[Commit] New Tiling code and Printing fixes

[Thomas E Deweese <th...@kodak.com>]

Hi all,

  Apache mailer bailed on me again.
  Below is the header from the commit message.

  Enjoy!

---

  Modified:    sources/org/apache/batik/ext/awt/image GraphicsUtil.java
               sources/org/apache/batik/ext/awt/image/renderable
                        BumpMap.java CompositeRable8Bit.java
                        FilterResRable8Bit.java FloodRable8Bit.java
                        GaussianBlurOp.java GaussianBlurRable8Bit.java
                        PointLight.java TurbulencePatternGenerator.java
                        TurbulenceRable8Bit.java
               sources/org/apache/batik/ext/awt/image/rendered
                        AbstractRed.java AffineRed.java
                        MultiplyAlphaRed.java SpecularLightingRed.java
                        TileGrid.java TileMap.java TileStore.java
               sources/org/apache/batik/gvt/filter
                        GraphicsNodeRable8Bit.java GraphicsNodeRed8Bit.java
  Added:       sources/org/apache/batik/ext/awt/image/rendered
                        AbstractTiledRed.java CompositeRed.java
                        FloodRed.java GaussianBlurRed8Bit.java
  Log:
  1) Tiled demand pull implementation of BumpMap, Gaussian, Composite, Flood,
     SpecularLighting
  
  2) GaussianBlurRable now uses an Affine if the StdDeviation of the gaussian
     gets large (>10).
  
  3) Fixed a bug where GaussianBlurRable wasn't accounting for surround needed
     by skew/scale affine.
  
     Fixed a similar bug where FilterResRable wasn't accoutning for the
     surround needed for resampleing filter.
  
     These caused white to appear at the edges of tiles (or the window).
  
  4) Turbulance now clips it's computation to it's bounds (it clipped the
     output anyway, but now it does less work).
  
  5) Added AbstractTiledRed which is a useful base class for Red's that want
     to use the TileCache.
  
  6) Loosened the BufferedImage hint requirement on Graphics2D.
     the Destination methods now return null if the destination is likely
     to be incorrect.  Most code now simply assumes the 'work arounds' aren't
     needed in these cases, which may or may not be true...
  
     Worked around a bug in Graphics2D.drawImage(BufferedImage img,
  			  BufferedImageOp op,
  			  int x, int y)
     that affected printing.
  
     Vincent's SVGPrinting example now works properly.