On Feb 24, 2007, at 2:10 AM, Ben Hood wrote:

> Hi,
>
> I was wondering if anybody knows of any approaches to encode and
> decode large object streams using Hessian.

Cool.  I've added it to a bug report.  It sounds like a good addition.

-- Scott

>
> The motivation is to be able to handle vast amounts of data without
> having to materialize it into memory.
>
> I've written a module to handle this (see attached maven project) as a
> suggestion of how one may do this (with an example test case).
>
> To create an input stream of encoded data, you supply an Iterator
> which wraps your application data source. You also supply the method
> for each data object to be invoked against. On the decoding side, the
> method to be invoked is read, an implementation thereof is registered
> and a stream handler decodes the objects in the stream, fires each one
> against the registered interface until the stream is exhausted.
>
> I've tried to reuse the hessian toolkit as much as possible. The only
> problem I had was the method mangling routine as used by the
> HessianSkeleton is  a private method, so I lifted that source into a
> separate class (homage goes to the author of that code).
>
> Cheers,
>
> Ben
> <hessian-streaming.tar.gz>
> _______________________________________________
> hessian-interest mailing list
> [email protected]
> http://maillist.caucho.com/mailman/listinfo/hessian-interest



_______________________________________________
hessian-interest mailing list
[email protected]
http://maillist.caucho.com/mailman/listinfo/hessian-interest

Reply via email to