Transforming Bit-Map Data

In general, graphics defined in device coordinates (bit maps and image primitives) cannot be transformed. For example, the size of an image primitive is specified in device coordinates, and cannot be altered. The size, therefore, remains unaltered down the viewing pipeline. The position of an image primitive, however, is specified in world coordinates. The image is therefore subject to translation transformations. Note, however, that GpiWCBitBlt permits the target rectangle to be specified in world coordinates, which are transformed.

Because the position of the image primitive is specified in world coordinates and its width is specified in device coordinates, positioning two images together on the screen causes special difficulties. The second image cannot be positioned without knowing the width, in world coordinates, of the first image. To get the width of the first image:

• Identify two coordinate positions, one on the image's left edge, and one on its right edge. For example, the two positions could be (10,80) and (150,80). These positions are in device coordinates.

• Convert these device coordinates to world coordinates using GpiConvert. GpiConvert converts an array of (x,y) coordinates that apply in one coordinate space to their corresponding values in another coordinate space.

• Subtract the lower x-coordinate from the higher x-coordinate. In the above example, the width of the image is the difference between the world-coordinate equivalents of 150 and 10.

When you have the width of the first image in world coordinates, you can calculate the start position of the second image.

Paths, although defined in world coordinates, are device-dependent and are bound in device coordinates when they are defined. Subsequent transformations (other than the windowing-system transformation) have no effect on paths. However, if a path is used to create a wide line, the width of the line is scaled as required.