For the DLF Aquifer portal (, we have 
implemented unAPI as follows.

In the html head there is a link:

<link rel="unapi-server" type="application/xml" title="unAPI" 
href="" />

which identifies the base URL for the unAPI service, then throughout 
various web pages we have abbr tags, such as this:

<abbr class='unapi-id' title='251639'></abbr>

My understanding is that an unAPI processor should take the base URL 
from the link and concatenate the title value from the abbr to create 
the URL used to retrieve first the list of formats available:

And then create another URL to retrieve the specific format of interest, 
such as this:

What Chad has described might also work, but I know that the above does 
work with Zotero, at least.

Kind regards,

Chad Fennell wrote:
> On Dec 2, 2008, at 4:21 PM, Karen Coyle wrote:
>> I've looked at unAPI, but I can't get a handle on what needs to be in 
>> the <abbr>. What would be useful?
>> kc
> I put together a pretty basic unAPI interface (technically, a Drupal 
> module) for our EthicShare project; you can see how the <abbr> tags are 
> used there.  Ex:
> embedded on the page, you'll find things like: <abbr class='unapi-id' 
> title=''>
> There are a few other pieces, but you can see these in action there as 
> well.  Ex,
> -tells clients what formats are available 
> on the sever - I've only implemented RIS so far...
> - tells the client what formats 
> are available for a given resource.
> - kind of 
> self-explanatory.
> Hope that helps a little...
> Cheers,
> -Chad

Thomas G. Habing
Research Programmer
Grainger Engineering Library Information Center
University of Illinois at Urbana-Champaign