Re: [PATCH] Adequately read mailboxes with s/@/ at / obfuscations
* On 06 May 2010, Bertrand Yvain wrote:
> On Tue, May 04, 2010 at 10:55:46PM -0500, David Champion wrote:
> > Adequately read mailboxes with s/@/ at / obfuscations.
>
> This kind of substitution is not reversible in the general case.
> Consider for instance this e-mail address: "works at home"@some.dom.ain
Of course. That is one of the reasons (digital signature invalidation
is another[1]) for only handling rfc822 parsing in headers, and not messing
with bodies. If you look at the code or try the patch, you see that it
handles your example in exactly the same way as unpatched mutt does.
This patch does not disrupt any signature system that relies on rfc822
address headers, because it does not modify the source text of the
message. It only changes what is parsed into the in-memory header
structure: what you can interact with in mutt. The point is to be able
to read a mangled mailbox -- currently mutt fails miserably -- without
changing it on disk.
> Address obfuscation/mangling cause much more problems that they solve
> (if any). The actual problem is the mangling
I agree completely, but I can't stop people from mangling. I can undo
their damage.
> and I don't think mutt
> should jump through hoops in order to handle them.
I just want to be able to read the mailboxes that manglers have created.
--
-D. dgc@xxxxxxxxxxxx IT Services University of Chicago