Commit graph

14025 commits

Author SHA1 Message Date
Dirk Hohndel
da1d6e97f6 Cleanup: avoid memory leak
Coverity CID 208337

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 14:10:53 -08:00
Dirk Hohndel
0e94c21a55 Cleanup: prevent potential out of bounds write
Since we cannot store tanks / gases past MAX_CYLINDERS (currently 20),
there is no point in analyzing those data.

Coverity CID 208339

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 14:04:10 -08:00
Dirk Hohndel
efb2640fc7 Cleanup: avoid memory leak
The JUMP macro includes a 'goto bail', so we need to free devdata there

Coverity CID 208340

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 13:49:49 -08:00
Dirk Hohndel
2c8614e43e Cleanup: check we have dive before dereferencing
Coverity CID: 208341

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 13:41:58 -08:00
Dirk Hohndel
652e382e68 Cleanup: avoid a few memory leaks
Coverity CID 215199
Coverity CID 215195
Coverity CID 215196
Coverity CID 215198

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 13:41:58 -08:00
Berthold Stoeger
65213f044e Update CHANGELOG.md to reflect new filter features
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-26 13:08:49 -08:00
Berthold Stoeger
b6bf57a13b Introduce negate-toggle buttons to filter lists
Introduce toggle buttons which mean "filter all dives except
those fulfilling the selected criteria".

The old code used to check for rowCount() == 0. This should never happen,
because there is always a row "empty field". This check was moved into
the preamble of the functions to seperate it from the actual logic.

Fixes #435

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-26 13:08:49 -08:00
Berthold Stoeger
b86c70ab2c Turn filter list options into context menu
Let the menu introduced in commit 5e86442bab
pop up on right-click instead of button-click.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-26 13:08:49 -08:00
Berthold Stoeger
931947b852 Remove Q_OBJECT macros and friend declaration from *Filter classes
Since these classes do not possess any signals/slots,
these macro are unnecessary.

Likewise, the friend declaration in TagFilter is useless,
since the corresponding fields are in the base class.

Fallout of commit 1a4e7ad0dd.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-26 13:08:49 -08:00
Jan Mulder
ff3b107a9d mobile: close GPS menu drawer when selected
This is mainly a cosmetic improvement. The GPS menu stayed
visible, when selecting most of the options. This looks weird,
as the close of the menu is also an indication of the selected
action carried out. So, just close it

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-26 13:08:37 -08:00
Dirk Hohndel
702ec37145 Android build: make androiddeployqt less verbose
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 12:32:33 -08:00
Dirk Hohndel
eb95cd02c1 Travis: try another work around
The previous hack got us to see that the error was a write error to
stdout and based on
https://github.com/travis-ci/travis-ci/issues/4704#issuecomment-321777557
setting filter_secrets: false will work around that issue.

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 12:32:33 -08:00
Dirk Hohndel
f4c1cef73b Android build: work around odd bug with cmake 3.10
For some reason it suddenly cannot figure out which build program
to use. This seems like a weird hack, but works.

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 09:27:02 -08:00
Dirk Hohndel
4346aaa333 Android build: no point making the docs
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-26 09:24:35 -08:00
Jan Mulder
8e9ba7bd7a CHANGELOG.md
Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-25 19:47:57 +01:00
Jan Mulder
37e98a8d89 Use correct date format
A very very trival fix, for a mysterious issue. When loading
GPS fix data from the server, the string date was parsed with
the format "yyy-M-d". And no, the "yyy" is no typo here, but
was the reason that data from the read from server got a
1/1/1970 data. And when a user decided to upload that data
to the server again, we ended up with 2 copies of the
GPS fix. One with correct data (as originally saved), and
one new with the bogus date.

In order to het rid of those weird 1/1/1970 GPS fixes, users
will have to remove them by hand.

Fixes: #567

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-25 19:47:57 +01:00
Robert C. Helling
ca38644005 Prevent overflow in effective GF calculation
For deep dives with long deco, the sum of deco stops could
overflow. This is prevent by turning it into long.

Signed-off-by: Robert C. Helling <helling@atdotde.de>
2017-12-25 19:47:45 +01:00
Dirk Hohndel
5eec251be8 Travis: hack around Android build failuers
For some reason accepting the licenses for the SDK appears to fail.

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-24 13:50:59 -08:00
Dirk Hohndel
9df88f0145 Android build: reorder downloading Qt vs SDK
Since it's the SDK where things are failing, doing it this way makes the
turnaround time of my attempts to fix this faster. And in the larger
scheme of things, the order is irrelevant.

