My question involves filtering of search results. Some relevant facts
for my application:
- Each document is one fragment
- The index for case-sensitive search is enabled
- Diacritic-sensitive searches are not necessary
- Wildcards are enabled
- The desired result is a list of documents that contain one or more
matches, and I don't care how many matches occur in each document
- Searches are performed on selected elements, rather than the entire
document
- Some elements are unique within a document (e.g., title), some
elements occur multiple times within a document (e.g., author)
- My system uses lib-parser
My understanding is that a set of fragments with potential matches is
returned, followed by an optional filtering step. The purpose of
filtering is two-fold:
- Find fragments with multiple matches (which I don't care about - I
only want the list of matching documents and each fragment is a document)
- Eliminate false positives (I expect false positives if there are
wildcard or phrase searches)
My thought is that I'd like to do a filtered search only when there is
the possibility of false positives (wildcard or phrase search only?),
and otherwise do an unfiltered search. I'd select a filtered search if
*, ? or " appears in the query; otherwise I'd do an unfiltered search.
I expect that what I'm try to achieve is not unusual, so I thought that
there might be an easier way to achieve by goal. I'd appreciate anyone's
thoughts on the topic. Thanks.
-Dave
_______________________________________________
General mailing list
[email protected]
http://xqzone.com/mailman/listinfo/general
