There is also a similar problem with the function read_wkt

MultiPolygon mp;

boost::geometry::read_wkt(s, mp);

line52 of multi/io/wkt/read.hpp l52 contains "geometry.back()" while i don't have a container ; a few lines above there is a line with traits::resize<MultiGeometry>::apply(geometry, ...) which suggests that geometry is not necessarly a container.

Renaud

MultiPolygon mp;

boost::geometry::read_wkt(s, mp);

line52 of multi/io/wkt/read.hpp l52 contains "geometry.back()" while i don't have a container ; a few lines above there is a line with traits::resize<MultiGeometry>::apply(geometry, ...) which suggests that geometry is not necessarly a container.

Renaud

De :Bruno Lalande <bruno.lalande@gmail.com>

À :Lepere Renaud <lepere.renaud@yahoo.com>; Boost.Geometry library mailing list <geometry@lists.boost.org>

Envoyé le :Mercredi 28 août 2013 15h09

Objet :Re: [geometry] problem with intersection on legacy geometry

Hi Renaud,Looks like the library is using this ring as a container while the Ring concept specifies that Range requirements are supposed to suffice. Either we can rewrite this algorithm in such a way that it only uses the capabilities of a Range, or we need to revise the Polygon and/or Ring concepts.

Barend, any thought on this?RegardsBruno