/*
Copyright 2006 by Sean Luke and George Mason University
Licensed under the Academic Free License version 3.0
See the file "LICENSE" for more information
*/
package sim.portrayal;
import java.awt.*;
import java.awt.geom.*;
import sim.util.*;
import sim.display.*;
/**
Superclass of all Field Portrayals in 2D. Implements default versions of the
necessary methods, which you should feel free to override (especially portray,
hit, and draw!).
<p>The default version of getDefaultPortrayal() returns an empty SimplePortrayal: you'll
might want to override that as well. These defaults enable you to use FieldPortrayal2D
as an "empty field portrayal" to be attached to a Display2D and draw arbitrary things
when its draw(...) method is called. In that case, all you need to do is override
the draw(...) method and you're set.
<p>The default versions of hitObjects and draw call a protected but empty method
called hitOrDraw. It's very common that a field's hitObjects and draw methods will
be identical save for a single line which calls hitObjects or draw on an underlying
SimplePortrayal. So most fields unify both methods into this single method
and unify the line as something like this:
<pre><tt>
if (graphics == null)
{
if (portrayal.hitObject(portrayedObject, newinfo))
putInHere.add(getWrapper(portrayedObject, portrayedObjectLocation));
}
else
portrayal.draw(portrayedObject, graphics, newinfo);
</tt></pre>
<p>...where <i>getWrapper(...)</i> returns a LocationWrapper appropriate to the Field.
*/
public abstract class FieldPortrayal2D extends FieldPortrayal implements Portrayal2D
{
/** Returns the position on-screen of an object at a given location in the field,
using another object's location and DrawInfo2D to perform the computation.
Returns null if we can't compute it -- this happens by default if the
two locations are neither Int2D nor Double2D. This method is largely used by TrailPortrayal2D. */
public Point2D.Double getRelativeObjectPosition(Object location, Object otherObjectLocation, DrawInfo2D otherObjectInfo)
{
double dx = 0;
double dy = 0;
if (location instanceof Int2D)
{
Int2D loc = (Int2D) location;
Int2D oloc = (Int2D) otherObjectLocation;
dx = loc.x - oloc.x;
dy = loc.y - oloc.y;
}
else
{
Double2D loc = (Double2D) location;
Double2D oloc = (Double2D) otherObjectLocation;
dx = loc.x - oloc.x;
dy = loc.y - oloc.y;
}
double xScale = otherObjectInfo.draw.width;
double yScale = otherObjectInfo.draw.height;
return new Point2D.Double(dx * xScale + otherObjectInfo.draw.x,
dy * yScale + otherObjectInfo.draw.y);
}
/** Draws the field with its origin at [info.draw.x,info.draw.y], relative to the
scaled coordinate system defined by [info.draw.width,info.draw.height]. The
only parts that need be drawn are those which fall within the [info.clip] rectangle.
Since your draw and hitObjects methods are likely to be nearly identical, you can choose
instead to override the hitOrDraw method to handle both of them. By default this
method simply calls hitOrDraw. FieldPortrayals will receive null for the object;
they should just draw their own fields. */
public void draw(Object object, Graphics2D graphics, DrawInfo2D info)
{
hitOrDraw(graphics, info, null);
}
/** Adds to the provided Bag LocationWrappers for any objects which
overlap the provided hit range. The hit range will usually
define a single point, but COULD be a range. The object
should perceive itself as located at the (range.draw.x,range.draw.y) origin and drawn
relative to the (range.draw.width,range.draw.height) scale. */
public void hitObjects(DrawInfo2D range, Bag putInHere)
{
hitOrDraw(null, range, putInHere);
}
/** Instead of overriding the draw and hitObjects methods, you can optionally override
this method to provide <i>both</i> the draw(...) and hitObjects(...)
functionality in a single method, as it's common that these two methods have nearly
identical code. You should test which operation to do
based on whether or not graphics is null (if it is, you're hitting, else you're drawing). */
protected void hitOrDraw(Graphics2D graphics, DrawInfo2D info, Bag putInHere)
{
}
SimplePortrayal2D simple = new SimplePortrayal2D();
public Portrayal getDefaultPortrayal()
{
return simple;
}
/** Moves (or tries to move) the object to an internal location equivalent to the given position on-screen of the
provided object, assuming that the object exists within the underlying field and that this
location is acceptable. <b>Optionally overridable</b>. The default implementation calls setObjectLocation to do its work. */
public void setObjectPosition(Object object, Point2D.Double position, DrawInfo2D fieldPortrayalInfo)
{
synchronized(fieldPortrayalInfo.gui.state.schedule)
{
if (getObjectLocation(object, fieldPortrayalInfo.gui) == null) return;
Object location = getPositionLocation(position, fieldPortrayalInfo);
if (location != null)
setObjectLocation(object, location, fieldPortrayalInfo.gui);
}
}
/** Returns the width and height, in pixels, of 1.0 x 1.0 units in the underlying field.
