"Larry E. Dixson" wrote:
> I agree. I think if one describes (in words) what the searches are
> attempting to "communicate" (and ignore the different Bath Truncation
> attribute values), you can represent all of the Bath searches with our SRW
> CQL strings. We just have to agree on a brief description of each of
> those four (instead of six) index types -- the "?" will take care of the
> two searches that are right truncated in the Bath specification, so
> we don't need brief descriptions for them.
In other words, don't even mention Z39.50 truncation (type 104 or whatever), but
specify that for this particular Bath search (one with right truncation) it can
include the masking character, but only at the end.
Ok, that sounds like the right approach (if I understand correctly). Thanks.