lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "david m" <dmat...@gmail.com>
Subject Re: Suspected problem in the QueryParser
Date Wed, 16 May 2007 16:16:57 GMT
I see my tables didn't come through so well. Here's (I hope) a plain
text version:

I understand that it would be best to have a UI that mapped lists of
terms to MUST, MUST_NOT and SHOULD, but I'm currently constrained to
using the QueryParser with boolean operators.

Given that, I was thinking that the addition of "*:*" (in 2.1) to
represent MatchAllDocsQuery might help solve:

> "  NOT Chair " is the problem ... you can't have a negative
> clause in isolation by itself -- it doesn't make sense because
> there isn't anything positively selecting results for you to then
> exclude results from.

by making that: "*:* NOT Chair"

To test, I indexed these documents:

DocID  Body Value
-----  ----------
  0     A B C
  1     A B C D
  2     A C D
  3     B C D
  4     C D
  5     B
                                          Expected Actual
Query           Parsed As                 Matches  Matches
-----           ---------                 -------- -------
NOT B           -body:B                     2, 4   <none>
*:* NOT B       MatchAllDocsQuery -body:B   2, 4   2, 4, 5
*.* AND NOT B   +MatchAllDocsQuery -body:B  2, 4   2, 4, 5

Can someone explain why doc 5 is matched for both "*:* NOT B" and
"*.* AND NOT B"?

Are my Expected Matches incorrect?

Thanks, david

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