]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/youku.py
Update changelog
[youtubedl] / youtube_dl / extractor / youku.py
index 48d47a24556666df12f3a1ea4c0d1f1d6a51a06b..e37f237c76c6880eb1d442e8302dcef558d0e9d8 100644 (file)
 # coding: utf-8
-
 from __future__ import unicode_literals
 
-import math
+import base64
+import itertools
 import random
 import re
+import string
 import time
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_urllib_parse_urlencode,
+    compat_ord,
+)
 from ..utils import (
     ExtractorError,
+    get_element_by_attribute,
 )
 
 
 class YoukuIE(InfoExtractor):
+    IE_NAME = 'youku'
+    IE_DESC = '优酷'
     _VALID_URL = r'''(?x)
         (?:
             http://(?:v|player)\.youku\.com/(?:v_show/id_|player\.php/sid/)|
             youku:)
         (?P<id>[A-Za-z0-9]+)(?:\.html|/v\.swf|)
     '''
-    _TEST = {
-        'url': 'http://v.youku.com/v_show/id_XNDgyMDQ2NTQw.html',
-        'md5': 'ffe3f2e435663dc2d1eea34faeff5b5b',
+
+    _TESTS = [{
+        # MD5 is unstable
+        'url': 'http://v.youku.com/v_show/id_XMTc1ODE5Njcy.html',
+        'info_dict': {
+            'id': 'XMTc1ODE5Njcy_part1',
+            'title': '★Smile﹗♡ Git Fresh -Booty Music舞蹈.',
+            'ext': 'flv'
+        }
+    }, {
+        'url': 'http://player.youku.com/player.php/sid/XNDgyMDQ2NTQw/v.swf',
+        'only_matching': True,
+    }, {
+        'url': 'http://v.youku.com/v_show/id_XODgxNjg1Mzk2_ev_1.html',
+        'info_dict': {
+            'id': 'XODgxNjg1Mzk2',
+            'title': '武媚娘传奇 85',
+        },
+        'playlist_count': 11,
+        'skip': 'Available in China only',
+    }, {
+        'url': 'http://v.youku.com/v_show/id_XMTI1OTczNDM5Mg==.html',
+        'info_dict': {
+            'id': 'XMTI1OTczNDM5Mg',
+            'title': '花千骨 04',
+        },
+        'playlist_count': 13,
+    }, {
+        'url': 'http://v.youku.com/v_show/id_XNjA1NzA2Njgw.html',
+        'note': 'Video protected with password',
+        'info_dict': {
+            'id': 'XNjA1NzA2Njgw',
+            'title': '邢義田复旦讲座之想象中的胡人—从“左衽孔子”说起',
+        },
+        'playlist_count': 19,
         'params': {
-            'test': False
+            'videopassword': '100600',
         },
+    }, {
+        # /play/get.json contains streams with "channel_type":"tail"
+        'url': 'http://v.youku.com/v_show/id_XOTUxMzg4NDMy.html',
         'info_dict': {
-            'id': 'XNDgyMDQ2NTQw_part00',
-            'ext': 'flv',
-            'title': 'youtube-dl test video "\'/\\ä↭𝕐'
+            'id': 'XOTUxMzg4NDMy',
+            'title': '我的世界☆明月庄主☆车震猎杀☆杀人艺术Minecraft',
+        },
+        'playlist_count': 6,
+    }]
+
+    def construct_video_urls(self, data):
+        # get sid, token
+        def yk_t(s1, s2):
+            ls = list(range(256))
+            t = 0
+            for i in range(256):
+                t = (t + ls[i] + compat_ord(s1[i % len(s1)])) % 256
+                ls[i], ls[t] = ls[t], ls[i]
+            s = bytearray()
+            x, y = 0, 0
+            for i in range(len(s2)):
+                y = (y + 1) % 256
+                x = (x + ls[y]) % 256
+                ls[x], ls[y] = ls[y], ls[x]
+                s.append(compat_ord(s2[i]) ^ ls[(ls[x] + ls[y]) % 256])
+            return bytes(s)
+
+        sid, token = yk_t(
+            b'becaf9be', base64.b64decode(data['security']['encrypt_string'].encode('ascii'))
+        ).decode('ascii').split('_')
+
+        # get oip
+        oip = data['security']['ip']
+
+        fileid_dict = {}
+        for stream in data['stream']:
+            if stream.get('channel_type') == 'tail':
+                continue
+            format = stream.get('stream_type')
+            fileid = stream['stream_fileid']
+            fileid_dict[format] = fileid
+
+        def get_fileid(format, n):
+            number = hex(int(str(n), 10))[2:].upper()
+            if len(number) == 1:
+                number = '0' + number
+            streamfileids = fileid_dict[format]
+            fileid = streamfileids[0:8] + number + streamfileids[10:]
+            return fileid
+
+        # get ep
+        def generate_ep(format, n):
+            fileid = get_fileid(format, n)
+            ep_t = yk_t(
+                b'bf7e5f01',
+                ('%s_%s_%s' % (sid, fileid, token)).encode('ascii')
+            )
+            ep = base64.b64encode(ep_t).decode('ascii')
+            return ep
+
+        # generate video_urls
+        video_urls_dict = {}
+        for stream in data['stream']:
+            if stream.get('channel_type') == 'tail':
+                continue
+            format = stream.get('stream_type')
+            video_urls = []
+            for dt in stream['segs']:
+                n = str(stream['segs'].index(dt))
+                param = {
+                    'K': dt['key'],
+                    'hd': self.get_hd(format),
+                    'myp': 0,
+                    'ypp': 0,
+                    'ctype': 12,
+                    'ev': 1,
+                    'token': token,
+                    'oip': oip,
+                    'ep': generate_ep(format, n)
+                }
+                video_url = \
+                    'http://k.youku.com/player/getFlvPath/' + \
+                    'sid/' + sid + \
+                    '_00' + \
+                    '/st/' + self.parse_ext_l(format) + \
+                    '/fileid/' + get_fileid(format, n) + '?' + \
+                    compat_urllib_parse_urlencode(param)
+                video_urls.append(video_url)
+            video_urls_dict[format] = video_urls
+
+        return video_urls_dict
+
+    @staticmethod
+    def get_ysuid():
+        return '%d%s' % (int(time.time()), ''.join([
+            random.choice(string.ascii_letters) for i in range(3)]))
+
+    def get_hd(self, fm):
+        hd_id_dict = {
+            '3gp': '0',
+            '3gphd': '1',
+            'flv': '0',
+            'flvhd': '0',
+            'mp4': '1',
+            'mp4hd': '1',
+            'mp4hd2': '1',
+            'mp4hd3': '1',
+            'hd2': '2',
+            'hd3': '3',
         }
-    }
+        return hd_id_dict[fm]
+
+    def parse_ext_l(self, fm):
+        ext_dict = {
+            '3gp': 'flv',
+            '3gphd': 'mp4',
+            'flv': 'flv',
+            'flvhd': 'flv',
+            'mp4': 'mp4',
+            'mp4hd': 'mp4',
+            'mp4hd2': 'flv',
+            'mp4hd3': 'flv',
+            'hd2': 'flv',
+            'hd3': 'flv',
+        }
+        return ext_dict[fm]
 
