tiled Map Case Study: Rendering with JPanel © Allan C. Milne v

Agenda.  Accessing a JPanel.  Map & tile representation.  Drawing the map.  Using the renderer.

Where are we now? we use a strategy pattern to include rendering behaviour; the MapRenderer interface; composition of tiledMap with a concrete MapRenderer object; the render() method mechanism. We have a concrete rendering class; ConsoleRenderer.

Where do we want to be? Add more concrete rendering implementations; –different ways of displaying on the console; –using different display technologies. In particular, rendering in a windows user interface; –for example, using a JPanel to draw some representation of the map.

A problem with JPanel. There is only one console and any code can access it; –thus any code writing to the console to render a map simply uses System.out … but a client application’s user interface may have many JPanel objects making it up; –which JPanel object will the map renderering code draw on?

Our approach. Make the concrete rendering class –implement the MapRenderer interface; and –inherit from JPanel; This means that our class can both –act as a concrete rendering class; and –be used as a JPanel within a client’s user interface.

class JPanelRenderer extends JPanel implements MapRenderer public void render (TiledMap aMap) { … … … } // end render protected void paintComponent (Graphics g) { … … … } // end paintcomponent method. … … … } // end JPanelRenderer class. The concrete renderer.

paintComponent(Graphics g) Renders a map by drawing a representation on the JPanel through its supplied Graphics object. Called by the panel’s redraw() method when this is called by a client. We require to have –a map object to draw; and –a representation to use for the drawing.

class JPanelRenderer extends JPanel implements MapRenderer { private TiledMap public void render (TiledMap aMap) { mMap = aMap; redraw(); } // end render protected void paintComponent (Graphics g) { … … … } // end paintcomponent method. … … … } // end JPanelRenderer class. So we have …

Constructor method(s). Since our rendering class is also a JPanel any constructor methods must reflect the requirements of a JPanel. In particular, we can supply the size of the panel in pixels. We should also initialize the mMap field.

public JPanelRenderer (int aWidth, int aHeight) { //TODO validate the arguments are >0 and set to 0 if not. setPreferredSize (new Dimension (aWidth, aHeight)); mMap = null; } // end constructor method. A constructor …

Map and tile representation. Consider the panel to be made up of a grid of squares: –size of grid = size of map; –Each rid square is drawn in a colour representing the terrain tile type of the corresponding map tile; –each grid square will have a size (resolution) in pixels.

Panel size (in pixels) and therefore shape will be determined by the client and Java layout manager when the JPanel object is created and added to the user interface. For a map to be rendered we must have –panel width >= map width; and –panel height >= map height. If dimension sizes are equal then the grid squares will be only 1 pixel in size. the shape of the grid is defined by the shape of the map and may not fill the panel if it has a different shape. Map and grid size.

Map tile colours. java.awt.Color can represent colours. The colour to draw a square on the grid is determined by the terrain tile type of the corresponding tile on the map. Use a switch statement; or adopt a similar approach to that used for determining the character to display for a tile in the console renderer.

TerrainType enumeration. Add a private java.awt.Colour attribute. Amend the constructor to initialize this attribute with a supplied colour. Amend the value definitions to include a specific colour parameter. Expose the asColour() method to return the colour attribute for an value.

Drawing a map. paintComponent (Graphics g) –super.paintComponent(g); –for each map tile: use g.setColor(…) to set colour; to draw the square use g.fillRect ( –x-coord of top left corner, –y-coord of top left corner, –x-size, –y-size )

We know … the size of the panel; –getWidth() and getHeight() panel methods. the map; –mMap field. the size of the map; –height() and width() methods of mMap. the colours to use for tiles; –asColour() method of TerrainType.

Outstanding problems. The resolution (pixels per grid square) must be determined. the coordinate system of a JPanel is different from that of the map.

Calculating resolution. Is panel big enough to represent the map even at 1 pixel per tile? –If not, then draw a message on the panel and return. Resolution can only be an integer. Is calculated as the smallest resolution of the width and height dimensions; –Maintains the square aspect ratio of the grid elements; –May result in unused areas of the panel if it is a different shape than the map.

Coordinate systems. In the map, –(0,0) is at the bottom left; –X coordinates go from left to right; –Y coordinates go from bottom to top. On the JPanel, –(0,0) is at the top left; –X coordinates go from left to right; –Y coordinates go from top to bottom (but still positive). We must apply a transform from map to JPanel coordinates.

Coordinate transformation. Need to transform tile position (x m,y m ) on the map to (x p,y p ) on the JPanel. Requires us to consider –the resolution of the panel grid squares; –the different origin positions; –the different Y coordinate directions. x p = x m * resolution y p = (map-height * resolution) – ( (y m + 1) * resolution)

Client usage. public static void main (…) { final TiledMap map = new TiledMap (…); … … … SwingUtilities.invokeLater(new Runnable() public void run() { createWindowAndRender (map); } }); … … … } // end main method.

… and the renderer … private static final void createWindowAndRender (TiledMap aMap) { JPanelRenderer panelRenderer = new JPanelRenderer(…); final JFrame frame = new JFrame(“ title "); … … … frame.add (panelRenderer, BorderLayout.CENTER); … … … frame.pack(); aMap.setRenderer (panelRenderer); aMap.render(); } // end createWindowAndRender method.