GitLab

liblber:ber_bvarray_{add,free}.

leftover ber_bvdup eliminated; rewritten ber_bvdup in terms of ber_dupbv; eliminated wrapper macro; other ber_* improvements

Misc cleanup / lint removal

Fix nameAndOptionalIdentifer syntax.

now dnValidate loops thru each value; added private member to LDAPAVA to hold the attribute description

liberally accept many LDAPv2/LDAPv3 stuff in DN (quoted parts, ';' as rdn separator and so); enforce exact case match of normalized dn in dnMatch (let's see what happens ...); fix a ldapava_free_rdn bug

Add flag to matching rules to indicate if DN folding is needed.
Don't specify dnPretty as pretty function for DN attribute types as
this works too well (it breaks tests).

A pretty routine may rewrite the representation of a value but
must not alter the value itself.

another step towards schema aware normalization: ava sorting in rdns; now by defining USE_LDAP_DN_PARSING both the new dn parsing and the server side normalization is used. There might be, every now and then, a flaw mostly due to naive normalization of pseudo-dn (thinking about some acl/regex stuff and so)

reimplement dn_validate/normalize/(match) using the ldap_str2dn/dn2str stuff; need to #define USE_LDAP_DN_PARSING