Incorrect file extension detection

Thanks for the patch. The problem with this whole thing is that users can name their files what ever they please and it is really hard to figure out what is valid and what needs to be removed. I took a different approach than you did and removed all strings that match '.12345' assuming that file extensions are always 5 or smaller (eg: the improper mpeg4). This not perfect- really, we should use mime detect's list of file extensions and strip off anything that matches the mime detect functions, but there isn't currently a function in mime detect to do this (see: http://drupal.org/node/342189). Here's the solution I came up with:

  // file name for demonstration purposes
  $current_file = 'myvideo.mynewvid.mov.avi.flv.mpeg4';
  // Construct the file name
  // we need to strip out any existing file extensions. The
  // problem with this approach is that it assumes that extension names
  // are always 5 or less characters. It could be the case that there
  // is a file name like: myvideo.char.mov.avi.flv.mpeg4 which
  // will have 'char' removed.
  $file_name_parts = explode('.', $current_file);
  foreach ($file_name_parts as $part) {
    if (strlen($part) >= 5) {
      $final_name[] = $part;      
    }
    else{
      break;
    }
  }
  $output_file = implode('.', $final_name);
  print $output_file;

Any thoughts?

Also, would be great if you could use the Transliteration module to normalize filenames:
http://drupal.org/project/transliteration

ckng- where you say:

Let's say for filename myvideo.mov.flv or myvideo.mov.avi.flv, 1st 'extention' .mov is detected and the file may not be processed

From what I understand of your concerns here, you want all extensions removed, not just the last one, correct? In this case we either have to guess what the extensions are, or we have to just use the last one.

So here's my propsal: we just do str_replace('.', '_', $filename) and append $new_extension on. This way we leave the "filename" intact, while avoiding the above concern.

How's that sound?

So would you be happy with something like:

  $file = 'myvideo.flv.sometext.avi.mov';
  $pattern = '/(.*)\..*$/';
  $replace = '$1.flv';
  print_r(preg_replace($pattern, $replace, $file));

While your solution is totally workable, I feel like this kind of approach is a bit cleaner- granted regexs are harder to read, but I think keeps things a bit more obvious

Slight change to change white space to '_' as well. Fix is committed to D6 and D5. Thanks for the help!