I want to connect several classes which are functioning separately but are related.
Lets say I am writing an app in which you can swipe to draw a chart. There are lots of classes in the app which are related and should be connected.
For example three of the classes are:
Swiper - responsible for interpreting the gesture of the user
Points - responsible for handling the points on the chart
ChartDrawer - responsible for drawing the chart on the screen
I want to know is there any design pattern such as a connector which can handle the relation and communication of these classes? Any way i can redesign in a better way or make mind more object oriented?
This is my ChartDraw class which extends a view:
public class ChartDraw extends View implements GestureReceiver {
int chartYPosition;
private int circleColor;
private int circleRadius;
int height;
private float lastPointOnChart;
private int lineColor;
private int lineWidth;
private Paint paint;
private float tempPoint;
int width;
public ChartDraw(Context context) {
super(context);
init();
}
public ChartDraw(Context context, AttributeSet attrs) {
super(context, attrs);
init();
}
public ChartDraw(Context context, AttributeSet attrs, int defStyleAttr) {
super(context, attrs, defStyleAttr);
init();
}
private void init() {
this.lineWidth = 15;
this.circleRadius = 20;
this.lineColor = Color.parseColor("#1976D2");
this.circleColor = Color.parseColor("#536DFE");
this.lastPointOnChart = 0.0f;
this.tempPoint = 0.0f;
this.paint = new Paint();
this.height = getHeight();
this.width = getWidth();
this.chartYPosition = this.height / 2;
}
protected void onDraw(Canvas canvas) {
super.onDraw(canvas);
this.chartYPosition = canvas.getHeight() / 2;
this.paint.setStrokeWidth((float) this.lineWidth);
this.paint.setColor(this.lineColor);
canvas.drawLine(0.0f, (float) this.chartYPosition, this.tempPoint, (float) this.chartYPosition, this.paint);
if (this.tempPoint > 20.0f) {
this.paint.setColor(this.circleColor);
canvas.drawCircle(20.0f, (float) this.chartYPosition, 20.0f, this.paint);
drawTriangle(canvas, this.paint, this.tempPoint, this.chartYPosition);
}
}
private void drawTriangle(Canvas canvas, Paint paint, float startX, int startY) {
Path path = new Path();
path.moveTo(startX, (float) (startY - 20));
path.lineTo(startX, (float) (startY + 20));
path.lineTo(30.0f + startX, (float) startY);
path.lineTo(startX, (float) (startY - 20));
path.close();
canvas.drawPath(path, paint);
}
public void onMoveHorizontal(float dx) {
this.tempPoint = this.lastPointOnChart + dx;
invalidate();
}
public void onMoveVertical(float dy) {
}
public void onMovementStop() {
this.lastPointOnChart = this.tempPoint;
}
}
And this is My SwipeManager which is handling user gesture:
public class SwipeManager implements View.OnTouchListener {
GestureReceiver receiver;
private int activePointer;
private float initX,
initY;
private long startTime,
stopTime;
private boolean resolving = false;
private boolean resolved = false;
private Direction direction;
#Override
public boolean onTouch(View view, MotionEvent motionEvent) {
if (receiver == null) throw new AssertionError("You must register a receiver");
switch (motionEvent.getActionMasked()) {
case ACTION_DOWN:
activePointer = motionEvent.getPointerId(0);
initX = motionEvent.getX(activePointer);
initY = motionEvent.getY(activePointer);
startTime = new Date().getTime();
break;
case ACTION_MOVE:
if (!resolving && !resolved) {
resolving = true;
float x = motionEvent.getX(activePointer);
float y = motionEvent.getY(activePointer);
direction = resolveDirection(x, y);
if (direction != Direction.STILL) {
resolved = true;
resolving = false;
} else {
resolving = false;
resolved = false;
}
break;
}
if (resolved) {
if (direction == Direction.HORIZONTAL)
receiver.onMoveHorizontal(motionEvent.getX(activePointer) - initX);
else receiver.onMoveVertical(motionEvent.getX(activePointer) - initY);
}
break;
case ACTION_UP:
resolved = false;
receiver.onMovementStop();
break;
}
return true;
}
private Direction resolveDirection(float x, float y) {
float dx = x - initX;
float dy = y - initY;
float absDx = Math.abs(dx);
float absDy = Math.abs(dy);
if (absDx > absDy + 10) {
return Direction.HORIZONTAL;
} else if (absDy > absDx + 10) {
return Direction.VERTICAL;
}
return Direction.STILL;
}
public void setReceiver(GestureReceiver receiver) {
this.receiver = receiver;
}
private enum Direction {HORIZONTAL, VERTICAL, STILL;}
}
And i didn't start the Points class because i was not sure about the architecture.
