This seems fine in JTS 1.11. There's been several bug
fixes/improvements to the buffer code since it was last ported to GEOS,
I think - and any of them could be causing the problem.
Ah, the joys of two code bases.... gcj, anyone?
GEOS wrote:
#241: GEOSBuffer produces an invalid geometry
------------------------+---------------------------------------------------
Reporter: pleroux | Owner: geos-devel@lists.osgeo.org
Type: defect | Status: new
Priority: major | Milestone: 3.1.1
Component: Default | Version: 3.0.0
Severity: Unassigned | Keywords:
------------------------+---------------------------------------------------
Here's a a valid polygon for which the GEOSBuffer function produces an
invalid geometry.
At the beginning, I didn't test the returned geometry but only wrote it in
WKT format (and it "worked"). I tested different values for dist (1 (m)
and 2) and, with 1, it is obvious than the result is not valid (a big hole
of the poygon disappears and two small polygons are created). With 2, it
"seems" good and the obvious error obtained with 1 don't occur any more.
Even if they are not valid, I think they're interesting to investigate
{{{
>>> from shapely.wkb import loads
>>> polygon =
loads(open('polygon.hexwkb').readline().rstrip().decode('hex'))
>>> polygon.is_valid
True
>>> buffer_1 = polygon.buffer(1.)
>>> buffer_1.is_valid
False
>>> buffer_2 = polygon.buffer(2.)
>>> buffer_2.is_valid
False
>>> print >> open('buffer_1.hexwkb','w'), buffer_1.wkb.encode('hex')
>>> print >> open('buffer_2.hexwkb','w'), buffer_2.wkb.encode('hex')
}}}
On MacOSX, libgeos 3.0.0, with an equivalent code in C, the problem occurs
too.
------------------------------------------------------------------------
_______________________________________________
geos-devel mailing list
geos-devel@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/geos-devel
--
Martin Davis
Senior Technical Architect
Refractions Research, Inc.
(250) 383-3022
_______________________________________________
geos-devel mailing list
geos-devel@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/geos-devel
