I have a file places.csv with the following contents:
name,lat,lon,type,population
Aalfang,48.8393882,15.0675236,village,410
Abern,48.0556773,13.1533381,village,203
Abersee,47.7266536,13.4169871,village,320
Abetzberg,48.0839721,14.76567,village,260
Abfaltersbach,46.7564042,12.5298515,village,616
Absam,47.2965003,11.5051409,village,6543
Abschlag,48.6241667,14.8638577,village,78
Absdorf,48.4001618,15.9803689,village,459
and an accompanying file places.csvt to specify the column types:
"String","CoordY","CoordX","String","Integer"
I am using ogr2ogr to filter the CSV file for entries in a spatial region with this command:
ogr2ogr -spat 12.0 47.0 15.0 49.0 -a_srs EPSG:4326 -oo X_POSSIBLE_NAMES=lon -oo Y_POSSIBLE_NAMES=lat places_filtered.csv places.csv
In the resulting CSV file, all values in the column population
are represented as Strings with double quotes:
name,lat,lon,type,population
Abern,48.0556773,13.1533381,village,"203"
Abersee,47.7266536,13.4169871,village,"320"
Abetzberg,48.0839721,14.76567,village,"260"
Abschlag,48.6241667,14.8638577,village,"78"
How can I get ogr2ogr to output unquoted integer values for the integer field?
I tried many options, but no success. I haven't found a way to specify field/column types explicitly, besides supplying the .csvt file with equal name as the input file. Also the -oo AUTODETECT_TYPE=YES
option gives the same result.