Apparently, in some Qt-versions the destructor of the base class
calls hide on child-objects (according to Qt's object hierarchy).
This is obviously called after the derived class has been destructed.
In a concrete case, the base class of the destructed MainWindow
would hide the TabDiveSite object. That would reset the filtering
if a dive site was selected, which would indirectly access the
MainWindow, which is already partially destroyed.
Therefore, destroy the MainTab before destroying the MainWindow.
Do this by keeping it as a std::unique_ptr subobject. Thus, it
will be destroyed before the MainWindow and remove itself from
Qt's object hierarchy.
Reported-by: Dirk Hohndel <dirk@hohndel.org>
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Properly implement the unsaved-changes flag(s). Since we currently have
two kinds of changes, there are two flags:
1) dive_list_changed in divelist.c marks non-undoable changes. This flag
is only cleared on save or load.
2) QUndoStack::isClean() is used to determine the state of undoable
changes. Every time the user returns to the state where they saved,
this flag is cleared.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The edit dive site button was connected to a *signal* of MainWindow,
which was connected to a slot of MainWindow. Remove the unnecessary
intermediate signal.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Add an edit column that calls the new editDiveSite() function
of MainWindow. The calling code is in DiveSiteSortedModel.
Quite illogical, but that's how TableView works, for now.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
In commit b0556abdd35f96b816ba11e40bf5707abe0c3ebf, the filter-widget
and the filter were connected by a direct function call. This led
to a rather obscure crash on application-close with Qt 5.12. The
crash is due to the Ui::MainWindow class being a sub-object of MainWindow,
but the FilterWidget2 being *not* a subobject.
What happens is that after calling the MainWindow destructor, the
subobjects are destructed, notably the Ui class. Then the base-class
destructor is called (which makes sense, as destructors are called
in reverse order of constructors).
But: the QObject destructor calls hide() on all still existing child-objects
according to Qt's object hierarchy, notably the visible FilterWidget2.
Now the FilterWidget2, on hiding, updates the MainWindow, which has already
destructed all its subobjects. Crash.
Prevent this crash by making FilterWidget2 a subobject of MainWindow
and thus have it destructed before running the QObject destructor.
Alternative ways would be:
1) Use signal/slot() instead of function calls, as these are automatically
removed if an object is destroyed.
2) Make the FilterWidget2 subobject a smart-pointer. Thus, we probably
wouldn't have to include the corresponding header.
3) Make the FilterWidget2 subobject a plain pointer and delete it
explicitly in the constructor.
Reported-by: Jan Mulder <jlmulder@xs4all.nl>
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Register the new FilterDive widget on the mainwindow
so we can trigger a shortcut to display it.
The shortcut currently doesn't exists.
Signed-off-by: Tomaz Canabrava <tcanabrava@kde.org>
Does not solve any problem, but might help users that are confused
about the next/prev DC menu items, to select a different profile
for the currently selected dive. So, enable these menu items only
for dives where more than one DC is used.
Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
Instead of the weirdly named "information" and the inconsistent
"dive_list" use the logical "mainTab" and the camel-cased
"diveList", respectively.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The keeps track of different sub widgets needed by other parts
of the code, notably:
MainTab
PlannerDetails
PlannerSettingsWidget
ProfileWidget2
DivePlannerWidget
DiveListView
Access to these widgets was provided with accessor functions.
Now these functions were very weird: instead of simply returning
pointers that were stored in the class, they accessed a data
structure which describes the different application states.
But this data structure was "duck-typed", so there was an
implicit agreement at which position the pointers to the
widgets were put inside. The widgets were then down-cast by
the accessor functions. This might make sense if the individual
widgets could for some reason be replaced by other widgets
[dynamic plugins?], but even then it would be strange, as one
would expect to get a pointer to some base class.
Therefore, directly store the properly typed pointers to the
widgets and simply remove the accessor functions. Why bother?
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The global object "displayed_dive_site" is used to store the
old dive site data for the edit-dive-site widget. The fields
of the widget were initialized from this object in the show
event. Therefore the object was updated in numerous parts of
the code to make sure that it was up-to-date. Instead, move
the initialization of the object to the function that also
initiatlizes the fields. Call this function explicitly before
showing the widget.
This makes the data-fow distinctly easier to understand.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
In DiveListView, we have a very fundamental problem: When
On the one hand, we get informed of user-selection in the
DiveListView::selectionChanged() slot. This has to set the
correct flags in the C-backend.
On the other hand, sometimes we have to set the selection
programatically, e.g. when selecting a trip. This is done
by calling QItemSelectionModel::select().
But: this will *also* call into the above slot, in which
we can't tell whether it was a user interaction or an
internal call. This can lead to either infinite loops or
very inefficient behavior, because the current dive
is set numerous times.
The current code is aware of that and disconnects the
corresponding signal. This is scary, as these signals are
set internally by the model and view. Replace this
by a global "command executing" flag in DiveListNotifier.
The flag is set using a "marker" class, which resets the flag
once it goes out of scope (cf. RAII pattern).
