Arbitrarily Qualified Dublin Core Element Names are based on RFC 5013 Element Names prefixed with the String aqdc_
.
curl -s "https://tools.ietf.org/html/rfc5013" \
| grep "Element Name" \
| sed -e 's/^Element Name: /Element Name: aqdc_/'
Element Name: aqdc_title
Element Name: aqdc_creator
Element Name: aqdc_subject
Element Name: aqdc_description
Element Name: aqdc_publisher
Element Name: aqdc_contributor
Element Name: aqdc_date
Element Name: aqdc_type
Element Name: aqdc_format
Element Name: aqdc_identifier
Element Name: aqdc_source
Element Name: aqdc_language
Element Name: aqdc_relation
Element Name: aqdc_coverage
Element Name: aqdc_rights
Elements names in Arbitrarily Qualified Dublin Core are the Element Names from RFC5013 prefixed with the string aqdc_
.
Content is as described in the RFC, and is encoded as per this specification.
Dublin core values require a simple string which may optionally be qualified by any or all of the following:
value_uri
: linked data style URI representing the same value as the value string.qualifier_uri
: URI to a qualifier, such asdcterms:
or other mapping analog from the source metadataqualifier_string
: arbitrarty label displayed to the end user along with this field
aqdc_creator: [ 'Brian Tingle' ]
aqdc_creator: [
'Brian Tingle', {
'value_uri': 'http://orcid.org/0000-0002-5309-7921'
}
]
(question, is authority source/type needed, or can it be infered by the URI value?)
{aqdc_creator: [
'Brian Tingle', {
'qualifier_uri': 'http://id.loc.gov/vocabulary/relators/ppt',
'qualifier_string': 'Puppeteer'
}
]}
aqdc_creator: [
{ 'value_uri': 'http://orcid.org/0000-0002-5309-7921' }
]
should this be valid?
aqdc_creator: [
{ 'qualifier_uri': 'http://id.loc.gov/vocabulary/relators/ppt' }
]
Fully normalized for indexing, this might look like
aqdc_creator: [ '[not supplied]', {
'qualifier_uri': 'http://id.loc.gov/vocabulary/relators/ppt',
'qualifier_string': 'Puppeteer'
}]
Which might display along the lines of
Creator (Puppeteer): [not supplied]
{"aqdc_creator": [
"Brian Tingle", {
"value_uri": "http://orcid.org/0000-0002-5309-7921",
"qualifier_uri": "http://id.loc.gov/vocabulary/relators/ppt",
"qualifier_string": "Puppeteer"
}
]}
Create one JSON-L file per record? Probably not. Could we use \n\n
as the record seperator, and a single \n
as a field seperator?
{"aqdc_creator":["Brian Tingle",{"value_uri":"http://orcid.org/0000-0002-5309-7921","qualifier_uri":"http://id.loc.gov/vocabulary/relators/ppt","qualifier_string":"Puppeteer"}]}
A file with metadata for a set of records.
{'aqdc_title':['Record 1']}
{'aqdc_identifier':['id:01']}
{'aqdc_title':['Record 2']}
{'aqdc_identifier':['id:02']}
ordering option 2 --
[
{"aqdc_title":["Record 1"]},
{"aqdc_identifier":["id:01"]}
],[
{"aqdc_title":["Record 2"]},
{"aqdc_identifier":["id:02"]}
]
// an array of single key objects?
A special parser option that will normalize input values such as:
'aqdc_creator':['http://orcid.org/0000-0002-5309-7921']
For Calisphere/DPLA use case, need to create thumbnail/reference image
Want to full text index everything
This will have links to the best viewers for the content described in the record.