You can achieve this by:
a) Modifying Lucene using something like the code I originally posted which stops reading a document's fields if the required list has been found (hoping the stored field order is favourable to all of your app's access patterns)
b) Storing any large-but-rarely-accessed fields externally eg in files or an RDBMS and keeping a primary key in Lucene to retrieve the content when required.
c) If you dont want to use an RDBMS and want to keep it all Lucene-based you could store any large-but-rarely-accessed fields in another index alongside your main one to avoid the hit.. Each large field could be a separate document in this index and you could retrieve it when required by accessing largeFieldStoreIndexReader.document(x) where x is
1) calculated by taking the mainSearchIndexReader document id and multiplying by the number of externally stored large fields per main document or
2) Querying largeFieldStoreIndex for a unique key which is stored in the mainSearchIndex.
I would favour b) as a and c seem messy.
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
