On Fri, 2005-04-08 at 13:22 -0400, Jeffrey Stedfast wrote:
> +void
> +em_utils_send_receipt (CamelFolder *folder, CamelMimeMessage *message)
> +{
> + /* See RFC #2298 for a description of message receipts */
> +
> + EAccount *account = guess_account (message, folder);
> +
> + CamelMimeMessage *receipt = camel_mime_message_new ();
>
> the extra blank lines should be removed as they will cause compile
> problems with some compilers (e.g. non-gcc)
>
> + /* We use camel_header_msgid_generate () to get a canonical
> + * hostname, then skip the part leading to '@' */
> + fake_msgid = camel_header_msgid_generate ();
> + for (hostname = fake_msgid; hostname && *hostname != '@';
> ++hostname);
> + ++hostname;
>
> you check that hostname != NULL in the for-loop but then never expect
> that it could be NULL afterward?
>
> I don't think it's possible for 'hostname' to ever be NULL so that check
> should be removed...
>
> also, because of the way it's written, that for-loop can easily be
> misread (e.g. it's easy to look at it and think that the following line
> is part of the for-loop if you are just skimming the code)
>
> 2 possible solutions to this might be:
>
> fake_msgid = camel_header_msgid_generate ();
> for (hostname = fake_msgid; *hostname != '@'; hostname++)
> ;
>
> or
>
> hostname = fake_msgid = camel_header_msgid_generate ();
> while (*hostname != '@')
> hostname++;
>
> I personally prefer the latter, but I suppose either would be fine.
er... what was I thinking? easier still to just use strchr()
I've committed a patch to make it simpy sue strchr() in HEAD.
>
> and then I have to wonder if there isn't a better way of getting the
> hostname than generating a msgid that we don't actually use.
>
--
Jeffrey Stedfast
Evolution Hacker - Novell, Inc.
fejj ximian com - www.novell.com
Attachment:
smime.p7s
Description: S/MIME cryptographic signature