Can't animate QTransform in qgraphicsItem - c++

I'm trying do some 3D animation in GraphicsScene, for example, to rotate pictures in GraphicsScene (using class, subclassed from qPixmapItem and QObject, if it matters) with Animation framework.
Everything works fine, until i want to rotate pictures around vertical axis.
There is no way doing so via item.rotate(), so i'm using QTranform.
The problem is that doing so does not animate anything at all. What am i doing wrong?
P.S. I do not want use OpenGl for this.
Here is the way i'm doing it. This way works for animating simpler properties like pos, rotation(via rotation, setRotation)
My code :
// hybrid graphicsSceneItem, supporting animation
class ScenePixmap : public QObject, public QGraphicsPixmapItem
{
Q_OBJECT
Q_PROPERTY(QTransform transform READ transform WRITE setTransform)
public:
ScenePixmap(const QPixmap &pixmap, QObject* parent = NULL, QGraphicsItem* parentItem = NULL):
QObject(parent),
QGraphicsPixmapItem(pixmap, parentItem)
{}
};
Here is how I setup scene and animation:
//setup scene
//Unrelated stuff, loading pictures, etc.
scene = new QGraphicsScene(this);
foreach(const QPixmap& image, images)
{
ScenePixmap* item = new ScenePixmap(image);
item->moveBy(70*i, 0);
i++;
this->images.append(item);
scene->addItem(item);
}
}
ui->graphicsView->setBackgroundBrush(QBrush(Qt::black, Qt::SolidPattern));
ui->graphicsView->setScene(scene);
//setup animation
QTransform getTransform()
{
QTransform transform;
transform.rotate(-30, Qt::ZAxis);//also tried transform = transform.rotate(...)
return transform;
}
QAbstractAnimation* SetupRotationAnimation(ScenePixmap* pixmapItem)
{
QPropertyAnimation* animation = new QPropertyAnimation(pixmapItem, "transform");
animation->setDuration(1400);
animation->setStartValue( pixmapItem->transform());
animation->setEndValue(getTransform());//here i tried to multiply with default transform , this does not work either
return animation;
}
here is the way i start animation:
void MainWindow::keyPressEvent ( QKeyEvent * event )
{
if((event->modifiers() & Qt::ControlModifier))
{
QAnimationGroup* groupAnimation = new QParallelAnimationGroup();
foreach(ScenePixmap* image, images)
{
groupAnimation->addAnimation( SetupRotationAnimation(image));
}
groupAnimation->start(QAbstractAnimation::DeleteWhenStopped);
}
}
EDIT[Solved] thx to Darko Maksimovic:
Here is the code that worked out for me:
QGraphicsRotation* getGraphicRotation()
{
QGraphicsRotation* transform = new QGraphicsRotation(this);
transform->setAxis(Qt::YAxis);
return transform;
}
QAbstractAnimation* SetupRotationAnimation(ScenePixmap* pixmapItem)
{
QGraphicsRotation* rotation = getGraphicRotation();
QPropertyAnimation* animation = new QPropertyAnimation(rotation, "angle");
animation->setDuration(1400);
animation->setStartValue( 0);
animation->setEndValue(45);
pixmapItem->setTransformOriginPoint(pixmapItem->boundingRect().center());
QList<QGraphicsTransform*> transfromations = pixmapItem->transformations();
transfromations.append(rotation);
pixmapItem->setTransformations(transfromations);
return animation;
}

I see you use QTransform. If you want only one rotation, and simple rotation that is, it is better that you use setRotation [don't forget about setTransformOriginPoint].
If you, however, want to remember many rotations, around different transform points for example, then you should use QGraphicsTransform, i.e. its specialised derived class, QGraphicsRotation, which you apply by calling setTransformations on a graphics object (you should first fetch the existing transformations by calling o.transformations(), append to this, then call setTransformations; if you keep the pointer to the added transformation you can also change it later directly).
From the code you posted I can't see where your error is coming from, but by using specialised functions you can avoid some of frequent problems.
P.S. I also see you didn't use prepareGeometryChange in the code you posted, so please be advised that this is necessary when transforming objects.

