to explain the absence of languages that are listed by 'Preloading'
because the languages handled by a remote LanguageTool are removed
from the languages handled with local data.
Change-Id: I203657f68a7b9399e39ebf9ab9447c7ae56efe68
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173448
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
(cherry picked from commit c5fc86e770901eb90356cb31d4ac71ac7dad9ce6)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173538
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Changed the slide layer message layout.
Sometime some text field layer was missing.
Try to skip empty layers.
Change-Id: I52f00e637da7ab705a1fcb52a70af88cf619c8f7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173540
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
* Update translations from branch 'master'
to 42ecd2971a60e08485dbafbace2f2ed7245cd48c
- update translations for master
and force-fix errors using pocheck
Change-Id: Ic6577d3caf2dfad168da2a61b78bf2bc26e5d361
To facilitate the debugging of the
ScriptForge Basic code called by the
python scripts, 2 methods have been
added to the ScriptForge() class:
- initializeRoot(force = False)
Initialize the global scriptforge
data structure.
- When force = False, only when
not yet done.
- When force = True, reinitialize it
whatever its status.
- errorHandling(standard = True)
Determine how errors in the ScriptForge
Basic code are handled. Either
- the standard mode, i.e. display a "crash" message
to the user
- the debugging mode, i.e. the execution stops
on the line causing the error
Example (cfr. tdf#162962)
from scriptforge import CreateScriptService, ScriptForge
sf = ScriptForge()
sf.errorHandling(False)
def testSF():
srv = CreateScriptService("Calc")
g_exportedScripts = (testSF,)
Change-Id: I25d7d8e36e71d95ff8a04191bb8344794886d69e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173458
Reviewed-by: Jean-Pierre Ledure <jp@ledure.be>
Tested-by: Jenkins
which is very expensive these days
Change-Id: Ia8485f38b7aeb763a49a1acb433e22d1360694e4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173513
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Load a document, see the 'signaturestatus: 4' callback
(SignatureState::NOTVALIDATED, "signature is OK, but certificate could
not be validated"), wait till the view is initialized, notice the
strange later arriving 'statechanged: .uno:Signature=3' callback, which
hints that the document was modified, but we're right after load, so
that should not happen.
I already tried to prevent doc modified status in commit
654f972a97c374fa90eb3984a44c7f54ddba9c61 (sw lok: make sure author name
change doesn't mark the doc as modified, 2023-01-16), but this is not
enough for the doc sign case, where SfxObjectShell::SetModified() ends
up refreshing the doc title, which calls
SfxObjectShell::GetDocumentSignatureState() and then caches the wrong
SignatureState::INVALID state.
Fix the problem by adding a flag to not mark the document as modified
during "init view" in the first place: that keeps the document unchanged
after updating fields, but doesn't break the doc sign status.
Now LOK clients get a first 'signaturestatus: 4' on load, and a
'statechanged: .uno:Signature=4' on initializing the view, which is a
bit redundant, but at least is consistent.
Change-Id: I42d50525ba96900eeeb927a50f1484c882347b13
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173514
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
... duplicates; the directory names of AutoCorrect entries are
user-editable, so this needs to be supported.
AutoCorrect uses an ODF package because the ODF document loading code
requires the ODF document to be in an ODF storage with a MediaType
property.
AutoCorrect writes an empty mimetype file, and if such is present in an
.odt file that is being loaded, existing checks will detect it as
corrupted, so we can use this to check that the file is an AutoCorrect
file and turn off the case-insensitive check.
(regression from commit 4833f131243bdb409ddfaff8b4db87d4ed2af98f)
Change-Id: I43887f7dad0c8cbb465b4c0f1c38bcc3244a7675
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173477
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Jenkins
Happens when using a "non-text-only" replacement (which can be created by
selecting text before opening AutoCorrect dialog).
asserts in SvxAutocorrWordList::WordMatches():
/usr/include/c++/14/string_view:256: constexpr const std::basic_string_view<_CharT, _Traits>::value_type& std::basic_string_view<_CharT, _Traits>::operator[](size_type) const [with _CharT = char16_t; _Traits = std::char_traits<char16_t>; const_reference = const char16_t&; size_type = long unsigned int]: Assertion '__pos < this->_M_len' failed.
