Closed proycon closed 1 year ago
I implemented this, we force "@container": "@list"
in the context internally for schema:author
and schema:contributor
, so these will always be interpreter as ordered lists. We do not explicitly output this deviant context in the serialisation, however (but the json-ld output will be serialised in the proper order).
People express authors as lists in JSON-LD but if this is done without the
@list
semantics (or without schema:position), the order remains undefined in RDF (unlike in JSON itself). We currently supportschema:position
in codemetapy but we may want actual support for@list
so people can express for instance authorship like:"author": [ "First Author", "Second Author" ]
This is in line with discussions in codemeta/codemeta#272 . The codemeta context indeed has
"@container": "@list"
for things like properties likeauthor
, as discussed in that issue. In our current implementation, however, we explicitly load (and even forcibly inject) the whole schema.org context as well, after the codemeta one, and the regular schema.org context does not use"@container": "@list"
.