OpenWork-NZ / core-geonetwork

GeoNetwork repository
http://geonetwork-opensource.org/
GNU General Public License v2.0
0 stars 3 forks source link

JSON ‘dataDownload’ #76

Open alcinnz opened 1 year ago

alcinnz commented 1 year ago

Every resource added to the catalogue maps to dataDownload on json export. This means that non-data resource links (e.g. reports/papers) are indistinguishable from actual data. See below: ‘Access File’ and ‘Download’ function The XML carries their identity from the metadata form to the schema, but are both handled as a dataDownload in the json.

Picture 1 Picture 1 Picture 1

ByronCinNZ commented 1 year ago

Confirm that this has been fixed

ghost commented 1 year ago

This is fine and can be removed from the board. The issue I had is the same as the JSON not updating where the browser cache needed to be cleared.

alcinnz commented 1 year ago

I don't see 'dataDownload' in the JSON export.

ghost commented 1 year ago

Agree. It's "DataDownload".

ByronCinNZ commented 1 year ago

4.5 hours 7-3-23 including nested structure as requested to support multiple March Inv

ByronCinNZ commented 1 year ago

Re: Request in Feature Additions word Doc: TITLE: Add “DataLink” field to the JSON Schema and Metadata Form. DESCRIPTION: POLDER and Google Datasets expect the JSON “DataDownload” field to link to the metadata site page which contains the link to the data (https://www.seanoe.org/data/00629/74128/) The Marine Geospatial Working Group’s DataMesh initiative requires access to the data file (.nc, .txt etc) Example: https://ant-nz.geodata.nz/geonetwork/srv/eng/catalog.search#/metadata/4a38ad3e-f0c9-4bf1-acd5-173ed41360a4

Current json schema:

This currently as two “DataDownload” fields which POLDER and the DataMesh can’t distinguish between.

"distribution": [ { "@type":"DataDownload", "contentUrl": "https://www.seanoe.org/data/00629/74128/", "encodingFormat": "WWW:LINK-1.0-http--link", "name": "Distribution Metadata", "description": "Seanoe site page" , }, "@type":"DataDownload", "contentUrl": "https://data- marineinsitu.ifremer.fr/glo_multiparameter_nrt/history/CT/GL_PR_CT_EXEI0031.nc", "encodingFormat": "WWW:LINK", "name": "Data Link", "description": "Download netcdf file" } ]

The current schema doesn’t distinguish between what POLDER is indexing, and what the DataMesh is indexing.

Would the creation of a custom property using the "additionalType" property solve this?

{ "@context": "http://schema.org/", "distribution": [{ "@type": "DataDownload", "contentUrl": "https://www.seanoe.org/data/00629/74128/", "encodingFormat": "WWW:LINK-1.0-http--link", "name": "Distribution Metadata", "description": "Seanoe site page", "additionalType": { "@type": "DataLink", "name": "Get Data", "description": " https://data-marineinsitu.ifremer.fr/glo_multiparameter_nrt/history/CT/GL_PR_CT_EXEI0031.nc " } }] }

For this to work the user would need the ability to specify the “DataLink” separately from the default “DataDownload”. As discussed, this could be done from the Distribution dropdown.