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-24 13:50:59 -08:00
Dirk Hohndel
f067d3b5d6 Fix build for cmake before 3.10
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-24 13:50:59 -08:00
Dirk Hohndel
726e231d64 Update translation source strings
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-24 12:00:52 -08:00
Dirk Hohndel
6a39e3ff20 Update CHANGELOG
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-24 11:45:26 -08:00
Dirk Hohndel
795f6e24ea Silence cmake 3.10 warnings
Explicitly turning of automoc and friends for every single auto
generated file seems a rather dumb approach...

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-24 11:34:49 -08:00
Jan Mulder
36c8a7e4ca mapwidget: QtQuick really to 2.5
It shall match the Qt version ... according to Dirk.

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-24 11:33:46 -08:00
Berthold Stoeger
130f109442 Remove superfluous QScopedPointer<>s in singletons
There was a curious pattern of singletons being implemented based on
QScopedPointer<>s. This is an unnecessary level of indirection:
The lifetime of the smart pointer is the same as that of the
pointed-to object. Therefore, replace these pointers by the respective
objects.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-24 11:04:10 -08:00
Jan Mulder
4c4222d611 mapwidget: revert QtQuick version to 2.0
In commit f3d978b8a5 the QtQuick version was upgraded
to 2.6 (from 2.0 for the mapwidget). This is, apparently,
too agressive, as there are still mainstream distributions
that are still on Qt 5.5.1 for which 2.6 of QtQuick is too
new. And as a sidenote: Qt 5.5.1 was released in October
2015.

So partially revert commit f3d978b8a5.

Fixes: #978

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-24 10:56:00 -08:00
Dirk Hohndel
eec69de7ab Small whitespace updates
We aren't really consistent. And I don't do this often enough. But based
on a few things that I saw in a recent commit, I wanted to at least fix
those. And then of course fixed everything in those two files.

Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-12-24 09:10:12 -08:00
Berthold Stoeger
734bd6d459 Re-add newlines to help message
The two final newlines in the help message were removed in commit
0c74f7a2c8.
Re-add them.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-24 08:28:26 -08:00
Jan Mulder
dbcf044b93 beginInsertRows requires last >= first
Found while compiled against Qt 5.10 build from source. The assert
Q_ASSERT(last >= first) is trapped by this beginInsertRows in
case there are no pictures. Just do not call this when there are
no pictures.

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-24 08:25:21 -08:00
Berthold Stoeger
7451517e4a Add select-all, deselect-all and invert-selection options to filters
To every filter list add a menu button that allows selection of all,
selection of none or inversion of selection.

Implements #435.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-24 08:24:11 -08:00
Berthold Stoeger
668635e98e Move *FilterModel functions into base class
The *FilterModels had a number of of virtual functions, which only
accessed members of the base class. Moreover, these functions were
identical and generated with macros. Therefore, move these functions
to the base class.

The one excption is data(), which uses different count functions
(passed as a macro parameter). Thus, introduce a virtual countDives()
function and likewise move data() to the base class. A function pointer
might be even more clear, but since the rest of the code/Qt relies
heavily on runtime polymorphism, let's do the same here.

The only macros left are those creating the singleton accessors.
This could be more clearly realized by templates, but let's
likewise keep it the way is.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-24 08:24:11 -08:00
Berthold Stoeger
7eed752ba1 Fold MultiFilterInterface into FilterModelBase
There were two classes, MultiFilterInterface and FiterModelBase.
The latter derives from the former and from QStringListModel.
The former was not used anywhere else. Moreover, in contradiction
to its name, MultiFilterInterface is not an interface (in the Java
sense), because it actually has (non-virtual) data members. All in
all, the data model is very weird.

Merge these two classes, since there seems to be no gain whatsoever
from keeping MultiFilterInterface separate.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-24 08:24:11 -08:00
Berthold Stoeger
1a4e7ad0dd Factor out *Filter code into FilterBase base class
The TagFilter, BuddyFilter, SuitFilter and LocationFilter classes
all did essentially the same thing. Therefore, factor out common
code / objects into a base class FilterBase.

The new base class stores a pointer to the filter model. It was
felt that this is simpler than introducing virtual methods.

