CGI download filename weirdness

JohannShunt has asked for the wisdom of the Perl Monks concerning the following question:

Hi all, I have an otherwise perfectly functioning PDF download script which is producing some serious weirdness on the filename front: despite the headers having the Content-disposition set to "attachment", with a correct filename declared, the filename always comes out when saving the file in a browser as "http__<domain name minus the TLD>.pdf". Anyone got any ideas as to what could be causing this? The problem code:

open (PDF_FILE,"<$pdffile") || die "Could not open file: $pdffile";
print $cgi->header(-type=>"application/pdf",-attachment=>"$pdfid.pdf")
+; # set the appropriate header
while (<PDF_FILE>) { # send the PDF bitstream to the client
  print;
}
close PDF_FILE;
[download]

Thanks.

Comment on CGI download filename weirdness Download Code

Replies are listed 'Best First'.
Re: CGI download filename weirdness by ikegami (Patriarch) on Feb 08, 2010 at 16:58 UTC
It works for me, for `$pdfid = "foo"`. What browser are you using? `.pdf` aren't text files, so you don't want them mangled. Add `binmode(PDF_FILE);`. `.pdf` aren't text files, so trying to read lines makes no sense. Add `local $/ = \(64*1024);`.	[reply] [d/l] [select]
Re: CGI download filename weirdness by kennethk (Abbot) on Feb 08, 2010 at 15:54 UTC
What is in `$pdffile`? What is in `$pdfid`? Are they supposed to be different? According to the documentation, specifically Creating a Standard HTTP Header, the suggested name is specified with the `-attachment` parameter.	[reply] [d/l] [select]
Re^2: CGI download filename weirdness by Anonymous Monk on Feb 08, 2010 at 17:18 UTC
Thanks for the response. $pdffile is a path to the "physical" file, and $pdfid is the name of that file, without the extension. I just tried pulling out this code into a separate script and giving these variables hard-coded values, and it works fine there, so it seems to be caused by something elsewhere than in the snippet I posted. This little PDF subroutine is part of a monstrous pile of code, thousands of lines long, and a real nightmare to debug. A perhaps related issue is that on some browsers, the download never quite gets wrapped up properly - it gets left with a ".part" extension which needs to be removed manually in order to use an otherwise fine file. Still digging away at this...	[reply]
Re^3: CGI download filename weirdness by kennethk (Abbot) on Feb 08, 2010 at 17:29 UTC
I was trying to imply that `$pdfid` does not contain what you think it does. You should be able to do a simple file search for `$pdfid` to see where it was last modified. I highly doubt this is a browser issue.	[reply] [d/l] [select]
Re^4: CGI download filename weirdness by JohannShunt (Novice) on Feb 08, 2010 at 21:23 UTC