cgi.py performance

[Jon Ribbens]

In article <a8i60e$csg$1 at web1.cup.hp.com>, jimh wrote:
>> You could try jon.cgi from jonpy, it might be more efficient.
>> http://jonpy.sourceforge.net/
> 
> Hey, that looks pretty cool.  I do have one question before I try it out: in
> mime.py, Entity.__init()__, which splits up the data on the boundaries,
> loads the entire file at once, if I am not mistaken ("msg = fp.read()").
> What will happen if I am transferring a very large file, 50-100MB?

In theory: your computer uses as much RAM as it has, then starts using
swap space instead. So it's like using a temporary file except it uses
as much RAM as possible first, which makes it more efficient.

That's the idea anyway. I haven't tested it with huge files because
I've never had the need to use even reasonably large files. I'd be
very interested in your experiences if you try it.