The problem is very simple. QVariantAnimation and QPropertyAnimation don't support QTransform. Heck, they don't even support unsigned integers at the moment. That's all there's to it. Per Qt documentation:
If you need to interpolate other variant types, including custom types, you have to implement interpolation for these yourself. To do this, you can register an interpolator function for a given type. This function takes 3 parameters: the start value, the end value and the current progress.
It'd might not be all that trivial to generate such an interpolator. Remember that you have to "blend" between two matrices, while maintaining the orthonormality of the matrix, and the visual effect. You'd need to decompose the matrix into separate rotation-along-axis, scaling, skew, etc. sub-matrices, and then interpolate each of them separately. No wonder the Qt folks didn't do it. It's probably much easier to do the inverse problem: generate the needed transformation matrix as a function of some parameter t, by composing the necessary rotations, translations, etc., all given as (perhaps constant) functions of t.
Just because a QVariant can carry a type doesn't mean it's supported by the default interpolator. There probably should be a runtime warning issued to this effect.

Related

SFML: How to check if a point is contained in a group of transformed drawables

Context
I have a class representing a text box. the text box contains a header, some text and a rectangle to enclose the box. It only displays itself (for now):
struct Textbox : public sf::Drawable, public sf::Transformable{
sf::Text header;
sf::Text text;
sf::RectangleShape border;
Textbox(){
// set relative locations of the members
header.setPosition(0,0);
auto header_bounds = header.getGlobalBounds();
// the text should be just below the header
text.setPosition(0, header_bounds.top + header_bounds.height);
auto source_bounds = text.getGlobalBounds();
// this function just returns a rectangle enclosing two rectangles
sf::FloatRect rect = enclosing_rect(header_bounds, source_bounds);
// this function sets the position, width and length of border to be equal to rect's.
setRectParams(border, rect);
}
void draw(sf::RenderTarget& target, sf::RenderStates states){
states.transform = getTransform();
target.draw(header,states);
target.draw(text,states);
target.draw(border,states);
};
The Problem
What I want
I want to add a contains method. It should return true if coor is inside the border of the box. Here is my naive implementation:
bool Textbox::contains(sf::Vector2i coor) const {
return border.getGlobalBounds().contains(coor.x, coor.y);
}
Why this implementation doesn't work
This implementation breaks when I move the Textbox via the Transformable non-virtual functions. The Textbox moves and it also draws the shapes as transformed. But! It does not actually transform them! it only displays them as transformed. So the border doesn't even know it has been moved.
Possible solutions
I can add all the functions of the Transformable API to this class, thus shadowing them and calling transform by myself on each of the members. I don;t like this because it make me write sooo much more code than I wanted. It also raises the question of how to tackle the double transforms (the one for the Textbox and the others for it's members).
I can write a completely different class Group that holds a vector of drawables and transformables and it has all that shadowing API mechanism. All that is left is to inherit from it. This doesn't actually sound that bad.
I heard about Entity System Component - it's just sound pretty overkill.
I can apply the transform when contains is called. The function is const - it's a query. Also, it's bad design to update your data on seemingly random calls.
just as before just that the transform applies to a function-local rectangle. This smells too - why do I call the transform functions on the whole Textbox just so it would apply them on every method call (so far just it's draw and contains but down the line who knows)
Make the members mutable and somehow transform them inside the draw method. This smell hackish.
The question
How do I group transformations onto multiple entities via an ergonomic API?
The only method that you really need to 'change', but to be fair add on your own is getGlobalBounds().
When you are inheriting from sf::Transformable, sf::Drawable you should treat the base class (your Textbox struct) as a shape itself therfore you just need to call myTextbox.getGlobalBounds().contains(x,y), where myTextbox is a Textbox.
Using your own code:
struct Textbox : public sf::Drawable, public sf::Transformable{
sf::Text header;
sf::Text text;
sf::RectangleShape border;
sf::FloatRect getGlobalBounds() const {
auto header_bounds = header.getGlobalBounds();
auto source_bounds = text.getGlobalBounds();
sf::FloatRect rect = enclosing_rect(header_bounds, source_bounds);
//Don't really know what it does but let say that it returns Top and Left as 0, and calculates Height, Width.
return sf::FloatRect(getPosition(), sf::Vector2f(rect.width,rect.height));
}
};
But you still have to manage the rotation, resizing,etc. when calculating globalBounds.
EDIT:
One way to implement rotation and scaling.
sf::FloatRect getGlobalBounds() const {
auto header_bounds = header.getGlobalBounds();
auto source_bounds = text.getGlobalBounds();
sf::FloatRect rect = enclosing_rect(header_bounds, source_bounds);
//Don't really know what it does but let say that it returns Top and Left as 0, and calculates Height, Width.
sf::RectangleShape textbox(sf::Vector2f(rect.width, rect.height));
//at this point textbox = globalBounds of Textbox without transformations
textbox.setOrigin(getOrigin());//setOrigin (point of transformation) before transforming
textbox.setScale(getScale());
textbox.setRotation(getRotation());
textbox.setPosition(getPosition());
//after transformation get the bounds
return textbox.getGlobalBounds();
}
The solution might be much more simple than you expect. Instead of applying all the transforms to the transformable children/members, just de-transform the point you want to check (take it to local space).
Try this:
bool Textbox::contains(sf::Vector2i coor) const {
// Get point in the local space of the rectangle
sf::Transform inverseTr = this->getInverseTransform();
sf::Vector2f pointAsLocal = inverseTr.transformPoint(coor.x, coor.y);
// Check if the point, now in local space, is containted in the rectangle
return border.getLocalBounds().contains(pointAsLocal);
// ^
// Important! Use local bounds here, not global
}
Why does this work?
Math!
When you work with transformation matrices, you can think of them as portals between spaces. You have a local space where no transformation have been applied, and you have a final space, where all transformations are applied.
The problem with global bounds of a transformable member is that they belong neither to the local space nor the final space. They are just a rectangle bounding the shape in a possibly intermediate space where this bounds doesn't even take rotation into account.
What we are doing here is taking the coordinates, that exist in the final space, and take them to the local space of the rectangle, thanks to the inverse transformation matrix. So no matter how many translations, rotations or scales (or even skews, if you have customized the matrix) you apply to the rectangle. The inverse matrix takes the point to a new space where you can just check if it belongs, as if no transformation have ever been applied.

