On 3/30/2013 5:25 AM, Laurent Bourgès wrote:
I would like first try / do simple things: skip segments totally out of the
clipping + margin (10%) to avoid the dasher create millions of invisible
segments.
it seams simple but check each segment could become a performance bottleneck
too !
Two caveats:
- For curves, just compare the control points, not the tight curve to save
time. Bezier curves are guaranteed to live inside the convex hull of their
control points.
- For miter joins, you have to adjust that 10%, or whatever margin you use based on the miter limit. I'm
assuming you meant "margin = line width + 10%", but it should be "margin = line width * miter
limit + 10%", unless the join style is not "MITER" in which case you can treat the miter limit
as 1.0.
For complex shapes (curves), it could be interesting to take into account the
spatial resolution: for example a curve that represents less than 1 (AA) pixel
should be skipped ... it can apply to caps (mitter, bevel, rounded) too!
For skipping in the renderer maybe, but if you try to apply that logic in the
pre-stroking stage then keep in mind that a very tiny curve could sweep a very
large angle and cause a very large number of pixels to be rendered. In the
simplest case stroking a zero length line segment with square caps will
minimally draw a rectangle that is lw x lw. Think of a tiny person holding a
very long pole across their chest. As they move the pole sweeps out the line's
area. They may spin in place, and you might think to ignore that move, but
that pole is going to be sweeping out a huge area around them.
Do you look at the java2D Pisces code, Jim ?
I think it quite difficult to understand (only few comments / explanations
...). Does somebody know if there is some docs on this implementation /
algorithms ?
I have worked extensively with it. Denis Lila did the major work to create
what it is today, but I was doing the code reviews every step of the way.
We've also integrated a version of it into JavaFX and ported it to native code
as part of that project.
...jim
