Re: [Boost-users] [serialization] deserializing chunk by chunk

Subject: Re: [Boost-users] [serialization] deserializing chunk by chunk
From: Jeff Flinn (Jeffrey.Flinn_at_[hidden])
Date: 2012-12-31 08:17:00

On 12/24/2012 5:33 PM, Arash Abghari wrote:
> Hi,
>
> I have an encrypted archive. I can decrypt the file chunk by chunk. I
> wonder if there is a way to deserialize each chunk on the fly. The main
> purpose is to reduce the memory footprint of the application since the
> archive can be a few hundred mega bytes. The problem as I mentioned is
> the file is encrypted and it has to be decrypted first in the memory and
> then deserialized.

Have you looked at boost iostreams library? You can create a filter to 
provide a decrypted stream which is then used by the input archive. See:

http://www.boost.org/doc/libs/1_52_0/libs/iostreams/doc/index.html

Jeff