I want this Connector to register all the listeners for the classes and wait for a change and inform the corresponding class of the change, like new point added or swipe started and finished or any other event in the app.
Chain of Responsibility might be what you are looking for.
It is a pattern to tie a series of 'processing objects' in a 'chain' that can handle 'command objects'.
I could see you making command objects that encapsulate the touch events and then get passed through several processors and finally get 'processed' by the 'processing objects' which handle input detection/output generation for that particular 'command object'.
I don't know if this is -ideal-, but it is potentially valid.
Other related patterns to look into might be:
https://en.wikipedia.org/wiki/Command_pattern
https://en.wikipedia.org/wiki/Observer_pattern
https://en.wikipedia.org/wiki/Bridge_pattern
Really what you're looking for here is an MVC-style architecture. Your application should (broadly speaking) be separated into 3 different areas:
the Model, which is completely divorced from your presentation or communication concerns. It provides an API for interaction and can be tested entirely independently with a simple framework such as JUnit.
the View, which is responsible for displaying the Model. It may be that a model can be displayed in different ways - in which case you get a single model and a few different views.
the Controller, which is responsible for making changes to the Model in response to user (or other) input.
The important thing is that the three sets of components are loosely-coupled and that responsibilities are clearly separated. All three should communicated via well defined interfaces (perhaps using the Observer, Command and ChainOfResponsibility patterns). In particular, the Model classes should have no direct knowledge of any of the View or Controller classes.
So, you might have some Model/View classes like this...
public interface ChartListener {
void notifyUpdate();
}
public interface Chart {
void newPoint(Point p);
Collection<Point> thePoints();
void addListener(ChartListener listener);
}
public class ChartModel implements Chart {
private final Collection<Point> points;
private final Collection<ChartListener> listeners;
public Collection<Point> thePoints() {
return Collections.unmodifiableCollection(points);
}
public void newPoint(Point p) {
thePoints.add(p);
listeners.stream().forEach(ChartListener::notifyUpdate);
}
public void addListener(ChartListener cl) {
listeners.append(cl);
}
}
public PieChartViewer implements ChartListener {
// All you colour management or appearance-related concerns is in this class.
private final Chart chart;
public PieChartView(Chart chart) {
this.chart = chart;
// set up all the visuals...
}
public void notifyUpdate() {
for (final Point p:chart.thePoints()) {
// draw a point somehow, lines, dots, etc,
}
}
}
Then you might have multiple different implementations of your View classes, utilising the ChartListener interface.
Your Swipe class seems like a Controller class, which would take a ChartModel implementation and then modify it in response to some input from the user.
Related
I'm using Dave Morrissey's Subsampling Scale Image View. I'm using his Pinview example (as shown here: https://github.com/davemorrissey/subsampling-scale-image-view/blob/master/sample/src/com/davemorrissey/labs/subscaleview/sample/extension/views/PinView.java)
What I've done differently is that I've created an ArrayList of Bitmaps that are Pins. But I want to make each pin clickable to set off an on click function. I know Bitmaps can not be clicked on. I have multiple pins on a map image and would like for each pin to be associated with an object.
What would be the best approach to accomplish this?
Note: I did override the setOnClickListener method inside of the Pinview class, but what happens that all pins that were dropped become associated to the same object. And that clearing 1 pin would then clear all pins.
