FW: commit: ldap/tests/scripts test012-glue defines.sh

["Howard Chu" <hyc@highlandsun.com>]

Expanding on the log message; the original DIT is broken into three subtrees
in data/slapd-glue.conf. The three parts are
  o=University of Michigan,c=us
  ou=Information Technology Division,ou=People,o=University of Michigan,c=us
  ou=Groups,o=University of Michigan,c=US

Each of the subtrees is searched completely before moving on to the next. So,
first all of the entries for "o=Univ..." that are not in the other two subtrees
will be returned. Then all of the entries in the "ou=Info..." subtree come
next.
Finally all the entries in the "ou=Groups..." subtree. If any entries are
returned out of this order, then something broke during the add or the search.

Since sorting the ldapsearch output would obscure these relationships, this
test will only work when the real backend is using a sorted database (e.g. BDB
Btree). (Now that I think about it, perhaps the test should use a smaller data
set that would be independent of backend sort order. That just requires that
most nodes in the DIT have no siblings...)

-----Original Message-----
From: owner-openldap-commit@OpenLDAP.org
[mailto:owner-openldap-commit@OpenLDAP.org] On Behalf Of
hyc@OpenLDAP.org

Update of /repo/OpenLDAP/pkg/ldap/tests/scripts

Modified Files:
	defines.sh  1.50 -> 1.51
Added Files:
	test012-glue  NONE -> 1.1

Log Message:
Added test for backend glue. Tests using slapadd/ldapsearch then starts
over using ldapadd/ldapsearch. Order of results is significant, and
differs from order in data/test-ordered.ldif (thus data/test-glued.ldif
is needed.)

  -- Howard Chu
  Chief Architect, Symas Corp.       Director, Highland Sun
  http://www.symas.com               http://highlandsun.com/hyc
  Symas: Premier OpenSource Development and Support