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

RE: Formal definition of LDAPOID values



At 01:30 PM 1/24/01 +1100, Steven Legg wrote:
>Is there some reason why you can't just reference the numericstring
>production
>in Section 4.1 of RFC 2252 ?

IMO, the reference should be the other way around.  Excepting DNs and
URLs, I belive RFC 2251 should detail syntaxes used directly
in the protocol.  RFC 2252 should define attribute type syntaxes,
pulling in whatever protocol syntaxes it needs to do this by reference.

>If not, then like RFC 2252, you should say that leading zeroes are not
>permitted in a numericstring, or else rewrite the ABNF to disallow them,
>e.g.

I concur.


>numericstring  = "0" / ( non-zero-digit *digit)
>digit          = %x30-39 ; "0" to "9"
>non-zero-digit = %x31-39 ; "1" to "9"

I suggest:
        numericoid = numericstring *( DOT numericstring )
        numericstring = DIGIT / ( LEADDIGIT 1*DIGIT )
        DIGIT = %x30-39 ;  "0"-"9" (digits)
        LEADDIGIT = 0x31-39 ; "1"-"9" (leading digits)
        DOT = %x2E ; "." (period)

Kurt