The model that stores the bitmap, pointF and point name:
public class CategoryPoint {
private String category;
private Bitmap image;
private PointF pointF;
public CategoryPoint(String category, Bitmap image, PointF pointF) {
this.category = category;
this.image = image;
this.pointF = pointF;
}
// getters/setters
}
View looks like this:
public class PinsView extends SubsamplingScaleImageView {
private OnPinClickListener onPinClickListener;
private final Paint paint = new Paint();
private List<CategoryPoint> categoryPoints;
public PinsView(Context context) {
this(context, null);
}
public PinsView(Context context, AttributeSet attr) {
super(context, attr);
categoryPoints = new ArrayList<>();
initTouchListener();
}
public void addCategories(List<CategoryPoint> categoryPoints) {
this.categoryPoints = categoryPoints;
invalidate();
}
public void removeCategories(List<CategoryPoint> categoryPoints) {
this.categoryPoints.removeAll(categoryPoints);
invalidate();
}
public void removeAllCategories() {
this.categoryPoints.clear();
invalidate();
}
public void setOnPinClickListener(OnPinClickListener listener) {
onPinClickListener = listener;
}
#Override
protected void onDraw(Canvas canvas) {
super.onDraw(canvas);
if (!isReady()) {
return;
}
paint.setAntiAlias(true);
for (CategoryPoint categoryPoint: categoryPoints) {
Bitmap pinIcon = categoryPoint.getImage();
if (categoryPoint.getPointF() != null && categoryPoint.getImage() != null) {
PointF point = sourceToViewCoord(categoryPoint.getPointF());
float vX = point.x - (pinIcon.getWidth()/2);
float vY = point.y - pinIcon.getHeight();
canvas.drawBitmap(pinIcon, vX, vY, paint);
}
}
}
private void initTouchListener() {
GestureDetector gestureDetector = new GestureDetector(getContext(), new GestureDetector.SimpleOnGestureListener() {
#Override
public boolean onSingleTapConfirmed(MotionEvent e) {
if (isReady() && categoryPoints != null) {
PointF tappedCoordinate = new PointF(e.getX(), e.getY());
Bitmap clickArea = categoryPoints.get(0).getImage();
int clickAreaWidth = clickArea.getWidth();
int clickAreaHeight = clickArea.getHeight();
for (CategoryPoint categoryPoint : categoryPoints) {
PointF categoryCoordinate = sourceToViewCoord(categoryPoint.getPointF());
int categoryX = (int) (categoryCoordinate.x);
int categoryY = (int) (categoryCoordinate.y - clickAreaHeight / 2);
if (tappedCoordinate.x >= categoryX - clickAreaWidth / 2
&& tappedCoordinate.x <= categoryX + clickAreaWidth / 2
&& tappedCoordinate.y >= categoryY - clickAreaHeight / 2
&& tappedCoordinate.y <= categoryY + clickAreaHeight / 2) {
onPinClickListener.onPinClick(categoryPoint);
break;
}
}
}
return true;
}
});
setOnTouchListener((v, event) -> gestureDetector.onTouchEvent(event));
}
}
Fragment:
pinView.setOnImageEventListener(this);
pinView.setOnPinClickListener(this);
// implementation
I'm a newbie in Android development, and I would just like to know a little bit about the Scroller widget (android.widget.Scroller). How does it animate the view? Can the Animation object, if it exists, be accessed? If so, how? I've read the source code, but could find no clues, or maybe I'm too new?
I just wanted to do some operations after a Scroller finishes scrolling, something like
m_scroller.getAnimation().setAnimationListener(...);
The Scroller widget doesn't actually do much of the work at all for you. It doesn't fire any callbacks, it doesn't animate anything, it just responds to various method calls.
So what good is it? Well, it does all of the calculation for e.g. a fling for you, which is handy. So what you'd generally do is create a Runnable that repeatedly asks the Scroller, "What should my scroll position be now? Are we done flinging yet?" Then you repost that runnable on a Handler (usually on the View) until the fling is done.
Here's an example from a Fragment I'm working on right now:
private class Flinger implements Runnable {
private final Scroller scroller;
private int lastX = 0;
Flinger() {
scroller = new Scroller(getActivity());
}
void start(int initialVelocity) {
int initialX = scrollingView.getScrollX();
int maxX = Integer.MAX_VALUE; // or some appropriate max value in your code
scroller.fling(initialX, 0, initialVelocity, 0, 0, maxX, 0, 10);
Log.i(TAG, "starting fling at " + initialX + ", velocity is " + initialVelocity + "");
lastX = initialX;
getView().post(this);
}
public void run() {
if (scroller.isFinished()) {
Log.i(TAG, "scroller is finished, done with fling");
return;
}
boolean more = scroller.computeScrollOffset();
int x = scroller.getCurrX();
int diff = lastX - x;
if (diff != 0) {
scrollingView.scrollBy(diff, 0);
lastX = x;
}
if (more) {
getView().post(this);
}
}
boolean isFlinging() {
return !scroller.isFinished();
}
void forceFinished() {
if (!scroller.isFinished()) {
scroller.forceFinished(true);
}
}
}
The details of using Scroller.startScroll should be similar.
