[ckan-dev] Handle System extensions for CKAN?

[John Erickson]

Sean Hammond <sean.hammond at okfn.org> wrote:
> If I understand right, you want to have different types of datasets, and
> the metadata fields that a type of dataset has (and presents to the user
> to be filled in) depends on an externally-hosted RDF file that defines
> the type?

That is *exactly* what we have in mind!

> This should be doable using a CKAN IDatasetForm plugin, I think.

Good to hear...

> However for the data files themselves (i.e. the "resources" in CKAN
> language) there is no corresponding IResourceForm plugin interface yet,
> CKAN's resource form is not customisable in the way that its dataset
> form is yet. You _can_ add arbitrary key, value pairs to resources just
> like you can to datasets, but for resources you have to do this over the
> API, because the web form isn't customisable. You could implement your
> own web form for creating and updating resources though, I believe this
> might be possible to do either with a CKAN extension or with a
> standalone web app (communicating with a CKAN instance via CKAN's API).

Can you provide a link into the API documentation
<http://docs.ckan.org/en/latest/api.html> to a place that discusses
this?

Thanks!

>
> On Thu, Jan 17, 2013 at 09:24:52AM -0500, John Erickson wrote:
>> Hi Everyone!
>>
>> We're working on extending the CKAN deposit workflow to including
>> object registration with esp. the Handle System. This will include
>> basic PID generation/registration, type-based metadata collection and
>> deposit, and of course data deposit (or "deposit" of external download
>> link).
>>
>> We're curious if anyone has any lessons-learned from hacking the basic
>> CKAN ingest workflow in that way: calling out to third party
>> registration systems, etc. One of the big issues we'll be facing is
>> implementing a type registry, in which the metadata template presented
>> to the user is based on a Handle-named object type, which will be
>> defined esp. by an RDF schema linked to from the Handle record (so the
>> CKAN metadata template will need to be driven by an external
>> definition, not hard-coded in an extension).
>>
>> Thoughts? Thanks!
>>
>> John
>> PS: In addition to depositing the typed data objects via CKAN (e.g. a
>> scientific dataset of a particular type), the data files (e.g. RDF)
>> that will define the object types will be deposited in this same way.
>>
>> --
>> John S. Erickson, Ph.D.
>> Director, Web Science Operations
>> Tetherless World Constellation (RPI)
>> <http://tw.rpi.edu> <olyerickson at gmail.com>
>> Twitter & Skype: olyerickson
>>
>> _______________________________________________
>> ckan-dev mailing list
>> ckan-dev at lists.okfn.org
>> http://lists.okfn.org/mailman/listinfo/ckan-dev
>> Unsubscribe: http://lists.okfn.org/mailman/options/ckan-dev
>
> _______________________________________________
> ckan-dev mailing list
> ckan-dev at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/ckan-dev
> Unsubscribe: http://lists.okfn.org/mailman/options/ckan-dev



-- 
John S. Erickson, Ph.D.
Director, Web Science Operations
Tetherless World Constellation (RPI)
<http://tw.rpi.edu> <olyerickson at gmail.com>
Twitter & Skype: olyerickson