...as for the by-default --enable-qt5 case we generate an additional
qt_soffice.html anyway (and for a --disable-gui --disable-qt5 build, client code
probably wants to embed the results into its own HTML document and wouldn't rely
on an soffice.html); this simplifies the build infrastructure a little
Change-Id: Ib50799b5dd2fcfeb680f257f46ff65775dcddd71
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174725
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
move this logic to the destructor, like we do elsewhere
Change-Id: Ie41c78b0a01d79020cb77fd5be70116f473e910f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174747
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
This introduces Office/Calc/Input/CycleNextPrevSheetTab
boolean property, that controls the behavior implemented
in commit 552170c9c8
(tdf#156311 Cycling through tabs in Calc, 2023-08-16).
Change-Id: I0510743b4ab6d92488ba73a7b8d5b83be9586b2f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174745
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Tested-by: Jenkins
When using the Signing tab on the "Export as PDF" dialog, the TSA URLs will be
appended to the ComboBox each time a certificate is selected. Thus, if a user
selects a certificate and then another (or the same) is selected the ComboBox
will end up with duplicated values.
This patch moves the code to append the TSA URLs to the function that sets up
the signing page, fixing the problem.
Change-Id: I93d9cb61e10f18b0c33765d63a5ce47c92e588bf
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174424
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Pasting paragraph with list should apply the source list to the target
paragraph if the target is empty or completely selected.
Change-Id: I27d462f121c03b6c5625f8c0967bc975d3ac2f88
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174606
Tested-by: Gabor Kelemen <gabor.kelemen.extern@allotropia.de>
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
This will start to fail on pdfium-6764 with:
pdfexport2.cxx:4186:Assertion
Test name: (anonymous namespace)::testPdfImageHyperlink::TestBody
equality assertion failed
- Expected: 0.0012626264
- Actual : 0.0012626261
But the original problem was that the repriprocal was just 0.00126. If
we require one less digit, 1/0.001262626 is 792.0001647360343 and
1/0.0012626264 is 791.9999138304095, so the original problem that the
rounded result is 794 points (and not 792) is still tested.
Change-Id: Ia4489625fd25eca971261b14005b10bbf34343eb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174724
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
Applying the zoom value works now as expected. The prior
fix mixed up zoom value and 100% and didn't always apply
special zoom values.
Change-Id: I7cf1f34fca5724eea14339535e7614cfecffdd79
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174688
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
Tested-by: Jenkins
which requires some care, because these objects are linked
together into a double-linked list, and they call back into
each other.
Change-Id: If5c128918a1ae901ab7789067dc47d30f7415b43
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174705
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Those were mandatory prior to commit d367ee2f8c,
and still mandatory e.g. in AOO. Restore writing those, for interoperability
and backward compatibility.
Change-Id: I84f33e13e6bade8b2f2dd0cc5e2e0f4a96d50e84
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174721
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
* chart2/uiconfig/ui/steppedlinesdlg.ui:
- Correct the swapped tooltips for stepped line types
"End with horizontal line" and "Step at the horizontal mean".
Change-Id: Ieeff2e71b2e924060fb5ef3fca313c0d0bc2502d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174700
Reviewed-by: Ilmari Lauhakangas <ilmari.lauhakangas@libreoffice.org>
Tested-by: Jenkins
Also, improve detection algorithm by replacing the limit
of 20 lines with a time limit of 500ms.
Change-Id: Iac519b6ebe675b91ce84b900646d9d320ea9ddc1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/165905
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Jenkins
This patch implements sensitivity analysis when using the LpSolve solver engine.
It also adds the infrastructure needed for future implementations in other solver engines via the css::sheet::SensitivityReport struct.
Change-Id: I74c2ed9c6201a0b2ffc29ef612d2b778d11a3bef
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173642
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
GetODFSaneDefaultVersion() is surprisingly hot, mostly because it is
called from every ScXMLExport::WriteCell call, so cache it
Change-Id: Ibc54c007f4979315487badc23889eb264b11d4ca
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174671
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
This builds on top of commit d09336fbdc
(tdf#125038 DOCX import: fix lost MERGEFIELD result inside an IF field,
2019-10-31), and extends it for DOCPROPERTY.
With this, all unexpected content is gone from the paragraph.
