import sys, csv, os, stat, re

agent = "Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.0.10) Gecko/2009042316 Firefox/3.0.10"

if len(sys.argv) < 3:
	print "syntax: %s list urls" % sys.argv[0]
	sys.exit()

list_csv = csv.reader(open(sys.argv[1]), delimiter='\t')
urls_csv = csv.reader(open(sys.argv[2]), delimiter='\t')

start = 0
if len(sys.argv) >= 4:
	start = int(sys.argv[3])

def dist(a, b):
	la, lb = len(a), len(b)
	c = [[0 for j in range(lb)] for i in range(la)]
	d0 = 0
	if a[0] != b[0]: d0 = 1
	for j in range(lb): c[0][j] = d0 + j
	for i in range(la): c[i][0] = d0 + i
	
	for i in range(1, la):
		for j in range(1, lb):
			ci = c[i-1][j] + 1
			cj = c[i][j-1] + 1
			cij = c[i-1][j-1]
			if a[i] != b[j]: cij += 1

			m = ci
			if m > cj:  m = cj
			if m > cij: m = cij
			c[i][j] = m
	
	return c[la-1][lb-1]

urls = {}
for row in urls_csv:
	url_track, name, url_mp3 = row
	results = urls.get(url_track, [])

	results.append([name, url_mp3])
	urls[url_track] = results

def cleanup(s): return s.lower().replace(" ", "")

rows = [row for row in list_csv]

for pos in range(start, len(rows)):
	row = rows[pos]
	url_track, url_artist, track, artist = row
	results = urls.get(url_track, [])
	list_name = "%s - %s" % (artist, track)

	print "saving %d (%s)" % (pos, list_name)
	
	for rpos in range(len(results)):
		result = results[rpos]
		result.append(rpos)
		result.append(rpos * 2 + dist(cleanup(list_name), cleanup(result[0])))

	results.sort(lambda x, y: cmp(x[3], y[3]))

	for result in results: print [result[0], result[2], result[3]]

	for result in results:
		url = result[1]
		safename = re.sub("[^a-zA-Z0-9-_ ]", " ", result[0])
		outfile = "data/%s.mp3" % safename

		referer = re.search("http://[^/]+", url).group(0)

		os.system('wget "%s" --referer="%s" --user-agent="%s" -O "%s"' % (url, referer, agent, outfile))
		size = 0
		try:
			st = os.stat(outfile)
			size = st[stat.ST_SIZE]
		except:
			size = 0

		if size < 1024*1024:
			os.remove(outfile)
		else:
			print "satisfied"
			break