When the cylinders became a dynamic data structure, a get_cylinder() call
suddenly could return a NULL pointer. So use get_or_create_cylinder() for the
first call when parsing the user's data.
Also, deal with an oddity where the string lists look different because an
empty list technically isn't the same as a list with one empty string.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
Add an image that was lost in the latest updates.
Add text about checking of hand-entered coordinates.
Signed-off-by: willemferguson <willemferguson@zoology.up.ac.za>
This seems to work much more reliably as it specifically compares the cursor
position to the visible bottom and top of the screen.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
The GridLayout isn't part of a Layout, so set its width and
have the combo box fill it's part of the grid.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
This still is way to fragile. Ordering of object completion doesn't appear
consistent. Hopefully bringing the properties all into one (now reasonably
named) object will help.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
I don't expect this to be complieted. I'll keep it in the tree
for the moment, but expect to remove all of the related code
eventually.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
With the columnSpan 3 (in a grid with only two columns) they were drawn
on top of the drop down for the theme.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
In a color swatch the color of the body of the swatch is the one that
matters. And we need to ensure readability. So for each pair of colors,
use the combination in the regular order (color/textcolor) for the
background swatch, and inverted for the text swatch.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
The TemplateLabel comes with a set background color which creates a very odd
effect and makes the text generally unreadable in the color swatches.
Fix this by matching the color of the parent rectangle.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
These are not color swatches like the rest, these are the column titles, so
don't draw them the same way as the color swatches.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
Added text dealing with problems in user cloud registration
and loss of password
Other small edits and corrections to these texts.
Signed-off-by: willemferguson <willemferguson@zoology.up.ac.za>
The whole point of having X::instance() functions is to solve
the infamous "Static Initialization Order Fiasco": When having
global objects in different translation units, their order
of initialization is undefined. Thus, when these objects access
each other one cannot guarantee the correct order of
initialization. The X::instance() functions generate the objects
on first use.
DivePlannerPointsModel has such an instance() function. However,
for convenience(?) in diveplanner.cpp we find the global variable
static DivePlannerPointsModel* plannerModel =
DivePlannerPointsModel::instance();
Thus, the DivePlannerPointsModel constructor is run before main(),
negating the whole purpose of the instance() function.
Let's remove this line to avoid hard-to-debug startup issues.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Some Wikipedia pages use special (non-ASCII) unicode symbols for
representing the " and ' separators. Before parsing, replace these
by the ASCII symbols to enable copy & paste from Wikipedia (and
other sources?).
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
On the dive site screen, when entering invalid GPS coordinates,
we cleared the location of the dive site. Don't do this. To
clear the location, the user now has to enter the empty string.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
On the dive site edit screen, when the user enters invalid
coordinates and saves, we treat this as "no location". This
is rather unfriendly, therefore warn the user with a visual
clue. This is performed by setting the background color of
the widget to red.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
As we do XSLT parsing for the CSV import, ampersand characters need to
be encoded with & for the parsing to succeed.
Fixes#2037
Signed-off-by: Miika Turkia <miika.turkia@gmail.com>
move initialWidth/initialHeight out of subsurfaceTheme to placeHolder,
this secures no change in behaviour.
Signed-off-by: jan Iversen <jan@casacondor.com>
Show all colors in a theme, allowing user to change theme.
This is preparation for allowing users to click on the individual
colors and change them.
Signed-off-by: jan Iversen <jan@casacondor.com>
Make variables and functions static where possible, this is done
to prevent different versions of theme with different values.
Signed-off-by: jan Iversen <jan@casacondor.com>
Use the TemplatePage and TemplateButton in the converted pages.
In order to avoid a bigger search/replace all other pages are left
untouched (for now).
Signed-off-by: jan Iversen <jan@casacondor.com>
Do "git mv SsrfButton TemplateButton", and search/replace
all uses.
The general idea of the templates are to secure common layout,
but also to isolate the Kirigami parts (slowly) in the templates.
Signed-off-by: Jan Iversen <jani@apache.org>
When editing a dive site, the user can search for close dive sites
to merge duplicates. Dive sites without location are treated as
being located at 0N0E. This makes no sense, because:
When selecting a dive site without location, we shouldn't list
dive sites close to 0N0E.
