[openbiblio-dev] Solr Search and Bibligraphica.org NQuads

[Ben O'Steen]

Big update - I have committed all the required code for the FTS Solr
search code to the openbiblio repository and made this the default
search in the URL routing.

The index is built and I've been keeping Friedrich(cc'd) up to date with
what I've been doing there (hopefully not disrupted you too much!) As I
write, it is optimising its index and the core should be responsive
quite soon.

Its the bibliographica.org core on http://solr.okfn.org/solr/  (requires
user/pass)

Admin/overview docs of the FTS in openbiblio:
https://bitbucket.org/okfn/openbiblio/src/409e66d4ca85/README_SOLR.txt

Solr schema.xml for use is at the top-level of the repo:
https://bitbucket.org/okfn/openbiblio/src/409e66d4ca85/example_solr_schema.xml

I adopted the search style and paging so no-one should be shocked by the
underlying change. However, as I mention in the doc above, this HTML
output can be enhanced and extended quite easily in the template.

The search.json response is the raw json which Solr can reply with and a
limited subset of Solr search params are passed on. The parameters can
be easily extended as mentioned in the doc.

However, to get the FTS search URIs to work, you'll need the NQuads
themselves, which brings me onto:

http://www.archive.org/details/Bibliographica.orgBnbDataset

Any ideas for how best to describe this or metadata to add to the page
will be *gratefully* received!

For deployment then:

1. Use a recent version of openbiblio.
2. python setup.py develop
3. paster make-config openbiblio development.ini
4. Edit solr params in the ini, including user/pass
  (just uncomment the bibliographica solr server example line and add
valid user/pass below it)
5. (Required for URIs to work, but optional if you just want to search)
	Load instance with NQuads from the archive.org entry.
6. Go to /search and give it a go.

(NB this can be done on any instance that can connect to solr.okfn.org
if you have the solr user/pass combo.)

Ben