[Top] [All Lists]

Re: NULL vs. ""

2005-07-14 18:24:39

[cleaning my mailbox of unanswered suggestions...]

Barry Leiba <leiba(_at_)watson(_dot_)ibm(_dot_)com> writes:
On Mon, 2005-05-30 at 09:21 -0700, Ned Freed wrote:
if anyof (header :is "Cc" "", not exists "Cc")

This is actually clearer to read IMO, and is probably how I would code it if
I ever wanted such a test.

Indeed.  And so why don't we wrap up this too-long discussion this way?:
1. *Document* that string comparisons against headers MUST NOT match 
absent headers.

I've updated 5.7p3, second sentence to now read:
    However, if the named header is not present, it does not match any
    key, including the empty key.

2. *Document* that the preferred way to test for a header that is either 
empty or absent is as Ned wrote above.

I've added as a new paragraph at the end of 5.7:
   The preferred way to test whether a given header is either empty or
   absent is to combine an "exists" test and a "header" test:

           anyof (header :is "Cc" "", not exists "Cc")

Philip Guenther

<Prev in Thread] Current Thread [Next in Thread>