#!/usr/bin/python

"""
Little script to convert Realaudio stream to mp3 & download podcasts

To use it you need:
- mplayer with win32 codecs
- lame to encode the wav in mp3
- please set the path to store the mp3 files

Copyright Jkx 2004 / Licence GPL

Modifications by James Minchin (http://chaucery.com/)
8-Feb-2006
- updated deprecated mplayer parameters
- changed some Unix functions to Python (OS-neutral) functions
- added support for .ra streams

25-Feb-2006
- fixed path parameter (temp file still where script runs though)

15-Oct-2006
- fix stream url containing containing querystring, ie. '?'

9-Apr-2007
- added direct naive podcast downloads (gets first entry in feed)
- added Ogg Vorbis encoding option

10-Apr-2007
- added optional parameter to set output filename

1-Jul-2007
- fixed podcast to always download first audio found

4-Nov-2007
- fix podcast filenames containing '?'
- add optional parameter to set filename prefix on podcasts
- fix Real streams containing '&'

25-Nov-2007
- fix podcast title containing '/'
- abort when temp filename already in use (eg. when 2 simultanous
streams using same name

"""


import os, sys, string, time 


path = "e:/multimedia/rip" #os.getcwd()   # destination folder
mplayer = "e:/multimedia/mplayer/mplayer" # location of MPlayer
lame = "e:/multimedia/lame/lame"          # location of Lame mp3 encoder
ogg = "e:/multimedia/oggenc2.exe"         # location of Ogg Vorbis encoder


hr = "*" * 78
def rip(url, filename):
    """ rip stream """
    os.system('%s -cache 320 -vc dummy -vo null -ao pcm:waveheader:file=%s "%s"' % (mplayer, filename, url))
    
def encode(filename, path):
    """ encode stream with lame and delete it """
    dateBit = time.strftime("%Y%b%d", time.localtime())
    dest= os.path.join(path, filename[:-4] + '_' + dateBit + '.mp3')
    # changed from 64bps constant to variable bit rate of quality 8 (9 = lowest)
    #os.system('%s -b 64 %s %s' % (lame, filename, dest) )
    os.system('%s -V 8 --vbr-new %s %s' % (lame, filename, dest) )
    #os.system('%s -q 0 -o %s %s' % (ogg, dest, filename) )
    os.remove(filename)


def parseSource(sourceFile):
    if sourceFile.startswith('http://'):
        # this is a url not a file
        import urllib
        f=urllib.urlopen(sourceFile)
    else:
        try:
            f = open(sourceFile)
        except IOError:
            print "Unable to open %s" % ramFile
            return None,None
    
    data = f.read()
    f.close()
    url = None

    import re

    #==== look for mp3 podcast
    matches = re.match(".*?<item>[^<]*<title\s*>(.*?)</title>.*?<enclosure[^>]+url=\"([^\"]*)\".*?</item>", data, re.DOTALL)
    if matches != None:
        title = matches.group(1)
        url = matches.group(2)
        print "Found podcast '%s' at\n%s" % (title, url)
        
        #==== fix title so it only contains valid chars for saving file
        #==== workaround url containing weird ampersand encoding - (for CBC Spark podcast)
        return url.replace("&amp;","&"), title.replace("/","_").replace(" ","_").replace(":","_").replace("__","_").replace("?","")

    #==== no podcast, check for Real stream instead
    urlMatches = re.match(".*(rtsp://.*)", data)
    if urlMatches != None:
        url = urlMatches.group(1).strip()

    if url:
        print 'Found an rtsp stream at:\n%s' % url

        filename=string.split(url,'/')[-1]
        qPos = filename.find('?')
        if qPos <> -1:
            filename = filename[:qPos]
        if filename.endswith('.rm') or filename.endswith('.ra'): 
            return url,filename[:-3]+'.wav'
        else:
            print "Unable to find the .ram file"
            return None,None

    else:
        print "Unable to find the podcast or rtsp stream"
    return None,None
        
def reportHook(block_count, block_size, total_size):
    total_kb = total_size/1024
    print "%d kB of %d kB downloaded\r" % (block_count * (block_size/1024), total_kb ),
    
def main():
    print hr
    url,filename = parseSource(sys.argv[1])

    if url and filename:
            
        if url.endswith(".mp3"):
            if len(sys.argv) == 3:
                filename = sys.argv[2] + "_" + filename
            filename = os.path.join(path, filename) + '.mp3'
            print hr
            print "Downloading to %s..." % filename
            import urllib
            # socket.setdefaulttimeout() 
            urllib.urlretrieve(url, filename, reporthook=reportHook)
            print "...Completed."
            print hr
        else:
            if len(sys.argv) == 3:
                filename = sys.argv[2] + '.wav'

            if os.path.exists(filename):    
                print "ERROR: temporary file %s already exists" % (filename)
                sys.exit(1)
            else:    
                print "Ripping file to %s" % (os.path.join(path,filename))
                print hr
                rip(url,filename)
                print hr
                encode(filename, path)
                print "SUCCESS: Download complete"
                print hr
        
    else:
        print "ERROR: Could not find audio to download"
        print hr


def usage():
    me = sys.argv[0]
    (head, tail) = os.path.split(me)
    print "Convert realaudio to mp3 or download latest mp3 podcast - usage:\n"
    print "    %s infile [outfile]\n" % tail
    print "infile: a file or url of a podcast or Real audio"
    print "outfile: optional filename to save to\n"
    print "eg. %s ramfile.ram AppleWeek" % tail
    print "    %s http://XXXX/stream.ram" % tail 
    print "    %s http://XXXX/podcast.xml\n" % tail 
    print "Files will be stored in %s" % path

if __name__ == '__main__':
    if len(sys.argv) < 2: 
        usage()
    else:
        main()