New section suggestion - Hacking Trove
Trove is a fantastic resource of course, but it's still rather inward focused. While it aggregates info from many sources, it generally only provides this aggregated info as html pages. Please correct me if I'm wrong, but I think the only publicly available machine-readable interfaces at the moment are the SRU & OpenSearch interfaces for People Australia. As a result, there's little opportunity for re-using, mashing-up, or extending the data.
It'd be great to provide a space where the possibilities for hacking and extending Trove are discussed. Topics might include APIs, RDF/RDFa, Linked Data, machine tags, and until we get something better we could also share screen-scraping recipes.
I know that there are discussions about these sorts of things going on within the NLA, but I also know that such developments are often demand driven. There's obviously a chicken/egg situation here because people aren't likely to demand things like RDF until they have an understanding of what it might make possible.
So how about it? A forum section on hacking Trove? Let's start a public discussion on the possibilities...
Last edited by wragge; 12-07-2010 at 04:10 PM.
Couldn't agree more...have just been suggesting we create a developers/hackers area on our website at work...the idea is that when people have the opportunity to 'do their stuff' they may do it! it's called community engagement and we need it to remain relevant.
Following up my own post, here's a list of current NLA APIs.
As well as the ones I mentioned above, there are OpenSearch interfaces to Libraries Australia and Picture Australia, and SRU and OpenSearch interfaces to ARO.
The NLA catalogue also has z39.50 and OpenSearch. You can also pass view=raw rather than view=rss argument to get more detailed XML search results -- not officially supported but unlikely to change in a breaking way, extra fields might be added though.
For our digital collections, while not really an API per se, there's a lot of stuff you can do by manipulating persistent identifiers (that document doesn't cover all the PI forms, but other collection types follow the same sort of pattern). You can combine that with a "pi:" prefixed catalogue search to get bibliographic data.
Excellent, thanks -- more to play with! Also seems like further justification for having a dedicated section on the forum for sharing such info.
Search and Retrieve via URL (SRU) Server documentation
Did you know that the National Library of Australia provides access to the Trove People and Organisation zone/NLA Party Infrastructure records using a web service based on the Search and Retrieve via URL (SRU) protocol? SRU is a standard XML-based search protocol which you can use to find and get records in a range of XML formats. It is similar to the OpenSearch protocol in a number of ways but far more powerful.
Documentation for the NLA Party Infrastructure SRU Server has just been released and is available via the National Library of Australia APIs page and directly at http://bit.ly/bndML1. This documentation provides general information about the protocol itself and specific information about the National Library of Australia (NLA) Party Infrastructure SRU Server. As the ARDC Party Infrastructure Project progresses minor changes to the SRU Server and other Party Infrastructure APIs will be made so it will be worth signing up to the trove-api-announce list to keep abreast of changes.
Last edited by bdewhurs [NLA]; 17-09-2010 at 01:57 PM.
I know that I am going the way of the dodo, but I have not the slightest idea what you are talking about. I tried out some of the links given, and on some I managed to get past the first page, but could not make any sense out of them, or get any useful information. Just about every time computer experts make some brilliant improvement, a site becomes more difficult for the general public to use. Especially the older folk. There is a German word for this: Verschlimmbesserung. An improvement that makes everything worse.
Still, as long as it does not interfere with the basic usage of Trove.
Tags for this Thread