lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jorge Luis Betancourt González <jlbetanco...@uci.cu>
Subject Re: Avoiding wildcard queries using edismax query parser
Date Thu, 22 Jan 2015 22:57:14 GMT
I was also suspecting something like that, the odd thing was that the with the dismax parser
this seems to work, I mean passing a single * in the query just like:

http://localhost:8983/solr/collection1/select?q=*&wt=json&indent=true&defType=dismax

Returns:

{
  "responseHeader":{
    "status":0,
    "QTime":3},
  "response":{"numFound":0,"start":0,"docs":[]
  },
  "highlighting":{}
}

Which is consisten with no "*" term indexed. 

Based on what I saw with dismax, I though that perhaps a configuration option existed to accomplish
the same with the edismax query parser, but I haven't found such option. 

I'm going to test with a custom search component. 

Thanks for the quick response Alex,

Regards,

----- Original Message -----
From: "Alexandre Rafalovitch" <arafalov@gmail.com>
To: "solr-user" <solr-user@lucene.apache.org>
Sent: Thursday, January 22, 2015 4:46:08 PM
Subject: Re: Avoiding wildcard queries using edismax query parser

I suspect the special characters get caught before the analyzer chains.

But what about pre-pending a custom search components?

Regards,
   Alex.
----
Sign up for my Solr resources newsletter at http://www.solr-start.com/


On 22 January 2015 at 16:33, Jorge Luis Betancourt González
<jlbetancourt@uci.cu> wrote:
> Hello all,
>
> Currently we are using edismax query parser in an internal application, we've detected
that some wildcard queries including "*" are causing some performance issues and for this
particular case we're not interested in allowing any user to request all the indexed documents.
>
> This could be easily escaped in the application level, but right now we have several
applications (using several programming languages) consuming from Solr, and adding this into
each application is kind of exhausting, so I'm wondering if there is some configuration that
allow us to treat this special characters as normal alphanumeric characters.
>
> I've tried one solution that worked before, involving the WordDelimiterFilter an the
types attribute:
>
> <filter class="solr.WordDelimiterFilterFactory" generateWordParts="0" generateNumberParts="0"
catenateWords="0"
> catenateNumbers="0" catenateAll="0" splitOnCaseChange="0" preserveOriginal="0" types="characters.txt"
/>
>
> and in characters.txt I've mapped the special characters into ALPHA:
>
> + => ALPHA
> * => ALPHA
>
> Any thoughts on this?
>
>
> ---------------------------------------------------
> XII Aniversario de la creación de la Universidad de las Ciencias Informáticas. 12 años
de historia junto a Fidel. 12 de diciembre de 2014.
>


---------------------------------------------------
XII Aniversario de la creación de la Universidad de las Ciencias Informáticas. 12 años
de historia junto a Fidel. 12 de diciembre de 2014.


Mime
View raw message