the encoding is often a source of confusion. I work in a mixed
environment, me on Linux (so UTF-8 is the "System"-default), the rest is
using on Windows (with Windows-1252 I assume)
Transfering shapes often results in crippled data, and I have not found
workflows/settings that this will not happen.
Is there anything I can do about it, so we do not have to adjust things
manually all the time, or first look if the codepage is correctly set ?
Am 07.07.2016, 12:58 Uhr, schrieb Even Rouault
Le jeudi 07 juillet 2016 12:16:17, Andrea Peri a écrit :
I have some shapefiles with an extra file having extension .cpg.
The .cpg extension is a optional extension for declaration of Code-Page.
Is this file knowed and supported from QGIS 2.14. ?
I don't know if the qgis when loading a shapefile is using the ogr or
instead is using an own shapefile provider, but however I don't know
if gdal is knowing and using this .cpg file.
OGR does use the .cpg file when present (I'm just updating
http://gdal.org/drv_shapefile.html since it only mentions the reading of
codepage byte in the DBF header, but the .cpg presence overrides that)
QGIS uses OGR to read shapefiles.
But, in QGIS, the default behaviour is to make OGR *not* use the encoding
detected by the OGR shapefile driver and use instead the user defined
in the GUI. Unless you go to Settings / Options / Data Sources and
"Ignore shapefile encoding declaration", in which case OGR will manage
Qgis-user mailing list
List info: http://lists.osgeo.org/mailman/listinfo/qgis-user