lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jack Krupansky" <j...@basetechnology.com>
Subject Re: field "title_ngram" was indexed without position data; cannot run PhraseQuery
Date Tue, 15 Oct 2013 16:01:45 GMT
Show us the field and field type from your schema.

Likely you are "omitting" position info for the field, and the field type 
has "autoGeneratePhraseQueries="true"" - the ngram analyzer generates a 
sequence of terms for a single source term and then the query parser 
generates a PhraseQuery for that sequence, but that requires position info 
in the index but you have omitted them. That's one theory.

So, if that theory is correct, either retain position info by getting rid of 
the "omit", or remove the autoGeneratePhraseQueries.

-- Jack Krupansky

-----Original Message----- 
From: Jason Hellman
Sent: Tuesday, October 15, 2013 11:19 AM
To: solr-user@lucene.apache.org
Subject: Re: field "title_ngram" was indexed without position data; cannot 
run PhraseQuery

If you consider what n-grams do this should make sense to you.  Consider the 
following piece of data:

White iPod

If the field is fed through a bigram filter (n-gram with size of 2) the 
resulting token stream would appear as such:

wh hi it te
ip po od

The usual use of n-grams is to match those partial tokens, essentially 
giving you a great deal of power in creating non-wildcard partial matches. 
How you use this is up to your imagination, but one easy use is in partial 
matches in autosuggest features.

I can't speak for the intent behind the way it's coded, but it makes a great 
deal of sense to me that positional data would be seen as unnecessary since 
the intent of n-grams typically doesn't collide with phrase searches.  If 
you need both behaviors it's far better to use copyField and have one field 
dedicated to standard tokenization and token filters, and another field for 
n-grams.

I hope that's useful to you.

On Oct 15, 2013, at 6:14 AM, MC <videmos0@gmail.com> wrote:

> Hello,
>
> Could someone explain (or perhaps provide a documentation link) what does 
> the following error mean:
> "field "title_ngram" was indexed without position data; cannot run 
> PhraseQuery"
>
> I'll do some more searching online, I was just wondering if anyone has 
> encountered this error before, and what the possible solution might be. 
> I've recently upgraded my version of solr from 3.6.0 to 4.5.0, I'm not 
> sure if this has any bearing or not.
> Thanks,
>
> M
> 

Mime
View raw message