Finny that you say that, I just wanted to measure if what you are suggesting has a significant impact on performance in general, not for/because of Lucene. It looks like it does not. Attached is a class with a bad name (was used for something else originally), but that will run a few things of this nature and print out the timings. Not a good way to benchmark, but I think it gives an idea.
Otis
--- Julien Nioche <[EMAIL PROTECTED]> wrote:
> Hello,
>
> In the code of BooleanQuery and PhraseQuery loops are made using a
> size()
> method to check the bound. Since the methods of these classes are
> used very
> often, it could be useful to change the code in order to compute the
> size of
> the object only once, store it in an integer and check the bound on
> that
> integer.
>
> For example, the method sumOfSquaredWeights of the class BooleanQuery
> currently use the following :
>
> for (int i = 0 ; i < clauses.size(); i++) {
> ...
> }
>
> which could be replaced by :
>
> int clausenumber = clauses.size();
>
> for (int i = 0 ; i < clausenumber; i++)
> {...}
>
>
> This kind of modification could be done in almost all the methods of
> the
> classes BooleanQuery and PhraseQuery, providing a small optimization
> (I did
> not mesure it - but even small optimizations can be useful).
>
> Supposing that the size of the clauses or terms Vectors is not likely
> to
> change during the execution of these methods, this kind of change
> must be
> pretty harmless...
>
> Any opinion on that? Could it have a side effect?
>
>
> Julien Nioche
> www.lingway.com
>
>
>
>
> --
> To unsubscribe, e-mail:
> <mailto:[EMAIL PROTECTED]>
> For additional commands, e-mail:
> <mailto:[EMAIL PROTECTED]>
>
__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com
SizeTest.java
Description: SizeTest.java
-- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
