| Age | Commit message (Collapse) | Author |
|---|
|
The code is rather complex. Firstly, we have different representations
of pictures throughout the code. Secondly, this tries to do add the
pictures in batches to the divepicture model and that is always rather
tricky.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
QVector doesn't have a function to insert a range of pictures,
which we will need for undo of image adding/deletion.
Moreover, std::vector gives us stronger guarantees. For example,
if capacity is large enough, it guarantees that there will be
no reallocation and thus iterators stay valid. I have not found
such a guarantee in the Qt docs.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
dive-pointers are stable and the dive picture model is reset
if a selected dive is removed, so there is no risk in keeping
pointers.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
The profile-widget doesn't use that information anymore.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Even though the functionality is seemingly trivial, this is a bit
invasive, as the code has to be split into two distinct parts:
1) Post undo command
2) React to changes to the divelist
Don't compile that code on mobile.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
For consistency with equipment, use our table macros for pictures.
Generally tables (arrays) are preferred over linked lists, because
they allow random access.
This is mostly copy & paste of the equipment code.
Sadly, our table macros are quite messy and need some revamping.
Therefore, the resulting code is likewise somewhat messy.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
For undo of picture manipulation, it will be crucial that the
model and the core have the same order of pictures. The first
sort criterion will be time, the second filename in the case
that two pictures have, for whatever reason, the same timestamp.
However in the core we us C-strings and thus sort byte-wise
using strcmp. In the Qt-part we use QStrings, which sort according
to unicode encoding. To enable consistent sorting, change the
Qt-part to std::string, which uses a C-style 0-terminated string
as its backing store.
One might argue that in general filenames should use system-encoding
and therefore use std::string instead of QString. However, a
broader conversion to std::string turned out to be very painful,
since Qt is (deliberately?) difficult to use with std::string.
Notable all the file-manipulation functions don't take std::string
by default. Thus, this commit only converts the internal data
of DivePictureModel, but continues to use QString for the Qt-facing
interface.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
This allows us to decouple dive.h and divelist.h, a small step in
include disentangling.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
The DivePictureModel kept a pointer to picture for each entry.
Firstly, this is dangerous from a data-consistency point of view.
Secondly, the entry wasn't even used anywhere. Remove it.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Instead of assigning to a QVariant ret and returning at the end,
return directly in the various switch-cases. This makes the code
more readable, and is more idiomatic C++, as it avoids unnecessary
copies.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Currently, move only those functions that do not access dive
structures.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
This commit adds an entry to the dive media context
menu which offers to write a subtitle file. This
creates an .ass file for the selected videos.
In an attempt to to clutter the screen too much, don't
show irrelevant entries (zero temperature or
NDL and show TTS only for dives with stops).
VLC is able to show these subtitles directly, they
can be integrated into the video file with ffmpeg.
Signed-off-by: Robert C. Helling <helling@atdotde.de>
|
|
The thumbnails were fetched in the background to achieve a
snappier UI. The problem with that is that on LaTeX etc.
export only placeholder thumbnails were shown.
Therefore, implement a synchronous mode. This only tries
to fetch cached thumbnails or calculate thumbnails for
images. Videos and remote files are not supported.
Fixes #1963
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
On the profile, the run-length of the videos is visualized by a bar.
Add the same information to video-thumbnails in the dive-photo-tab.
Though in this case, render it as text on top of the thumbnails.
Fixes #359
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Paint a rectangle on top of thumbnails indicating the run-time
of the video.
Use the z=100.0-101.0 range for painting the thumbnails, whereby
the z-value increases uniformly from first to last thumbnail
(sorted by timestamp). The duration-bars are placed at z-values
midway between those of the thumbnails.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
9efb56e2d43161d952efb444d1f13d87bfdd45b5 introduced rather complex
logic for picture drag'n'drop events onto the profile. Among other
things, the code had to check whether the picture actually belongs
to the displayed dive.
This can be simplified by transporting the dive-id in the drag'n'drop
event structure. The flow goes like this:
DivePictureModel--(1)-->DivePictureWidget--(2)-->ProfileWidget
For (1), we can use the Qt::UserRole role. This was used to transport
the picture-offset, but this is not needed anymore since ProfileWidget
was decoupled from DivePictureModel.
For (2), we simply replace the "position" value, which was never used.
Why would the receiver care which pixel was pressed in the media-tab?
