Tim Thompson

Results 11 comments of Tim Thompson

Thanks, that's what I thought. My problem seems to have been with DBpedia's default Turtle output, which interweaves prefixes and triples. For example, querying `http://dbpedia.org/sparql` with `DESCRIBE WHERE { ?p...

@arturisimo You can find a similar approach for the Authority and Holdings Formats here: https://github.com/timathom/marc-schema. I hope to align this with @pkiraly's work in the future.

@pkiraly It's a bit of a corner case, but for completeness, you might consider adding fixed-length values ("static values") to your definition for applicable subfields, such as [533$7](https://github.com/timathom/marc-schema/blob/837ea61ab9b2fe176cb3d4e1ec9188266f043b35/marc21_json_schema.json#L21748).

It seems the issue is caused by a schema that defines a subdocument class that inherits from an abstract class. This schema, with the data that follows it, causes the...

@hoijnet Good catch, the namespace was a typo--I had set the `@schema` to the CRM IRI, but didn't remove all the prefixes. Adding the `@subdocument` declaration to the abstract class...

In DFRNT, the following error is thrown when trying to fetch the doc with ID `person/f51cf591-1ddb-4c56-ba93-1200db8d4edb`: ``` Invalid document returned from TerminusDB, missing @id: null ``` @hoijnet, you also see...

@matko Indeed, I can retrieve with `client.get_document("lc:names/n2007180387")` but **not** with `client.get_document("person/f51cf591-1ddb-4c56-ba93-1200db8d4edb")`, which was the first command referenced above.

@matko, thanks--that makes sense! @hoijnet, is this something you can accommodate in DFRNT?

@matko I notice that with `unfold='false'` subdocuments are also not expanded. Example: ``` { "@id": "digital/003e5c9f-2f78-49b7-b9c0-5f5ace8e4cdb", "@type": "dig:D1_Digital_Object", "crm:P129i_is_subject_of": [ "Blank_Node/KFaAAyt9prA_sppF1", "Blank_Node/KFaAAyt9prA_sppF10", "Blank_Node/KFaAAyt9prA_sppF4", "Blank_Node/KFaAAyt9prA_sppF7" ], "crm:P1_is_identified_by": [ "Blank_Node/KFaAAyt9prA_sppF13", "Blank_Node/KFaAAyt9prA_sppF14", "Blank_Node/KFaAAyt9prA_sppF16",...

Ah, interesting. So, it works for label lookups but not yet with existing $0 IRIs