Transformations igrnores sf::Sprite's origin

Transforming a sprite in SFML, does not regard it's new origin.
In my case sf::Sprite is rotating around the axis that is in the left top corner ({0,0}) regardless its origin. Setting new origin with .setOrigin() earlier takes no effect.
I am sure that sprite is getting the right origin position earlier which is center of rectangle.
My code:
In each of my Card class constructors I set the origin of my sprite.
card_sprite.setOrigin(Card::get_default_single_card_size().x*Game::get_scale()/2,Card::get_default_single_card_size().y*Game::get_scale()/2);
And then in my Deck class which behaves like std::stack of Cards I use function:
void Deck::push(const Card& crd)
{
push_back(crd);
..//
std::default_random_engine generator;
std::uniform_real_distribution<float> distributor(0,360);
top().setRotation(distributor(generator));
}
Card::setRotaion() looks like this, (which stil rotates card around top left corner) :
void Card::setRotation(float angle)
{
card_sprite.setRotation(angle);
}
Thanks for help in advance
Edit: Actually most methods in sf::Transform accept extra arguments to specify a center for the transformation, as per https://stackoverflow.com/users/7703024/super 's comment on my question on the same theme : How to "set the origin" of a Transform in sfml
I'm not too sure from your code, but I might've come up against a similar problem.
I "solved" it (in a very not ideal way) by replacing every call to a sfml drawing function with a call to a custom function when using sf::Transforms.
eg: instead of doing something like:
window.draw(thing, my_transform);
I had to do :
draw_transformed(thing, my_transform, window)
Where the code of draw_transformed looks like this:
void draw_transformed (sf::Shape const& thing, sf::Transform const& t, sf::RenderWindow& window) // cf note (1)
{
sf::Vector2f pos = thing.getPosition();
sf::Transform go_to_zero;
go_to_zero.translate(-pos);
sf::Transform go_back;
go_back.translate(pos);
sf::Transform conjugated_transform = go_back * t * go_to_zero ;
window.draw(thing, conjugated_transform);
}
(1) we can't use sf::Drawable as the type of thing because in sfml not all drawable things have a getPosition method, so we have to overload the function or do something "complicated" to go beyond this example.

