[turba] How do Turba LDAP filters work?

[Kevin Myer]

I'm trying to resolve some performance issues with my LDAP address book lookups.
 I have specified the following filter:

objectclass=person

However, from my LDAP server logs, the filter that is being used is not that
but:

filter="(|(objectClass=top)(objectClass=person)(objectClass=inetOrgPerson)(objectClass=calEntry)(objectClass=organizationalPerson))"

Those object classes are what are in my objectclass array in my Turba
sources.php config, so I'm a little confused as to what the purpose of the
objectclass array is.  It appears to actually be ANDing any filter value I
specify, with the objectclass array, but then whats the purpose of the filter
value in sources.php?

And I'm finding that for all the results returned (approximately 1400 users),
there is a search occuring against base="uid=uid,ou=people.." for each user. 
This is very inefficient.

What should be generated:

SRCH base="ou=People,dc=iu13,dc=k12,dc=pa,dc=us" scope=1
filter="(objectClass=person)" attrs="dn uid cn mail sn title o postalAddress
telephoneNumber facsimileTelephoneNumber homePostalAddress homePhone mobile
description roomNumber ou displayName labeledUri object_pgppublickey calFBURL
l"

What is happening:

(the equivalent of foreach the_uid):

SRCH base="uid=the_uid, ou=People, dc=iu13, dc=k12, dc=pa, dc=us" scope=0
filter="(|(objectClass=person))"
attrs="dn uid cn mail sn title o postalAddress telephoneNumber
facsimileTelephoneNumber homePostalAddress homePhone mobile description
roomNumber ou displayName labeledUri object_pgppublickey calFBURL l"

The former is one search, with 1400 results and takes .239s .

The latter is 1400 searches with 1 result per search and takes over 4 seconds.

Kevin

-- 
Kevin M. Myer
Senior Systems Administrator
Lancaster-Lebanon Intermediate Unit 13  http://www.iu13.org