lucenenet-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel Wertheim <dan...@wertheim.se>
Subject Re: Prefix query performance
Date Sun, 08 Aug 2010 19:46:43 GMT
The first query is parsed as a TermQuery whilst the second as a PrefixQuery
(since it only contains a trailing asterisk). Don't know about PrefixQueries
if they are the same as WildcardQueries in the sence where a larger prefix
decreases the number of terms it needs to enumerate to find matches.

Look at 3.4.7 & 3.5.5 in Lucene in Action.

//Daniel

2010/8/8 Ayende Rahien <ayende@ayende.com>

> I run into an interesting perf issue just now.
>
> 100 runs of this query: "Name:WILLY" @ 224ms
> 100 runs of this query: "Name:WIL*" @ 790ms
>
> There are ~5000 documents in the index, two fields, Name is using
> StandardAnalyzer.
>
> Is there supposed to be that big a perf difference between the two?
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message