<b>Optionally overridable</b>. The default version thows an error if called. */
public Double2D getScale(DrawInfo2D fieldPortrayalInfo)
{
throw new RuntimeException("getScale not implemented in " + this.getClass());
}
/** Returns the Location, in the parlance of the underlying Field, of the given position.
If there is no such Location, then null is returned. <b>Optionally overridable</b>.
By default null is returned.
*/
public Object getPositionLocation(Point2D.Double position, DrawInfo2D fieldPortrayalInfo)
{
return null;
}
/** Returns the position on-screen of the provided location in the underlying field. Negative positions are acceptable.
If null is returned, then the portrayal is unable to perform the requested action on the given location.
<b>Optionally overridable</b>. The default implementation returns null. */
public Point2D.Double getLocationPosition(Object location, DrawInfo2D fieldPortrayalInfo)
{
return null;
}
/** Returns an object representing the location in the field of the origin of the clip of the DrawInfo2D.
This method calls getPositionLocation, which may or may not be implemented by the FieldPortrayal2D.
@deprecated use getPositionLocation
*/
public Object getClipLocation(DrawInfo2D info)
{
return getPositionLocation(new Point2D.Double(info.clip.x, info.clip.y), info);
}
/** Returns the position-onscreen of the provided object, assuming that the object exists within the underlying field.
Negative locations are acceptable. If null is returned, then the portrayal is unable to perform the requested
action on the given object. */
public Point2D.Double getObjectPosition(Object object, DrawInfo2D fieldPortrayalInfo)
{
synchronized(fieldPortrayalInfo.gui.state.schedule)
{
Object location = getObjectLocation(object, fieldPortrayalInfo.gui);
if (location == null) return null;
return getLocationPosition(location, fieldPortrayalInfo);
}
}
/** Default buffering: let the program decide on its own (typically in a platform-dependent fashion) */
public static final int DEFAULT = 0;
/** Use a buffer */
public static final int USE_BUFFER = 1;
/** Don't use a buffer */
public static final int DONT_USE_BUFFER = 2;
int buffering = DEFAULT;
Object bufferingLock = new Object();
/** Returns whether or not the FieldPortrayal2D will use a buffering "trick" to draw quickly. This optional
property is in FieldPortrayal2D but is only taken advantage of by one or two subclasses. Some FieldPortrayal2Ds
primarily draw lots of rectangles in a grid. Certain ones can draw this in one of two ways: either by drawing
each rectangle separately, or by filling a buffer (an image) with individual points, then stretching the buffer
over the area, causing the points to enlarge into rectangles. The second is often faster, especially if given
lots of memory (and always faster on Macs regardless). The FieldPortrayal2D's behavior in this regard will
depend on the value of the buffering property. The property can take on one of three values:
DEFAULT (let the machine decide on its own in a platform-dependent fashion -- the default),
USE_BUFFER, or DONT_USE_BUFFER. */
public int getBuffering() { synchronized(bufferingLock) { return buffering; } }
/** Sets whether or not the FieldPortrayal2D will use a buffering "trick" to draw quickly. This optional
property is in FieldPortrayal2D but is only taken advantage of by one or two subclasses. Some FieldPortrayal2Ds
primarily draw lots of rectangles in a grid. Certain ones can draw this in one of two ways: either by drawing
each rectangle separately, or by filling a buffer (an image) with individual points, then stretching the buffer
over the area, causing the points to enlarge into rectangles. The second is often faster, especially if given
lots of memory (and always faster on Macs regardless). The FieldPortrayal2D's behavior in this regard will
depend on the value of the buffering property. The property can take on one of three values:
DEFAULT (let the machine decide on its own in a platform-dependent fashion -- the default),
USE_BUFFER, or DONT_USE_BUFFER. */
public void setBuffering(int val) { synchronized(bufferingLock) { buffering = val; } }
}