In DiveListView, only process a selection if the flag is not
set. Otherwise simply call the QTreeView base class, to reflect
the new selection in the UI.
To have a common point for notifications of selection changes,
add such a signal to DiveListNotifier. This signal will be
used by the DiveListView as well as the Command-objects.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
We have to avoid that undo/redo removes the currently edited
dive from under our feet. This code can be removed once proper
undo/redo (including editing) is implemented.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
This refactors the undo-commands (which are now only "commands").
- Move everything in namespace Command. This allows shortening of
names without polluting the global namespace. Moreover, the prefix
Command:: will immediately signal that the undo-machinery is
invoked. This is more terse than UndoCommands::instance()->...
- Remove the Undo in front of the class-names. Creating an "UndoX"
object to do "X" is paradoxical.
- Create a base class for all commands that defines the Qt-translation
functions. Thus all translations end up in the "Command" context.
- Add a workToBeDone() function, which signals whether this should be
added to the UndoStack. Thus the caller doesn't have to check itself
whether this any work will be done. Note: Qt5.9 introduces "setObsolete"
which does the same.
- Split into public and internal header files. In the public header
file only export the function calls, thus hiding all implementation
details from the caller.
- Split in different translation units: One for the stubs, one for
the base classes and one for groups of commands. Currently, there
is only one class of commands: divelist-commands.
- Move the undoStack from the MainWindow class into commands_base.cpp.
If we want to implement MDI, this can easily be moved into an
appropriate Document class.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Poseidon MkVI logs (.txt) were special cased in MainWindow.cpp,
which led to a user-interface inconsistency. In some cases
[user chooses ".txt" (non-Poseidon) and ".csv"], *two*
import-dialogs were shown.
Move handling of Poseidon MkVI logs into DiveLogImportDialog.
There are already other "special" cases handled in this dialog.
At the moment, this shows the first 10 depth-values, which is
kind of useless, as this will all be at surface level. We
might think about something more useful.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Nowadays, we edit dives just by starting to enter data for the dive.
There is no need to explicitly ask to start editing the dive, using the
now removed menu option. This was a left-over of a long past history.
This is fallout from PR #1673.
Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
Actually remove the Subsurface webservice UI from the source, and
deal with all the fallout.
Notice that a part of the change in subsurfacewebservices.cpp is
a block of code that becomes unused, but might contain some valid
logic to be used later. Very similar code is in core/gpslocation.cpp.
And as I earlier broke something here, the unused code is ifdef-ed for
now.
Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
This reverts commit 321a920a98.
It appears that the load_xxx functions aren't called, so while the correct
values are stored to the settings, they aren't retrieved. Let's revert while
this gets fixed.
Fixes#1609
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
The variables referenced are moved into qPref in earlier commits
so in general all QSettings calls are replaced by qPref*:: calls
Signed-off-by: Jan Iversen <jani@apache.org>
The keyword "virtual" signalizes that the function is virtual,
i.e. the function of the derived class is called, even if the
call is on the parent class.
It is not necessary to repeat the "virtual" keyword in derived
classes. To highlight derived virtual functions, the keyword
"override" should be used instead. It results in a hard compile-
error, if no function is overridden, thus avoiding subtle bugs.
Replace "virtual" by "override" where appropriate. Moreover,
replace Q_DECL_OVERRIDE by override, since we require reasonably
recent compilers anyway. Likewise, replace /* reimp */ by
"override" for consistency and compiler support.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
WindowsTitleUpdate is such a trivial object (a QObject with a single
signal and no own state), that it's not really understandable why
it would need all that "singleton" boiler-plate. Just make it
a default constructed/destructed global object.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The currentDiveChanged signal was emitted by the DiveListView
to inform the MainWindow of a change of current dive. The
new current dive was passed as a parameter. The slot in MainWindow
then called select_dive() on the dive.
This seems pointless because:
1) In both emits, selected_dive dive was passed as argument. But
MainWindow can read this global variable itself.
2) Calling select_dive() again is a no-op, because obviously,
this already *was* the selected dive.
Moreover it seems conceptually wrong to set the current dive in the
slot that is informed of the change of the current dive.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Move the find-moved-images functions into a new translation unit
and present the user with the identified matches before applying
them.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The lastUsedDir() functions of MainWindow and Smrtk2ssrfcWindow don't
use any member-objects and are only used in their respective translation
units. Therefore, remove them from the class and made of static linkage.
The lastUsedImageDir() function was declared as a slog, which makes
no sense. Make it a normal static function (though one might argue
why it is assiociated with the DiveListView class in the first place).
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
In commit d21d42b691 helpView was made
a child-object of MainWindow, which is Qt's idiomatic way of having
helpView deleted with MainWindow.
As an unintended consequence, the helpView didn't show. The reason
is that UserManual derives directly from QObject. In contrast, UserSurvey
derives from QDialog and is correctly shown. Therefore also derive
UserManual from QDialog.
Reported-by: Willem Ferguson <willemferguson@zoology.up.ac.za>
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
If "Edit dive" is selected from the dive list or the map view, switch
to a new mode, which shows the dive infos and the profile.
After the edit, switch back to the previous state.
