Subject: [ggl] Snapping
From: Stephen Leary (sleary)
Date: 2009-11-22 13:38:06
On Sun, Nov 22, 2009 at 6:19 PM, Barend Gehrels <Barend.Gehrels_at_[hidden]>wrote:
> Hi Stephen,
> Welcome to the list!
> Just started looking at GGL and it strikes me that the feature that is
>> always missing from geometry libraries is snapping. Point to point, point to
>> line, line to line, line to poly and poly to poly. Line to line is the most
>> complex of these where a section of a line may snap to another line.
> That sounds as very interesting!
>> I recently wrote algorithms to do this in 3D using OpenCascade for a UK
>> based spatial data management company. I'm wondering if there would be any
>> interest in adding something like this to GGL or is it just me that wants
> You mean something like: ggl::snap(geometry_from, geometry_to,
Something like that yeah. Most of the actual work involved is in finding
points on a line that are at the snapping tolerance distance from "snap to"
> I think it can be very useful. For many things it is difficult to determine
> where it fits. I think it can be an "extension" (GGL has now an extension
> model). But it is applicable for more than GIS only. So might be an
> "snapping" extension or "editing" extension. Ideas?
I'll have a look at the extension model then and see how this can fit. For
me a snap feature is just as important as a "buffer" feature.
> The only thing I don't understand exactly, is why line/line is more complex
> than poly/poly.
poly/poly breaks down to multiple line/line snaps. line line is a
complicated beast because all parts of the line that are within a snap
tolerance of the "snap to" line get snapped. I'll produce some diagrams of
this and post them.
> p.s. for your information,
> 1) GGL is currently under Boost Formal Review, which officially ends today
> (but the report will take somewhat longer)
> 2) until GGL is accepted or rejected by Boost, the most recent sources are
> hosted at SVN_at_Geodan, which is protected, you can get an account if you
> 3) on acceptance, the SVN will be moved to SVN_at_boost, else (probably) to
> ggl mailing list
-- Stephen Leary -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/ggl/attachments/20091122/c492f554/attachment.html
Geometry list run by mateusz at loskot.net