i m thinkking another solution : it could also possible building additional
info in relation to the query to use :).
2016-08-03 19:27 GMT+02:00 Cristian Lorenzetto <
cristian.lorenze...@gmail.com>:
> yes internally lucene could make a lot of things for reaching target:
> - it could add info map {-1,0,1,undefined} 2 bit for every combination of
> concrete fields. With additional short array it could make possible it. for
> ordinable comparation. i think it is not so difficult to do for lucene
> developers.
> - more difficult is to add pattern matching info. probably it could be
> possible applying patterni matching machine to a global suffix tree index .
>
> 2016-08-03 15:09 GMT+02:00 Adrien Grand <jpou...@gmail.com>:
>
>> Lucene cannot run such a query without doing a linear scan over all
>> documents in order to compare the values of both fields, this is why we do
>> not have a query for it. The recommended approach in such cases is to
>> index
>> a third field where the difference between value1 and value2 is computed
>> at
>> index time.
>>
>> Le mer. 3 août 2016 à 14:42, Cristian Lorenzetto <
>> cristian.lorenze...@gmail.com> a écrit :
>>
>> > hi i need a help.
>> >
>> > I want create a lucene query parser for adding evaluation of expression
>> in
>> > lucene
>> > value1:>= value2 where value1 and value2 are 2 field names in the
>> > document.
>> >
>> > how to do it?
>> > for me it is sufficient also to work programatically creating a new
>> Query
>> > Object making it
>> >
>>
>
>
