Flat File Active Sync changing format from ldif to csv - any tricks?

Hellos.

Currently we maintain an IdM Organization by running a FF Async using an LDIF dump of the source. The source makers do not allow access to their network but kindly ftp files to the IdM host.

The source owners are switching from LDAP to Oracle DB so we will soon be receiving CSV files rather than LDIF.

What I am concerned about is the change-over from ldif to csv.

The first CSV dump could well have a different number of records than the last LDIF dump.

I realise that I *could* empty the ldif source file to delete all current entries and let the first CSV FF AS run recreate them naturally. This will work. But, Can I do better than this?

[689 byte] By [greenfan88] at [2007-11-26 11:40:07]
# 1

Its not so daunting as I expected.

Its a matter of identifying the ids that exist on LDIF but dont exist in first CSV dump. Finding the ids to zap is a matter of a q&d perl script. Just set up a new FFAS resource and manually/bulk delete those ids from IdM.

IdM seems to convert creates to updates (fortunately in this case) so I can keep the LDIF generated entry.

greenfan88 at 2007-7-7 11:41:12 > top of Java-index,Web & Directory Servers,Directory Servers...