Re: Simple Wildcard Search Question

[Erick Erickson]

Hmmm. What do you get when you use ?debugQuery=true? Have you
gone in through the SOLR admin page and tries queries that way? what
do you see?

Puzzles me too

Erick

On Fri, Dec 18, 2009 at 6:52 AM, QBasti  wrote:

>
> Thanks for hanging in there and helping me
>
> when wildcard-queries aren't analysed it makes sense for this one. But i'm
> still stuck at the problem mentioned above with "AndererName09". I'm not
> using any wildcards, the query-string and the index-value clearly show - a
> lot of - matches but still I get no results search for that very same
> query-string.
> e.G.
> Index-Value: "Herrn AndererName09 Vorname09" split into
> "herrn|ander|name|09|nochmal|name|09
>anderernam"
>
> Query-Value: "AndererName09" split into:
> "ander|name|09
> andernam"
>
> but no results?!
>
> --
> View this message in context:
> http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26842359.html
> Sent from the Solr - User mailing list archive at Nabble.com.
>
>

Re: Simple Wildcard Search Question

[QBasti]

Thanks for hanging in there and helping me
 
when wildcard-queries aren't analysed it makes sense for this one. But i'm
still stuck at the problem mentioned above with "AndererName09". I'm not
using any wildcards, the query-string and the index-value clearly show - a
lot of - matches but still I get no results search for that very same
query-string.
e.G.
Index-Value: "Herrn AndererName09 Vorname09" split into 
"herrn|ander|name|09|nochmal|name|09
anderernam"

Query-Value: "AndererName09" split into:
"ander|name|09
 andernam"

but no results?!

-- 
View this message in context: 
http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26842359.html
Sent from the Solr - User mailing list archive at Nabble.com.

Re: Simple Wildcard Search Question

[Erick Erickson]

I think your problem is WordDelimiterFilterFactory. For reference, see:
http://wiki.apache.org/solr/AnalyzersTokenizersTokenFilters#solr.WordDelimiterFilterFactory

As
I understand it, your input name01 gets split into
"name" and "01" given the parameters to WordDelimiterFactory
at index time.

Then, at query time, as Ankit said, wildcard queries aren't
analyzed. So "name0*" looks for *single *terms that start with
"name0". But you don't have any, so get no results

As an aside, having different parameters for WordDelimiterFactory
in your index and query clauses is just *asking* for trouble unless
you thoroughly understand what the effects of the differences
are (catenatewords and catenatenumbers are different).

Setting CatenateAll=1 might bring you joy. But it begs the
question whether you want WordDelimiterFilterFactory at all,
but that depends on whether you ever want something like
"name" to match "name01"

HTH
Erick

On Thu, Dec 17, 2009 at 3:02 PM, QBasti  wrote:

>
> Hey there,
>
> I put the complete configuration for the "text"-type fields (this field is
> a
> "text"-type) at the bottom.
> but now it got me totally confused. i switched off the WordDelimiterFactory
> and the index value is:
> "Herrn AndererName09 NochmalName09", which is getting analysed into:
> "herrn|anderername09|nochmalname09", my query therefore is:
> "AndererName09",
> which is getting analysed into "anderername09".
> So the analyser even shows that there is a match!
> But searching for that query returns no results? how can that be possible,
> if it finds a match?
>
> here's the config (with WordDelimiterFactory still on:) hope it's not to
> long
>
>
> 
>  
>
>ignoreCase="true"
>words="stopwords.txt"
>enablePositionIncrements="true"
>/>
> generateWordParts="1" generateNumberParts="1" catenateWords="1"
> catenateNumbers="1" catenateAll="0" splitOnCaseChange="1"/>
>
> protected="protwords.txt"/>
>
>  
>  
>
> ignoreCase="true" expand="true"/>
> words="stopwords.txt"/>
>generateNumberParts="1" catenateWords="0" catenateNumbers="0"
> catenateAll="0" splitOnCaseChange="1"/>
>
> protected="protwords.txt"/>
>
>  
>
>
> --
> View this message in context:
> http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26834096.html
> Sent from the Solr - User mailing list archive at Nabble.com.
>
>

Re: Simple Wildcard Search Question

[QBasti]

Hey there,

I put the complete configuration for the "text"-type fields (this field is a
"text"-type) at the bottom.
but now it got me totally confused. i switched off the WordDelimiterFactory
and the index value is:
"Herrn AndererName09 NochmalName09", which is getting analysed into:
"herrn|anderername09|nochmalname09", my query therefore is: "AndererName09",
which is getting analysed into "anderername09".
So the analyser even shows that there is a match!
But searching for that query returns no results? how can that be possible,
if it finds a match?

here's the config (with WordDelimiterFactory still on:) hope it's not to
long



  






  
  



   



  


-- 
View this message in context: 
http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26834096.html
Sent from the Solr - User mailing list archive at Nabble.com.

Re: Simple Wildcard Search Question

[Erick Erickson]

Still not enough information. Please post the exact
definition of your analyzers from your schema file.

