Bug report #20541

"Topology Checker" and multipart (shapefiles)

Added by Thomas Nogatz about 4 years ago. Updated almost 4 years ago.

Category:C++ plugins/Topology checker
Affected QGIS version:3.7(master) Regression?:No
Operating System: Easy fix?:No
Pull Request or Patch supplied:No Resolution:
Crashes QGIS or corrupts data:No Copied to github as #:28361


New description:
In QGIS 3 shapefiles are now always multipart, so the Topology Checker rule "must not have muultipart feature" always return errors for such datasource.

Old description:

Prerequisite: Polygon shape (see attached shapefile) with clean topology after im- and export with GRASS (v.in.ogr.qgis, snapping threshold 0.1 m):
without invalid/overlapping/multipart geometries, without duplicates, with some allowed gaps.

QGIS 2.18 - Topology Checker (check on invalid/overlapping/multipart/duplicates/gaps) recognizes the gaps. That's correct.
QGIS 3.4.1 - same topology check: The gaps are recognized and a lot of multipart features are reported. But there are no multipart features - so this should be an error.

FFH_326.zip (164 KB) Thomas Nogatz, 2018-11-18 02:44 PM


#1 Updated by Giovanni Manghi about 4 years ago

  • Status changed from Open to Feedback

in QGIS 3 shapefiles are always forced as multipart, so I guess that the tool should consider shapefiles as a special case.

#2 Updated by Giovanni Manghi almost 4 years ago

  • Operating System deleted (Windows 7)
  • Affected QGIS version changed from 3.4.1 to 3.7(master)
  • Status changed from Feedback to Open
  • Description updated (diff)
  • Subject changed from Plugin "Topology Checker" - false error report: Multipart Features to "Topology Checker" and multipart (shapefiles)
  • Category changed from Vectors to C++ plugins/Topology checker

#3 Updated by Dario Bevilacqua almost 4 years ago

I have the same problem. I need to check the singlepart and the multipart features and since qgis 3.x it is no more possible.
The command in python 3 (geom.isMultipart()), same as for qgis gui and plugin, return always multipart, even if the geom is single part.
The same file in qgis 2.x and python 2 return the check as "single part".

It's pretty ridiculous that I have to check the geometry with an old and outdated version of qgis. It will be fixed?

#4 Updated by Giovanni Manghi almost 4 years ago

Dario Bevilacqua wrote:

pretty ridiculous


have you seen #20541-1 ?

also if there is some bug/regression that is a blocker for your workflow please consider supporting the effort for fixing it, rather than ranting about it.

Also available in: Atom PDF