code4lib solr4lib Preconference
Erik Hatcher author of Lucene in Action gave a wonderful all day pre-conference on Solr at code4lib 2007. He provided many examples of Solr in action. Most of these were written in Ruby or Ruby on Rails. I was very lucky and got most of the examples to work. The item I had the most problem with was getting Flare (a user-search interface to Solr) to work.
Even though I didn’t get Flare to work, the preconference was an extremely useful experience. I got to see a couple different examples of Solr in action. Some interesting examples included:
- NINES
- OurOntario
- Villanova is also working on a project using Solr but it isn’t live yet. So I can’t link to it.
Additionally, I spent most of the day getting my MARCXML data into a format the Solr could index. I’m still working on getting a Solr schema properly written. This in particular was much more difficult than I anticipated. Thanks to Kevin Clarke and Andrew Nagy I’m a lot closer though.
The surprise of the day was that I got to exchange tips with Dan Chudnov. I helped him with an XSLT and he showed me how to set environmental variables. Overall it was a very good but exhausting day. I want some time to sit down and absorb all the information thrown at me today.
This surprises you why? :)
Just an update that the new and improved OurOntario site is located at http://search.ourontario.ca. Solr facets were the big thing at this code4lib if I recall. Some cool advances since then are:
* the improving ability to highlight matching search terms in the full text
* Did You Mean for spelling corrections
* Data Import Handler (DIH) for when you ever take a stab at importing enormous amounts of data again. Also good for indexing pdfs and rss feeds.
* Solr 1.4 should allow for the ability for near real-time searching which people were asking about at the pre-conference for things like book availability for their catalogs.