[ietf-dkim] canonicalized null body and dkim

Tue Jan 9 03:23:44 PST 2007

On Mon, 08 Jan 2007 17:27:49 -0000, Eric Allman <eric+dkim at sendmail.org>  
wrote:

> --On January 8, 2007 11:51:13 AM +0000 Charles Lindsey  
> <chl at clerew.man.ac.uk> wrote:

>> Indeed there is no ambiguity in that, but that is because you have
>> only  quoted half the text. The full text is:
>>
>>     The "simple" body canonicalization algorithm ignores all empty
>> lines
>>     at the end of the message body.  An empty line is a line of zero
>>     length after removal of the line terminator.  If there is no
>> trailing
>>     CRLF on the message, a CRLF is added.  It makes no other
>> changes to
>>     the message body.  In more formal terms, the "simple" body
>>     canonicalization algorithm converts "0*CRLF" at the end of the
>> body
>>     to a single "CRLF".
>>
>> Observe carefully that the text some times tells you to consider
>> the  "message", and somtimes the "message body" (which I take to
>> mean exactly  the <body>, if any, defined by RFC 2822).
>
> OK, I'll change the single instance of "message" to "message body".

Well that certainly fixes the case I gave, but then the question arises as  
to whether it has fixed it in the "right" direction. I still, cannot see  
why one would want to canonicalize a genuinely empty <body> intpo "CRLF",  
nor do some others on this list. I gather that some implementations do not  
do so either, and neither did the original DK, and many on the list  
thought they were standardizing the DK behaviour.

Moreover, there remains another case that is ambiguous. Consider:

      Field: foobar<CRLF>
      .<CRLF>

That is a valid RFC 2822 message with NO <body> at all (which is NOT the  
same thing as an empty <body>). Let us apply your revised wording.

    The "simple" body canonicalization algorithm ignores all empty lines
    at the end of the message body.

There is no body, so no action is needed.

    An empty line is a line of zero
    length after removal of the line terminator.

Not needed.

    If there is no trailing
    CRLF on the message BODY, a CRLF is added.

That is your modified wording. There is no <body> so no action occurs.  
(But it would have been the same with your original wording.)

    It makes no other changes to
    the message body.

There is no <body>, so there is nothing to not make any changes to. But we  
are still finished.

So what do we pass to the canonicalization? It doesn't say, but the only  
reasonable intpretation would be to pass <empty>. So it appears that an  
absent body canonicalizes differently to an empty body.

Now let us compare that with the final sentence:

    In more formal terms, the "simple" body
    canonicalization algorithm converts "0*CRLF" at the end of the body
    to a single "CRLF".

Since there is no <body>, there is nothing to do, so it indeed agrees with  
the first 4 sentences in this case.

But we still have the bizarre situation that an absent body is treated  
differently from an empty body. Can you please confirm that this was your  
intention?

>
>> Consider the example, in DATA format:
>>
>>     Field: foobar<CRLF>.
>>     <CRLF>
>>     <CRLF>
>>     .<CRLF>

> As Hector points out, this is not a valid 2822 message to begin with, ...

Sure, there was a typo, as various people spotted.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131                       
   Web: http://www.cs.man.ac.uk/~chl
Email: chl at clerew.man.ac.uk      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5