cgi download and delete file

[teddyttas]

Hi Monks!

I would like to have your advice on the following problem:

I am making a small webapp and my purpose is to create a number of text files, compress them, download as zip file and erase.

For the moment I've been able to create the directory with the file, put them in a zip file and to download, but I can't erase the zip file. Well..It's not totally true. The problem is that if I do it, it compromises the download. My question would then be : Ho can I postpone the file elimination or, how can I check that the file has been totally completed before I erase it?

I know about the existence of File::Temp, but as I need this zip file just once, I think it would be simpler to just delete it. Here is a part of the code

sub write_and_download{

    [...]

    $self->generate_zip();

    $self->download_file();

    my $filename = "temporary/" . $projectname . ".zip";
    unlink $filename;

};

sub download_file{
my $self = shift;
    my $downloadfile = $projectname . ".zip";
    my $fullpathdownloadfile = "temporary/$downloadfile";
    my $output = '';
    my $buffer = '';
    open my $fh, '<', $fullpathdownloadfile
      or return error_handler("Error: Failed to download file <b>$down
+loadfile</b>:<br>$!<br>");

    while (my $bytesread = read($fh, $buffer, 1024)) {
      $output .= $buffer;
    }

    close $fh
      or return error_handler("Error: Failed to download file <b>$down
+loadfile<b>:<br>$!<br>");

    my  $downloadfilesize = (stat($fullpathdownloadfile))[7]
      or return error_handler("Error: Failed to get file size for <b>$
+downloadfile<b>:<br>$!<br>");
      
    $self->header_props(
                        '-type'                => 'application/x-downl
+oad',
                        '-content-disposition' => "attachment;filename
+=$downloadfile",
                        '-content_length'      => $downloadfilesize,
                       );
    return $output;
};

I thank you from now for your help

[poj]

You could create the zip 'on-the-fly' so no file to delete

#!perl
use strict;
use Archive::Zip;
use CGI qw(:standard);

my $zip = Archive::Zip->new();
$zip->addFile('c:/temp/inet/test.pdf','test.pdf');

print header(-type=>'application/zip',-attachment=>'download.zip');
binmode(STDOUT);
$zip->writeToFileHandle(*STDOUT);

poj

[flexvault]

teddyttas,

I have to guess that another script is going to use the zip file you create, since you don't call a subroutine or use the zip file in any way. A couple things can be done:

• Let the other script delete the zip file.
• Delete old zip files when creating new zip files. (see 'stat')
• Call the other script before doing the 'unlink'.
• Use 'sleep' to wait to do the 'unlink'.
• etc. etc. etc.
I think this comes down to "you know what you want to do", but we don't. Maybe a description of the timing consideration would help give you a better answer.

Regards...Ed

[sundialsvc4]

You know that Private SNAFU is always on duty:   the mere fact that you zipped-up a bunch of files and pushed them to an HTML stream doesn’t mean that they arrived, any more than it does at the Post Office.   I would therefore design this program to build the Zip files, then remove the files from whence they were made (if appropriate), then put that file into a “to be downloaded” directory under some appropriate name.   From these, when the user requests a download, the file is renamed in some slight way to indicate that it “has been or is-being downloaded.”   (Renaming ahead-of-time avoids any file contention issues.   Moving it to a different directory ahead-of-time does the same thing.)   But in every step you have thus given yourself a way out.   If the download did not for whatever reason succeed, the data is still there and the process could be repeated.

[teddyttas]

Thank you very much! Anyway, I think there is something I am getting wrong... When I use the cgi::application's header_prop (in the download_file subroutine) what I'm actually doing is a redirection to the download window, is it right? Cause I tried to make a redirection after the execution of the subroutine, but in this case the download doesn't start. If I put a sleep command after the subroutine, the script executes before the sleep command and then launches the download. So...I don't really understand this behaviour. I tried :

     $self->download_zip();
    sleep 5;
    my $filename = "temporaneo/" . $self->session->param("nom_projet")
+ . ".zip";
    unlink $filename;

and

    
    $self->generate_zip();
    $self->download_zip();
    
    return $self->redirect("http://delete.cgi");

What I was trying to do was following your advice: put the zip file in another directory and, if succeed, redirect to another run mode that erases the file, otherwise stay on the page and relaunch the download.

[teddyttas]

Okay, I think I got the solution to my last question... it was simple. In fact I solved it in the following way...

 
    $self->generate_zip();
    my $output = $self->download_zip();
    #here I can do what I want. unlink and so on..
    my $filename = "temporaneo/" . $self->session->param("nom_projet")
+ . ".zip";
    unlink $filename;    

    #and only then I stream the output
    return $output;

I can now redirect to another runmode passing the $output variable and solve the problem as you suggested. thank you very much!

[Anonymous Monk]

If you're using a variant of Unix for the web server, you can delete the file as soon as the download is initiated. The OS won't fully delete the file as long as a process has it open. (Of course, resuming the download will not work.)