Fixes#1213
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Only allow to enable maximum one of both items tissue heatmap or
heartrate in profile.
This is done by always switching off the other one at the moment you
turn on one of the two items (heatmap or heartrate).
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
Up to now, errors produced by threads were not directly shown in
the MainWindow. Code running in the GUI thread had to manually
show the errors.
This can be simplified by using Qt's queued connection as message
passing facility.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Enable the menu item cloud-online only if the current file is the
cloud storage. Since this has to be checked every time the current
file is set, factor this out into the new MainWindow::setCurrentFile()
function.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Replace the "Take cloud storage online" menu entry by a "Cloud online"
checkbox. After this change, the user can also force going offline.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
There was no reason to do this in a macro. Let the compiler decide
if it wants to inline or not. Note that for consistency with the
Qt functions, collapsAble was replaced by collabsIble.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The MainWindow::setTitle() function was passed an enum, which depended
on whether existing_file is set or not. The check can be (and was!) done
directly in setTitle(). Therefore, remove the whole enum.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
This reenables the computation of plan variations but now in a separate
thread. Once finieshed, a signal is sent to update the notes.
Signed-off-by: Robert C. Helling <helling@atdotde.de>
Instead of using four recent files actions defined in mainwindow.ui,
generate these actions dynamically depending on the macro NUM_RECENT_FILES.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Instead of handing the QSession object down, simply create it in
MainWindow::checkSurvey()
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The old code used to be unnecessarily complex: the recent files list
was extended for each file and shrunk if a load failed.
By adding a file to the recent file list only if the load succeeded, a
whole method could be removed.
Other changes: keep track of the recent files using a QStringList and
clearly separate the actions:
- Read recent files from settings [loadRecentFiles()]
- Write recent files to settings [updateRecentFiles()]
- Update the recent files actions in the menu [updateRecentFilesMenu()]
- Add a file to the list of recent files [addRecentFile()]
With this reorganization the code hopefully became more clear.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Correct spelling and typos in file filters.
Unify and translate file filter names.
Don't pass a file filter to a directory open dialog - not needed.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
Calls to report_error() crashed if not called from GUI thread.
Fix this by postponing error message display if not in GUI thread.
Code that creates a thread which possibly calls report_error()
is responsible for calling MainWindow::showErrors() to flush
the accumulated messages.
Note that there is a race condition in report_error() and
get_error_string(). Nevertheless, hitting it should be rather
unlikely (two threads producing error messages at the same time)
and hopefully it can be fixed rather easily.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
When enabling the dive list filter via the menu entry
"Log->Filter dives" and then switch off the filter via the small "close"
button of the filter:
Set the checked status of the menu entry correctly.
Also set it correctly when switching on/off via the menu entry to avoid
any situation where it is not synced.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
report_error() now does this automatically. So all these odd places in which we tried
to make sure that we show errors are no longer needed.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
Also change the menu entry text itself to "Map" instead of "Globe".
"Map" covers both Marble and Qt Location in terms of "map" solutions.
Signed-off-by: Lubomir I. Ivanov <neolit123@gmail.com>
See https://github.com/Subsurface-divelog/subsurface/issues/253. The original
behaviour (greyed out when disconnected) is restored. Not sure where and when
got lost, but this simple change fixes it.
Fixes#253
Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
This way the menu entry becomes the togglee it appears it was designed to be in
the first place.
This closes#129
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
If an attempt to contact the cloud storage fails, Subsurface switches into
offline mode. This allows us to go back online again.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
Having subsurface-core as a directory name really messes with
autocomplete and is obviously redundant. Simmilarly, qt-mobile caused an
autocomplete conflict and also was inconsistent with the desktop-widget
name for the directory containing the "other" UI.
And while cleaning up the resulting change in the path name for include
files, I decided to clean up those even more to make them consistent
overall.
This could have been handled in more commits, but since this requires a
make clean before the build, it seemed more sensible to do it all in one.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
start of the QSettinsg Object Wrapper usage on the code
this first patch removes two macros that generated around
200 lines in runtime for something like a quarter of it
Basically, whenever we changed anything we called the
PreferencesDialog::settingsChanged and connected everythign
to that signal, now each setting has it's own changed signal
and we can call it directly.
The best thing about this approach is that we don't trigger
repaints for things that are not directly profile related. (
actually we still do, but the plan is to remove them in due time)
this commit breaks correct atualization of the profile (because
everything was connected to PreferencesDialog::settingsChanged)
and now I need to hunt a bit for the correct connections
Signed-off-by: Tomaz Canabrava <tomaz.canabrava@intel.com>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
Since we have now destkop and mobile versions, 'qt-ui' was a very
poor name choice for a folder that contains only destkop-enabled
widgets.
Also, move the graphicsview-common.h/cpp to subsurface-core because
it doesn't depend on qgraphicsview, it merely implements all the
colors that we use throughout Subsurface, and we will use colors on both
desktop and mobile versions
Same thing applies for metrics.h/cpp
Signed-off-by: Tomaz Canabrava <tomaz.canabrava@intel.com>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
