: > Would highly appreciate if someone can suggest other efficient ways to : > address this kind of a requirement.
one approach would be to index each attachment as it's own document and search those. you could then use things like the group collapsing features to return onlly the "main" type documents when multiple attachments match. similarly: you could still index each "main" document with a giant text field containing all of the attachment text, *and* you could indx each attachment as it's own document. You would search on the main docs as you do now, but then your app could issue a secondary request searching for all "attachment" docs that match on one of the main docIds in a special field, and use the results to note which attachment of each doc (if any) caused the match. -Hoss
