My apologies to Jonathan, who will get this twice since I replied only to
him the first time

***

My Solr index is used for the storage of LCSH or localized subject terms.  I
have an XForms application for editing EAD, and the subject element utilizes
TermsComponent in the index for autosuggest.  It is quite fast, especially
with about 400,000 terms in the index.  I'll create another XForm app for
editing these Solr documents directly in a web form and creating new
localized ones.  Since Orbeon can post to Solr or any other REST interface,
making modifications directly to an index is quite simple.

A doc generated from the LOC's rdf file looks like:

<doc>
  <field name="created">1986-02-11T00:
00:00-04:00</field>
  <field name="id">sh85069593</field>
  <field name="modified">1992-08-07T14:52:50-04:00</field>
  <field name="source">loc</field>
  <field name="subject">Japanese</field>
</doc>

The index can also contain other controlled vocabularies, such as personal
or corporate names and places.

Ethan


On Mon, May 17, 2010 at 10:39 AM, Jonathan Rochkind <[log in to unmask]>wrote:

> I'm interested in hearing more about what you're doing with your solr index
> of LCSH terms.  Do you have an application with documents _using_ those LCSH
> terms in Solr?  I'm trying to figure out how to deal with an index of LCSH
> terms updated from 'authorities' like id.loc.gov, in an application who's
> primary index is documents which may USE those LCSH terms.  Having
> essentially two different types of 'thing' indexed seperately (documents and
> LCSH terms) in Solr, and how to deal with that in an actual application...
> is not entirely clear to me.
>
> Jonathan
>
>
> Ed Summers wrote:
>
>> On Fri, May 14, 2010 at 3:06 PM, Ethan Gruber <[log in to unmask]> wrote:
>>
>>
>>>  I found it to be pretty fast.  I
>>> can easily integrate this into an Orbeon pipeline to keep my Solr index
>>> of
>>> LCSH terms up to date.
>>>
>>>
>>
>> That's awesome Ethan, thanks for giving it a whirl.
>>
>> //Ed
>>
>>
>>
>