This commit also contains a minor cleanup in DivePictureWidget:
QListView::mousePressEvent(event) was called in both branches of an
if and can therefore be removed from the if. This is so trivial,
that it doesn't warrant its own commit.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Gracefully handle drag & drop to the profile, which changes the
offset of the pictures. To do this, keep the pictures in the
DivePictureModel and the ProfileWidget2 sorted by offset and
re-arrange if needed to keep the list sorted. This needs some
code reshuffling.
Introduce a helper-function that moves ranges in arrays.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
|
|
As long as ProfileWidget2 and DivePictureModel showed the same set of
pictures and any change would lead to a full recalculation of the set,
it made sense to let ProfileWidget2 use DivePictureModel's data.
Recently, keeping the two lists in sync become more and more of a
burden. Therefore, disconnect ProfileWidget2 and DivePictureModel. This
will lead to some code-duplication and perhaps a temporary drop in
UI-performance, but in the end the code is distinctly simpler and also
more flexible.
Thus, for example the DivePhotoTab could be changed to support headings
without having to touch ProfileWidget2 at all.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Users might have edited their pictures. Therefore, instead of identifying
pictures by the hash of the file-content, use the file path. The match
between original and new filename is graded by a score. Currently, this
is the number of path components that match, starting from the filename.
Camparison is case-insensitive.
After having identified the matching images, write the caches so that they
are saved even if the user doesn't cleanly quit the application.
Since the new code uses significantly less resources, it can be run in a
single background thread. Thus, the multi-threading can be simplified.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
This fixes a bug introduced in fbe1144eaf7e800a014c7a97b846835ba9f3bc7f:
For an empty log, in DivePictureModel::updateDivePictures()
beginResetModel() would be called without a corresponding endResetModel().
It is unclear whether this can ever be hit, because in the no-dives
case, at least in the desktop version no profile is shown.
Note, that this makes the check double-unnecessary.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
In DivePictureModel, rowDDEnd and rowDDStart specify the range of
pictures in the profile plot. Obviously, these have to be adjusted
when pictures are deleted.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
C++ permits use of parameters without name, which signals unused
Signed-off-by: Jan Iversen <jani@apache.org>
|
|
On deletion of a single or multiple pictures, the whole DivePictureModel
was repopulated, which was clearly visible in the UI, owing to the
reconstructing of all images in the profile plot.
To avoid this vexing behavior, implement proper deletion routines in
DivePictureModel and ProfileWidget2. Since this needs sensible erase()
semantics the QList<PictureEntry> member of DivePictureModel was
replaced by a QVector. A QVector should be the default anyway, unless
there are very specific reasons to use a QList (which actually is
a deque, not a classical linked list).
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
The function removePicture() had a flag "last", which would indicate
that the called had finished removing pictures. Only then would
the model be recalculated.
This is a strange interface and, matter of fact, the caller was buggy:
if the last picture to be removed didn't have a proper url, removePicture()
was never called with "last" being set.
Change the interface to take a list of pictures to be deleted. This
will allow us to make picture deletion smarter in follow-up commits.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
In the old code a combination of removeRows()/insertRows() was used
to signal a model reset.
Replace this by a single, modelReset signal. This saves a call to
plotPictures() and will allow us to be smarter in the future,
when removing pictures.
Reset-model and remove-items are semantically different.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
In the old code, we used to reload the whole picture list on drag & drop
to the profile. Instead, only update the drag&dropped picture and repaint
the profile-pictures.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
The size of the to-be-created thumbnails was passed from DivePictureModel
to Thumbnailer. This became more and more bothersome, because the size
had to be stored with the request. Calling from Thumbnailer into
DivePictureModel was not an option, since this is not linked to all tests.
Therefore, move these functions to the Thumbnailer class.
Since the maximum thumbnail size is now known to the thumbnailer, the
dummy and failure images can be precalculated, which makes switching
between dives faster.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Connect the thumbnailer signal to the dive picture model slot.
This needs some code-reshuffling in the dive picture model.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Create a new class, which performs all thumbnailing code.
This is mostly code reshuffling. Thumbnails are extracted
either from a cache or thumbnail calculation is started in
a worker thread.
Since getHashedImage() is called from a worker thread it
makes no sense to call subfunctions in yet another worker
thread. Remove these calls.
In contrast to the previous code, on error the background
thread produces a failure image, but it is not yet shown.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
In imagedownloader.cpp the only thing we need from the picture struct
is the filename. Therefore, use QStrings instead of the picture struct.
