[gdal-dev] Coordinate system transfromations in dxf driver

Even Rouault even.rouault at spatialys.com
Wed Mar 23 01:55:54 PDT 2016


Le mercredi 23 mars 2016 00:02:46, Pero Brbora a écrit :
> > Are there other entites you identified to be incorrectly handled ?
> 
> It depends which revision of dxf reference you read. If you're
> refering to R2000 (AutoCAD 2000 which ogr supports for writing) POINT
> and LINE entity can safely be removed from coordinate transformations.
> 
> POINT:
> 100 Subclass marker (AcDbPoint)
> 10 Point location (in WCS) DXF: X value; APP: 3D point
> 20, 30 DXF: Y and Z values of point location (in WCS)
> 
> LINE:
> 100 Subclass marker (AcDbLine)
> 10 Start point (in WCS) DXF: X value; APP: 3D point
> 20, 30 DXF: Y and Z values of start point (in WCS)
> 11 End point (in WCS) DXF: X value; APP: 3D point
> 21, 31 DXF: Y and Z values of end point (in WCS)
> 
> Point and line entities were always considered 3D entities in AutoCAD
> (from R10 of dxf reference). Other entities I will have to investigate
> further.

Interesting. Would be probably worth noting that in a Trac ticket. Though I'd 
wait for fixing for real-world data confirming the need to remove the OCS 
transformation.

> 
> One question: What is the logic behind flattening circle, arc and
> ellipse entities?

At the time the driver was developed there was no support in OGR for curve 
geometries. Now we could probably expose circle and arc as CircularString. For 
ellipse, there isn't an adequate geometry in the OGC Simple Feature standard.

> 
> >> RTFM, bye
> > 
> > Not the most friendly way to say good bye IMHO...
> 
> My apologies to all who read my previous post. That last line was
> meant mainly for me as I spent few hours reading unhelpful stuff
> instead of dxf reference. I understand that it was a wrong place for it.

Got it ;-)

> 
> Bye

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com


More information about the gdal-dev mailing list