From f2b9bed34652812d570486266fe7b9681cc0f1e4 Mon Sep 17 00:00:00 2001
From: jackyzy823 <jackyzy823@gmail.com>
Date: Mon, 1 Sep 2014 20:19:42 +0800
Subject: [PATCH] fix support for iqiyi #334

some explanation:
    1.Download work should be complete in 10 minutes because urls are generated before downloading and the key may be expired after 10 minutes.You may retry to continue downloading
---
 src/you_get/extractors/iqiyi.py | 100 ++++++++++++++++++++++++++------
 1 file changed, 83 insertions(+), 17 deletions(-)

diff --git a/src/you_get/extractors/iqiyi.py b/src/you_get/extractors/iqiyi.py
index 0bfec350..53f81ad5 100644
--- a/src/you_get/extractors/iqiyi.py
+++ b/src/you_get/extractors/iqiyi.py
@@ -3,8 +3,62 @@
 __all__ = ['iqiyi_download']
 
 from ..common import *
+from uuid import uuid4
+from random import random,randint
+import json
+from math import floor
+import hashlib
+
+'''
+com.qiyi.player.core.model.def.DefinitonEnum
+bid meaning for quality
+0 none
+1 standard
+2 high
+3 super
+4 suprt-high
+5 fullhd
+10 4k
+96 topspeed
+
+'''
+
+
+def getVRSXORCode(arg1,arg2):
+    loc3=arg2 %3
+    if loc3 == 1:
+        return arg1^121
+    if loc3 == 2:
+        return arg1^72
+    return arg1^103
+
+
+def getVrsEncodeCode(vlink):
+    loc6=0
+    loc2=''
+    loc3=vlink.split("-")
+    loc4=len(loc3)
+    # loc5=loc4-1
+    for i in range(loc4-1,-1,-1):
+        loc6=getVRSXORCode(int(loc3[loc4-i-1],16),i)
+        loc2+=chr(loc6)
+    return loc2[::-1]
+
+def getVMS(tvid,vid,uid):
+    tm=randint(1000,2000)
+    vmsreq='http://cache.video.qiyi.com/vms?key=fvip&src=p'+"&tvId="+tvid+"&vid="+vid+"&vinfo=1&tm="+str(tm)+"&enc="+hashlib.new('md5',bytes('ts56gh'+str(tm)+tvid,"utf-8")).hexdigest()+"&qyid="+uid+"&tn="+str(random())
+    return json.loads(get_content(vmsreq))
+
+def getDispathKey(rid):
+    tp=")(*&^flash@#$%a"  #magic from swf
+    time=json.loads(get_content("http://data.video.qiyi.com/t?tn="+str(random())))["t"]
+    t=str(int(floor(int(time)/(10*60.0))))
+    return hashlib.new("md5",bytes(t+tp+rid,"utf-8")).hexdigest()
+
 
 def iqiyi_download(url, output_dir = '.', merge = True, info_only = False):
+    gen_uid=uuid4().hex
+
     html = get_html(url)
 
     tvid = r1(r'data-player-tvid="([^"]+)"', html)
@@ -12,26 +66,38 @@ def iqiyi_download(url, output_dir = '.', merge = True, info_only = False):
     assert tvid
     assert videoid
 
-    info_url = 'http://cache.video.qiyi.com/vj/%s/%s/' % (tvid, videoid)
-    info = get_html(info_url)
-    raise NotImplementedError('iqiyi')
+    info = getVMS(tvid,videoid,gen_uid)
 
-    from xml.dom.minidom import parseString
-    doc = parseString(info_xml)
-    title = doc.getElementsByTagName('title')[0].firstChild.nodeValue
-    size = int(doc.getElementsByTagName('totalBytes')[0].firstChild.nodeValue)
-    urls = [n.firstChild.nodeValue for n in doc.getElementsByTagName('file')]
-    assert urls[0].endswith('.f4v'), urls[0]
+    title = info["data"]["vi"]["vn"]
 
-    for i in range(len(urls)):
-        temp_url = "http://data.video.qiyi.com/%s" % urls[i].split("/")[-1].split(".")[0] + ".ts"
-        try:
-            response = request.urlopen(temp_url)
-        except request.HTTPError as e:
-            key = r1(r'key=(.*)', e.geturl())
-        assert key
-        urls[i] += "?key=%s" % key
 
+    #for highest qualities
+    bid=0
+    for i in info["data"]["vp"]["tkl"][0]["vs"]:
+        if int(i["bid"])<=10 and int(i["bid"])>=bid:
+            bid=int(i["bid"])
+            video_links=i["fs"] 
+    #todo support choose quality  with cmdline
+
+    urls=[]
+    size=0
+    for i in video_links:
+        vlink=i["l"]
+        # print(vlink)
+        if not vlink.startswith("/"):
+            #vlink is encode
+            vlink=getVrsEncodeCode(vlink)
+        assert vlink.endswith(".f4v")
+        size+=i["b"]
+        key=getDispathKey(vlink.split("/")[-1].split(".")[0])
+        baseurl=info["data"]["vp"]["du"].split("/")
+        baseurl.insert(-1,key)
+        url="/".join(baseurl)+vlink+'?su='+gen_uid+'&client=&z=&bt=&ct=&tn='+str(randint(10000,20000))
+        urls.append(json.loads(get_content(url))["l"])
+
+    #download should be complete in 10 minutes
+    #because the url is generated before start downloading
+    #and the key may be expired after 10 minutes
     print_info(site_info, title, 'flv', size)
     if not info_only:
         download_urls(urls, title, 'flv', size, output_dir = output_dir, merge = merge)