like Bill Phillips said, Scroller is just an Android SDK class helping with calculating scrolling positions. I have a full working example here:
public class SimpleScrollableView extends TextView {
private Scroller mScrollEventChecker;
private int mLastFlingY;
private float mLastY;
private float mDeltaY;
public SimpleScrollableView(Context context) {
this(context, null, 0);
}
public SimpleScrollableView(Context context, AttributeSet attrs) {
this(context, attrs, 0);
}
public SimpleScrollableView(Context context, AttributeSet attrs, int defStyleAttr) {
super(context, attrs, defStyleAttr);
}
#Override
public boolean onTouchEvent(MotionEvent event) {
if (mScrollEventChecker != null && !mScrollEventChecker.isFinished()) {
return super.onTouchEvent(event);
}
final int action = event.getAction();
switch (action) {
case MotionEvent.ACTION_DOWN:
mLastY = event.getY();
return true;
case MotionEvent.ACTION_MOVE:
int movingDelta = (int) (event.getY() - mLastY);
mDeltaY += movingDelta;
offsetTopAndBottom(movingDelta);
invalidate();
return true;
case MotionEvent.ACTION_UP:
mScrollEventChecker = new Scroller(getContext());
mScrollEventChecker.startScroll(0, 0, 0, (int) -mDeltaY, 1000);
post(new Runnable() {
#Override
public void run() {
if (mScrollEventChecker.computeScrollOffset()) {
int curY = mScrollEventChecker.getCurrY();
int delta = curY - mLastFlingY;
offsetTopAndBottom(delta); // this is the method make this view move
invalidate();
mLastFlingY = curY;
post(this);
} else {
mLastFlingY = 0;
mDeltaY = 0;
}
}
});
return super.onTouchEvent(event);
}
return super.onTouchEvent(event);
}
}
The demo custom view above will scroll back to original position after the user release the view. When user release the view, then startScroll() method is invoked and we can know what the distance value should be for every single message post.
Full working example: Github repository
Great answer above. Scroller#startScroll(...) indeed works the same way.
For example, the source for a custom scrolling TextView at:
http://bear-polka.blogspot.com/2009/01/scrolltextview-scrolling-textview-for.html
Sets a Scroller on a TextView using TextView#setScroller(Scroller).
The source for the SDK's TextView at:
http://grepcode.com/file/repository.grepcode.com/java/ext/com.google.android/android/2.2_r1.1/android/widget/TextView.java#TextView.0mScroller
Shows that TextView#setScroller(Scroller) sets a class field which is used in situations like bringPointIntoView(int) where Scroller#scrollTo(int, int, int, int) is called.
bringPointIntoView() adjusts mScrollX and mScrollY (with some SDK fragmentation code), then calls invalidate(). The point of all this is that mScrollX and mScrollY are used in methods like onPreDraw(...) to affect the position of the drawn contents of the view.
We can extend the Scroller class then intercept corresponding animation start methods to mark that was started, after computeScrollOffset() return false which means animation finished's value, we inform by a Listener to caller :
public class ScrollerImpl extends Scroller {
...Constructor...
private boolean mIsStarted;
private OnFinishListener mOnFinishListener;
#Override
public boolean computeScrollOffset() {
boolean result = super.computeScrollOffset();
if (!result && mIsStarted) {
try { // Don't let any exception impact the scroll animation.
mOnFinishListener.onFinish();
} catch (Exception e) {}
mIsStarted = false;
}
return result;
}
#Override
public void startScroll(int startX, int startY, int dx, int dy) {
super.startScroll(startX, startY, dx, dy);
mIsStarted = true;
}
#Override
public void startScroll(int startX, int startY, int dx, int dy, int duration) {
super.startScroll(startX, startY, dx, dy, duration);
mIsStarted = true;
}
#Override
public void fling(int startX, int startY, int velocityX, int velocityY, int minX, int maxX, int minY, int maxY) {
super.fling(startX, startY, velocityX, velocityY, minX, maxX, minY, maxY);
mIsStarted = true;
}
public void setOnFinishListener(OnFinishListener onFinishListener) {
mOnFinishListener = onFinishListener;
}
public static interface OnFinishListener {
void onFinish();
}
}
i'm trying to fix this problem for more than 2 days now and have become quite desperate.
