I think he means programmatically (ldapsearch is a command, not a routine; at least that's how I differentiate between them). It is difficult to work with the OpenLDAP schema when using a library API such as JNDI because the data is all under the cn=subschema entry which is why the novell library is good because there are routines that let you easily parse/retrieve information from the schema so it doesn't all get returned as 1 large entry that needs manual parsing. I realize though that using cn=subschema is the standard but, again, when working with JNDI it can be difficult to work with.
--On Wednesday, August 30, 2006 9:23 PM -0400 Michael B Allen <email@example.com> wrote:
On Wed, 30 Aug 2006 17:03:14 -0700 "Kurt D. Zeilenga" <Kurt@OpenLDAP.org> wrote:
slapd(8) supports the LDAP subschema discovery mechanism detailed in RFC 4512.
Actually I think I'm going to opt for shipping a file of known binary attributes. Searching the schema is clearly not portable.
Also, I'm just curious by why didn't you present the schema information like all other information in the directory so that it's searchable using regular ldap routines?
What are you talking about? The schema is easily searched by ldapsearch, etc...