Re: [gdal-dev] S-57 driver Generic attribute DSNM

2020-05-03 Thread Andreas Oxenstierna
Note that the same logic should be applied to the spatial features, if these are handled (i.e. RETURN_PRIMITIVES=ON). And for feature-feature links, i.e. RETURN_LINKAGES=ON Thank you all for your explanations. I'm sad that this is a mistake in the doc :-) because as you pointed out S57 spec

Re: [gdal-dev] S-57 driver Generic attribute DSNM

2020-05-03 Thread loïc
Thank you all for your explanations. I'm sad that this is a mistake in the doc :-) because as you pointed out S57 spec has no GUID definition. With the combination of the LNAM + DSNM, it allowed to have one that would be convenient in a PostGIS database that contains many S-57 charts. On my fo

Re: [gdal-dev] S-57 driver Generic attribute DSNM

2020-05-03 Thread kusala nine
DSNM is dataset name. I'ts defined in the DSID which only occurs in the first field/subfield group in the dataset - not sure what the driver does in respect of it. In terms of identifiers s-57 has "FOID"s, so each feature has a unique id made up of a combination of the agency code and two integers

Re: [gdal-dev] S-57 driver Generic attribute DSNM

2020-05-03 Thread Even Rouault
On dimanche 3 mai 2020 09:54:22 CEST kusala nine wrote: > DSNM is dataset name. I'ts defined in the DSID which only occurs in the > first field/subfield group in the dataset - not sure what the driver does > in respect of it. Right. The driver doc was incorrect. I've just fixed it per https://gith

Re: [gdal-dev] S-57 driver Generic attribute DSNM

2020-05-03 Thread Andreas Oxenstierna
Hi I have not checked the GDAL implementation but DSNM is an unneccessary addition to the S57 spec to generate a (potential unique) GUID. S57 spec has no GUID definition, IDs needs only be to unique inside a dataset/ENC cell. DSNM belongs to the dataset-wide DSID fields, see 6.3.2 in https:/

[gdal-dev] S-57 driver Generic attribute DSNM

2020-04-30 Thread loïc
Hello, After reading this documentation concerning the S-57 driver : https://gdal.org/drivers/vector/s57.html I thought the attribute DSNM ( Dataset name) should be defined on all features, no ? But on the source code, the method S57GenerateStandardAttributes (s57featuredefns.cpp) always a