I want to write a 'Checkers-like' board game for android. The game engine itself is kinda complete but i have problems with updating the views.
I wrote a little example class to demonstrate my problem:
public class GameEngineView extends View {
private static final String TAG = GameEngineView.class.getSimpleName();
private int px;
private int py;
private int cx;
private int cy;
private boolean players_move;
private int clickx;
private int clicky;
Random rgen;
private RefreshHandler mRedrawHandler = new RefreshHandler();
class RefreshHandler extends Handler {
#Override
public void handleMessage(Message msg) {
GameEngineView.this.update();
GameEngineView.this.invalidate();
Log.d(TAG, "invalidate()");
}
public void sleep(long delayMillis) {
this.removeMessages(0);
sendMessageDelayed(obtainMessage(0), delayMillis);
}
};
public GameEngineView(Context context) {
super(context);
setFocusable(true);
players_move = true;
rgen = new Random();
}
public void update() {
updateGame();
Log.d(TAG, "update -> sleep handler");
mRedrawHandler.sleep(100);
}
public void updateGame() {
if(players_move) {
px = clickx;
py = clicky;
} else {
calcAIMove();
switchMove();
}
}
public void switchMove() {
players_move = !players_move;
}
public void calcAIMove() {
for(int i = 0; i < 20000; i++) {
cx = rgen.nextInt(getWidth());
cy = rgen.nextInt(getHeight());
}
}
#Override
public boolean onTouchEvent(MotionEvent event) {
Log.d(TAG, "event");
int eventaction = event.getAction();
if(eventaction == MotionEvent.ACTION_DOWN) {
Log.d(TAG, "action_down");
clickx = (int) event.getX();
clicky = (int) event.getY();
switchMove();
update();
}
return super.onTouchEvent(event);
}
#Override
public void onDraw(Canvas canvas) {
super.onDraw(canvas);
Paint green = new Paint();
green.setColor(Color.GREEN);
Paint red = new Paint();
red.setColor(Color.RED);
canvas.drawColor(Color.BLACK);
canvas.drawCircle(px, py, 25, green);
canvas.drawCircle(cx, cy, 25, red);
}
}
The function calcAIMove() just burns time to simulate a real evaluation of the position in a board game.
Now my Problem is: If the player clicks(makes a move) the green ball is first drawn when the ai move calculation has been complete. So both moves are drawn at the same time.
I wonder HOW to accomplish this:
-Player clicks
-green Ball is drawn
-AI calculates
-red ball is drawn
-and so on..
When searching the web i found a lot of game loop examples but they all need a Thread with constant polling.. it should be possible without this since the whole program runs sequentially .. right?
Hoping for advice.
thanks,
Dave
Here is how your game could work:
user makes a move
game view is updated
game switches to CPU's turn
sleep to simulate CPU player thinking (if move computation is trivial)
compute CPU's move
game view is updated
game switches to player's turn
There is no need for constant polling in a turn-based game like this. The only place you should have sleep is in step 4, so you can remove it from the other areas (no need for the old update() method which is just a delayed call to updateGame()).
One way to implement this is to simply delay the call to calcAIMove() and switchMove() by putting it into a Runnable and using Handler.postDelayed() or similar.
I don't see how you are disabling touch events when it is the CPU's turn, which can lead to a host of other problems if switchMove() and update() are still being called...
In your game loop you can use a instance of TimerTask to ensure certain delay between greenBall and some other drawing task. Game tutorials like Snake and LunarLander are great references on when and if you need to invalidate your View. Hope this helps a little!
ok so the answer provided by antonyt helped me solve this.. I provide the corrected code for other interested ones.
public class GameEngineView extends View {
private static final String TAG = GameEngineView.class.getSimpleName();
private int px;
private int py;
private int cx;
private int cy;
private boolean players_move;
private int clickx;
private int clicky;
Random rgen;
/**
* Create a simple handler that we can use to cause animation to happen. We
* set ourselves as a target and we can use the sleep()
* function to cause an update/invalidate to occur at a later date.
