[Email-SIG] Reconstituting stored messages and quoted printable encoding

[Stuart Bishop]

Hi.

I'm currently looking into persistent email storage for some internal 
applications.

One of our requirements is that we need to be able to produce on demand 
a byte-by-byte identical copy of the original message (we expect GPG 
signed messages to be quite common). My understanding of this means that 
to do this we would either need to store the original message, or split 
it up into chunks but store these chunks in their encoded form. This is 
because quoted printable encoding  does not necessarily round trip:

 >>> '=66=6f=6f'.decode('quopri').encode('quopri')
'foo'

Anyone know if this assumption is correct for email messages? I suspect 
I'm going to have to double our storage requirements and store a copy of 
the original message as well as decoded text for indexing and decoded 
attachments for easy retrieval. Or just decide to break GPG signed 
messages for the pathalogical cases.

-- 
Stuart Bishop <stuart at stuartbishop.net>
http://www.stuartbishop.net/