flightclub.framework3d
Class Obj3d

java.lang.Object
  |
  +--flightclub.framework3d.Obj3d

All Implemented Interfaces:

CameraSubject

Direct Known Subclasses:

Obj3dDir

public class Obj3d

extends java.lang.Object

implements CameraSubject

This class implements a 3d object made up of polygons. We have a list of points (x, y, z) in model space and a second list of the those points after they have been mapped into screen space (x_, y_, z_). The x_ coord is handy for fogging and sorting; it represents the 'depth' of the projected point. (y_, z_) is the location of the point in screen space.

The mapping from model space to screen space is as follows: 1. A translation such that the camera focus becomes the origin. 2. A rotation such that the camera eye is on the +ve x_ axis. 3. Forshorten y_ and z_ by one over x_, the depth of the point. 4. Scale y_ and z_ according to the height of the display area in pixels. We use a flat array to store the coords as follows: (x0, y0, z0, x1, y1, z1, x2, ...). If npoints * 3 exceeds the array size then we double the array size and do a System.copyarray(). I got the flat array idea from the WireFrame demo applet.

Field Summary

static int	CONCAVE
static int	CONVEX
protected ModelViewer	modelViewer

Constructor Summary

Obj3d(ModelViewer modelViewer, int npolygons)
Creates an Obj3d and registers it with the 3d object manager.

Obj3d(ModelViewer modelViewer, int npolygons, boolean register)

Obj3d(Obj3d from, boolean register)
Creates a copy of obj.

Obj3d(java.io.StreamTokenizer st, ModelViewer modelViewer, boolean register)
Parses a text file and creates an Obj3d.

Method Summary

int	addPolygon(float[][] vs) Adds a single sided gray polygon.
int	addPolygon(float[][] vs, java.awt.Color c) Adds a polygon with one side visible.
int	addPolygon(float[][] vs, java.awt.Color c, boolean doubleSided) Adds a polygon.
int	addPolygon2(float[][] vs) Adds a gray polygon that is visible from both sides.
int	addPolygon2(float[][] vs, java.awt.Color c) Adds a polygon that is visible from both sides.
void	addPolygonBent(float[][] ps, java.awt.Color color, int concaveFlag) Adds a bent polygon (made out of two triangles).
void	destroyMe()
void	draw(java.awt.Graphics g) Draws a 2d representation of the object onto the screen.
float	getDepthMax()
float	getDepthMin()
float[]	getEye()
float[]	getFocus()
int	getPointIndex(int poly, int vertex) Gets the index of a point on a polygon.
static Obj3d	makeCube(ModelViewer modelViewer) Creates a unit cube whose centre of mass lies at the origin.
static Obj3d	parseFile(java.io.InputStream is, ModelViewer modelViewer, boolean register) This method is a hack.
void	scaleBy(float s)
void	setBB() Creates unit axes at the origin.
void	setColor(java.awt.Color c) Gives all polygons the specifed color.
void	setNormals() Loops thru' surfaces setting their dirtyNormal flags so that the surface normals will be computed again.
void	setPoint(int index, float x, float y, float z) Sets the co-ords of a point.
java.lang.String	toString() Dumps the data of this 3d shape.
void	transform() Uses the camera to map all points from model space, (x, y, z) to screen space (x_, y_, z_).
void	translateBy(float dx, float dy, float dz)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

modelViewer

protected ModelViewer modelViewer

CONCAVE

public static final int CONCAVE

See Also:

Constant Field Values

CONVEX

public static final int CONVEX

See Also:

Constant Field Values

Constructor Detail

Obj3d

public Obj3d(ModelViewer modelViewer,
             int npolygons,
             boolean register)

Obj3d

public Obj3d(ModelViewer modelViewer,
             int npolygons)

Creates an Obj3d and registers it with the 3d object manager.

Obj3d

public Obj3d(Obj3d from,
             boolean register)

Creates a copy of obj.

