Merge branch 'the-daily-show-podcast' of https://github.com/fstirlitz/youtube-dl...
[youtube-dl] / youtube_dl / extractor / mixcloud.py
1 from __future__ import unicode_literals
2
3 import re
4 import itertools
5
6 from .common import InfoExtractor
7 from ..compat import (
8     compat_urllib_parse,
9 )
10 from ..utils import (
11     ExtractorError,
12     HEADRequest,
13     str_to_int,
14 )
15
16
17 class MixcloudIE(InfoExtractor):
18     _VALID_URL = r'^(?:https?://)?(?:www\.)?mixcloud\.com/([^/]+)/([^/]+)'
19     IE_NAME = 'mixcloud'
20
21     _TESTS = [{
22         'url': 'http://www.mixcloud.com/dholbach/cryptkeeper/',
23         'info_dict': {
24             'id': 'dholbach-cryptkeeper',
25             'ext': 'mp3',
26             'title': 'Cryptkeeper',
27             'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.',
28             'uploader': 'Daniel Holbach',
29             'uploader_id': 'dholbach',
30             'thumbnail': 're:https?://.*\.jpg',
31             'view_count': int,
32             'like_count': int,
33         },
34     }, {
35         'url': 'http://www.mixcloud.com/gillespeterson/caribou-7-inch-vinyl-mix-chat/',
36         'info_dict': {
37             'id': 'gillespeterson-caribou-7-inch-vinyl-mix-chat',
38             'ext': 'mp3',
39             'title': 'Caribou 7 inch Vinyl Mix & Chat',
40             'description': 'md5:2b8aec6adce69f9d41724647c65875e8',
41             'uploader': 'Gilles Peterson Worldwide',
42             'uploader_id': 'gillespeterson',
43             'thumbnail': 're:https?://.*/images/',
44             'view_count': int,
45             'like_count': int,
46         },
47     }]
48
49     def _get_url(self, track_id, template_url, server_number):
50         boundaries = (1, 30)
51         for nr in server_numbers(server_number, boundaries):
52             url = template_url % nr
53             try:
54                 # We only want to know if the request succeed
55                 # don't download the whole file
56                 self._request_webpage(
57                     HEADRequest(url), track_id,
58                     'Checking URL %d/%d ...' % (nr, boundaries[-1]))
59                 return url
60             except ExtractorError:
61                 pass
62         return None
63
64     def _real_extract(self, url):
65         mobj = re.match(self._VALID_URL, url)
66         uploader = mobj.group(1)
67         cloudcast_name = mobj.group(2)
68         track_id = compat_urllib_parse.unquote('-'.join((uploader, cloudcast_name)))
69
70         webpage = self._download_webpage(url, track_id)
71
72         preview_url = self._search_regex(
73             r'\s(?:data-preview-url|m-preview)="([^"]+)"', webpage, 'preview url')
74         song_url = preview_url.replace('/previews/', '/c/originals/')
75         server_number = int(self._search_regex(r'stream(\d+)', song_url, 'server number'))
76         template_url = re.sub(r'(stream\d*)', 'stream%d', song_url)
77         final_song_url = self._get_url(track_id, template_url, server_number)
78         if final_song_url is None:
79             self.to_screen('Trying with m4a extension')
80             template_url = template_url.replace('.mp3', '.m4a').replace('originals/', 'm4a/64/')
81             final_song_url = self._get_url(track_id, template_url, server_number)
82         if final_song_url is None:
83             raise ExtractorError('Unable to extract track url')
84
85         PREFIX = (
86             r'm-play-on-spacebar[^>]+'
87             r'(?:\s+[a-zA-Z0-9-]+(?:="[^"]+")?)*?\s+')
88         title = self._html_search_regex(
89             PREFIX + r'm-title="([^"]+)"', webpage, 'title')
90         thumbnail = self._proto_relative_url(self._html_search_regex(
91             PREFIX + r'm-thumbnail-url="([^"]+)"', webpage, 'thumbnail',
92             fatal=False))
93         uploader = self._html_search_regex(
94             PREFIX + r'm-owner-name="([^"]+)"',
95             webpage, 'uploader', fatal=False)
96         uploader_id = self._search_regex(
97             r'\s+"profile": "([^"]+)",', webpage, 'uploader id', fatal=False)
98         description = self._og_search_description(webpage)
99         like_count = str_to_int(self._search_regex(
100             r'\bbutton-favorite\b.+m-ajax-toggle-count="([^"]+)"',
101             webpage, 'like count', fatal=False))
102         view_count = str_to_int(self._search_regex(
103             [r'<meta itemprop="interactionCount" content="UserPlays:([0-9]+)"',
104              r'/listeners/?">([0-9,.]+)</a>'],
105             webpage, 'play count', fatal=False))
106
107         return {
108             'id': track_id,
109             'title': title,
110             'url': final_song_url,
111             'description': description,
112             'thumbnail': thumbnail,
113             'uploader': uploader,
114             'uploader_id': uploader_id,
115             'view_count': view_count,
116             'like_count': like_count,
117         }
118
119
120 def server_numbers(first, boundaries):
121     """ Server numbers to try in descending order of probable availability.
122     Starting from first (i.e. the number of the server hosting the preview file)
123     and going further and further up to the higher boundary and down to the
124     lower one in an alternating fashion. Namely:
125
126         server_numbers(2, (1, 5))
127
128         # Where the preview server is 2, min number is 1 and max is 5.
129         # Yields: 2, 3, 1, 4, 5
130
131     Why not random numbers or increasing sequences? Since from what I've seen,
132     full length files seem to be hosted on servers whose number is closer to
133     that of the preview; to be confirmed.
134     """
135     zip_longest = getattr(itertools, 'zip_longest', None)
136     if zip_longest is None:
137         # python 2.x
138         zip_longest = itertools.izip_longest
139
140     if len(boundaries) != 2:
141         raise ValueError("boundaries should be a two-element tuple")
142     min, max = boundaries
143     highs = range(first + 1, max + 1)
144     lows = range(first - 1, min - 1, -1)
145     rest = filter(
146         None, itertools.chain.from_iterable(zip_longest(highs, lows)))
147     yield first
148     for n in rest:
149         yield n