uml - 为 UML 图表创建 "scruffy"纸效果的算法？

Question

我喜欢 http://yuml.me 的脏纸效果UML图，有没有一种算法最好不是在Ruby中，而是在PHP、java或C#中，我想看看在Rebol中做同样的事情是否容易:

http://reboltutorial.com/blog/easy-yuml-dialect-for-mere-mortals2/

Answer 1

效果结合

对角线渐变填充

阴影

线条不是笔直的，而是有一些明显随机的小偏差，给人一种“邋遢”的感觉。

您可以使用输入的散列作为随机数生成器的种子，以便每次获得相同的图像。

这似乎可以用于整理行:

public class ScruffyLines {
    static final double WOBBLE_SIZE = 0.5;
    static final double WOBBLE_INTERVAL = 16.0;

    Random random;

    ScruffyLines ( long seed ) {
        random = new Random(seed);
    }


    public Point2D.Double[] scruffUpPolygon ( Point2D.Double[] polygon ) {
        ArrayList<Point2D.Double>   points = new ArrayList<Point2D.Double>();
        Point2D.Double              prev   = polygon[0];

        points.add ( prev ); // no wobble on first point

        for ( int index = 1; index < polygon.length; ++index ) {
            final Point2D.Double    point = polygon[index];
            final double            dist = prev.distance ( point );

            // interpolate between prev and current point if they are more
            // than a certain distance apart, adding in extra points to make 
            // longer lines wobbly
            if ( dist > WOBBLE_INTERVAL ) {
                int    stepCount = ( int ) Math.floor ( dist / WOBBLE_INTERVAL );
                double step = dist / stepCount;

                double x  = prev.x;
                double y  = prev.y;
                double dx = ( point.x - prev.x ) / stepCount;
                double dy = ( point.y - prev.y ) / stepCount;

                for ( int count = 1; count < stepCount; ++count ) {
                    x += dx;
                    y += dy;

                    points.add ( perturb ( x, y ) );
                }
            }

            points.add ( perturb ( point.x, point.y ) );

            prev = point;
        }

        return points.toArray ( new Point2D.Double[ points.size() ] );
    }   

    Point2D.Double perturb ( double x, double y ) {
        return new Point2D.Double ( 
            x + random.nextGaussian() * WOBBLE_SIZE, 
            y + random.nextGaussian() * WOBBLE_SIZE );
    }
}

example scruffed up rectangle http://img34.imageshack.us/img34/4743/screenshotgh.png