TwistedでRedTubeのビデオを全てダウンロードする

む。誰も興味ないのか。
じゃあ全部ダウンロードする奴書いてみるか。
一応20〜30同時に落としても怒られません。
100とかになるとわからんけど。
403が出るのでそいつをケツに突っ込みます。
あとはretryってやつはログ代わりに出してるだけです。

from twisted.web import client
from twisted.internet import reactor

URL = 'http://dl.redtube.com/_videos_t4vn23s9jc5498tgj49icfj4678/%s/%s.flv'

map = ["R", "1", "5", "3", "4", "2", "O", "7", "K", "9", 
        "H", "B", "C", "D", "X", "F", "G", "A", "I", "J", 
        "8", "L", "M", "Z", "6", "P", "Q", "0", "S", "T", 
        "U", "V", "W", "E", "Y", "N"]


def get_video_id(id):
    dir_id = str(int(id / 1000)).zfill(7)
    id = str(id).zfill(7)
    my_int = 0
    for x in xrange(7):
        my_int = my_int + int(id[x]) * (x + 1)

    my_char = str(my_int)
    my_int = 0
    for x in my_char:
        my_int = my_int + int(x)
    new_char = str(my_int).zfill(2)
    mapping = ''
    mapping = mapping + map[ord(id[3]) - 48 + my_int + 3]
    mapping = mapping + new_char[1]
    mapping = mapping + map[ord(id[0]) - 48 + my_int + 2]
    mapping = mapping + map[ord(id[2]) - 48 + my_int + 1]
    mapping = mapping + map[ord(id[5]) - 48 + my_int + 6]
    mapping = mapping + map[ord(id[1]) - 48 + my_int + 5]
    mapping = mapping + new_char[0]                          
    mapping = mapping + map[ord(id[4]) - 48 + my_int + 7]
    mapping = mapping + map[ord(id[6]) - 48 + my_int + 4]
    return dir_id, mapping

ids = []

retry = []

def download(list):
    if len(list):
        id = list.pop(0)
        downloadFlv(list, id, ('/Users/xxxxxx/video/%s.flv' % str(id).zfill(8)))
    else:
        reactor.stop()

def downloadFlv(list, id, file):
   
    def finish(data):
        print "%s finished." % id
        print "retry %s " % retry
        download(list)     
    
    def err(e):
        print id, e
        r = e.value
        if r:
            status = r.status
            if status == 403 or status == '403':
                ids.append(id)
                retry.append(id)
        download(list)

    url = URL % get_video_id(id)
    print id, url
    client.downloadPage(url, file, supportPartial=1).addCallback(finish).addErrback(err);

def start():
    ids = range(1,9000)
    for x in xrange(20):
        download(ids)

if __name__ == '__main__':
    start()
    reactor.run()
 

まあ9000とか適当ですけど。
2000ファイルぐらいで20Gいかないぐらいなので一日で全部落とせるはずです。
うくく。