summaryrefslogtreecommitdiffhomepage
path: root/youtube_dl/extractor/laola1tv.py
blob: 1f91ba0173429aa8b86969b9dc40e45695213e6d (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
# coding: utf-8
from __future__ import unicode_literals

import json

from .common import InfoExtractor
from ..utils import (
    ExtractorError,
    unified_strdate,
    urlencode_postdata,
    xpath_element,
    xpath_text,
    update_url_query,
    js_to_json,
)


class Laola1TvEmbedIE(InfoExtractor):
    IE_NAME = 'laola1tv:embed'
    _VALID_URL = r'https?://(?:www\.)?laola1\.tv/titanplayer\.php\?.*?\bvideoid=(?P<id>\d+)'
    _TESTS = [{
        # flashvars.premium = "false";
        'url': 'https://www.laola1.tv/titanplayer.php?videoid=708065&type=V&lang=en&portal=int&customer=1024',
        'info_dict': {
            'id': '708065',
            'ext': 'mp4',
            'title': 'MA Long CHN - FAN Zhendong CHN',
            'uploader': 'ITTF - International Table Tennis Federation',
            'upload_date': '20161211',
        },
    }]

    def _extract_token_url(self, stream_access_url, video_id, data):
        return self._download_json(
            stream_access_url, video_id, headers={
                'Content-Type': 'application/json',
            }, data=json.dumps(data).encode())['data']['stream-access'][0]

    def _extract_formats(self, token_url, video_id):
        token_doc = self._download_xml(
            token_url, video_id, 'Downloading token',
            headers=self.geo_verification_headers())

        token_attrib = xpath_element(token_doc, './/token').attrib

        if token_attrib['status'] != '0':
            raise ExtractorError(
                'Token error: %s' % token_attrib['comment'], expected=True)

        formats = self._extract_akamai_formats(
            '%s?hdnea=%s' % (token_attrib['url'], token_attrib['auth']),
            video_id)
        self._sort_formats(formats)
        return formats

    def _real_extract(self, url):
        video_id = self._match_id(url)
        webpage = self._download_webpage(url, video_id)
        flash_vars = self._search_regex(
            r'(?s)flashvars\s*=\s*({.+?});', webpage, 'flash vars')

        def get_flashvar(x, *args, **kwargs):
            flash_var = self._search_regex(
                r'%s\s*:\s*"([^"]+)"' % x,
                flash_vars, x, default=None)
            if not flash_var:
                flash_var = self._search_regex([
                    r'flashvars\.%s\s*=\s*"([^"]+)"' % x,
                    r'%s\s*=\s*"([^"]+)"' % x],
                    webpage, x, *args, **kwargs)
            return flash_var

        hd_doc = self._download_xml(
            'http://www.laola1.tv/server/hd_video.php', video_id, query={
                'play': get_flashvar('streamid'),
                'partner': get_flashvar('partnerid'),
                'portal': get_flashvar('portalid'),
                'lang': get_flashvar('sprache'),
                'v5ident': '',
            })

        _v = lambda x, **k: xpath_text(hd_doc, './/video/' + x, **k)
        title = _v('title', fatal=True)

        token_url = None
        premium = get_flashvar('premium', default=None)
        if premium:
            token_url = update_url_query(
                _v('url', fatal=True), {
                    'timestamp': get_flashvar('timestamp'),
                    'auth': get_flashvar('auth'),
                })
        else:
            data_abo = urlencode_postdata(
                dict((i, v) for i, v in enumerate(_v('req_liga_abos').split(','))))
            stream_access_url = update_url_query(
                'https://club.laola1.tv/sp/laola1/api/v3/user/session/premium/player/stream-access', {
                    'videoId': _v('id'),
                    'target': self._search_regex(r'vs_target = (\d+);', webpage, 'vs target'),
                    'label': _v('label'),
                    'area': _v('area'),
                })
            token_url = self._extract_token_url(stream_access_url, video_id, data_abo)

