Megaupload-dl
-------------

Megaupload-dl helps on the painful process of downloading files hosted
on the popular Megaupload site (http://www.megaupload.com) if you don't
have a premium account. The process is done from the command line (the
captcha is obtained with an external OCR Tesseract).

= Requeriments =

The script should work on any GNU/Linux and BSD. Make sure you have
installed:

  * Python (>= 2.4): The finest programming language ever! Read more
    impartial opinions at http://www.python.org.  
  
  * Beautiful Soup: A nice HTML parser for Python.  
  
  * Python Image Library: Using Python a great pleasure thanks to this 
    kind of libraries.  
   
  * Tesseract: Optical Character Recognition engine developed at HP 
    and now at Google.
      
  * Some command-line downloader: wget, curl, ...

If you use Debian/Ubuntu, simply:

apt-get install python-beautifulsoup python-imaging tesseract-ocr tesseract-ocr-eng

= Install =

Download the last version source code tarball, extract it and run the
setup script:

python setup.py install

= Usage =

Megaupload-dl returns only the link; use your favorite web downloader
to actually download the file. For example, you could do from the shell:

FILEURL=$(megaupload_dl -vv SOME_MEGAUPLOAD_URL) wget -c "$FILEURL"

Let's say you have a file with a bunch of links to download (one per line):

cat links.txt | xargs -t -n1 megaupload_dl | xargs -d"\n" -r -t wget -c 

= Some comments =

  * Tesseract is a pretty decent OCR, but it will fail some times to get
  the captcha image (though it is really simple). The retry mechanism
  should be enough to make the script still useful.

  * Although the script gets inmediately the file URL from the javascript
  code (run the script with verbose enabled to see the whole process),
  it has to sleep the compulsory time anyway, as any earlier access will
  effectively disable the link.

  * As it is obvious, the script is extremelly fragile: it will stop
  working if Megaupload guys do any change on their web (captcha, URL
  encoding mechanism, ...).

  * The script has been only tested on Debian Sid, but it should also
  work on any other GNU/Linux distros and **BSD. With some effort,
  I imagine you could also run it on Microsoft Windows.

= TODO =

  * Output HTML files (for debug) 
  
  * Use a premium account (but I don't
  have any to test... did anyone get the message?)  
  
  * Eventually extend the script to more downloading pages 
    (Rapidshare, ...) with a plug-in mechanism.  
  
  * GUI

= Contact =

My e-mail: tokland AT gmail.com