-    def _gen_sid(self):
-        nowTime = int(time.time() * 1000)
-        random1 = random.randint(1000,1998)
-        random2 = random.randint(1000,9999)
-
-        return "%d%d%d" %(nowTime,random1,random2)
-
-    def _get_file_ID_mix_string(self, seed):
-        mixed = []
-        source = list("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\:._-1234567890")
-        seed = float(seed)
-        for i in range(len(source)):
-            seed  =  (seed * 211 + 30031) % 65536
-            index  =  math.floor(seed / 65536 * len(source))
-            mixed.append(source[int(index)])
-            source.remove(source[int(index)])
-        #return ''.join(mixed)
-        return mixed
-
-    def _get_file_id(self, fileId, seed):
-        mixed = self._get_file_ID_mix_string(seed)
-        ids = fileId.split('*')
-        realId = []
-        for ch in ids:
-            if ch:
-                realId.append(mixed[int(ch)])
-        return ''.join(realId)
+    def get_format_name(self, fm):
+        _dict = {
+            '3gp': 'h6',
+            '3gphd': 'h5',
+            'flv': 'h4',
+            'flvhd': 'h4',
+            'mp4': 'h3',
+            'mp4hd': 'h3',
+            'mp4hd2': 'h4',
+            'mp4hd3': 'h4',
+            'hd2': 'h2',
+            'hd3': 'h1',
+        }
+        return _dict[fm]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
+        video_id = self._match_id(url)
 
-        info_url = 'http://v.youku.com/player/getPlayList/VideoIDS/' + video_id
+        self._set_cookie('youku.com', '__ysuid', self.get_ysuid())
+
+        def retrieve_data(req_url, note):
+            headers = {
+                'Referer': req_url,
+            }
+            headers.update(self.geo_verification_headers())
+            self._set_cookie('youku.com', 'xreferrer', 'http://www.youku.com')
 
-        config = self._download_json(info_url, video_id)
+            raw_data = self._download_json(req_url, video_id, note=note, headers=headers)
 
-        error_code = config['data'][0].get('error_code')
-        if error_code:
-            # -8 means blocked outside China.
-            error = config['data'][0].get('error')  # Chinese and English, separated by newline.
-            raise ExtractorError(error or 'Server reported error %i' % error_code,
-                expected=True)
+            return raw_data['data']
 
-        video_title = config['data'][0]['title']
-        seed = config['data'][0]['seed']
+        video_password = self._downloader.params.get('videopassword')
 
-        format = self._downloader.params.get('format', None)
-        supported_format = list(config['data'][0]['streamfileids'].keys())
+        # request basic data
+        basic_data_url = 'http://play.youku.com/play/get.json?vid=%s&ct=12' % video_id
+        if video_password:
+            basic_data_url += '&pwd=%s' % video_password
 
