Re: bug in testing of message id's

Chris Niekel <[email protected]> wrote:
> Hi,
> 
> Ocassionally, I get a message that some messageid's are illegal, and
> that it has been replaced. I've been searching for the cause, and am
> looking at some strange code in int addr_domain(char* buf) (addr.c)
> 
> The behaviour:
>     [email protected] is good,
>     [email protected] is bad
>     [email protected] is good.
> 
> The code treats the '-' about the same as a '.', and thus doesn't allow
> multiple dashes after another. For the dot ('.') that makes sense, but
> for the '-', it's not right.
> 
> If I change the order of the '.' and the '-' in the case statement, my
> test-cases behave as expected.

Good catch. I've applied a shorter patch which should fix this, see below.

(As you may notice from the patch format, I'm trying out Subversion. :) )


Index: addr.c
===================================================================
--- addr.c      (revision 17)
+++ addr.c      (working copy)
@@ -45,7 +45,7 @@
             if (lb)
                return 0; /* else fall through */
          C('.')
-            if (p == buf || *p == p[-1] || !p[1])
+            if (p == buf || p[-1] == '.' || !p[1])
                 return 0;
              else
                 continue; /* Bugfix: no fall through */


-- 
  Patrik Rådman  ·  patrik at iki dot fi  ·  http://www.iki.fi/patrik/
                "I know this because Tyler knows this."