Minor documentation error with GMT_Docs on OGR/GMT Data
GMT v5.1.0: in section 28.4 which describes OGR/GMT data I believe have found one error and one essential point not mentioned:
1. The @N field must not contain quoted fields, so this example (from section 28.5)
@VGMT1.0 @GPOLYGON @N"Polygon name"|substrate|id @Tstring|string|integer
does not work. It leads to error messages for each subsequent @T and @D saying that they do not match the @N format. Replace by
@VGMT1.0 @GPOLYGON @NPolygon_name|substrate|id @Tstring|string|integer
and all is well.
2. It's perhaps obvious to everyone but me, but I had a problem making a doughnut polygon (one with a hole in it) when creating data from scratch, ie not from an existing shp file via ogr2ogr. The essential piece of information of which I was ignorant, so which would have helped were it to be noted in section 28.4, is that the handedness of the @P and the @H polygons must be different. When I made the @P one CCW and the @H one CW it then worked. But if both were CCW or both CW the hole was not taken as a hole but as a separate enclosing polygon.
Hope these observations useful to someone else, j
I think your first point was a bug in 5.1.0 and was fixed awhile ago. I will check, if to it will be fixed. Quotes should be allowed. As for 2: Yes, if you build these by hand then the handedness is important. Of course, we could add a check when finding such files to make sure the handedness of P and H differ. I will make an issue from that.
BTW, I have updated the OGR/GMT section to explicitly mention the need to manage opposite handedness. In r20189.