Integration of Data Fragment Selectors

[floWetzels]

Data Fragment Selectors for data outputs of processes can be represented in the RO-Crate by adding the data format and data selector format information to the json object representing the output.

The output is a schema.org/MediaObject, with its @id being the URL of a file or a directory containing the data selector. Information on how to interpret the data selector (given in the DataSelectorFormat column as a URL) is added through the usageInfo property. Information on the data format (given in the DataFormat column as a URL) is added through the encodingFormat property.

Example

For example, consider the following process table:

Input [Sample Name]	Output [Data]	Data Format	Data Selector Format
input1	result.csv#col=1	text/csv	https://datatracker.ietf.org/doc/html/rfc7111
input2	result.csv#col=2	text/csv	https://datatracker.ietf.org/doc/html/rfc7111

The corresponding jsonld objects should look like this:

{
  "@id": "#some_process_id",
  "@type": "LabProcess",
  "object": ["#Sample_input1","#Sample_input2"],
  "result": ["result.csv#col=1","result.csv#col=2"]
},
{
  "@id": "#result.csv#col=1",
  "@type": "MediaObject",
  "encodingFormat": "text/csv",
  "usageInfo": "https://datatracker.ietf.org/doc/html/rfc711"
},
{
  "@id": "#result.csv#col=2",
  "@type": "MediaObject",
  "encodingFormat": "text/csv",
  "usageInfo": "https://datatracker.ietf.org/doc/html/rfc711"
},
{
  "@id": "#Sample_input1",
  "@type": "Sample"
},
{
  "@id": "#Sample_input2",
  "@type": "Sample"
}

[HLWeil]

• https://github.com/nfdi4plants/ARC-specification/blob/main/ISA-XLSX.md#inputs-and-outputs

@floWetzels

[HLWeil]

I think this is ready to be incorporated? @floWetzels

[HLWeil]

Actually, maybe we could model this more closely to Web Annotation Model:
https://www.w3.org/TR/annotation-model/#selectors

[HLWeil]

Something like:

{
  "@id": "#result.csv#col=2",
  "@type": {"MediaObject","Annotation"}
  "encodingFormat": "text/csv",
  "body": {
    "source" : "result.csv",
    "selector" : {
       "type" : "FragmentSelector",
       "conformsTo" : "https://datatracker.ietf.org/doc/html/rfc711",
       "value" : "col=2"
    }    
  }
}

[floWetzels]

Integrated the initial version (without the json model from Web Annotation Model) in PR #28