[rte] Improve thumbnail extraction (Closes #9085)
[youtube-dl] / youtube_dl / extractor / rte.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from ..utils import (
8     float_or_none,
9     parse_iso8601,
10     unescapeHTML,
11 )
12
13
14 class RteIE(InfoExtractor):
15     IE_NAME = 'rte'
16     IE_DESC = 'Raidió Teilifís Éireann TV'
17     _VALID_URL = r'https?://(?:www\.)?rte\.ie/player/[^/]{2,3}/show/[^/]+/(?P<id>[0-9]+)'
18     _TEST = {
19         'url': 'http://www.rte.ie/player/ie/show/iwitness-862/10478715/',
20         'info_dict': {
21             'id': '10478715',
22             'ext': 'flv',
23             'title': 'Watch iWitness  online',
24             'thumbnail': 're:^https?://.*\.jpg$',
25             'description': 'iWitness : The spirit of Ireland, one voice and one minute at a time.',
26             'duration': 60.046,
27         },
28         'params': {
29             'skip_download': 'f4m fails with --test atm'
30         }
31     }
32
33     def _real_extract(self, url):
34         video_id = self._match_id(url)
35         webpage = self._download_webpage(url, video_id)
36
37         title = self._og_search_title(webpage)
38         description = self._html_search_meta('description', webpage, 'description')
39         duration = float_or_none(self._html_search_meta(
40             'duration', webpage, 'duration', fatal=False), 1000)
41
42         thumbnail = None
43         thumbnail_meta = self._html_search_meta('thumbnail', webpage)
44         if thumbnail_meta:
45             thumbnail_id = self._search_regex(
46                 r'uri:irus:(.+)', thumbnail_meta,
47                 'thumbnail id', fatal=False)
48             if thumbnail_id:
49                 thumbnail = 'http://img.rasset.ie/%s.jpg' % thumbnail_id
50
51         feeds_url = self._html_search_meta('feeds-prefix', webpage, 'feeds url') + video_id
52         json_string = self._download_json(feeds_url, video_id)
53
54         # f4m_url = server + relative_url
55         f4m_url = json_string['shows'][0]['media:group'][0]['rte:server'] + json_string['shows'][0]['media:group'][0]['url']
56         f4m_formats = self._extract_f4m_formats(f4m_url, video_id)
57         self._sort_formats(f4m_formats)
58
59         return {
60             'id': video_id,
61             'title': title,
62             'formats': f4m_formats,
63             'description': description,
64             'thumbnail': thumbnail,
65             'duration': duration,
66         }
67
68
69 class RteRadioIE(InfoExtractor):
70     IE_NAME = 'rte:radio'
71     IE_DESC = 'Raidió Teilifís Éireann radio'
72     # Radioplayer URLs have two distinct specifier formats,
73     # the old format #!rii=<channel_id>:<id>:<playable_item_id>:<date>:
74     # the new format #!rii=b<channel_id>_<id>_<playable_item_id>_<date>_
75     # where the IDs are int/empty, the date is DD-MM-YYYY, and the specifier may be truncated.
76     # An <id> uniquely defines an individual recording, and is the only part we require.
77     _VALID_URL = r'https?://(?:www\.)?rte\.ie/radio/utils/radioplayer/rteradioweb\.html#!rii=(?:b?[0-9]*)(?:%3A|:|%5F|_)(?P<id>[0-9]+)'
78
79     _TESTS = [{
80         # Old-style player URL; HLS and RTMPE formats
81         'url': 'http://www.rte.ie/radio/utils/radioplayer/rteradioweb.html#!rii=16:10507902:2414:27-12-2015:',
82         'info_dict': {
83             'id': '10507902',
84             'ext': 'mp4',
85             'title': 'Gloria',
86             'thumbnail': 're:^https?://.*\.jpg$',
87             'description': 'md5:9ce124a7fb41559ec68f06387cabddf0',
88             'timestamp': 1451203200,
89             'upload_date': '20151227',
90             'duration': 7230.0,
91         },
92         'params': {
93             'skip_download': 'f4m fails with --test atm'
94         }
95     }, {
96         # New-style player URL; RTMPE formats only
97         'url': 'http://rte.ie/radio/utils/radioplayer/rteradioweb.html#!rii=b16_3250678_8861_06-04-2012_',
98         'info_dict': {
99             'id': '3250678',
100             'ext': 'flv',
101             'title': 'The Lyric Concert with Paul Herriott',
102             'thumbnail': 're:^https?://.*\.jpg$',
103             'description': '',
104             'timestamp': 1333742400,
105             'upload_date': '20120406',
106             'duration': 7199.016,
107         },
108         'params': {
109             'skip_download': 'f4m fails with --test atm'
110         }
111     }]
112
113     def _real_extract(self, url):
114         item_id = self._match_id(url)
115
116         json_string = self._download_json(
117             'http://www.rte.ie/rteavgen/getplaylist/?type=web&format=json&id=' + item_id,
118             item_id)
119
120         # NB the string values in the JSON are stored using XML escaping(!)
121         show = json_string['shows'][0]
122         title = unescapeHTML(show['title'])
123         description = unescapeHTML(show.get('description'))
124         thumbnail = show.get('thumbnail')
125         duration = float_or_none(show.get('duration'), 1000)
126         timestamp = parse_iso8601(show.get('published'))
127
128         mg = show['media:group'][0]
129
130         formats = []
131
132         if mg.get('url'):
133             m = re.match(r'(?P<url>rtmpe?://[^/]+)/(?P<app>.+)/(?P<playpath>mp4:.*)', mg['url'])
134             if m:
135                 m = m.groupdict()
136                 formats.append({
137                     'url': m['url'] + '/' + m['app'],
138                     'app': m['app'],
139                     'play_path': m['playpath'],
140                     'player_url': url,
141                     'ext': 'flv',
142                     'format_id': 'rtmp',
143                 })
144
145         if mg.get('hls_server') and mg.get('hls_url'):
146             formats.extend(self._extract_m3u8_formats(
147                 mg['hls_server'] + mg['hls_url'], item_id, 'mp4',
148                 entry_protocol='m3u8_native', m3u8_id='hls', fatal=False))
149
150         if mg.get('hds_server') and mg.get('hds_url'):
151             formats.extend(self._extract_f4m_formats(
152                 mg['hds_server'] + mg['hds_url'], item_id,
153                 f4m_id='hds', fatal=False))
154
155         self._sort_formats(formats)
156
157         return {
158             'id': item_id,
159             'title': title,
160             'description': description,
161             'thumbnail': thumbnail,
162             'timestamp': timestamp,
163             'duration': duration,
164             'formats': formats,
165         }