Looks like th new parser performs worse on the exceptional case which is way way OK since this happens less than 1% of the time. Just perfect. I see the 2-2.2x improvement.

Alex

On Mon, Apr 27, 2009 at 2:35 PM, Stefan Seelmann <seelmann@apache.org> wrote:
Hi,

here are some other numbers, using different DNs, on my laptop. The last
one switches to the Antlr based parser and this is muuuuuch slower then
the old parser.


Old Parser:

Parsing '' 100000 times takes 8ms.

Parsing 'cn=test' 100000 times takes 247ms.

Parsing 'CN=Marshall T. Rose, O=Dover Beach Consulting, L=Santa Clara,
ST=California, C=US' 100000 times takes 1449ms.

Parsing 'cn=Jrme,cn=,cn=' 100000 times takes 788ms.

Parsing 'a=\,\=\+\<\>\#\;\\\"\C4\8D,cn=Jrme,cn=,cn='
100000 times takes 1164ms.


New Parser:

Parsing '' 100000 times takes 8ms.

Parsing 'cn=test' 100000 times takes 127ms.

Parsing 'CN=Marshall T. Rose, O=Dover Beach Consulting, L=Santa Clara,
ST=California, C=US' 100000 times takes 708ms.

Parsing 'cn=Jrme,cn=,cn=' 100000 times takes 364ms.

Parsing 'a=\,\=\+\<\>\#\;\\\"\C4\8D,cn=Jrme,cn=,cn='
100000 times takes 6368ms.


Regards,
Stefan


Emmanuel Lecharny schrieb:
> Hi guys,
>
> so i'm back ! I did a very quick test in the train this afternoon,
> comparing the old DN parser and the fast one : the fast one is 2.20
> times faster than the old one for DN like "uid=admin, dc=com".
>
> This is just great !
>
> Thanks Stefan !
>




--
Alex Karasulu
My Blog :: http://www.jroller.com/akarasulu/
Apache Directory Server :: http://directory.apache.org
Apache MINA :: http://mina.apache.org