| Age | Commit message (Collapse) | Author |
|---|
|
See instructions in solenv/gbuild/Trace.mk . This generates a file than
can be viewed e.g. in the Chromium tracing view.
Change-Id: I5f90647c58ca729375525b6daed2d4918adc8188
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/88754
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
Previously, once 'm_aContents' had been assigned in
'VclQt5Clipboard::setContents()', its value (or the one set
in a subsequent call to the same method) was always
returned in 'VclQt5Clipboard::getContents()', thus
ignoring all system clipboard updates done by any other
third-party applications, preventing copy-paste from other
applications.
In order to take external clipboard updates into account,
add a slot for the 'QClipboard::changed' signal and
drop the current own clipboard content if the clipboard
has been updated by another application.
In order to detect whether the clipboard update was made
by this 'VclQt5Clipboard' itself or elsewhere, a custom MIME type
"application/x-libreoffice-clipboard-uuid" is added, whose
value is set to the clipboard's (randomly generated) UUID.
If the entry is present and has the correct value, the clipboard
content was added by this clipboard and can be kept.
Otherwise, clear 'm_aContents', so that it's newly assigned
in 'VclQt5Clipboard::getContents()', taking into account
the external clipboard update.
[Side note: Testing showed that, on Wayland, more
'QClipboard::changed' events were emitted without the clipboard
content actually having changed (e.g. when switching focus between
windows), which is why an approach of simply setting a flag and
checking for that one is not enough, like "wrapping" the
'QClipboard::setMimeData()' call in 'VclQt5Clipboard::setContents()'
as follows
m_bIsFillingClipboard = true;
clipboard->setMimeData(pMimeData.release(), m_aClipboardMode);
m_bIsFillingClipboard = false;
and then evaluating the 'm_bIsfillingClipboard' flag in
'VclQt5Clipboard::handleClipboardChange' instead of using the UUID-based
approach. These additional 'QClipboard::changed' events did not
show up the same way while testing on X11.]
Change-Id: Ib3a6a4f9b7f5ca3573666fb9c072ae97cf2e0049
Reviewed-on: https://gerrit.libreoffice.org/68214
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
|
|
Destroy Qt5Frame after all events are received.
Update Qt5Object to use delayed destruction too.
Change-Id: I07c7251e097ad56f3616f325fdc394d00fec4f4f
Reviewed-on: https://gerrit.libreoffice.org/68316
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
|
|
This gets us into the document itself, not only the widgets.
Change-Id: Id0cf1294817309c889729d01242cae69d2391130
Reviewed-on: https://gerrit.libreoffice.org/61094
Tested-by: Jenkins
Reviewed-by: Samuel Mehrbrodt <Samuel.Mehrbrodt@cib.de>
|
|
Widget tree is displayed and basic role mapping is available.
Still needs quite some work (positions, actions, texts, etc.)
Change-Id: I9d26a762f1d9684f33bbb80a384cf2a0b8a905c0
Reviewed-on: https://gerrit.libreoffice.org/60853
Tested-by: Jenkins
Reviewed-by: Samuel Mehrbrodt <Samuel.Mehrbrodt@cib.de>
|
|
This is just a minimal fix for Qt errors from JunitTest_svx_unoapi,
which was manipulating Qt5Timer and other timers, resulting in a
spew of timer error messages like:
QObject::killTimer: Timers cannot be stopped from another thread
QObject::startTimer: Timers cannot be started from another thread
QBasicTimer::start: QBasicTimer can only be used with threads started with QThread
Eventually all the QWidget manipulation in the Qt5Frame must be
redirected to the main thread, just like Qt5FilePicker already does.
Change-Id: I66054e6c90f99d27bd5818dcaa5876c515867f77
Reviewed-on: https://gerrit.libreoffice.org/60672
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
|
|
this is meant to solve the problem of native menu bar overlapping
w/ non-native, as well as the inability to place an object or select
text dragging the mouse cursor w/ LMB pressed
Change-Id: I29f590ebf79d1ecc7e17b402125384cf13774bf3
Reviewed-on: https://gerrit.libreoffice.org/58171
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
|
|
Change-Id: If3a134f67f59d3238c27eb0fef99bf49a8970433
|
|
they will be used by kde5 vcl plugin
Change-Id: Ibcece683802c7a448b165da4c2515408d641927c
|
|
Change-Id: I99d8213d23b09b6cb1e398359ad05c8fa702a8e0
|
|
CommonSalLayout doesn't rally have an interface. It's cluttered
with #ifdefs. Currently we have to move the Qt5Font into the
VCL library. Someone should refactor this...
Doen't render any text yet, but reports some sizes.
Eventually that would cut down the public interface again.
Change-Id: I12f32affb05b37e070c6cbc80db01779f84590b6
|
|
Move out of unx, as this will eventually compile on other
OS platforms. At least currently it doesn't contain platform
dependant code.
Change-Id: Iea0bebf574201881ea158381fe7ba8af2a9a6488
|
