Re: [URN] Re: RFC 2141 on URN Syntax

["Ryan Moats" <[e-mail]>]

On Fri, 09 May 1997 10:08:42 -0600, Ron Daniel, Jr. wrote:

>>I have a similar concern regarding the usage of "%" character within the
>><NSS> subsection. This is what the rule resolution says 
>><NSS> ::= 1*<URN chars>
>><URN chars> ::= <trans> | "%" <hex> <hex>
>><trans> ::= <upper> | <lower> | <number> | <other> | <reserved>
>><reserved> ::= '%" | "/" | "?" | "#"
>>
>>Here we observe that 
>>
>>  <NSS> ::= %    could be a valid rule according to the  rule definition. At
>>the same time the system has another defination that says wheneever "%" is
>>used two  <hex> characters should be used.
>>
>>  <NSS> ::=%%12  -> could also be a valid syntax for NSS.  
>>
>>My suggestion here is to remove "%" from the <reserved> subsection of <NSS>
>>rules.
>>
>>Please let me know if my thinking is along the lines of what the group has
>>in mind or am I way out of wack.
>
>Single occurances of '%' are disallowed. '%' is supposed to ALWAYS mean
>that the next two characters will be hex digits that are taken to
>form an octet.

This is a case of the RFC being ahead of an ABNF specification.
draft-ietf-urn-syntax-05.txt has the ABNF specification "aligned"
with draft-ietf-drums-abnf-02.txt.  There, the definition is

URN chars = trans | ( "%" hex hex)

Otherwise Ron is right.  There is text in RFC 2141 that specifically
says "%" by itself is not allowed and that a "%" character MUST
be followed by two "hex" characters so both NSS containing "%" above
are incorrect syntax.

Ryan