The "IsTextOnly()" branch adjusts nEndPos, clearly it's needed in the
else branch too.
Change-Id: I442042a5560d8143f1293ac29476aff6e93d7f3b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173476
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Length of tabulator portions wasn't taken into account
during calculating overhanging lines, resulting missing
space shrinking.
Regression from commit 17eaebee279772b6062ae3448012133897fc71bb
"tdf#119908 sw smart justify: fix justification by shrinking".
Change-Id: I34bb007940b91e861b24413e76c910ee62aed158
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173451
Reviewed-by: László Németh <nemeth@numbertext.org>
Tested-by: Jenkins
Unify the handling to initialize the control state
in `FileDialogHelper_Impl::preExecute` and set the
state of the controls synchronously (right away)
for all platforms.
This was already done for Windows and macOS, but
not for Linux, where this was delayed by doing
so asynchronously in a user event.
The asynchronous way of doing it was introduced
in
commit e1d6dc26677c9ef24360cd3b2b40acfc7dee13a0
Author: Frank Schönheit <fs@openoffice.org>
Date: Tue Oct 30 13:12:46 2001 +0000
#93977# do an asynchronous InitControls
and the the comment gives this reason:
// allow for dialog implementations which need to be executed before they return valid values for
// current filter and such
However, this doesn't work everyhwere, so this
was already switched back to being synchronous
for macOS in
commit 420c7215ad52ce0a2fc652c9b537af20c80e86f0
Author: Kurt Zenker <kz@openoffice.org>
Date: Wed Mar 5 15:44:58 2008 +0000
INTEGRATION: CWS aquafilepicker02_DEV300 (1.134.22); FILE MERGED
2008/02/11 17:38:41 fheckl 1.134.22.2: RESYNC: (1.134-1.138); FILE MERGED
2008/01/14 08:40:47 fheckl 1.134.22.1: Allowing for corrrect fpicker init on MacOSX
and for Windows in
commit 9ebe7641b7f5f5ceebbe0b2dad73e8bb73997967
Author: Oliver Bolte <obo@openoffice.org>
Date: Fri May 30 10:34:06 2008 +0000
INTEGRATION: CWS hro35_DEV300 (1.142.32); FILE MERGED
2008/05/26 09:16:01 hro 1.142.32.1: #87244# Dialog initialization sequence has to be done synchronous for modal Vista dialog
. The asnyc way also doesn't work for the Qt-based VCL
plugins as mentioned in
Change-Id: I9edf65a31f8268feb13fd6abf627744d68fd79e3
Author: Michael Weghorn <m.weghorn@posteo.de>
Date: Mon Sep 16 14:06:39 2024 +0200
qt: Add "Sign with default certificate" to QtFilePicker
:
> When no signing key is set there, the checkbox currently
> isn't disabled as expected, but that's a separate issue,
> also seen with the "Edit filter settings" checkbox that
> is disabled with gtk3, but not with qt6 independent of this
> commit.
Unify to set the state synchronously everywhere again, which
makes the Qt-based VCL plugins work, and also works as
expected in my tests with the gen and gtk3 VCL
plugins on Linux, even when locally disabling the
`CommonFilePickerElementIds::LISTBOX_FILTER` handling in
`FileDialogHelper_Impl::handleControlStateChanged` for
testing. (That gets called for these implementations
when the dialog shows and would otherwise take care of
setting the proper state as well.)
The gtk4 file dialog currently doesn't shown these checkboxes at
all, independent of this commit.
Since the first commit mentioned above only
refers to an issue in the internal StarDivision
bug tracker, it's not clear on what platform the
synchronous handling was causing problems in the
past, but it seems that the issue is gone by now.
