sw.mk,
so that gbuildtojson can properly set last included
makefile for test jsons.
Change-Id: Ie8ed3296ae97cf4a33d652599673f389b224993e
Reviewed-on: https://gerrit.libreoffice.org/32502
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Matúš Kukan <matus@libreoffice.org>
|
|
- also skip extra GetMaster() null-check as it is only used for
nullchecked dynamic_casts
Change-Id: Iab51a6a3d5e40aa96f8f252d4ef5dce7d805549f
Reviewed-on: https://gerrit.libreoffice.org/32466
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|
|
Change-Id: Icea70da002f88752603ad21482975d1707d29520
|
|
Change-Id: I303e993429e40147579c77623b554ef10d87dbfb
|
|
In dbtree.cxx and dbtree.hxx manual allocated pImpl is changed to
std::unique_ptr.
Change-Id: I19e76147f79733139c2535269431eeb94f85f557
Reviewed-on: https://gerrit.libreoffice.org/32463
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Tamás Zolnai <tamas.zolnai@collabora.com>
|
|
Change-Id: I5917aa91021a8ddbb14ac7ccee1ba06ab6a514ab
|
|
Change-Id: I2d4ac010ff5818e673567cee05700872588918e4
Reviewed-on: https://gerrit.libreoffice.org/32384
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Jochen Nitschke <j.nitschke+logerrit@ok.de>
|
|
... as the old code did not limit itself to SwDrawContacts either
(possibly fixing regressions from d2be1f90faa64cbbfe789eb62370555eb5400ae3)
Change-Id: Ia821a025979d7eb2b5e244802c708a42ab1256ed
Reviewed-on: https://gerrit.libreoffice.org/32393
Reviewed-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
Tested-by: Björn Michaelsen <bjoern.michaelsen@canonical.com>
|