This simplifies memory management.
Remove the clone_picture() function, which is not needed anymore.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
SHashedImage was a subclass of QImage, which fetched the image according
to the filename hashes. Turn this into a function, as this is much more
idiomatic and flexible.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
The old code loaded all thumbnails into memory at once. This does
not scale to logs with thousands of pictures. Therefore, save
the pictures to individual files and only load the currently
needed pictures.
Currently, this will make changing switching between dives slower,
because the thumbnails are loaded from disk. In the future, it
is planned to do this in a background thread without blocking
the user interface.
A notable difference to the old code: Thumbnails are now indexed
by the image-hash (i.e. the content of the raw image) and not
by the filename of the image. Thus, different paths to the same
image should only be saved once.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
To potentially conserve memory, don't keep copies of scaled thumbnails.
Scale the thumbnails on demand.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
To ease trouble-shooting of the picture thumbnailer add a number
of debug- and info-messages.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Signed-off-by: Lubomir I. Ivanov <neolit123@gmail.com>
|
|
Replace constructs of the kind
s.toUtf8().data(),
s.toUtf8().constData(),
s.toLocal8Bit().data(),
s.toLocal8Bit.constData() or
qUtf8Printable(s)
by
qPrintable(s).
This is concise, consistent and - in principle - more performant than
the .data() versions.
Sadly, owing to a suboptimal implementation, qPrintable(s) currently
is a pessimization compared to s.toUtf8().data(). A fix is scheduled for
new Qt versions: https://codereview.qt-project.org/#/c/221331/
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
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>
|
|
This implements different zoom levels for the dive photos tab as
suggested by Stefan Fuchs <sfuchs@gmx.de> in #898.
The zoom level can be changed using a slider or CTRL+mousewheel.
Zoom levels range from a third of the standard thumbnail size to
thrice the standard thumbnail size.
Thumbnails are cached in maximum resolution and scaled down on
the fly. Because the profile widget took its pictures from the
photo list model, an extra picture copy with a fixed size had
to be introduced.
The UI is still a bit crude.
Reported-by: Stefan Fuchs <sfuchs@gmx.de>
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
It was only used by the DivePictureModel class, no need to export it.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Correct way of using indices for rowDDstart and rowDDend.
Reset rowDDstart and rowDDend at beginning of updating dive pictures.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
|
|
Suggested-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
|
|
In the dive picture tab show pictures of all selected dive.
But at the same moment take care that in the profile only
pictures from displayed_dive are displayed.
Signed-off-by: Stefan Fuchs <sfuchs@gmx.de>
|
|
The thumbnailing in qt-models/divepicturemodel.cpp was performed
concurrently, but the thumbnailCache was not protected from races.
Resolve this by guarding the thumbnalCache accesses with mutexes.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
The code of DivePictureModel used a QHash to keep track of thumbnails.
Not only was the code rather complex - it also had the consequence that
pictures are sorted according to the hash function, i.e. seemingly
random.
This commit replaces the QHash by a simple QList which keeps track
of thumbnails and some meta-data.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
Note that constructs like QList<QFuture<void>> is incompatible with
pre-C++11 compilers. But the code base is so full of C++11isms that
we may just as well remove asymmetric eye-sores of the kind
"QList<QFuture<void> >".
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
|
|
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>
|
|
Signed-off-by: Tomaz Canabrava <tomaz.canabrava@gmail.com>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
|
|
This drastically improves the time it takes to select a dive
with several pictures at the expense of longer startup and
bigger hash files.
Signed-off-by: Robert C. Helling <helling@atdotde.de>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
|
|
This required a bit more untangling, but with this it seems we can build
subsurface-mobile again (at least on the desktop).
Interesting is the removal from inside the ImageDownloader of the call to
DivePictureModel::instance()->updateDivePictures() - which actually could
cause some interesting recursion issues. If it turns out we did indeed
need this, it needs to be re-architected.
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
|
|
Altering DivePicture model to allow deleting images
from the QListView without immediate updating of the
list. Updating is determined by an additioanl parameter
Signed-off-by: Guido Lerch <guido.lerch@gmail.com>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
|
Berthold Stoeger
Robert C. Helling
jan Iversen
Jan Mulder
Stefan Fuchs
Dirk Hohndel
Tomaz Canabrava
Guido Lerch