> > I didn't read the specs of JSON-LD but are those always present in the keys 
> > of json dicts?
>
> I suppose its the old lexers lack of context issue, IIUC the compact IRIs 
> need to be specified by `@context` specifications, otherwise they are just 
> strings with `:` in them.

>From what I can make of [RFC 3987], IRIs are basically URLs that can validly 
>contain literal Unicode characters, hence the *internationalized* in the 
>expanded acronym.

Plain JSON documents usually don't have URLs in the semantic role of a key. Not 
because any specification forbids it; a property name is just a string, and 
short strings naturally make better keys than verbose URLs.

In JSON-LD, URLs are first-class data types, generalized as graph nodes 
according to [the RDF spec], which JSON-LD implements, incidentally using a 
JSON-like syntax.

> > Did someone raise it on Lexilla?

It [was remarked][0] *à propos* an unrelated issue that JSON-LD semantics might 
be mutually exclusive with plain JSON, though without giving a concrete example 
like the one above.

[0]: https://github.com/ScintillaOrg/lexilla/issues/72#issuecomment-1093150057
[RFC 3987]: https://datatracker.ietf.org/doc/html/rfc3987
[the RDF spec]: https://www.w3.org/TR/rdf11-concepts/#dfn-iri


-- 
Reply to this email directly or view it on GitHub:
https://github.com/geany/geany/pull/3647#issuecomment-1778957356
You are receiving this because you are subscribed to this thread.

Message ID: <geany/geany/pull/3647/[email protected]>

Reply via email to