PHP and downloading..

Dolemite50

Hello,

I have a site with mp3's and I would like to be able to track download stats. I ran in to two question;
1. Is there a way to only count a download if it is fully downloaded?
2. Is there a way to dynamicaly change the filename for download? They are currently stored as12345.mp3, I would like to be able to change it to the song title, is that possible?

Thank you.

drew010

I think this simple file example will answer your questions, and help you make your own. If you have other questions, please feel free to ask.

<?php

header("Content-type: application/octet-stream");
header("Content-disposition: attachment; filename=anythingyouwant.mp3");
header("Content-length: " . filesize("mp3file.mp3"));

echo file_get_contents("mp3file.mp3");

//if the user cancels before the dl finishes, this wont run
$fp = fopen("log.txt", "a");
fwrite($fp, date("r") . " mp3file.mp3\n");
fclose($fp);

?>

Dolemite50

Oh man, that looks fantastic.

So on my page I curretly have a link for "Download this song". Would I have that link to a seperate page that has the above code? Am I in the ballpark with this code,...?

$songid=$_GET['songid'];
$sql = "Select song_title FROM,...."
....
$song_title = mysql_result($query, 0);

header("Content-type: application/octet-stream");
header("Content-disposition: attachment; filename=<?php echo $song_title ?>.mp3");
header("Content-length: " . filesize("<?php echo $songid ?>.mp3"));

echo file_get_contents("<?php echo $songid ?>.mp3");

//if the user cancels before the dl finishes, this wont run
UPDATE TABLE SET counter = counter +1 where songid='$songid'

I'm a little uncertain on the top part, am I using the song title and the song id in the correct spots? In other words, which of those needs the actual file name and which is the new title of the song?

Thank you very much for the help.

drew010

yeah i think it is pretty close.
it is a separate file all by itself. if the user clicks a link their browser will attempt to fetch the page and see that its just trying to get it to dl content so they will actually still be seeing the same page they clicked the link from, they will just get a save as dialog.

the header() calls are actually php code so there is no need for the <?php echo... parts inside. you can just use the variables directly like
header("Content-disposition: attachment; filename=$song_title.mp3");

the file_get_contents and the filesize functions will take the actual mp3 filename as it is on disk, not the $song_title one, the filename= part will tell the browser what to save it as. those functions should point to the location of the file on disk.

after the file get contents part you can just do another mysql_query() doing the update statement to increment the count.

in all it should be a short script, not too much longer than what you have.

i found this script i wrote a while back that i made as part of a file dump where people could upload files and stuff onto a main page and it counted download similar to what you are doing. the main difference is that i actually stored the binary file data in the database. but this should still be useful.

<?php
if(!isset($_GET['postid'])) {
  echo "No file ID specified";
  exit;
}
$postid = $_GET['postid'];
include("./config.inc.php");
dbconnect();

$query = "SELECT * FROM ${dbpref}dump WHERE postid = '$postid'";
$result = mysql_query($query);
$data = mysql_fetch_object($result);
if(!$result) {
echo mysql_error();
}


header("Content-type: application/octet-stream");
header("Content-disposition: attachment; filename=$data->filename");
header("Content-Length: " . strlen($data->filedata));
echo $data->filedata;  //output file data

//here i actually used a cookie to prevent counting a user downloading the same file twice but its optional
if(!isset($_COOKIE['download_' . $postid])) {
  $cupdate = "UPDATE ${dbpref}dlcount SET count = count + 1 WHERE postid='$postid'";
  $cupres = @mysql_query($cupdate);
  $value = "downloaded";
  @setcookie("download_" . $postid,$value,time()+31536000);
}
?>

Dolemite50

Thanks so much Drew,..good stuff. I'm going to go give it a whirl and let you know how it goes.

Dolemite50

I think I'm in the ballpark.

When I click on the Download link, the song title comes up just as it should. (the-name-of-song.mp3)

But when I click on download it goes really fast and the file exixts but it is 1k. Any thoughts?