        formats = self._extract_formats(token_url, video_id)

        categories_str = _v('meta_sports')
        categories = categories_str.split(',') if categories_str else []
        is_live = _v('islive') == 'true'

        return {
            'id': video_id,
            'title': self._live_title(title) if is_live else title,
            'upload_date': unified_strdate(_v('time_date')),
            'uploader': _v('meta_organisation'),
            'categories': categories,
            'is_live': is_live,
            'formats': formats,
        }


class Laola1TvIE(Laola1TvEmbedIE):
    IE_NAME = 'laola1tv'
    _VALID_URL = r'https?://(?:www\.)?laola1\.tv/[a-z]+-[a-z]+/[^/]+/(?P<id>[^/?#&]+)'
    _TESTS = [{
        'url': 'http://www.laola1.tv/de-de/video/straubing-tigers-koelner-haie/227883.html',
        'info_dict': {
            'id': '227883',
            'display_id': 'straubing-tigers-koelner-haie',
            'ext': 'flv',
            'title': 'Straubing Tigers - Kölner Haie',
            'upload_date': '20140912',
            'is_live': False,
            'categories': ['Eishockey'],
        },
        'params': {
            'skip_download': True,
        },
    }, {
        'url': 'http://www.laola1.tv/de-de/video/straubing-tigers-koelner-haie',
        'info_dict': {
            'id': '464602',
            'display_id': 'straubing-tigers-koelner-haie',
            'ext': 'flv',
            'title': 'Straubing Tigers - Kölner Haie',
            'upload_date': '20160129',
            'is_live': False,
            'categories': ['Eishockey'],
        },
        'params': {
            'skip_download': True,
        },
    }, {
        'url': 'http://www.laola1.tv/de-de/livestream/2016-03-22-belogorie-belgorod-trentino-diatec-lde',
        'info_dict': {
            'id': '487850',
            'display_id': '2016-03-22-belogorie-belgorod-trentino-diatec-lde',
            'ext': 'flv',
            'title': 'Belogorie BELGOROD - TRENTINO Diatec',
            'upload_date': '20160322',
            'uploader': 'CEV - Europäischer Volleyball Verband',
            'is_live': True,
            'categories': ['Volleyball'],
        },
        'params': {
            'skip_download': True,
        },
        'skip': 'This live stream has already finished.',
    }]

    def _real_extract(self, url):
        display_id = self._match_id(url)

        webpage = self._download_webpage(url, display_id)

        if 'Dieser Livestream ist bereits beendet.' in webpage:
            raise ExtractorError('This live stream has already finished.', expected=True)

        conf = self._parse_json(self._search_regex(
            r'(?s)conf\s*=\s*({.+?});', webpage, 'conf'),
            display_id, js_to_json)

        video_id = conf['videoid']

        config = self._download_json(conf['configUrl'], video_id, query={
            'videoid': video_id,
            'partnerid': conf['partnerid'],
            'language': conf.get('language', ''),
            'portal': conf.get('portalid', ''),
        })
        error = config.get('error')
        if error:
            raise ExtractorError('%s said: %s' % (self.IE_NAME, error), expected=True)

        video_data = config['video']
        title = video_data['title']
        is_live = video_data.get('isLivestream') and video_data.get('isLive')
        meta = video_data.get('metaInformation')
        sports = meta.get('sports')
        categories = sports.split(',') if sports else []

        token_url = self._extract_token_url(
            video_data['streamAccess'], video_id,
            video_data['abo']['required'])

        formats = self._extract_formats(token_url, video_id)

        return {
            'id': video_id,
            'display_id': display_id,
            'title': self._live_title(title) if is_live else title,
            'description': video_data.get('description'),
            'thumbnail': video_data.get('image'),
            'categories': categories,
            'formats': formats,
            'is_live': is_live,
        }