[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Randy Presuhn <rpresuhn@dorothy.bmc.com>] Re: Call forcensensus on path forward

>>>>> On Wed, 09 Oct 2002 12:12:54 -0700, Wes Hardaker <hardaker@tislabs.com> said:

Randy> Since the eos list won't let me post, perhaps one of you might
Randy> consider forwarding a strawman proposed filter syntax:

That's very close to what I was considering putting in.  The only
thing that I was going to do differently would be to not use an OBJECT
IDENTIFY to specify attribute values, at least in the PDUs I've
defined and rather use the straight suffix encoding beyond the base
instead for better compression.

Randy> initialString: true when the specified attribute's value begins
Randy> with the specified octets (meaningful only for OCTET STRINGs)
Randy> subString: true when the specified attribute's value has the specified
Randy> octet string as a substring (meaningful only for OCTET STRINGs)
Randy> finalString: true when the specified attribute's value ends with
Randy> the specified string (meaningful only for OCTET STRINGs)

These are easily replaced with a more expressive regexp syntax.  I'd
rather stick to regexp (and if people only wanted to support a subset
of regexp, we can talk, but it's probably just as easy to say
/^something/ or /something/ or /something$/ for the above 3 choices.

Randy> present: the specified attribute instance exists

Thanks for reminding me of that.  I forgot about the SQL equivalent of
"is null" (and the negation of "is not null").

Wes Hardaker
Network Associates Laboratories