Note that the original question was when working from a custom update request handler, you're not doing that at all...
It's not clear to me whether doing queries ahead of time like you're doing is more or less speedy than a custom update request handler given that in the one case you're querying a bunch, and in the other you're transmitting data across the wire then throwing lots of it away. I guess it depends on whether the file is sent across before deciding whether to add it or not in a custom update processor. (hint, hint, someone who knows can answer here <G>). It seems like it would have to be in order to get the document ID out of the doc in the first place, at least in the XML case.... Anyway, your query could probably be made more efficient just from the perspective of fewer search requests if you asked for a bunch of IDs at once, something like: http://solr/select?&q=id:(myid.doc1 OR myid.doc2 OR myid.doc3)&rows=3&fl=id and you'd have to look inside the results for IDs if the result count > 0, but I don't know whether it's enough faster to matter. Best Erick On Mon, Nov 28, 2011 at 7:26 PM, chadsteele.com <c...@chadsteele.com> wrote: > oops... the query looks more like this > > http://solr/select?&q=*id:*myid.doc&rows=0 > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/conditionally-update-document-on-unique-id-tp3119302p3543871.html > Sent from the Solr - User mailing list archive at Nabble.com.
