It's quite common to build a Query by API rather than through
QueryParser, exactly as you mentioned. Manually run through an
Analyzer's output to build whatever type of query clauses desired.
Erik
On Oct 30, 2009, at 8:55 PM, Ron Grabowski wrote:
Is there a way to avoid using QueryParser altogether? I don't need
all the text parsed, I just want to run the name field through an
analyzer and use the output of that along with TermQuerys for the
zip and city. I'm wanting to do ~2,500 specific searches on an
index with between 26m documents. I know ahead of time what most of
my fields will be. I thought I could save some time by not having to
separate my fields, making a giant string, then having QueryParser
split the string back into terms to search.
----- Original Message ----
From: Digy <digyd...@gmail.com>
To: lucene-net-user@incubator.apache.org
Sent: Fri, October 30, 2009 7:45:21 PM
Subject: RE: Is it generally a good idea to avoid using QueryParser
if you know how to make the Query yourself?
You can combine queries . (such as your custom "TermQuery" + the
result of a
"QueryParser.Parse")
DIGY
-----Original Message-----
From: Ron Grabowski [mailto:rongrabow...@yahoo.com]
Sent: Saturday, October 31, 2009 1:12 AM
To: lucene-net-user@incubator.apache.org
Subject: Is it generally a good idea to avoid using QueryParser if
you know
how to make the Query yourself?
I can convert this text into a Query using just TermQuery and
BooleanQuery:
name:"ACME Produce" AND city:Anytown AND zip:90210
If I create the Query object on my own will I loose all the
benefits from
things like StandardAnalyzer? Is there a way I can submit an object
graph
to QueryParser so I can get the benefits of its analyzer without
making it
parse the raw string? My input is coming from well defined areas;
its not
free form where the user can enter anything they want.
Can I use a StandardAnalyzer myself then take its output and create
the
appropriate Query objects then join all of them into a BooleanQuery?
I'm using /tags/Lucene.Net_2_4_0.