[Date Prev][Date Next] [Chronological] [Thread] [Top]

Multiple blank lines in LDIF files (ITS#88)



Full_Name: Christian Kirsch
Version: 1.2
OS: Linux
URL: ftp://ftp.openldap.org/incoming/
Submission from: (NULL) (195.35.2.98)


When I have several consecutive lines in an LDIF2 file,
the second one kinda puts an entry into the database. This shows up as
a number without any data when I use ldbmcat on the generated number.
 
However, the LDIF draft says that entries should be separated by
"1*SEP", and RFC 822 explains :
       "1*element" requires at  least  one (element);
which for me means that there can be more than one SEP between
entries.  

Sample LDIF file:

--- Cut here ---
dn: dc=heise, dc=de
objectclass: organization
o: Heinz Heise Verlag

dn: ou=ct, dc=heise, dc=de
objectClass: organizationalUnit
ou: ct

dn: ou=ix, dc=heise, dc=de
objectclass: organizationalUnit
ou: ix

dn: uid=ck, ou=ix, dc=heise, dc=de
objectclass: person
ou: ix
cn: Christian Kirsch
uid: ck

dn: uid=cp, ou=ct, dc=heise, dc=de
objectclass: person
ou: ct
cn: Christian Persson
uid: cp

dn: uid=hb, ou=ix, dc=heise, dc=de
objectclass: person
ou: ix
cn: Henning Behme
uid: hb
------------ Cut Here ---------

ldif2ldbm happily feeds this file into the database. However, 
	ldapsearch "(&(ou=ix)(objectclass=person))"
returns only the first entry (uid=ck).
If I remove the second empty line between "uid: cp" and "dn: uid=hb..." 
use ldif2ldbm again on this file, ldapsearch returns the two entries it 
should (uid=ck and uid=hb).