[ic] Fixed rare [read-cookie] bug
David Christensen
david at endpoint.com
Wed Mar 9 20:44:10 UTC 2011
On Mar 9, 2011, at 2:24 PM, Jon Jensen wrote:
> On Wed, 9 Mar 2011, David Christensen wrote:
>
>> I recently hunted down a very obscure set of circumstances in which [read-cookie <name>] can result in an incorrect value for the specific cookie. This came up for a client in which ~ 1% of orders were having an invalid affiliate field, which itself had been populated by this cookie.
>
> Nice find! :)
>
>> Any thoughts/opinions on whether [read_cookie foocookie] should be case-insensitive?
>
> It appears to me that RFC 2965 specifies that the attribute names are to be case-insensitive:
>
> http://www.apps.ietf.org/rfc/rfc2965.html#sec-3.1
>
> If I'm reading that correctly, then the /i option as it stands gives the correct behavior.
Okay, great; I think I was even looking at that document, so not sure why I didn't process that correctly :-). It looks like the parsing could be improved further based on the definition of the LHS, specifically either a token or a quoted-string (as defined by http://www.apps.ietf.org/rfc/rfc2616.html), but it's definitely an improvement as it stands.
Regards,
David
--
David Christensen
End Point Corporation
david at endpoint.com
More information about the interchange-users
mailing list