Likewise when having a dive site close to 0N0E, we shouldn't list
dive sites that have no location.
Therefore, ignore these cases.
This also means that now dive sites without location are not
considered as close to other dive sites without location. That
might be a debatable point.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
After selecting dives, the selected dive sites are collected.
This was done using the selectionModel()->selection().indexes(),
which is wrong, because it gives one index per row *and* column.
Accordingly, every dive site was added numerous times to the
array of dive sites to be selected. Change this to
selectionModel()->selectedRows(), which gives one entry per row.
Moreover, if multiple dives with the same site were selected,
this site was also added to the array multiple times. Therefore,
check the array before adding sites.
Note that all this should not change the user experience in
any way, it is only a code-hygiene thing.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Usually, we use PascalCase (i.e. camelCase with a capital
letter at the start) for class names. For consistency, let's
do it here as well.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
When starting / changing the dive-site filter, inform the map of
the changed dive site selection by calling
MapWidget::instance()->selectionChanged();
This fixes a bug, where on clicking dive sites in the dive site
tab the dive sites from the *previous* click were highlighted.
Perhaps the selectionChanged() call should be put into the
setSelected() call. But the data flow between the different
parts of the dive-site and map code are so convoluted that I
don't want to risk anything!
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
We tend to use lower-case filenames. Let's do it for these files
as well. Simple search & replace.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Something is seriously wrong with the default Ubuntu 18.04 environment
on GitHub Actions. C++ builds fail with very confusing messages about
C++11 support.
Switching to building in a Ubuntu 18.04 based container - that seems
redundant, but it fixes the problem.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
A recent merge upstream causes build faiilures because the
PROJECT_VERSION variiable isn't defined. I can't figure out how to make
sure that it is defined correctly, so hack around that by using the
version prior to that merge.
See https://github.com/LaurentGomila/qt-android-cmake/issues/33
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
This class contains only static functions (i.e. it does not contain
any state). There does not seem to be a reason to have an instance
of that class. Therefore, remove the instance() function.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
QMLInterface::instance() is only used in one single place. This
makes the whole notion of having a global instance of the object
moot, isn't it?
Simply make the object static to the function that uses it, which
guarantees that the object will be created when the function is
called. I.e. the same behavior is retained with less complexity.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
It appears weird to connect the QMLInterface signal/slot combination
in a static helper function. This generates a bunch of lambdas that
call the instance() function.
Instead, simply do the connections in the constructor as we do it
in numerous other places.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
When either length or volume unit is changed, the text needs to be
changed.
Let signal lengthChanged and volumeChanged cause a recalcation and therefore
new text.
Signed-off-by: jan Iversen <jan@casacondor.com>
In order to be able to roll out new betas, we need to first increment the
version number. Given the magnitude of the changes, incrementing the minor
version (not that we have ever been really consistent with how we do the
numbering in the first place).
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
The gas change context menu was filled from the CylindersModel.
This means that even before saving new cylinders, the user could
add a gas change event to these cylinders.
Instead, fill from the current dive.
Fixes#2552.
Reported-by: Doug Junkins <junkins@foghead.com>
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
When adding a gas change event via a context menu, the gas-id and
timestamp were passed in two distinct ways.
1) The gas id was extracted from the text of the action. This meant
doing rather complicated parsing.
2) The timestamp was passed via the "user data" of the action, which
means transporting via "QVariant".
There is a much simpler way to pass arbitrary data, that is strongly
typed: lambdas. Instead of shoehorning the data onto the action in
an archaic way, we can simply connect to a stateful lambda. That's
what they're for after all.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
It shouldn't happen, but currently we overwrite the displayed_dive
without updating the CylindersModel. Thus, CylindersModel may now
crash when the new displayed_dive has less cylinders than the old
one.
For now, catch this condition. Treat the root cause later.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Emit signals for each standard color when theme changes.
The iconStyle property was changed from being a constant, because it can
change, and thus a signal was added.
Signed-off-by: jan Iversen <jan@casacondor.com>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
