Update: the upload hook that you pass into new() only takes effect on file uploads, not on generic POST requests.

Here's a little script that lets you test it. You will soon notice that the upload hook - while working as intended - is completely useless. The reason for this is that under regular CGI, you simply don't get to send output to the browser before the entire post has been received. To give feedback while the upload is in progress, you must trick the webserver into letting you, and that's why Apache2::UploadProgress is a much better tool than rolling your own.

use strict;
use warnings;
use CGI;

my $hook_called;
my $cgi = CGI->new( \&_hook );

print $cgi->header(),
      $cgi->start_html("FOO"),
      $cgi->start_multipart_form,
      $cgi->filefield('file'),
      $cgi->submit,
      $cgi->end_form;

sub _hook
{
    my ( $filename, $buffer, $bytes_read, $data ) = @_;
    if( !$hook_called )
    {
        print "Content-Type: text/html; charset=ISO-8859-1\n\n";    # 
+$cgi is not defined here
        $hook_called = 1;
    }
    else
    {
        print "Read $bytes_read bytes of filename\n";
    }
    $hook_called += $bytes_read;
}

print $cgi->end_html;
[download]

Even without change, your code outputs a telling warning.

Adding use strict makes the error crystal clear.

Update: Apparently, that's not the only problem. I can't figure it out at the moment.

I don't think there's any provision in CGI.pm's new function for a coderef as argument. It won't be called, it will just be stringified and taken as a field name.

Calling new with arguments overrides GET or POST input, leaving requests unhandled. That may be confusing your idea of what's happening.

Clearly, you're looking for a callback to handle some sort of runtime condition - at least that's what your question sounds like. It's probably one of our famous X-Y questions. What are you really trying to accomplish?

I don't think there's any provision in CGI.pm's new function for a coderef as argument.

You should have checked the docs before replying.

You can set up a callback that will be called whenever a file upload is being read during the form processing. This is much like the UPLOAD_HOOK facility available in Apache::Request, with the exception that the first argument to the callback is an Apache::Upload object, here it's the remote filename.

$q = CGI->new(\&hook [,$data [,$use_tempfile]]);

sub hook
{
   my ($filename, $buffer, $bytes_read, $data) = @_;
   print  "Read $bytes_read bytes of $filename\n";         
}
[download]

In my CGI.pm, that passage reads,

           You can set up a callback that will be called whenever a
       file upload is being read during the form processing. This
       is much like the UPLOAD_HOOK facility available in
       Apache::Request, with the exception that the first
       argument to the callback is an Apache::Upload object, here
       it's the remote filename.

        $q = CGI->new();
        $q->upload_hook(\&hook,$data);

        sub hook
        {
               my ($filename, $buffer, $bytes_read, $data) = @_;
               print  "Read $bytes_read bytes of $filename\n";
        }

The hook is not an argument to new, but is set in the upload_hook method.

After Compline,
Zaxo

I want to be able to upload a file and perform an action within the _hook subroutine every 4000 bytes or so.

Calling new with arguments overrides GET or POST input, leaving requests unhandled.

Just to niggle, CGI->new will handle whatever you call it with, including any GET or POST input:

if ($ENV{'REQUEST_METHOD'} eq "GET" and $ENV{'QUERY_STRING'}) {
  (my $query = $ENV{'QUERY_STRING'}) =~ s/$munge//;
   my $q = CGI->new($query);
}
[download]