*/
private RefreshHandler mRedrawHandler = new RefreshHandler();
class RefreshHandler extends Handler {
#Override
public void handleMessage(Message msg) {
GameEngineView.this.update();
}
public void sleep(long delayMillis) {
this.removeMessages(0);
sendMessageDelayed(obtainMessage(0), delayMillis);
}
};
public GameEngineView(Context context) {
super(context);
setFocusable(true);
players_move = true;
rgen = new Random();
}
public void update() {
if(players_move) {
Log.d(TAG, "new player x");
px = clickx;
py = clicky;
GameEngineView.this.invalidate();
switchMove();
mRedrawHandler.sleep(100);
} else {
Log.d(TAG, "new ai x");
calcAIMove();
GameEngineView.this.invalidate();
switchMove();
}
Log.d(TAG, "update -> sleep handler");
}
public void switchMove() {
players_move = !players_move;
}
public void calcAIMove() {
for(int i = 0; i < 100000; i++) {
cx = rgen.nextInt(getWidth());
cy = rgen.nextInt(getHeight());
}
}
#Override
public boolean onTouchEvent(MotionEvent event) {
Log.d(TAG, "event");
int eventaction = event.getAction();
if(players_move && eventaction == MotionEvent.ACTION_DOWN) {
Log.d(TAG, "action_down");
clickx = (int) event.getX();
clicky = (int) event.getY();
update();
}
return super.onTouchEvent(event);
}
#Override
public void onDraw(Canvas canvas) {
super.onDraw(canvas);
canvas.drawColor(Color.BLACK);
Paint green = new Paint();
green.setColor(Color.GREEN);
Paint red = new Paint();
red.setColor(Color.RED);
canvas.drawCircle(px, py, 25, green);
canvas.drawCircle(cx, cy, 25, red);
}
}
I'm not a programmer, but I'm trying to learn android development, and having a blast.
Lately I've hit a fork in the road, and I don't know that any of the directions I can identify will meet all my needs. This is where you (the experts) can help ;)
Endpoint: I want to have the user touch a ball, and drag it to any location on the screen. I would also like to animate this ball when it gets drawn, and when it gets dropped.
I can accomplish the dragging part using a custom class (that doesn't extend anything... just a class like is found in this tutorial: basic drag & drop, however, I don't know how to apply the tween animation to it since it's not a view.
I have also developed an animated version of an ImageView with my image in it, however, I can't manage to apply the same drag & drop functionality without using AbsoluteLayout, which I know is a no-no.
So... how do I move an ImageView around a ??? layout using MotionEvents, or how do I animate (using tweens defined in XML) a non-view based custom class?
Please ask questions if this is not clear. I don't know all the terminology as well as most of you might.
There is also a copy of this question on the anddev.org forums, so I'll keep this post updated with any responses I get over there.
Why can't you extend View? Then, you have complete control over how it draws because you can override the OnDraw() method. Just make the ColorBall class extend View. Change its position when you move and then invalidate just that one view and have it draw itself instead of having the DrawView class draw it.
Edit - Here is an example class
public class Card extends View
{
private Bitmap mImage;
private final Paint mPaint = new Paint();
private final Point mSize = new Point();
private final Point mStartPosition = new Point();
public Card(Context context)
{
super(context);
}
public final Bitmap getImage() { return mCardImage; }
public final void setImage(Bitmap image)
{
mImage = image;
setSize(mCardImage.getWidth(), mCardImage.getHeight());
}
#Override
protected void onDraw(Canvas canvas)
{
Point position = getPosition();
canvas.drawBitmap(mCardImage, position.x, position.y, mPaint);
}
public final void setPosition(final Point position)
{
mRegion.set(position.x, position.y, position.x + mSize.x, position.y + mSize.y);
}
public final Point getPosition()
{
Rect bounds = mRegion.getBounds();
return new Point(bounds.left, bounds.top);
}
public final void setSize(int width, int height)
{
mSize.x = width;
mSize.y = height;
Rect bounds = mRegion.getBounds();
mRegion.set(bounds.left, bounds.top, bounds.left + width, bounds.top + height);
}
public final Point getSize() { return mSize; }
#Override
public boolean onTouchEvent(MotionEvent event)
{
// Is the event inside of this view?
