OGR Shapefile to CSV: Using Python to get WKT for multipart Polys

Question

Using ogr2ogr and python, I'm running into a wall with trying to create a CSV with WKTs for a shapefile that contains multipart polygons. Currently this is this code I'm using (found it on internet):

#Purpose: To export a shapefile to WKT

from osgeo import ogr
import sys, os

input = ogr.Open(sys.argv[1])


layer_in = input.GetLayer()
layer_in.ResetReading()
feature_in = layer_in.GetNextFeature()

outfile = open(sys.argv[1] + ".wkt", "w")

while feature_in is not None:

    geom = feature_in.GetGeometryRef()
    geom_name = geom.GetGeometryName()

    outfile.write(str(geom)+ '\n')

    feature_in = layer_in.GetNextFeature()

So, do I need to use a different geometry container or do I need to evaluation each feature and if it's a multipart use a different set of geometry classes? I've looked around in the gdal/ogr documentation but I'm having a hard time reconciling it.

Answer 1

If you can't use ogr2ogr, and you are looking for wkt, use ogr.Geometry.ExportToWkt():

from osgeo import ogr
import sys, os
input = ogr.Open(sys.argv[1])
layer_in = input.GetLayer()
layer_in.ResetReading()
outfile = open(sys.argv[1] + ".wkt", "w")
for f in layer:
    geom = feature_in.GetGeometryRef()
    geom_name = geom.GetGeometryName()
    wkt = geom.ExportToWkt()
    outfile.write(wkt + '\n')

With a few code changes for brevity. I tested it briefly and was able to get wkt.