lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Erick Erickson" <>
Subject Re: BooleanQuery TooManyClauses in wildcard search
Date Sat, 01 Dec 2007 14:55:01 GMT
See below:

On Dec 1, 2007 1:16 AM, Ruchi Thakur <> wrote:

>  Erick/John, thank you so much for the reply. I have gone through the
> mailing list u have redirected me  to. I know i need to read more, but some
> quick questions. Please bear with me if they appear to be too simple.
>  Below is the code snippet of my current search. Also i need to get score
> info of each of my document returned in search, as i display the search
> result in the order of scroing.
>    {
>   Directory fsDir = FSDirectory.getDirectory(aIndexDir, false);
>   IndexSearcher is = new IndexSearcher(fsDir);
>   ELSAnalyser elsAnalyser = new ELSStopAnalyser();
>   Analyzer analyzer = elsAnalyser.getAnalyzer();
>     QueryParser parser = new QueryParser(aIndexField, analyzer);
>     Query query = parser.parse(aSearchStr);
>     hits =;
>  }

EOE: Minor point that you probably already know, but opening a searcher is
         I'm assuming you put it in here for clarity, but in case not be
aware you should
         open a reader and re-use it as much as posslble.

         Also, it looks like you're using an older version of Lucene, since
         getDirecotory(dir, bool) is deprecated.....

>  Now as i have understood, through the mail archives you have suggsted,
> below is what we need to do.
>  1)The second was to build a *Filter* that uses WildcardTermEnum -- not a
> Query.
>  because it's a filter, the scoring aspects of each document are taken out
> of the equation (I am worried abt it , as i need scoring info)

This is true *for the wildcard clause*. It's a legitimate question to ask
scoring means for a wildcard clause. Rather, it's legitimate to ask whether
that adds much value. I managed to convince my product manager that
the end user experience didn't suffer enough to matter, but it can be argued
either way.

That said, I'm pretty sure that if you make this a sub-clause of a boolean
you still get scoring for the *other* parts of the query. That is,
BooleanQuery bq = ....
bq.add(regular query);
bq.add(filtered wildcard query);

search (bq);

(note, really sloppy pseudo code there) will give you scoring for
the "regular query" part of the bq. Of course that requires you to
break up the incoming query to the wildcard parts and the not
wildcard parts.......

>  2)Once you have a "WildcardFilter" wrapping it in a ConstantScoreQuery
> would give you a drop in replacement for WildCardQuery that would sacrifive
> the TF/IDF scoring factors for speed and garunteed execution on any pattern
> in any index regardless of size. (Does that mean it will solve my scoring
> issue and i will get scoring info)

I'm pretty sure that you don't get scoring here. ConstantScoreQuery is
named that way on purpose <G>.

>  Also it suggests "SpanNearQuery on a wildcard". I am kinda cofused which
> is the approach that should be actually used. Please suggest. At the same
> time i am studing more abt it. Thanks a lot for ur help on this.

I think I was looking at this for a method of highlighting, but span queries
won't fix up wildcard queries.

handling arbitrary wildcard queries, that is queries with, say, only
one or two leading letters is an area of Lucene that requires that
one really dig into the guts of querying and do some custom work.
We've had quite reasonable results by imposing the restriction that
wildcard queries MUST have three leading characters. That is,
a*, ab* are illegal, but abc* is legal.

What I'd suggest is to start by imposing that restriction, bumping the
maxbooleanclauses number and just let Lucene do its tricks.
Perhaps catching the TooManyClauses exception and informing
the user with some message like "query to general" or some such.
Then, if the product manager says that is unacceptable let them
know what the cost is and go from there. You'll get push-back that
"of course we must allow one character wildcards", but in my
experience that's often just a knee-jerk reaction and not as
much of a requirement as people think. Your situation may vary....

Yet another possibility, depending upon the size and space
requirements is to play interesting games with your index, for
instance for any word index special one and two letter tokens,
e.g. abcde gets a$, ab$ and abcde all indexed. Then you
pre-process your query and turn a* into a$, ab* into ab$
and just use these terms (you have to take some car
that your query analyzer doesn't strip these out). Now, you've
turned arbitrary wildcards into your three-letter-prefix rule
or simple term queries, preserving relevance etc. You have
to take some care to index your special tokens with 0 increments
so phrase and span queries work, but that's not very hard
especially since Lucene In Action demonstrates the
SynonymAnalyzer which does the same thing.

But note the implicit restriction here that you really
don't search for ab*cd*e* with this scheme, but
neither does Google.

Whew! enough for Saturday morning <G>....


>  Best Regards,
>  Ruchika
> Erick Erickson <> wrote:
>  John's answer is spot-on. There's a wealth of information in the user
> group
> archives that you should be able to search on discussing ways of providing
> the functionality. One thread titled "I just don't get wildcards at all"
> is one where the folks who know generously helped me out.
> Once you find out how to search for that you'll know you're in the right
> place.
> Here's the searchable archive.....
> Make sure you select the "java user" from the top drop-down labeled
> "Search".
> Best
> Erick
> On Nov 30, 2007 2:07 PM, John Byrne wrote:
> > Hi,
> >
> > Your problem is that when you do a wildacrd search, Lucene expands the
> > wildacrd term into all possible terms. So, searching for "stat*"
> > produces a list of terms like "state", "states", "stating" etc. (It only
> > uses terms that actually occur in your index, however). These terms are
> > all added as OR clauses of a boolean query.
> >
> > The thing is, be defult, there is a limit of 1024 caluses for a boolean
> > query. If yuor wildacrd term expands into more than this, (which happens
> > very easily), you get that exception you described. You can solve the
> > issues by setting the maximum clause count yourself, using
> >
> > BooleanQuery.setMaxClauseCount(int maxClauseCount)
> >
> > See
> >
> >
> > for mroe info.
> >
> > Bear in mind that putting a wildcard near the start of the term results
> > in a large number of boolean clauses, which increases memory usage. This
> > is the reason for the default limit. This limit will also affect fuzzy
> > queries, because they are expanded in the same way.
> >
> > Regards,
> > JB
> >
> > Ruchi Thakur wrote:
> > >
> > > Hi there.
> > > I am a new Lucene user and I have been searching the group archives
> but
> > couldn't solve the problem. I have just joined a project that uses
> Lucene.
> > > We use the StandardAnalyzer for indexing our documents and our query
> is
> > as
> > > follows when we issue a search string of t* for example:
> > > +t* +cont_type:pa
> > >
> > > We get an Exception when we issue some of our wildcard text searches
> > we get following Exception
> > >$TooManyClauses Exception : Max
> > clause if 1024
> > >
> > > Please suggest.
> > >
> > > Regards,
> > > Ruchi
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > ---------------------------------
> > > Never miss a thing. Make Yahoo your homepage.
> > >
> > >
> ------------------------------------------------------------------------
> > >
> > > No virus found in this incoming message.
> > > Checked by AVG Free Edition.
> > > Version: 7.5.503 / Virus Database: 269.16.11/1161 - Release Date:
> > 30/11/2007 12:12
> > >
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail:
> > For additional commands, e-mail:
> >
> >
> ---------------------------------
> Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See
> how.

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