if(!mRegion.contains((int)event.getX(), (int)event.getY()))
{
return super.onTouchEvent(event);
}
if(event.getAction() == MotionEvent.ACTION_DOWN)
{
mStartPosition.x = (int)event.getX();
mStartPosition.y = (int)event.getY();
bringToFront();
onSelected();
return true;
}
else if(event.getAction() == MotionEvent.ACTION_MOVE)
{
int x = 0, y = 0;
if(mLock == DirectionLock.FREE || mLock == DirectionLock.HORIZONTAL_ONLY)
{
x = (int)event.getX() - mStartPosition.x;
}
if(mLock == DirectionLock.FREE || mLock == DirectionLock.VERTICAL_ONLY)
{
y = (int)event.getY() - mStartPosition.y;
}
mRegion.translate(x, y);
mStartPosition.x = (int)event.getX();
mStartPosition.y = (int)event.getY();
invalidate();
return true;
}
else
{
return super.onTouchEvent(event);
}
}
}
I'm not a programmer, but I'm trying to learn android development, and having a blast.
Lately I've hit a fork in the road, and I don't know that any of the directions I can identify will meet all my needs. This is where you (the experts) can help ;)
Endpoint: I want to have the user touch a ball, and drag it to any location on the screen. I would also like to animate this ball when it gets drawn, and when it gets dropped.
I can accomplish the dragging part using a custom class (that doesn't extend anything... just a class like is found in this tutorial: basic drag & drop, however, I don't know how to apply the tween animation to it since it's not a view.
I have also developed an animated version of an ImageView with my image in it, however, I can't manage to apply the same drag & drop functionality without using AbsoluteLayout, which I know is a no-no.
So... how do I move an ImageView around a ??? layout using MotionEvents, or how do I animate (using tweens defined in XML) a non-view based custom class?
Please ask questions if this is not clear. I don't know all the terminology as well as most of you might.
There is also a copy of this question on the anddev.org forums, so I'll keep this post updated with any responses I get over there.
Why can't you extend View? Then, you have complete control over how it draws because you can override the OnDraw() method. Just make the ColorBall class extend View. Change its position when you move and then invalidate just that one view and have it draw itself instead of having the DrawView class draw it.
Edit - Here is an example class
public class Card extends View
{
private Bitmap mImage;
private final Paint mPaint = new Paint();
private final Point mSize = new Point();
private final Point mStartPosition = new Point();
public Card(Context context)
{
super(context);
}
public final Bitmap getImage() { return mCardImage; }
public final void setImage(Bitmap image)
{
mImage = image;
setSize(mCardImage.getWidth(), mCardImage.getHeight());
}
#Override
protected void onDraw(Canvas canvas)
{
Point position = getPosition();
canvas.drawBitmap(mCardImage, position.x, position.y, mPaint);
}
public final void setPosition(final Point position)
{
mRegion.set(position.x, position.y, position.x + mSize.x, position.y + mSize.y);
}
public final Point getPosition()
{
Rect bounds = mRegion.getBounds();
return new Point(bounds.left, bounds.top);
}
public final void setSize(int width, int height)
{
mSize.x = width;
mSize.y = height;
Rect bounds = mRegion.getBounds();
mRegion.set(bounds.left, bounds.top, bounds.left + width, bounds.top + height);
}
public final Point getSize() { return mSize; }
#Override
public boolean onTouchEvent(MotionEvent event)
{
// Is the event inside of this view?
if(!mRegion.contains((int)event.getX(), (int)event.getY()))
{
return super.onTouchEvent(event);
}
if(event.getAction() == MotionEvent.ACTION_DOWN)
{
mStartPosition.x = (int)event.getX();
mStartPosition.y = (int)event.getY();
bringToFront();
onSelected();
return true;
}
else if(event.getAction() == MotionEvent.ACTION_MOVE)
{
int x = 0, y = 0;
if(mLock == DirectionLock.FREE || mLock == DirectionLock.HORIZONTAL_ONLY)
{
x = (int)event.getX() - mStartPosition.x;
}
if(mLock == DirectionLock.FREE || mLock == DirectionLock.VERTICAL_ONLY)
{
y = (int)event.getY() - mStartPosition.y;
}
mRegion.translate(x, y);
mStartPosition.x = (int)event.getX();
mStartPosition.y = (int)event.getY();
invalidate();
return true;
}
else
{
return super.onTouchEvent(event);
}
}
}