lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Erik Hatcher <e...@ehatchersolutions.com>
Subject Re: free text search with numbers
Date Tue, 05 Jul 2005 09:58:47 GMT

On Jul 5, 2005, at 2:26 AM, BOUDOT Christian wrote:

> :-) I changed the main lines and compiled the QueryParser.java  
> after that I
> compiled the entire package, build a new jar file and installed it  
> on my
> server. (I followed the same procedure I used when I added a new  
> analyzer)

Sorry if these are dumb questions, but you changed QueryParser.jj and  
then recompiled QueryParser.java?  Did you run JavaCC to regenerate  
QueryParser.java?

You don't need to rebuild Lucene to add a new Analyzer.   
QueryParser.jj is quite a different story as it is a JavaCC-based  
parser and needs another step of generation besides just compilation  
of the .java file.

     Erik


>
> Cheers
>
> Chris
>
> -----Original Message-----
> From: Erik Hatcher [mailto:erik@ehatchersolutions.com]
> Sent: 04 July 2005 18:33
> To: java-user@lucene.apache.org
> Subject: Re: free text search with numbers
>
> A couple of sanity checks... you changes the main lines, not the
> comments I hope :)  And also, you need to rebuild the parser code by
> running JavaCC on it.  If you're tinkering with a copy of Lucene's
> source code you can run the Ant target "javacc" and you must have
> JavaCC installed per the build instructions.
>
>      Erik
>
>
>
> On Jul 4, 2005, at 11:38 AM, BOUDOT Christian wrote:
>
>
>> I have found in the QueryParser.jj those lines of comments:
>>
>> // OG: to support prefix queries:
>> // http://nagoya.apache.org/bugzilla/show_bug.cgi?id=12137
>> // Change from:
>> // | <WILDTERM:  <_TERM_START_CHAR>
>> //              (<_TERM_CHAR> | ( [ "*", "?" ] ))* >
>> // To:
>> //
>> // | <WILDTERM:  (<_TERM_CHAR> | ( [ "*", "?" ] ))* >
>>
>> So as indicated I changed my line to the second option but it
>> didn't solve
>> the problem I still getting the lexical error. Do I have to do
>> something
>> special to this .jj file or is it read when the QueryParser.java
>> file is
>> compiled?
>>
>> Cheers
>> Chris
>>
>>
>> -----Original Message-----
>> From: Erik Hatcher [mailto:erik@ehatchersolutions.com]
>> Sent: 04 July 2005 16:15
>> To: java-user@lucene.apache.org
>> Subject: Re: free text search with numbers
>>
>>
>> On Jul 4, 2005, at 9:02 AM, BOUDOT Christian wrote:
>>
>>
>>
>>> Hi,
>>>
>>> I modified the analyzer (it is now vegetarian and won't eat numbers
>>> anymore
>>> :-) but I have hit a new problem. The parser won't accept a keyword
>>> to start
>>> with a wildcard character. (*/12/2003) Any hints to solve this new
>>> issue?
>>>
>>>
>>
>> This is by-design "issue" with QueryParser (to avoid WildcardQuery's
>> from being created that run through every term in the index).  It is
>> the way the grammar has been defined.  You would need to build your
>> own QueryParser to change this behavior.   The relevant QueryParser
>> grammar piece is this (from QueryParser.jj):
>>
>> | <WILDTERM:  <_TERM_START_CHAR>
>>                (<_TERM_CHAR> | ( [ "*", "?" ] ))* >
>>
>>      Erik
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
>> For additional commands, e-mail: java-user-help@lucene.apache.org
>>
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
>


---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org


Mime
View raw message