The only thing the *Filter classes now do is setting a label and
in one case a tooltip. Thus, in principle, they could be removed
completely, but let's keep them for now.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-24 08:24:11 -08:00
Stefan Fuchs
603e43f46e Use function emitDataChanged in diveplannermodel consistently
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
2017-12-24 00:01:39 +01:00
Stefan Fuchs
337f82c876 Code cleanup: Indentation mistake plus simplification
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
2017-12-24 00:01:39 +01:00
Stefan Fuchs
d703ba99c1 Simplify update of gflow and gfhigh values in the code
The more complex handling is no longer needed because:
- Keyboard tracking for gfhigh/low UI fields was switched off here:
  030c094854
- GFhigh was limited to 40 here:
  53fffe0ce3

Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
2017-12-24 00:01:39 +01:00
Stefan Fuchs
3ad398e3a7 Remove obsolete declaration of void gfLowAtMaxDepthChanged
in ./core/subsurface-qt/SettingsObjectWrapper.h

Code for this was removed here: 5b080bedde

Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
2017-12-24 00:01:39 +01:00
Stefan Fuchs
4db9f5dbbd Correcty typo ghflow -> gflow in diveplannermodel.cpp
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
2017-12-24 00:01:39 +01:00
Jan Mulder
60e82cc2ac cleanup: remove unused include
And why this one? Well, while this include is renamed in Qt 5.10
and gives deprecated compile warnings. And as it unused anyway, just
remove it.

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-23 11:17:47 +01:00
Berthold Stoeger
935fb3c3df Fix BuddyFilterModel
Commit 6343515fed introduced equality
instead of substring comparison for filters. This broke the buddy
filter in the case of more than one buddy, because in such a case
the buddy list is a comma-separated string.

Fix this by splitting the buddy string, trimming the individual
strings and search in the list.

Fixes #969

Reported-by: <yrevawerd@gmail.com>
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-21 21:09:35 +01:00
Jan Mulder
e4530cd5ef cleanup: remove empty function
subsurface_OS_pref_setup() is not not used. Remove it.

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-21 15:11:15 +01:00
Berthold Stoeger
28ae35e7df Make QMap<> access in deleteGpsFix() more idiomatic
To access a QMap<> entry, the value() function is used with a sentinel
as default value. If the sentinel is returned, the code assumes that
the searched for entry doesn't exist.

Make this code more idiomatic by using an iterator and testing for
end().

This fixes a compiler warning, because only one of the elements of
the sentinel was initialized, but the remaining elements were
copied. Harmless, because the code would exit early if it found
the sentinel. Still not nice.

While redoing this function, the entry-not-found message was improved
(adding of function name, space between massage and timestamp) and
elevated from debug to warning level.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-21 00:04:30 +01:00
Berthold Stoeger
22dc7b84f0 Handle plot_info->nr <= 0 gracefully
plot_info->nr should always be > 0. If this is not the case, write a
message to stderr instead of crashing in add_plot_pressure(). This
silences an use-of-uninitialized-variable warning.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-21 00:04:30 +01:00
Berthold Stoeger
0b2d22494c Fix use of uninitialized variable in core/profile.c
last_ceiling was used before initialization in the first iteration
of the loop in calculate_deco_information().

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-21 00:04:30 +01:00
Berthold Stoeger
3c346bb341 Check return of fread() in core/ostctools.c
Since the corresponding error message appears thrice, it is translated
once at the beginning of the function (even in the non-error case).

A single-byte fread() was transformed into getc().

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-21 00:04:30 +01:00
Berthold Stoeger
1c1ca8c1c7 Fix order of initilization list in MainTab constructor.
Besides being the right thing to do (code reflects reality), it
silences a compiler warning.

Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
2017-12-21 00:04:30 +01:00
Jan Mulder
1fee77487b core: ignore erroneous quotes as well
A small redo of 78bafe8f62. The quotes cause the original
functionality not to work. Ignore them as well.

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-20 15:39:54 +01:00
Jan Mulder
6e411cc67d core: fix commit 46004c39e2 for real (erroneous quotes)
Unfortunately, in my commit 48d9c8eb6e, I fixed only half of
the problems related to the functionality introduced by Stefan in
commit 46004c39e2. The lonely m (that was fixed) caused
a parsing error, but forgotten where the single quotes around
the depth value. These quotes simply causes the new functionality
not to work. Again, the fix is simple: do not erroneously save
quotes. And as the new functionality is pretty obscure
(replanning a non-planned dive, and manually entering a gas switch
depth), another bug that could go unnoticed for years.

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-20 15:39:54 +01:00
Jan Mulder
b7dcbb58fa CHANGELOG.md
Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-19 05:13:23 -08:00
Jan Mulder
8c77c15dd8 git storage: invalidate cache on merge dive site
In hindsight a very simple bug to fix, but it requires some
knowledge on the inner workings of our git storage. The changes
on merge of dive sites were simply not saved (completely) because
the git storage code has a cache that we need to invalidate
selectively (ie. for the dive we just gave a new dive site uuid)
to get things finally embedded in the overall commit.

The main reason this bug went unnoticed for more than 2 years is
that most people use the XML/SSRF format (where this problem is
non exsistent), and dive site merging is probably not a very
much used feature either.

Fixes: #939

Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
2017-12-19 05:13:23 -08:00