#A utility script to fetch all of the mp3s linked to by the given
#Web page.
#
#Usage: python fetchMp3s.py <Base URL> <Save Directory>
#where <Base URL> is an absolute URL pointing to a Web page and
#<Save Directory> is the location for the fetched files. Currently,
#if <Base URL> points to a directory, you must ensure that it ends with
#a "/" for this script to function properly.
#
#David Faden
#dfaden@iastate.edu
#Spring 2004

from HTMLParser import HTMLParser
from os.path import basename, expanduser, join
from sets import Set
from sys import argv
from urllib import urlopen, urlretrieve
from urlparse import urljoin

class Mp3Fetcher(HTMLParser):
	"""I fetch all of the MP3s from the given page, storing them in
	   the given directory."""
	
	def __init__(self, baseUrl, saveDirectory):
		HTMLParser.__init__(self)
		self.baseUrl = baseUrl
		self.saveDirectory = expanduser(saveDirectory)
		self.fetchedUrls = Set()
		self.feed(urlopen(baseUrl).read())
	
	def handle_starttag(self, tag, attrs):
		if ("a" == tag):
			href = dict(attrs).get("href", "")
			if href.endswith(".mp3"):
				#This method seems to require that if baseUrl refers
				#to a directory, then it should end with a slash.
				#Otherwise, the last directory is dropped.
				#XXX! How should we fix this?
				mp3Url = urljoin(self.baseUrl, href)
				if mp3Url not in self.fetchedUrls:
					self.fetchedUrls.add(mp3Url)
					localFilename = join(self.saveDirectory, basename(href))
					urlretrieve(mp3Url, localFilename)
					print "Fetched ", href
					
if __name__ == '__main__':
    if len(argv) == 3:
	   Mp3Fetcher(argv[1], argv[2])
    else:
	   print "Usage: python fetchMp3s.py <Base URL> <Save Directory>"