[PostGIS] #5872: CG_StraightSkeletonPartition should return a geometry collection instead of a multipolygon

[PostGIS]

#5872: CG_StraightSkeletonPartition should return a geometry collection instead of
a multipolygon
---------------------+-----------------------------
  Reporter:  robe    |      Owner:  Loïc Bartoletti
      Type:  defect  |     Status:  new
  Priority:  medium  |  Milestone:  PostGIS 3.6.0
 Component:  sfcgal  |    Version:  master
Resolution:          |   Keywords:
---------------------+-----------------------------
Comment (by robe):

 Replying to [comment:1 Loïc Bartoletti]:
 > You're right, a MultiPolygon will always invalid.
 >
 > I think that I used
 [https://gitlab.com/sfcgal/SFCGAL/-/blob/master/src/algorithm/straightSkeleton.cpp?ref_type=heads#L492
 MultiPolygon] type because it was simpler (and what we have done for
 subdivde in QGIS).
 > But it's true that we have an SRF in PostGIS. In fact, we can also
 return a SRF for CG_StraightSkeleton. However, in the case of SFCGAL, I
 think it's probably more appropriate to do a PolyhedralSurface (moreover
 PolyhedralSurface and TIN are well-supported in QGIS now). What do you
 think?


 Agree that polyhedral surface makes most sense.  And given you can reduce
 that to polygons using ST_Dump I think that should be fine.  We can always
 add a doc example showing using ST_Dump to explode into polygons for those
 that can't operated with polyhedral surfaces.
-- 
Ticket URL: <https://trac.osgeo.org/postgis/ticket/5872#comment:3>
PostGIS <http://trac.osgeo.org/postgis/>
The PostGIS Trac is used for bug, enhancement & task tracking, a user and developer wiki, and a view into the subversion code repository of PostGIS project.