How to set a custom point labels format for QLineSeries/QXYSeries?

Context: I try to plot a QLineSeries on a QChart in order to print the progression of a value over time.
Therefore the X axis (abscissa) is a QDateTimeAxis and the Y axis (ordinate) is a QValueAxis.
Issue: I want to display the point labels. But I cannot find how to set the desired format for the datetime.
By default, the labels can only draw the integer values of the point which is what I want for the ordinate value.
But for the abscissa (datetime) value, it prints the number of milliseconds that have passed since the last epoch (1970-01-01T00:00:00.000).
I want to change the format for the datetime to match "hh:mm:ss" (which is the format I use to display the ticks on the QDateTimeAxis).
I know there exists a QXYSeries::setPointlabelsFormat() that allows to specify the format but it only accepts the #xPoint and #yPoint format tags (as you can see in the documentation).
Here you can find a picture of the issue:
As you can see, I can set the format for the QDateTimeAxis but not on the points labels.
The code sample that produces this output is based on the one provided here. I just have added more points and uncommented the line //ls->setPointLabelsVisible(true);.
Question: Is there a way to print #xPoint labels with a custom format (ideally to match QDateTime::toString("hh:mm:ss")) ? If yes, how ?
It's not supported by QXYSeries. You can derive a new class from QAbstractSeries for your so-called QXYDatetimeSeries following the same thing implemented by QXYSeries by reviewing the code. You may also want to change QXYSeriesPrivate::drawSeriesPointLabels source code to support your behavior. Don't forget to push your changes to the Qt git repository so that other people can also use it.
As noted by Soheim, the point label format provided by QtCharts is too limited for this.
If you don't want to get your hands dirty with QtCharts internals, you can solve your problem in another way: Draw the text labels yourself.
Derive a custom graphics item from QGraphicsObject:
class PointLabelsItem : public QGraphicsObject
{
Q_OBJECT
public:
PointLabelsItem(QtCharts::QChart *parent = nullptr);
virtual QRectF boundingRect() const;
virtual void paint(QPainter*, const QStyleOptionGraphicsItem*, QWidget*);
public slots:
void setRect(const QRectF &area);
void setAxis(const QAbstractAxis *axis, Qt::Alignment alignment);
// setter for either source Q*Series or points / point labels
protected:
QRectF area;
// const axes*
// const series* or points/labels
};
In the implementation:
PointLabelsItem::PointLabelsItem(QtCharts::QChart *parent)
: QGraphicsObject(parent)
{
setZValue(10); // put above series
connect(parent, &QtCharts::QChart::plotAreaChanged,
this, &PointLabelsItem::setRect);
}
QRectF PointLabelsItem::boundingRect() const
{
return area;
}
void PointLabelsItem::paint(QPainter *painter, const QStyleOptionGraphicsItem *, QWidget *)
{
// translate between data coordinates and scene coordinates
// by making use of `area` and the axes' ranges, OR via
// qobject_cast<QtCharts::QChart>(parent())->mapToPosition()
// then draw custom point label texts at the corresp. positions
}
void PointLabelsItem::setRect(const QRectF &newArea)
{
area = newArea;
update();
}
void setAxis(const QAbstractAxis *axis, Qt::Alignment alignment)
{
disconnect(/* old axis member variable */);
// set corresp. axis member variable here
connect(axis, &QtCharts::QValueAxis::rangeChanged, this, [this] { update(); });
}
From outside:
auto pointLabels = new PointLabelsItem(chart);
// set axes, series
That's the barebone.
You may also decide to have QGRaphicsTextItems as children of your PointLabelsItem (ie., members that have this set as their parent). This way you can enable user interaction with the labels.

