[appletrailers] Rework extraction (fixes #1387)
[youtube-dl] / youtube_dl / extractor / appletrailers.py
1 import re
2 import xml.etree.ElementTree
3 import json
4
5 from .common import InfoExtractor
6 from ..utils import (
7     compat_urlparse,
8     determine_ext,
9 )
10
11
12 class AppleTrailersIE(InfoExtractor):
13     _VALID_URL = r'https?://(?:www\.)?trailers.apple.com/trailers/(?P<company>[^/]+)/(?P<movie>[^/]+)'
14     _TEST = {
15         u"url": u"http://trailers.apple.com/trailers/wb/manofsteel/",
16         u"playlist": [
17             {
18                 u"file": u"manofsteel-trailer4.mov",
19                 u"md5": u"d97a8e575432dbcb81b7c3acb741f8a8",
20                 u"info_dict": {
21                     u"duration": 111,
22                     u"title": u"Trailer 4",
23                     u"upload_date": u"20130523",
24                     u"uploader_id": u"wb",
25                 },
26             },
27             {
28                 u"file": u"manofsteel-trailer3.mov",
29                 u"md5": u"b8017b7131b721fb4e8d6f49e1df908c",
30                 u"info_dict": {
31                     u"duration": 182,
32                     u"title": u"Trailer 3",
33                     u"upload_date": u"20130417",
34                     u"uploader_id": u"wb",
35                 },
36             },
37             {
38                 u"file": u"manofsteel-trailer.mov",
39                 u"md5": u"d0f1e1150989b9924679b441f3404d48",
40                 u"info_dict": {
41                     u"duration": 148,
42                     u"title": u"Trailer",
43                     u"upload_date": u"20121212",
44                     u"uploader_id": u"wb",
45                 },
46             },
47             {
48                 u"file": u"manofsteel-teaser.mov",
49                 u"md5": u"5fe08795b943eb2e757fa95cb6def1cb",
50                 u"info_dict": {
51                     u"duration": 93,
52                     u"title": u"Teaser",
53                     u"upload_date": u"20120721",
54                     u"uploader_id": u"wb",
55                 },
56             }
57         ]
58     }
59
60     _JSON_RE = r'iTunes.playURL\((.*?)\);'
61
62     def _real_extract(self, url):
63         mobj = re.match(self._VALID_URL, url)
64         movie = mobj.group('movie')
65         uploader_id = mobj.group('company')
66
67         playlist_url = compat_urlparse.urljoin(url, u'includes/playlists/itunes.inc')
68         playlist_snippet = self._download_webpage(playlist_url, movie)
69         playlist_cleaned = re.sub(r'(?s)<script[^<]*?>.*?</script>', u'', playlist_snippet)
70         playlist_cleaned = re.sub(r'<img ([^<]*?)>', r'<img \1/>', playlist_cleaned)
71         # The ' in the onClick attributes are not escaped, it couldn't be parsed
72         # with xml.etree.ElementTree.fromstring
73         # like: http://trailers.apple.com/trailers/wb/gravity/
74         def _clean_json(m):
75             return u'iTunes.playURL(%s);' % m.group(1).replace('\'', '&#39;')
76         playlist_cleaned = re.sub(self._JSON_RE, _clean_json, playlist_cleaned)
77         playlist_html = u'<html>' + playlist_cleaned + u'</html>'
78
79         doc = xml.etree.ElementTree.fromstring(playlist_html)
80         playlist = []
81         for li in doc.findall('./div/ul/li'):
82             on_click = li.find('.//a').attrib['onClick']
83             trailer_info_json = self._search_regex(self._JSON_RE,
84                 on_click, u'trailer info')
85             trailer_info = json.loads(trailer_info_json)
86             title = trailer_info['title']
87             video_id = movie + '-' + re.sub(r'[^a-zA-Z0-9]', '', title).lower()
88             thumbnail = li.find('.//img').attrib['src']
89             upload_date = trailer_info['posted'].replace('-', '')
90
91             runtime = trailer_info['runtime']
92             m = re.search(r'(?P<minutes>[0-9]+):(?P<seconds>[0-9]{1,2})', runtime)
93             duration = None
94             if m:
95                 duration = 60 * int(m.group('minutes')) + int(m.group('seconds'))
96
97             first_url = trailer_info['url']
98             trailer_id = first_url.split('/')[-1].rpartition('_')[0]
99             settings_json_url = compat_urlparse.urljoin(url, 'includes/settings/%s.json' % trailer_id)
100             settings_json = self._download_webpage(settings_json_url, trailer_id, u'Downloading settings json')
101             settings = json.loads(settings_json)
102
103             formats = []
104             for format in settings['metadata']['sizes']:
105                 # The src is a file pointing to the real video file
106                 format_url = re.sub(r'_(\d*p.mov)', r'_h\1', format['src'])
107                 formats.append({
108                     'url': format_url,
109                     'ext': determine_ext(format_url),
110                     'format': format['type'],
111                     'width': format['width'],
112                     'height': int(format['height']),
113                 })
114             formats = sorted(formats, key=lambda f: (f['height'], f['width']))
115
116             info = {
117                 '_type': 'video',
118                 'id': video_id,
119                 'title': title,
120                 'formats': formats,
121                 'title': title,
122                 'duration': duration,
123                 'thumbnail': thumbnail,
124                 'upload_date': upload_date,
125                 'uploader_id': uploader_id,
126                 'user_agent': 'QuickTime compatible (youtube-dl)',
127             }
128             # TODO: Remove when #980 has been merged
129             info['url'] = formats[-1]['url']
130             info['ext'] = formats[-1]['ext']
131
132             playlist.append(info)
133
134         return {
135             '_type': 'playlist',
136             'id': movie,
137             'entries': playlist,
138         }