SeaDataNet values mapping

[pzaborowski]

in the block like this:
https://ogcincubator.github.io/iliad-apis-features/bblock/ogc.hosted.iliad.api.features.zarr_attrs_sdn/examples

wouldn't it be nice to map URNs like this value
:sdn_parameter_urn "SDN:P01::ALATZZ01"
to http://vocab.nerc.ac.uk/collection/P01/current/ terms ( in this case http://vocab.nerc.ac.uk/collection/P01/current/ALATZZ01/)

[rob-metalinkage]

What do you mean "map"?

Once uplifted to RDF there are ways to map, and validate using "closure" graphs, but if it requires an api call to a service for a large vocab that's trickier, could be added as a function, but we would be making the BB do more than standards support.

JQ transforms could do some work, again it's a contract with implementation layer the BB could have, but needs to be described.

Imho it's best to uplift as a literal, and say it's the job of the implementation to perform content mappings.

[avillar]

The main problem in this case is that JSON-LD can't work with mappings (property names, values) that have colons (:) in them, so there's no way to map "SDN:P01:" to http://vocab.nerc.ac.uk/collection/P01/current/.

That said, one thing that could be done is, as Rob mentions, use a light JQ transform that "prefix" into something that we can use with JSON-LD. For example, we can turn all SDN:(P[\d]+):: prefixes into http://vocab.nerc.ac.uk/collection/$1/current/ when uplifting. Let me see if I can get an example working.

[avillar]

Sorry, I forgot to push my changes. You can see them here: if there's a value for sdn_parameter_urn and/or sdn_uom_urn that starts with SDN:P...::, then whe replace that with the URI prefix for the vocabulary.

This is what it looks like in Turtle. If the value is not recognized, then it's kept as a string, but otherwise it's bound as a URI. More patterns could be added here, too.

[pzaborowski]

Hi, @rapw3k so if I would be the aim to support this SeaDataNet profile in the OIM, the point would be to have definitions for all the items mapped currently to defaults like here:

[rapw3k]

hi @pzaborowski sorry i missed this one.
we need to align these terms and/or add new terms to OIM in some extension.

this is about metadata right?
I can create quickly extension. would need just short description of the properties, do you have it?

[pzaborowski]

'sdn' are the SeaDataNet but the definitions are coming from the CF convention https://www.bodc.ac.uk/resources/delivery_formats/cfnetcdf_format/

Attribute 'sdn_parameter_urn': this is the SeaDataNet vocabulary code that best fits the variable in question, codes shall be lookedup in the http://vocab.nerc.ac.uk/collection/P01/current/'sdn_parameter_urn'.
Attribute 'sdn_parameter_name': this is the official SeaDataNet text string that is linked to the vocabulary code in 'sdn_parameter_urn'
Attribute 'sdn_uom_urn': this is the SeaDataNet code for the units being used by the file originator to quantify the variable
Attribute 'sdn_uom_name': this is the official SeaDataNet text string for the units described in 'sdn_uom_urn'

_ARRAY_DIMENSION is the dimension, in case of x,y,z it is the axis. In case of trajectories it is the id of the particle.
units - also CF convention somehow redundant to the sdn_uom_urn but required by the convention allowed values I'm aware are degrees_north, degrees_east, degrees_N, degrees_E

standard_name - values are the coming from http://vocab.nerc.ac.uk/standard_name/
long_name - maybe is the text correlated to standard_name

POSITION_SEADATANET_QC:flag_meanings = "no_quality_control good_value probably_good_value
probably_bad_value bad_value changed_value value_below_detection value_in_excess interpolated_value
missing_value value_phenomenon_uncertain value_below_limit_of_quantification"