@Project update.
Arc's are included in contour recognize today.
In relation to raw dxf data, the arc's can predict the cw or ccw so simple. (g2 or g3)
The cw and cww recognize has to have a algoritme...
A few lines of C++ algoritme is not enough.
A good algoritme looks like auto nesting. Then
it can recognize inner and outher contours without any problem.
Colorcode 1 = cad red.
The arc is swapped from g2 to g3 by the algoritme to match the contour direction, see terminal output for this arc.
For this contour we know it's a ccw contour. But if it was a reqtangular shape without the arc, we only have line's to define a cw or ccw shape. That makes it more difficult.
And a final predict algoritme solution (code) is not alway's working with very complex shapes for lines. So i have to find a solution for this. 90% of the algoritme will work, but 10% will not work by complex shapes that are multiple inlined.
I will think about the auto nesting option coming day's. This will solve any contour
direction cw or ccw recognize problem at once.
We can put out the g-code for now. But that will provide a basic program. It's my job to bring something new.
Will try to investegate auto nesting, product rotation, mirroring and optimalisation for plates.
Why not try to write some code from scratch. Try to translate your vision into output. In fact you only have to write the template.
The following linked program (deepnest) can not mirror the product and optimalisize it in the nesting... It's open source.
deepnest.io/
(open source, so far so good).
It's used for nesting Dassault Systems, a program called Fusion. Only available for Mac and Windows.
Today i looked by customer at deepnest, mirroring would be more effiecient. But deepnest can not mirror the product and
update the nesting.
So Linux has to use a VM (virtual machine). Strange isn't it? We are going to show something to Dassault Systems coming time.
And the final solution, that is not for sale and no hiring licences.
I won't look at the source code of this program. Better is to translate your own imagine into code.
Take for example the stress test in previous post's. I was advised to improve the python code dxf2gcode. But i would stay
in a problem situation, even when i would upgrade the python code.
Part of the message is hidden for the guests. Please log in or register to see it.