Obj3d

public Obj3d(java.io.StreamTokenizer st,
             ModelViewer modelViewer,
             boolean register)
      throws java.io.IOException,
             FileFormatException

Parses a text file and creates an Obj3d. The file format is as follows:

 
         # a simple kite made of two triangular polygons
         # by foo bar, 27 Aug 2002
         2
         f CC99FF 12.1 2.4 3.5, 12.4 2.2 3.5, 3.2 5 7.4, 
         t 336699 9.7 0 3.5, 2.2 2.2 3.5, 3.2 5 7.3,
 
         

The above represents an Obj3d with two polygons, the second of which is double sided; each is a differnt color; each is made up of 3 points.

Method Detail

parseFile

public static Obj3d parseFile(java.io.InputStream is,
                              ModelViewer modelViewer,
                              boolean register)
                       throws java.io.IOException,
                              FileFormatException

This method is a hack. I really want another constructor.

java.io.IOException

FileFormatException

destroyMe

public void destroyMe()

draw

public void draw(java.awt.Graphics g)

Draws a 2d representation of the object onto the screen.

See Also:

ModelCanvas.paintModel()

transform

public void transform()

Uses the camera to map all points from model space, (x, y, z) to screen space (x_, y_, z_). Remember that x_ represents the depth of the point and (y_, z_) maps to the screen (x, y) co-ords of the point.

See Also:

CameraMan

translateBy

public void translateBy(float dx,
                        float dy,
                        float dz)

scaleBy

public void scaleBy(float s)

setColor

public void setColor(java.awt.Color c)

Gives all polygons the specifed color.

addPolygon

public int addPolygon(float[][] vs,
                      java.awt.Color c,
                      boolean doubleSided)

Adds a polygon. Note that the vertices of the polygon are passed using a float[][] and NOT a float[]. Eg. {{x0, y0, z0}, {x1, y1, z1}, ...} We *flatten* the data once it is encapsulated inside this class; outside this class we want clarity; inside this class we want speed !

addPolygon

public int addPolygon(float[][] vs,
                      java.awt.Color c)

Adds a polygon with one side visible. Which of the two sides is visible is determined by the order of the points. If as you look at the polygon the points go clockwise then the normal points away from you.

addPolygon

public int addPolygon(float[][] vs)

Adds a single sided gray polygon.

addPolygon2

public int addPolygon2(float[][] vs,
                       java.awt.Color c)

Adds a polygon that is visible from both sides.

addPolygon2

public int addPolygon2(float[][] vs)

Adds a gray polygon that is visible from both sides.

addPolygonBent

public void addPolygonBent(float[][] ps,
                           java.awt.Color color,
                           int concaveFlag)

Adds a bent polygon (made out of two triangles). We are passed four points that do NOT lie in a plane.

 
        p3 .-------.  p2
           |     / |
           |   /   |
           | /     |
        p0 .-------.  p1
 
        

makeCube

public static Obj3d makeCube(ModelViewer modelViewer)

Creates a unit cube whose centre of mass lies at the origin.

setBB

public void setBB()

Creates unit axes at the origin. We color code as follows: x - red, y - green, z - blue.

getEye

public float[] getEye()

Specified by:

getEye in interface CameraSubject

getFocus

public float[] getFocus()

Specified by:

getFocus in interface CameraSubject

getPointIndex

public int getPointIndex(int poly,
                         int vertex)

Gets the index of a point on a polygon.

setPoint

public void setPoint(int index,
                     float x,
                     float y,
                     float z)

Sets the co-ords of a point.

getDepthMin

public final float getDepthMin()

getDepthMax

public final float getDepthMax()

toString

public java.lang.String toString()

Dumps the data of this 3d shape.

Overrides:

toString in class java.lang.Object

setNormals

public void setNormals()

Loops thru' surfaces setting their dirtyNormal flags so that the surface normals will be computed again.

See Also:

Obj3dDir.updateRotated