When computing the variations in a background thread, the
code has to work on a copy of the dive plan and the deco
state. Instead of passing a copy via a unique_ptr<>, simply
use value semantics when calling computeVariations().
This does an unnecessary copy of the deco state, when
computeVariations is not run in the background, but so what.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
In 8704a8b the code in cloneDiveplan() was replaced by a simple
assignment statement.
Alas, the original code was more complex: it copied only up
to a certain point (it stopped at automatically generated
steps).
The new behavior made the calculations of variations fail,
because a call to plan() adds deco stops.
Therefore, copy the plan _before_ calling plan().
Fixes#4368
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
We had a pattern where doubles were converted to long with
lrint() and then down-cast to a narrower int type.
Because this is unwieldy, introduce a function encapsulating
this.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The goal here is to add general addition and scalar multiplication
functions to the unit types.
Thereto, we need a CRTP
(https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern)
base class.
However, this breaks compound initialization, so we have to use
named initializers:
weight_t { 2000 } -> weight_t { .grams = 2000 }
The good thing is that this is exactly how these classes were
supposed to be used: make the unit explicit!
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
This was quite ominous: a 60-element fixed size table was
passed as argument to plan(). But there was no check for 60
anywhere? Use a dynamic vector instead.
The whole thing is weird, as the depth of the decostop table
doesn't seem to be used.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
I don't get the point of these calls to dp.clear().
The plan is overwritten immediately afterwards anyway.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
A reference to a unique_ptr<> was captured by a lambda used
to calculate variations in the background.
This is of course disastrous, because if the caller runs
first it will delete the object. It's a wonder that this
didn't crash regularly!?
The problem is that capturing unique_ptr<>s in lambdas
works, but makes the lambda non-copyable. Sadly, the
QtConcurrent::run() function wants to copy the lambda.
For now, do this by a release/reaquire pair. This is
not exception safe. However, sine Qt doesn't support
exceptions, we can live with that.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Improve the event loop architecture by making it set the divecomputer in
the constructor - using the same loop for multiple dive computers is not
intended to work.
Also change `next()` in `divemode_loop` to `at()` to make the name more
aligned with its function.
Signed-off-by: Michael Keller <github@ike.ch>
Fix an interger overflow warning when parsing setpoints.
@bstoeger: In the end it turned out that this parser was only used in
one place in the planner UI, and it was simplest to switch this to
using `QVariant.toFloat()` in the model itself, which is consistent how
the rest of the input values is parsed and validated.
Signed-off-by: Michael Keller <github@ike.ch>
This has become a bit of a catch-all overhaul of a large portion of the
planner - I started out wanting to improve the CCR mode, but then as I
started pulling all the other threads that needed addressing started to
come with it.
Improve how the gas selection is handled when planning dives in CCR
mode, by making the type (OC / CCR) of segments dependent on the gas use
type that was set for the selected gas.
Add a preference to allow the user to chose to use OC gases as diluent,
in a similar fashion to the original implementation.
Hide gases that cannot be used in the currently selected dive mode in
all drop downs.
Include usage type in gas names if this is needed.
Hide columns and disable elements in the 'Dive planner points' table if
they can they can not be edited in the curently selected dive mode.
Visually identify gases and usage types that are not appropriate for the
currently selected dive mode.
Move the 'Dive mode' selection to the top of the planner view, to
accommodate the fact that this is a property of the dive and not a
planner setting.
Show a warning instead of the dive plan if the plan contains gases that
are not usable in the selected dive mode.
Fix the data entry for the setpoint in the 'Dive planner points' table.
Fix problems with enabling / disabling planner settings when switching
between dive modes.
Refactor some names to make them more appropriate for their current
usage.
One point that is still open is to hide gas usage graphs in the planner
profile if the gas isn't used for OC, as there is no way to meaningfully
interpolate such usage.
Signed-off-by: Michael Keller <github@ike.ch>
In 1af00703b3 the constructor of
WSInfoModel was removed, not realizing that it contains a crucial
call to setHeaderDataStrings().
Fixes#4294
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Feels natural in a C++ code base.
The commit is somewhat complex, because it also changes the
return type to volume_t. The units system really needs some
work. :(
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Feels natural in a C++ code base.
This removes a nullptr-check so some care has to be taken.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Feels natural in a C++ code base.
Change the function to return a weight_t. Sadly, use of the
units.h types is very inconsistent and many parts of the code
use int or double instead. So let's try to make this consistent.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Feels natural in a C++ code base.
Moreover, remove the fulltext-unregistration, as this is a
layering violation.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
It is unclear why this was declared in divelist.h.
Moreover, rename it to get_maximal_gas() to better reflect
what it does.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
This accesses the global dive_table, so make this explicit.
Since force_fixup_dive() and default_dive() use fixup_dive(),
also move them to struct dive_table.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Before, a non-owning pointer was passed and the dive moved
away from the dive. Instead, let the caller decide if they
still want to keep a copy of the dive, or give up ownership:
In MainWindow and QMLManager new dives are generated, so
one might just as well give up ownership. In contrast,
the planner works on a copy (originally the infamous
"displayed_dive") and now moves the data manually.
This commit also removes duplicate code, by moving the
"create default dive" code from MainWindow and QMLManager
to struct dive.
Finally, determination of the "time zone offset" is not done
in POSIX, since we want to avoid calls form the core into
Qt.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
These functions accessed the global divelog make this explicit.
I'm still not happy about the situation, because these functions
access global state, such as the selection. I think these
should be moved up the call-chain.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
This function implicitely accessed the global divelog. To make
that explicit make it a member of dive_table, such that the
caller must access it via the global variable.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The decostate was generated in the main thread and passed down to
a worker thread. To make that explicit, use an std::unique_ptr<>
and std::move().
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
This is a messy commit, because the "qPref" system relies
heavily on QString, which means lots of conversions between
the two worlds. Ultimately, I plan to base the preferences
system on std::string and only convert to QString when
pushing through Qt's property system or when writing into
Qt's settings.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Since all code can now directly access C++ structures these
accessor functions were not necessary.
Split out the table from the filterconstraint source file
and include it directly into the divelog.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Having this as a pointer is an artifact from the C/C++ split.
The divesitetable header is small enough so that we can
include it directly.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Not strictly necessary, but a "natural" thing to do in a classical
C++ code base.
Move the tiny trip-table into its own source file, since it also
has its own header.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
This is a humongous commit, because it touches all parts of the
code. It removes the last user of our horrible TABLE macros, which
simulate std::vector<> in a very clumsy way.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Nothing against free-standing functions, but in the case
of dc_watertemp(), dc_airtemp(), endtime() and totaltime(),
it seems natural to move this into the dive class and avoid
polution of the global name space.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
The dive_table will be converted into a table of owning pointers.
Since the trip has only non-owning pointers to dives, turn
its dive_table into an std::vector<dive *>.
Add a helper functions to add/remove items in a sorted list.
These could be used elsewhere.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Since the sorted_owning_table depends on the fact that
different elements never compare as equal, make the
comparison function safer in that respect. If all failes,
compare the pointers.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
