GEOS Support

[Steve Lime]

Good catch, thanks for fixing. Very few of the others take non-shapeObj arguments so hopefully that was the only one.

Steve

>>> "Umberto Nicoletti" <umberto.nicoletti at gmail.com> 06/28/06 8:46 AM >>>
Steve,
the buffer function in mapscript is only accepting int values for
width instead of double values. I have already fixed it, but there
could be other functions with similar problems.

I'll take a look at the contains method.

Regards,
Umberto

On 6/27/06, Steve Lime <Steve.Lime at dnr.state.mn.us> wrote:
> Hi all: I did commit the port of the C++ support for GEOS to the new C-API. Seems to work well in testing. I wrapped all the meaningful GEOS methods:
>
>   - buffer, intersection, union, difference, symDifference, boundary, getCentroid, convexHull
>
>   - contains, within, overlaps, intersects, disjoint, equals, touches, crosses
>
>   - area, length, distance
>
> and have included those in the SWIG binding for shapeObj's.
>
> To do's or gotchas:
>
>   - configure needs to detect a minimum version of GEOS (2.2.2 or higher - that's when length and area were exposed to C API)
>
>   - need to handle z coordinates if enabled in MapServer
>
>   - error handling ain't great, it's not as easy to get at GEOS errors as before. We could define a GEOS error handler and register it when initializing
>     the library but I worry about conflicts with PostGIS or OGR that also use it.
>
>   - there's one MapScript issue. There already is a contains method that takes a pointObj. I created an overloaded method for contains that takes
>     a shapeObj. In perl that's ok, but I'm not sure about other languages.
>
>   - PHP/MapScript needs to be updated.
>
> I'll update all necessary bugs in Bugzilla with details. Let me know if you run into issues beside the one already fixed...
>
> Steve
>