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

Re: Root DSE

Julio Sánchez Fernández wrote:
> I blush.  I did not read the syntax carefully enough.  From RFC2252,
> section 4.2:
>       AttributeTypeDescription = "(" whsp
>             numericoid whsp              ; AttributeType identifier
>           [ "NAME" qdescrs ]             ; name used in AttributeType
>         [etc.]
> And from 4.1:
>     qdstrings       = qdstring / ( whsp "(" qdstringlist ")" whsp )
> Scrap all that, we have a standard syntax for alias definition.

Good catch... teach me not to go back to the standard...
I suggest now using:

  typedef struct ldap_attributetype {
	 char *at_oid;		/* REQUIRED */
         char **at_names;	/* OPTIONAL */
         char *at_desc;		/* OPTIONAL */
         char *at sup_oid;	/* OPTIONAL */
         char *at_equality_oid;	/* OPTIONAL */
         char *at_ordering_oid;	/* OPTIONAL */
         char *at_substr_oid;	/* OPTIONAL */
         char *at_syntax_oid;	/* OPTIONAL */
         int  *at_single-value; /* 0=no, 1=yes */

> Notice, however, that NAME is optional.

Yes... life with OIDs!  I suggest writing the parser per the spec.
Then, for now, when slapd builds it's internal structure, it can require
a name.  We'll lift this restriction once we have OID support in place.
> I guess you want this because some tools don't need the whole syntax
> bang, right?

It to separate parsing/generation from use.  Parsing/Generation of schema
specification in standard formats is something generally useful to both
LDAP clients and servers.  However, each client/server will likely use
the schema information in different ways.  If there is enough commonality
between uses, some additional utilities could be made sharable.