Giuseppe Bilotta <[EMAIL PROTECTED]> writes:
> Saturday, March 22, 2003 Hans Hagen wrote:
>
> >>I would like to know if it is possible in MetaPost (in particular
> >>MetaFun) to get all the possible intersection points of the paths,
>
> [snip]
>
> >>Anything like this?
>
> HH> not that i know, but assuming that those path have some distance, i can
> HH> imagine dividing a path in parts and determining the intersectionpoints iof
> HH> the pieces;
>
> HH> delta := .25 ;
>
> HH> for i=0 step delta upto length(p)-delta :
> HH> if intersection_found(1,subpath(i,i+.1) of p) :
> HH> store point
> HH> fi ;
> HH> endfor ;
>
> I'll see if I can make something of this ... (wow, looks like I'm
> gonna need to learn MetaPost too ...)
I implemented it by finding some intersectiontime and the recursively
finding the intersectionstimes of the two subpaths (each with some
neighbourhood of the first intersection point cut off). For the time
being I cannot find my code, but here it is in pseude code.
let p and q be paths;
let (s,t) be a pair of intersection times of p and q;
recursively find the intersection times of the
subpath (0, s - small number) of p and q;
store (s,t);
recursively find the intersection times of the
subpath (s + small number, infinity) of p and q;
feel happy;
When done this way, the intersection times will be sorted according to
the first path.
--
Emil Hedevang Lohse <http://home.imf.au.dk/emil/>
Alle sp�rgsm�l er lige dumme.
Og sp�rgsm�let "Kan �nder flyve?" er ikke dumt.
_______________________________________________
ntg-context mailing list
[EMAIL PROTECTED]
http://www.ntg.nl/mailman/listinfo/ntg-context