if (! $userdata['session_logged_in']) 
{ 
$url = "/forum/login.php?redirect=/signin2.php&outside=1";
header("Location: $url"); 
} 
$songid=$_GET['songid']; 
$bandid=$_GET['bandid']; 
$sql5 = "SELECT userid FROM download_tracker WHERE songid ='$songid'";
$result3 = mysql_query($sql5) or die(mysql_error());
if(mysql_num_rows($result3)){
$please = "Please use your browser's back button to return to the Song Page.<br>";
$error = "Sorry! You have already downloaded that song today.<br>$please";
include ('errorinc.php'); 
}
$sql5 = "SELECT stitle, bandName FROM table WHERE songid ='$songid'";
$result3 = mysql_query($sql5) or die(mysql_error());
$stitle= mysql_result($result3, 0);
$stitle = make_url_friendly($stitle);
$mp3 = "/mp3/$songid";

header("Content-type: application/octet-stream"); 
header("Content-disposition: attachment; filename=$stitle.mp3"); 
header("Content-length: " . filesize("$mp3.mp3")); 

echo file_get_contents("$mp3.mp3"); 

$sql5 = "INSERT INTO table2 SET songid ='$songid', bandid='$bandid', userid='$userid', dldate=now()";
$result3 = mysql_query($sql5) or die(mysql_error());

drew010

my first thought is this:

$mp3 = "/mp3/$songid";
and
header("Content-length: " . filesize("$mp3.mp3"));

echo file_get_contents("$mp3.mp3");

$mp3 is a URI, but filesize() and file_get_contents both require absolute paths from your servers root.

id try this:
header("Content-length: " . filesize($_SERVER['DOCUMENT_ROOT'] . "$mp3.mp3"));

echo file_get_contents($_SERVER['DOCUMENT_ROOT'] . "$mp3.mp3");

Dolemite50

Drew,

I'm so sorry man. The first code was perfect. I made a mistake and forgot that I'd updated the mp3 filenames to be $bandid.$songid.mp3 in a seperate column called "mp3".

Once I pointed to $mp3.mp3 instead of songid.mp3 the first code worked like a champ. Sorry to have wasted your time and THANKS!!! This is an awesome feature.

I have one final issue to deal with here if you have the time. Don't sweat it if you don't, you've already been a heap of help.

I'm trying to implemement this system to track downloads per month, per week and all-time. I've already been taking a stab at tracking streaming plays.

Tracking alltime is easy because I just do "dlcounter = dlcounter +1"

But for monthly tracking, I've built a seperate table where it inserts the userid, bandid, songid, and DL_date as now()

I then run a query every night to DELETE all entries >30 days old in the monthly_tracker table. It then sets all the monthly_plays in the song table to 0 WHERE monthly_plays is != 0. THen the script counts all the entries for each GROUPed songid and UPDATES the song table monthly_plays with the result.
The problem is,..even after 5 days of tracking I have over 300,000 entries and the script is incredibly slow. Could you please shed some advice on this matter?

Here is the original post:
http://www.phpbuilder.com/board/showthread.php?s=&threadid=10301825

And THANKS AGAIN!!! 😃 😃 😃 😃 😃

suntra

Originally posted by drew010

<?php
...
echo file_get_contents("mp3file.mp3");

//if the user cancels before the dl finishes, this wont run
$fp = fopen("log.txt", "a");
fwrite($fp, date("r") . " mp3file.mp3\n");
fclose($fp);

?>

I don't think that is true. The script will execute itself, the output will be sent to the webserver cache and the script will terminate very shortly. If the script waited for the user to have downloaded the file, we would get an error about the maximum execution time (30 seconds by defaults). So, with this code, any download will always be successful. PHP isn't able to know if a download was succesful or not.

drew010

Originally posted by suntra
I don't think that is true. The script will execute itself, the output will be sent to the webserver cache and the script will terminate very shortly. If the script waited for the user to have downloaded the file, we would get an error about the maximum execution time (30 seconds by defaults). So, with this code, any download will always be successful. PHP isn't able to know if a download was succesful or not.

its tested and works.
the simplest way to find out is to just try to run it and you will see.
instead of using file_get_contents just run a loop to output a fairly large amount of random data.
cancel it the first few times, then let it finish. you will see the log file isnt created unless you let it finish 100%

Dolemite50

On the script that Drew gave me, the last query
$sql5 = "INSERT INTO table2 SET songid ='$songid', bandid='$bandid', userid='$userid', dldate=now()";
$result3 = mysql_query($sql5) or die(mysql_error());

...won't run if I cancel the download before it finishes.

Thanks again Drew, that is a great chunk of code.

Zoruglu

Don't forget to clean the incomming data.

$songid=$_GET['songid'];  

$bandid=$_GET['bandid'];

would better be:

$songid = intval( $_GET['songid'] );
$bandid = intval( $_GET['bandid'] );

this will enshure the id's are integers, it's more secure when using them in queries 🙂

Dolemite50

Cool tip, thanks. 🙂