A proposal on the json test data stuff (though I imagine could be used
for other things):
bib json:
A spec for a simple JSON representation of bib data, based on the CSL
model, and for use primarily in testing implementations. The data structures
consist of the following:
type: string
variables: string
names: array, by role (authors, editors, etc.)
Names are represented as a dictionary/hash of the following structures:
family: string
given: string
suffix: string
articular: string
display: "western" | "eastern" # something to configure display and
sorting rules, with western the default; need better terms
in addition, optional URIs can be attached to different objects for
richer transformation to, say, RDF:
names (where the URI identifies the agent)
base (where the URI identifies the main document subject)
container (where the URI identifies the containing book, or whatever)
An example may be:
{
“type”: “article-journal”,
“uri”: “http://example.org/1”,
“authors”: [
{ “family”: “Doe”, “given”: “Jane”, “uri”: “http://people.org/doej” },
{ “family”: “Smith”, “given”: “John”, “uri”:
“http://people.org/smithj” }
],
“issued”: “2000-10”,
“volume”: “22”,
“issue”: “2”,
“pages”: “23-45”,
“doi”: “1202302209”,
“container-title”: “Journal of ABC”,
“container-uri”: “urn:issn:2343-8734”
}