[In a previous related commit, I wrote:
Actually this is all somewhat irrelevant since DOCX doesn't import
IF fields at all anyway - they are just left blank.
So AFAICS, we should ALWAYS return false at this point.]
make CppunitTest_sw_ooxmlfieldexport \
CPPUNIT_TEST_NAME=testConditionalText4
Not surprisingly, there were no existing unit tests that matched
(because I already did a crash-testing assert last year).
Change-Id: I4d737172ee278c07596eaf47e1bc320e4f42be55
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174699
Tested-by: Jenkins
Reviewed-by: Justin Luth <jluth@mail.com>
When using the Dvorak - QWERTY keyboard, the
event's charactersIgnoringModifiers string causes
pasting to fail so replace both the event's
characters and charactersIgnoringModifiers strings
with this menu item's key equivalent.
Also, fix the following related bugs when using the
Dvorak - QWERTY keyboard:
- When pressing Command-V with a Dvorak - QWERTY
keyboard, that single event passes through the
-[SalNSMainMenu performKeyEquivalent:] selector twice
which causes content to be pasted twice in any text
fields in the Find and Replace dialog.
- When using the Dvorak - QWERTY keyboard and the
Command key is pressed, any key events that match a
disabled menu item are handled in the
-[SalFrameView noop:]. However, the Dvorak - QWERTY
event's charactersIgnoringModifiers string can cause
cutting and copying to fail in the Find toolbar and
the Find and Replace dialog so replace the event's
charactersIgnoringModifiers string with the event's
character string.
Lastly, fix temporarily disabled menu items failing to
be reenabled when a modal dialog closes caused by the
fix for tdf#126638 by returning the last enabled state
set by the LibreOffice code. Apparently whatever is
returned by -[SalNSMenuItem validateMenuItem:] will be
passed to -[NSMenuItem setEnabled:] which can cause
the enabled state to be different than the enabled
state that the LibreOffice code expoects. This results
in menu items failing to be reenabled after being
temporarily disabled such as when a native modal dialog
is closed. So, return the last enabled state set by the
LibreOffice code.
Change-Id: Iae9cc6f1b94484c1529b22ea3a7acdac2009a58b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174462
Tested-by: Jenkins
Reviewed-by: Patrick Luby <guibomacdev@gmail.com>
The problem here is that coverity thinks that ParaPortion::GetLineNumber
can return -1, which it can, but only if things are already broken,
so add an assert tbere.
Change-Id: Iceea035edb49bb3e1e0a39a9b63a9176308346ea
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174621
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
The following patch appears to be upstreamed now
incomplete.patch.0
Change-Id: I36a4c62928c050ed4f8103a580899cdb3cf237cf
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174654
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
the SkImageGenerator patch seems to have been upstreamed
Change-Id: Ic2af62b9479c36394c6c0cf19d1c6f61edfff6b0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174653
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Ensure that a read of the pdfdocument
does not use info from failed previous reads.
The result in the problematic documents
was that GetCatalog failed because it was looking
for an m_aTrailerOffset[0] from the failed read,
and thus only empty pages were exported.
I failed to create a clean-room document,
and the single existing unit test that needs a fixup
has the trailer at the same GetPosition,
which replaced the m_aTrailerOffset
instead of creating a second entry.
Change-Id: I470b922323a2b2ebb6d4bbaa58c98f5cc492b750
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174652
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
Reviewed-by: Justin Luth <jluth@mail.com>
Load the 5 pages bugdoc to Writer, notice the image on the last page,
select-all, copy, new document, paste the selection: no image on the
last page.
Given the new document may have different compat flags, this may not be
a bug, but bisect shows this was working before commit
591673d163 (use more SwPosition::Assign,
2022-08-29). Also note that the image in question is anchored to-char
and it has a non-zero content (character) index, that's why it's on the
last page, and not on page 4. The old code kept the content index
unchanged while changing the node index, the new code avoids touching
the nNode member directly.
Fix the problem by still not touching the nNode member directly (to keep
the benefits of the above commit), but explicitly keep the content index
unchanged.
Changing the node index without changing the char index in general is
not safe (the new node may have less characters), but it does work in
this case, and this fixes the regression.
Change-Id: I2a90e4a88608d7435543880f46f09bea21eba48a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/174655
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
