[Dovecot] New full text search indexer

[Charles Marcus]

Timo Sirainen wrote:
> As described earlier
> (http://dovecot.org/list/dovecot/2006-December/018055.html), Dovecot
> nowadays has full text search indexing support in CVS HEAD.

<snip>

> So it takes somewhat more space, but definitely less than having both
> Squat + Lucene. 
> 
> No substring indexing, words up to 10 chars (Lucene replacement):
> 
> Index time: 1.39 CPU seconds (22.28MB/CPUs)
> Node memory: 3297663 bytes (3220 kB) in 548540 nodes
> UID memory: 1803735 bytes (1761 kB) in 213554 lists
> Total: 5101398 / 32474899 bytes = 15.71%
> 
> Is there a need for Lucene anymore with this kind of a speed and disk
> space usage? :)

<snip>

Wow - as usual impressive work, Timo! I wish I was skilled enough to 
help you stress test this, but alas, I'm not... I'll have to just wait 
and rely on those more capable. But it looks to make a huge difference 
in Clients that use the new 'Virtual Folder' technology, where the 
folders are based on multiple search criteria. I manage one Courier-imap 
server (am still trying to convince the boss to let me switch to 
dovecot, but he wants to wait until 1.0 has been released and out for a 
while, and on that server, people that have lots of messages in a single 
folder (they use Thunderbird) and create simple Virtual Folders - well, 
lets just say that they can't use them. Clicking on the Virtual Folder 
always takes forever and more often than not results in a time-out error.

Dovecot is already better than Courier in this respect, but it sounds 
like this will make it even better...

Thanks again!

-- 

Best regards,

Charles