Re: [art] Seeking clarification on boundary delimiters in RFC 2046

[Ned Freed <[e-mail]>]

> RFC 2046 sec. 5.1 contains the following:

>    NOTE TO IMPLEMENTORS:  Boundary string comparisons must compare the
>    boundary value with the beginning of each candidate line.  An exact
>    match of the entire candidate line is not required; it is sufficient
>    that the boundary appear in its entirety following the CRLF.

> Say a message has the following media type:

>    multipart/mixed; boundary="b1"

> It seems obvious that the following all work the same way as marking the next
> body part in the message:

>    --b1

>    --b1BOUNDARY

>    --b1 (followed by whitespace)

> What is not clear is whether any or all of the following work as
> the boundary delimiter marking the end of all body parts (no example
> is included in RFC 2045 clarifying this matter):

This is why we have ABNF. And the ABNF is very clear - the dash-boundary must
immediately follow the delimiter to be considered to be a closing delimiter:

   close-delimiter := delimiter "--"

>    --b1--

This is a close-delimiter.

>    --b1BOUNDARY--

This is a dash-boundary but not a close-delimiter.

>    --b1--BOUNDARY

This is a close-delimiter.

>    --b1--  (followed by whitespace)

This is a close-delimiter.

>    --b1--BOUNDARY (followed by whitespace)

This is a close-delimiter.

> Can this be clarified?

Why? The ABNF spells this out very clearly.

				Ned