-        # TODO proper format selection
-        if format is None or format == 'best':
-            if 'hd2' in supported_format:
-                format = 'hd2'
+        data = retrieve_data(basic_data_url, 'Downloading JSON metadata')
+
+        error = data.get('error')
+        if error:
+            error_note = error.get('note')
+            if error_note is not None and '因版权原因无法观看此视频' in error_note:
+                raise ExtractorError(
+                    'Youku said: Sorry, this video is available in China only', expected=True)
+            elif error_note and '该视频被设为私密' in error_note:
+                raise ExtractorError(
+                    'Youku said: Sorry, this video is private', expected=True)
             else:
-                format = 'flv'
-            ext = 'flv'
-        elif format == 'worst':
-            format = 'mp4'
-            ext = 'mp4'
-        else:
-            format = 'flv'
-            ext = 'flv'
-
-        fileid = config['data'][0]['streamfileids'][format]
-        keys = [s['k'] for s in config['data'][0]['segs'][format]]
-        # segs is usually a dictionary, but an empty *list* if an error occured.
-
-        files_info=[]
-        sid = self._gen_sid()
-        fileid = self._get_file_id(fileid, seed)
-
-        #column 8,9 of fileid represent the segment number
-        #fileid[7:9] should be changed
-        for index, key in enumerate(keys):
-            temp_fileid = '%s%02X%s' % (fileid[0:8], index, fileid[10:])
-            download_url = 'http://k.youku.com/player/getFlvPath/sid/%s_%02X/st/flv/fileid/%s?k=%s' % (sid, index, temp_fileid, key)
-
-            info = {
-                'id': '%s_part%02d' % (video_id, index),
-                'url': download_url,
-                'uploader': None,
-                'upload_date': None,
-                'title': video_title,
-                'ext': ext,
-            }
-            files_info.append(info)
+                msg = 'Youku server reported error %i' % error.get('code')
+                if error_note is not None:
+                    msg += ': ' + error_note
+                raise ExtractorError(msg)
+
+        # get video title
+        title = data['video']['title']
+
+        # generate video_urls_dict
+        video_urls_dict = self.construct_video_urls(data)
+
+        # construct info
+        entries = [{
+            'id': '%s_part%d' % (video_id, i + 1),
+            'title': title,
+            'formats': [],
+            # some formats are not available for all parts, we have to detect
+            # which one has all
+        } for i in range(max(len(v.get('segs')) for v in data['stream']))]
+        for stream in data['stream']:
+            if stream.get('channel_type') == 'tail':
+                continue
+            fm = stream.get('stream_type')
+            video_urls = video_urls_dict[fm]
+            for video_url, seg, entry in zip(video_urls, stream['segs'], entries):
+                entry['formats'].append({
+                    'url': video_url,
+                    'format_id': self.get_format_name(fm),
+                    'ext': self.parse_ext_l(fm),
+                    'filesize': int(seg['size']),
+                    'width': stream.get('width'),
+                    'height': stream.get('height'),
+                })
+
+        return {
+            '_type': 'multi_video',
+            'id': video_id,
+            'title': title,
+            'entries': entries,
+        }
+
+
+class YoukuShowIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?youku\.com/show_page/id_(?P<id>[0-9a-z]+)\.html'
+    IE_NAME = 'youku:show'
+
+    _TEST = {
+        'url': 'http://www.youku.com/show_page/id_zc7c670be07ff11e48b3f.html',
+        'info_dict': {
+            'id': 'zc7c670be07ff11e48b3f',
+            'title': '花千骨 未删减版',
+            'description': 'md5:578d4f2145ae3f9128d9d4d863312910',
+        },
+        'playlist_count': 50,
+    }
+
+    _PAGE_SIZE = 40
+
+    def _find_videos_in_page(self, webpage):
+        videos = re.findall(
+            r'<li><a[^>]+href="(?P<url>https?://v\.youku\.com/[^"]+)"[^>]+title="(?P<title>[^"]+)"', webpage)
+        return [
+            self.url_result(video_url, YoukuIE.ie_key(), title)
+            for video_url, title in videos]
+
+    def _real_extract(self, url):
+        show_id = self._match_id(url)
+        webpage = self._download_webpage(url, show_id)
+
+        entries = self._find_videos_in_page(webpage)
+
+        playlist_title = self._html_search_regex(
+            r'<span[^>]+class="name">([^<]+)</span>', webpage, 'playlist title', fatal=False)
+        detail_div = get_element_by_attribute('class', 'detail', webpage) or ''
+        playlist_description = self._html_search_regex(
+            r'<span[^>]+style="display:none"[^>]*>([^<]+)</span>',
+            detail_div, 'playlist description', fatal=False)
+
+        for idx in itertools.count(1):
+            episodes_page = self._download_webpage(
+                'http://www.youku.com/show_episode/id_%s.html' % show_id,
+                show_id, query={'divid': 'reload_%d' % (idx * self._PAGE_SIZE + 1)},
+                note='Downloading episodes page %d' % idx)
+            new_entries = self._find_videos_in_page(episodes_page)
+            entries.extend(new_entries)
+            if len(new_entries) < self._PAGE_SIZE:
+                break
 
-        return files_info
+        return self.playlist_result(entries, show_id, playlist_title, playlist_description)