Are the results different if the user-supplied parameters are different? Perhaps you could use a proxy front-end module from your web server to provide this function. I believe that Apache mod_rewrite has an example of this for automatically generating files in a cache.

Alternatively, you could save the output, before sending it to the client, as a file based on some unique, per-request generated file name. Do not base the file name on any unsanitized data, or it could become an avenue for exploitation.

A third option might be to store the response text in a blob field in a database.

Each option has benefits and drawbacks, but without further information, it is difficult to even recommend an approach to take.

Thanks for the reply, the users upload a .dat file which gets extracted and parsed and the script spits out a bunch of information based on the content. Since they do have to upload files to the server and it has to be extracted, I have the files uploaded and named based on a sequential counter so that each are unique and I know how many of times it's been used (I have the script delete the files at the end of the script)--the trouble is I don't know how to save the output at all other than brute-force doubling every print statement and pushing it to a file.

How about, instead of printing at every step, collecting the data to print, and then printing to both locations destinations in a single location?

Update: minor wording change

--MidLifeXis

I'm not sure if this is able to be done due to the nature of the input which is a file ... I tried something like: ... However the output html file I get just gives me my invalid file message so it doesn't look like it takes the input in that way.

The input is not a file, but an HTTP request. The HTTP request content includes the file embedded in it.

As you are suggesting running the script again using the command-line, this might be the simplest approach. Most (Plack-based) modern Perl web frameworks provide a command-line tool for doing this, so if you are using one of those, then check the documentation for it. For generic web servers such as a LAMP setup, you can use wget, curl, or lwp-download. Check the documentation on those for how to POST a request with a file attached. Windows has similar tools available. Watch out for recursion if you do this!

To avoid running the script twice, a more generic approach is to use a filter/handler. Again, this depends on the web framework, as any framework that buffers output can make this very easy to do. For Apache, check out: http://httpd.apache.org/docs/current/handler.html#examples

Would something like File::Tee or even the following work?

open (STDOUT, "| tee -ai log.txt");
[download]