you-get/src/you_get/extractors/qq.py

#!/usr/bin/env python

__all__ = ['qq_download']

from ..common import *

import xml.etree.ElementTree as ET
import urllib.parse
import random
import base64
import struct
import uuid

USER_AGENT = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:33.0) Gecko/20100101 Firefox/33.0'
PLAYER_PLATFORM = 11
PLAYER_VERSION = '3.2.18.285'
KLIB_VERSION = '2.0'

def pack(data):
    target = []
    target.extend(struct.pack('>I', data[0]))
    target.extend(struct.pack('>I', data[1]))
    target = [c for c in target]
    return target

def unpack(data):
    data = ''.join([chr(b) for b in data])
    target = []
    data = data.encode('latin')
    target.extend(struct.unpack('>I', data[:4]))
    target.extend(struct.unpack('>I', data[4:8]))
    return target

def tea_encrypt(v, key):
    delta = 0x9e3779b9
    s = 0
    v = unpack(v)
    rounds = 16
    while rounds:
        s += delta
        s &= 0xffffffff
        v[0] += (v[1]+s) ^ ((v[1]>>5)+key[1]) ^ ((v[1]<<4)+key[0])
        v[0] &= 0xffffffff
        v[1] += (v[0]+s) ^ ((v[0]>>5)+key[3]) ^ ((v[0]<<4)+key[2])
        v[1] &= 0xffffffff
        rounds = rounds - 1
    return pack(v)

def qq_encrypt(data, key):
    temp = [0x00]*8
    enc = tea_encrypt(data, key)
    for i in range(8, len(data), 8):
        d1 = data[i:]
        for j in range(8):
            d1[j] = d1[j] ^ enc[i+j-8]
        d1 = tea_encrypt(d1, key)
        for j in range(len(d1)):
            d1[j] = d1[j]^data[i+j-8]^temp[j]
            enc.append(d1[j])
            temp[j] = enc[i+j-8]
    return enc

def strsum(data):
    s = 0
    for c in data:
        s = s*131 + ord(c)
    return 0x7fffffff & s

def ccc(platform, version, timestamp):
    key = [1735078436, 1281895718, 1815356193, 879325047]
    s1 = '537e6f0425c50d7a711f4af6af719e05d41d8cd98f00b204e9800998ecf8427e8afc2cf649f5c36c4fa3850ff01c1863d41d8cd98100b204e9810998ecf84271'
    d = [0x3039, 0x02]
    d.append(timestamp)
    d.append(platform)
    d.append(strsum(version))
    d.append(strsum(s1))
    data = [0xa6, 0xf1, 0xd9, 0x2a, 0x82, 0xc8, 0xd8, 0xfe, 0x43]
    for i in d:
        data.extend([c for c in struct.pack('>I', i)])
    data.extend([0x00]*7)
    enc = qq_encrypt(data, key)
    return base64.b64encode(bytes(enc), b'_-').replace(b'=', b'')

def to_dict(json_object):
    class global_dict(dict):
        def __getitem__(self, key):
            return key
    return eval(json_object, global_dict())

def get_from(url):
    return 'v1001'

def qq_get_final_url(url, fmt_name, type_name, br, sp, vkey, level):
    params = {
        'stdfrom': get_from(url),
        'type': type_name,
        'vkey': vkey,
        'level': level,
        'platform': PLAYER_PLATFORM,
        'br': br,
        'fmt': fmt_name,
        'sp': sp,
    }
    form = urllib.parse.urlencode(params)
    return "%s?%s" % (url, form)

def load_key():
    url = 'http://vv.video.qq.com/checktime'
    tree = ET.fromstring(get_content(url))
    t = int(tree.find('./t').text)
    return ccc(PLAYER_PLATFORM, PLAYER_VERSION, t)

def qq_download_by_vid(vid, title = None, output_dir = '.', merge = True, info_only = False):
    player_pid = uuid.uuid4().hex.upper()
    params = {
        'vids': vid,
        'vid': vid,
        'otype': 'xml',
        'defnpayver': 1,
        'platform': PLAYER_PLATFORM,
        'charge': 0,
        'ran': random.random(),
        'speed': 8096, #random.randint(2048, 8096),
        'pid': player_pid,
        'appver': PLAYER_VERSION,
        'fhdswitch': 0,
        'defn': 'shd',  # default to super hd
        'defaultfmt': 'shd', # default to super hd
        'fp2p': 1,
        'utype': 0,
        'cKey': load_key(),
        'encryptVer': KLIB_VERSION,
    }

    form = urllib.parse.urlencode(params)
    url1 = '%s?%s' % ('http://vv.video.qq.com/getvinfo', form)
    content = get_content(url1, headers = {'User-Agent': USER_AGENT})
    tree = ET.fromstring(content)
    fmt_id = None
    fmt_name = None
    fmt_br = None
    for fmt in tree.findall('./fl/fi'):
        sl = int(fmt.find('./sl').text)
        if sl:
            fmt_id = fmt.find('./id').text
            fmt_name = fmt.find('./name').text
            fmt_br = fmt.find('./br').text

    video = tree.find('./vl/vi')
    filename = video.find('./fn').text
    filesize = video.find('./fs').text

    cdn = video.find('./ul/ui')
    cdn_url = cdn.find('./url').text
    filetype = int(cdn.find('./dt').text)
    vt = cdn.find('./vt').text

    if filetype == 1:
        type_name = 'flv'
    elif filetype == 2:
        type_name = 'mp4'
    else:
        type_name = 'unknown'

    clips = []
    for ci in video.findall('./cl/ci'):
        clip_size = int(ci.find('./cs').text)
        clip_idx = int(ci.find('./idx').text)
        clips.append({'idx': clip_idx, 'size': clip_size})

    size = 0
    for clip in clips:
        size += clip['size']

    user_agent = 'Mozilla/5.0 TencentPlayerVod_1.1.91 tencent_-%s-%s' % (vid, fmt_id)
    fns = os.path.splitext(filename)

    urls =[]
    for clip in clips:
        fn = '%s.%d%s' % (fns[0], clip['idx'], fns[1])
        params = {
            'vid': vid,
            'otype': 'xml',
            'platform': PLAYER_PLATFORM,
            'format': fmt_id,
            'charge': 0,
            'ran': random.random(),
            'filename': fn,
            'vt': vt,
            'appver': PLAYER_VERSION,
            'cKey': load_key(),
            'encryptVer': KLIB_VERSION
        }

        form = urllib.parse.urlencode(params)
        url2 = '%s?%s' % ('http://vv.video.qq.com/getvkey', form)
        content = get_content(url2, headers = {'User-Agent': user_agent})
        tree = ET.fromstring(content)

        vkey = tree.find('./key').text
        level = tree.find('./level').text
        sp = tree.find('./sp').text

        clip_url = '%s%s' % (cdn_url, fn)

        urls.append(qq_get_final_url(clip_url, fmt_name, type_name, fmt_br, sp, vkey, level))

    print_info(site_info, title, type_name, size)
    if not info_only:
        download_urls(urls, title, type_name, size, output_dir = output_dir, merge = merge)

def qq_download(url, output_dir = '.', merge = True, info_only = False):
    content = get_html(url)
    video_info = to_dict(match1(content, r'var\s+VIDEO_INFO\s?=\s?({[^}]+})'))
    vid = video_info['vid']
    title = video_info['title']
    assert title
    qq_download_by_vid(vid, title, output_dir, merge, info_only)

site_info = "QQ.com"
download = qq_download
download_playlist = playlist_not_supported('qq')