add hover/enter/leave capabilities to element in my QGraphicsScene

basically I am trying to implement the Hover event on a QPointF, I need it to be clickable or at least being highlighted in some way.
What I've tried is to create a custom point starting by inheriting from QPointF, and then added hover,enter and leave methods as described in this tutorial, but it is not working. Also, the tut is based on generic QWidget and not specifically on points.
have you guys some hints/resources? google points me on a few tuts that are not useful, and the Qt page referred to HoverEvent is not tailored with examples.
thanks!
EDIT
More info needed. I am trying to draw a rectangle in a scene. The rectangle is of type QPolygonF and the scene is a QGraphicsScene type.
below the code block used to create a polygon from a list of points and its inclusion in the scene:
void
MyDialog::paintEvent(QPaintEvent *event)
{
Q_UNUSED(event);
QPolygonF poly;
QPointF first;
for(int i = 0;i < points->size();i++)
{
double length = points->at(i).split(",").at(0).toDouble();
double rad = qDegreesToRadians( points->at(i).split(",").at(1).toDouble());
QPointF pt(length * qCos(rad),length * qSin(rad));
if(i == 0) first = pt;
poly << pt;
}
poly << first;
scene->addPolygon(poly,QPen(),QBrush(Qt::green,Qt::SolidPattern));
}
QPointF is a data object; it does not provide any graphical representation whatsoever (and I would strongly advise you against inheriting from it to add one).
One way to possibly do it is to override the QGraphicsScene::mouseMoveEvent() and QGraphicsScene::mousePressEvent() functions, allowing you to listen-in on where the user moves and clicks, and reacting on that. There are probably better ways to do this, though - I've only needed to react on clicks (and not on hovers) so far, so my experiences are limited.

Why is there no QTransform::fromRotate?

Qt's QTransform offers some more optimized ways to construct a translated/scaled QTransform matrix using these static methods:
QTransform::fromScale
QTransform::fromTranslate
Now I need a rotatated transform and I thought it would be nice to also have a QTransform::fromRotate. But this one does not exist.
In my case I am modifying a existing transform accordingly to mouse interaction like paning, zooming and also rotating.
void MapDrawingItem::wheelEvent(QWheelEvent* event)
{
//Moving the hovered point to the top left point on screen
m_view_transform *= QTransform::fromTranslate(-event->posF().x(), -event->posF().y());
//Apply transformations accordingly
if((event->modifiers() & Qt::ControlModifier) == Qt::ControlModifier)
m_view_transform *= QTransform().rotate(event->delta() / 30.);
else
{
auto factor = qPow(1.001, event->delta());
m_view_transform *= QTransform::fromScale(factor, factor);
}
//Move the hovered point back to the mouse cursor
m_view_transform *= QTransform::fromTranslate(event->posF().x(), event->posF().y());
emit signalViewTransformChanged(m_view_transform);
update();
}
The code works correctly, but I would like to replace the QTransform().rotate(...) with a QTransform::fromRotate(...)
Why does this method not exist already? I just can't imagine the Qt developers forgot for this one. Is there anything that makes this impossible?
The most probable reason for that is that creating and using a static function to create a rotation transformation forces to always link the mathematical library (for the sin(3) and cos(3) mathematical functions) so instead you can use the specific constructor for that and use them yourself.
By the way, you can use one of the constructors to specify the constants to use for the elements of the matrix. I'ts quite common to conserve the matrices for reuse when using the transformations library.