But I'm 90% sure your WordDelimterFactory is the
problem, it's probably splitting our input on letter->
number transitions. Please see the page I referenced
above.

Best
Erick

On Mon, Dec 14, 2009 at 6:58 AM, QBasti  wrote:

>
> But then, how can I search for a name, when i don't know the exact number,
> i
> only know "name" and the number begins with "0", so what expression instead
> of "nam0*" would i have to use and why cant I use the wildcard expression?
> --
> View this message in context:
> http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26776651.html
> Sent from the Solr - User mailing list archive at Nabble.com.
>
>

Re: Simple Wildcard Search Question

[QBasti]

But then, how can I search for a name, when i don't know the exact number, i
only know "name" and the number begins with "0", so what expression instead
of "nam0*" would i have to use and why cant I use the wildcard expression?
-- 
View this message in context: 
http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26776651.html
Sent from the Solr - User mailing list archive at Nabble.com.

Re: Simple Wildcard Search Question

[QBasti]

Hey,

thanks for the help

Need analyzer definitions BOTH for indexing and querying, the correlation
between the two is important./quote>

The Query Analyzers are: 
LowerCaseFilterFactory
RemoveDuplicatesFilterFactory
WordDelimiterFilterFactory
StopFilterFactory and
SynonymFilterFactory

So it says, the query "name0*" gets analyzed into "name" and "0" without the
*.
Is that the problem, that with my query "name0*" he is really searching for
"name0" which, doesn't exist?
-- 
View this message in context: 
http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26756663.html
Sent from the Solr - User mailing list archive at Nabble.com.

Re: Simple Wildcard Search Question

[Erick Erickson]

Nothing came through showing the filters

Need analyzer definitions BOTH for indexing and querying, the correlation
between the two is important.

Storing and indexing are orthogonal. When you *index* a field, you are
putting the tokens that come from the input stream into the inverted index,
possibly applying transformations such as lowercasing, splitting various
ways, etc.. The result of *all* these transformations is what is searched
against.

When you *store* a field, the raw data with *no* processing
is put in the index. This data is never used for searching.

So, say you index and store a field. You can think of it as producing
two fields, the searchable tokens and the raw copy of the
data.

HTH
Erick

On Fri, Dec 11, 2009 at 12:48 PM, QBasti  wrote:

>
> Hey,
>
> well, the field is stored as well as indexed, and these are the filters
> that
> are used while indexing:
> W
> --
> View this message in context:
> http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26748588.html
> Sent from the Solr - User mailing list archive at Nabble.com.
>
>

Re: Simple Wildcard Search Question

[QBasti]

Hey,

well, the field is stored as well as indexed, and these are the filters that
are used while indexing:
W
-- 
View this message in context: 
http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26748588.html
Sent from the Solr - User mailing list archive at Nabble.com.

Re: Simple Wildcard Search Question

[Erick Erickson]

In addition to Ankit's comments, SOLR manipulates the
input for each field according to the rules that govern
the tokenizers and filters that you define in your schema.

This page is very valuable, although it takes a while to
get comfortable with it...

http://wiki.apache.org/solr/AnalyzersTokenizersTokenFilters

So your question is hard to answer unless you tell us what
your analyzers (index and query) for the field in question
are

Best
Erick



On Fri, Dec 11, 2009 at 11:39 AM, QBasti  wrote:

>
> Hey there,
>
> I'm pretty new to SOLR and I tried to understand the essentials of
> searching, analyzing etc..
> As far as I understand, when I index the string "name01" it gets split into
> "name" as well as "01". So when I'm searching for either "name" or "01"
> oder
> "name01", i'm getting the right result.
> But since, i think, SOLR stores the string also under "name01", why don't a
> get any result when searching for "name0*" oder "name0?"
>
> thanks for any help
> QBasti
> --
> View this message in context:
> http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26747482.html
> Sent from the Solr - User mailing list archive at Nabble.com.
>
>

RE: Simple Wildcard Search Question

[Ankit Bhatnagar]

Firstly wild card queries are not analysed by Solr.

Also there is a difference between stored and indexed fields.


Ankit

-Original Message-
From: QBasti [mailto:sebastian.f...@gmail.com] 
Sent: Friday, December 11, 2009 11:40 AM
To: solr-user@lucene.apache.org
Subject: Simple Wildcard Search Question


Hey there,

I'm pretty new to SOLR and I tried to understand the essentials of
searching, analyzing etc..
As far as I understand, when I index the string "name01" it gets split into
"name" as well as "01". So when I'm searching for either "name" or "01" oder
"name01", i'm getting the right result.
But since, i think, SOLR stores the string also under "name01", why don't a
get any result when searching for "name0*" oder "name0?"

thanks for any help
QBasti
-- 
View this message in context: 
http://old.nabble.com/Simple-Wildcard-Search-Question-tp26747482p26747482.html
Sent from the Solr - User mailing list archive at Nabble.com.