Change-Id: I4bbc7f8d03061f9858b2e355b6d1874a0b71adfe
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173456
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
commit 426c641976688e3e4d1ce66f76b27ccbd2dca55a
Date: Wed Jul 17 12:47:22 2024 +0200
tdf#159040: add sign with default certificate to save dialog
added a call to `FileDialogHelper_Impl::updateSignByDefault`
to `FileDialogHelper_Impl::InitControls` that gets asynchronously
called for non-Windows and non-macOS, see
`FileDialogHelper_Impl::preExecute`.
Also explicitly call it for the Windows and macOS case, like
all of the other methods that also get called there in that
code path.
This makes the "Sign with default certificate" checkbox
disabled when no GPG key is set up in "Tools" -> "Options"
-> "User Data" -> "Signing Key" even if commenting the call to
`FileDialogHelper_Impl::updateSignByDefault`
in `FileDialogHelper_Impl::handleControlStateChanged`
on Windows, i.e. the correct state of the checkbox no longer
depends on an implicit filter change when the dialog
gets started.
This is mostly in preparation of using the same code
path for the Qt-based VCL plugins on Linux, where
no implicity filter changed signal happens and the
"Sign with default certificate" and "Edit filter settings"
checkboxes are currently always enabled after
starting Writer and opening the "Save As" dialog.
Change-Id: Ibc6c2a16dc04f0fe7e41317aeb98ea159e7aec8c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173455
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Extend `QtFilePicker` to implement the new feature
added and implemented for other file picker implementations
in
commit 426c641976688e3e4d1ce66f76b27ccbd2dca55a
Author: Sarper Akdemir <sarper.akdemir@allotropia.de>
Date: Wed Jul 17 12:47:22 2024 +0200
tdf#159040: add sign with default certificate to save dialog
This makes the checkbox show up when opening the
"Save As" dialog in Writer using Ctrl+Shift+S and
behaves as expected when a GPG key has previously been
set in "Tools" -> "Options" -> "User Data"
-> "Signing Key".
When no signing key is set there, the checkbox currently
isn't disabled as expected, but that's a separate issue,
also seen with the "Edit filter settings" checkbox that
is disabled with gtk3, but not with qt6 independent of this
commit.
Change-Id: I9edf65a31f8268feb13fd6abf627744d68fd79e3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173453
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Be consistent and just use the `m_xValue` member directly,
as is already done elsewhere in the same methods.
Change-Id: I9b7a1e91cbe224115863cde9bb8d4fed14448436
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173447
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Tested-by: Jenkins
Use m_ and x prefixes, as it's a class member and
a uno::Reference.
Change-Id: I3b345749c256bccb3c6c59b54eed4db8f38c5a7c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173446
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Merge the `CAccComponentBase` class into its
only subclass `CAccComponent`.
While at it, make the `m_xComponent` member
private, which was previously protected for no reason
anyway.
Change-Id: I029348c51aefd1b7cdc3db68bac6abc31de15d61
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173445
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Tested-by: Jenkins
... in preparation of merging `CAccComponent`
and `CAccComponentBase`.
Change-Id: Ic7f436ec1bf1d4cbebfe27e86a29710719a33785
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173444
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Just use the `m_xComponent` member directly, as
already happens elsewhere in these methods.
Change-Id: I1f3020e568d0d9b29576d0f446dea2f159119880
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173443
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Use an `m_ ` prefix because it's a class member, and `x`
because it's a uno::Reference.
Change-Id: I084f41f4bca97b39d3bf365dd9025b0a9d145d24
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173442
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
It's unclear to me what these
//special query
comments above the null checks refer to and I
find them more confusing than helpful.
Remove them.
Change-Id: I3a3d886a664d03ec37085e2f074928c2a3055820
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173441
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Tested-by: Jenkins
No need to manually use the pointer, and have
special handling for an empty Reference.
Change-Id: I414cfdccacecc79594c0bc5b57d782ea98b811fc
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173440
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
which is very expensive these days
Change-Id: I26da48f3b29bc3029d359214e0c50fa51e9e2194
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/173479
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Jenkins