yt-dlp/youtube_dl/extractor/channel9.py

301 lines
12 KiB
Python
Raw Normal View History

2014-01-28 20:00:56 +01:00
from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..utils import (
ExtractorError,
parse_filesize,
qualities,
)
2014-11-23 20:41:03 +01:00
class Channel9IE(InfoExtractor):
'''
Common extractor for channel9.msdn.com.
The type of provided URL (video or playlist) is determined according to
meta Search.PageType from web page HTML rather than URL itself, as it is
2014-01-28 20:00:56 +01:00
not always possible to do.
'''
2014-01-28 20:00:56 +01:00
IE_DESC = 'Channel 9'
IE_NAME = 'channel9'
_VALID_URL = r'https?://(?:www\.)?channel9\.msdn\.com/(?P<contentpath>.+?)(?P<rss>/RSS)?/?(?:[?#&]|$)'
_TESTS = [{
'url': 'http://channel9.msdn.com/Events/TechEd/Australia/2013/KOS002',
'md5': 'bbd75296ba47916b754e73c3a4bbdf10',
'info_dict': {
'id': 'Events/TechEd/Australia/2013/KOS002',
'ext': 'mp4',
'title': 'Developer Kick-Off Session: Stuff We Love',
'description': 'md5:c08d72240b7c87fcecafe2692f80e35f',
'duration': 4576,
'thumbnail': 're:http://.*\.jpg',
'session_code': 'KOS002',
'session_day': 'Day 1',
'session_room': 'Arena 1A',
'session_speakers': ['Ed Blankenship', 'Andrew Coates', 'Brady Gaster', 'Patrick Klug',
'Mads Kristensen'],
},
}, {
'url': 'http://channel9.msdn.com/posts/Self-service-BI-with-Power-BI-nuclear-testing',
'md5': 'b43ee4529d111bc37ba7ee4f34813e68',
'info_dict': {
'id': 'posts/Self-service-BI-with-Power-BI-nuclear-testing',
'ext': 'mp4',
'title': 'Self-service BI with Power BI - nuclear testing',
'description': 'md5:d1e6ecaafa7fb52a2cacdf9599829f5b',
'duration': 1540,
'thumbnail': 're:http://.*\.jpg',
'authors': ['Mike Wilmot'],
},
}, {
# low quality mp4 is best
'url': 'https://channel9.msdn.com/Events/CPP/CppCon-2015/Ranges-for-the-Standard-Library',
'info_dict': {
'id': 'Events/CPP/CppCon-2015/Ranges-for-the-Standard-Library',
'ext': 'mp4',
'title': 'Ranges for the Standard Library',
'description': 'md5:2e6b4917677af3728c5f6d63784c4c5d',
'duration': 5646,
'thumbnail': 're:http://.*\.jpg',
},
'params': {
'skip_download': True,
},
}, {
'url': 'https://channel9.msdn.com/Niners/Splendid22/Queue/76acff796e8f411184b008028e0d492b/RSS',
'info_dict': {
'id': 'Niners/Splendid22/Queue/76acff796e8f411184b008028e0d492b',
'title': 'Channel 9',
},
'playlist_count': 2,
}, {
'url': 'https://channel9.msdn.com/Events/DEVintersection/DEVintersection-2016/RSS',
'only_matching': True,
}, {
'url': 'https://channel9.msdn.com/Events/Speakers/scott-hanselman/RSS?UrlSafeName=scott-hanselman',
'only_matching': True,
}]
_RSS_URL = 'http://channel9.msdn.com/%s/RSS'
def _formats_from_html(self, html):
FORMAT_REGEX = r'''
(?x)
<a\s+href="(?P<url>[^"]+)">(?P<quality>[^<]+)</a>\s*
<span\s+class="usage">\((?P<note>[^\)]+)\)</span>\s*
(?:<div\s+class="popup\s+rounded">\s*
<h3>File\s+size</h3>\s*(?P<filesize>.*?)\s*
</div>)? # File size part may be missing
'''
quality = qualities((
'MP3', 'MP4',
'Low Quality WMV', 'Low Quality MP4',
'Mid Quality WMV', 'Mid Quality MP4',
'High Quality WMV', 'High Quality MP4'))
formats = [{
'url': x.group('url'),
'format_id': x.group('quality'),
'format_note': x.group('note'),
2014-01-28 20:00:56 +01:00
'format': '%s (%s)' % (x.group('quality'), x.group('note')),
'filesize_approx': parse_filesize(x.group('filesize')),
'quality': quality(x.group('quality')),
'vcodec': 'none' if x.group('note') == 'Audio only' else None,
} for x in list(re.finditer(FORMAT_REGEX, html))]
self._sort_formats(formats)
return formats
def _extract_title(self, html):
2014-01-28 20:00:56 +01:00
title = self._html_search_meta('title', html, 'title')
2014-11-04 23:20:34 +01:00
if title is None:
title = self._og_search_title(html)
2014-01-28 20:00:56 +01:00
TITLE_SUFFIX = ' (Channel 9)'
if title is not None and title.endswith(TITLE_SUFFIX):
title = title[:-len(TITLE_SUFFIX)]
return title
def _extract_description(self, html):
DESCRIPTION_REGEX = r'''(?sx)
<div\s+class="entry-content">\s*
<div\s+id="entry-body">\s*
(?P<description>.+?)\s*
</div>\s*
</div>
'''
m = re.search(DESCRIPTION_REGEX, html)
if m is not None:
return m.group('description')
2014-01-28 20:00:56 +01:00
return self._html_search_meta('description', html, 'description')
def _extract_duration(self, html):
2014-11-04 23:20:34 +01:00
m = re.search(r'"length": *"(?P<hours>\d{2}):(?P<minutes>\d{2}):(?P<seconds>\d{2})"', html)
return ((int(m.group('hours')) * 60 * 60) + (int(m.group('minutes')) * 60) + int(m.group('seconds'))) if m else None
def _extract_slides(self, html):
m = re.search(r'<a href="(?P<slidesurl>[^"]+)" class="slides">Slides</a>', html)
return m.group('slidesurl') if m is not None else None
def _extract_zip(self, html):
m = re.search(r'<a href="(?P<zipurl>[^"]+)" class="zip">Zip</a>', html)
return m.group('zipurl') if m is not None else None
def _extract_avg_rating(self, html):
m = re.search(r'<p class="avg-rating">Avg Rating: <span>(?P<avgrating>[^<]+)</span></p>', html)
return float(m.group('avgrating')) if m is not None else 0
def _extract_rating_count(self, html):
m = re.search(r'<div class="rating-count">\((?P<ratingcount>[^<]+)\)</div>', html)
return int(self._fix_count(m.group('ratingcount'))) if m is not None else 0
def _extract_view_count(self, html):
m = re.search(r'<li class="views">\s*<span class="count">(?P<viewcount>[^<]+)</span> Views\s*</li>', html)
return int(self._fix_count(m.group('viewcount'))) if m is not None else 0
def _extract_comment_count(self, html):
m = re.search(r'<li class="comments">\s*<a href="#comments">\s*<span class="count">(?P<commentcount>[^<]+)</span> Comments\s*</a>\s*</li>', html)
return int(self._fix_count(m.group('commentcount'))) if m is not None else 0
def _fix_count(self, count):
return int(str(count).replace(',', '')) if count is not None else None
def _extract_authors(self, html):
m = re.search(r'(?s)<li class="author">(.*?)</li>', html)
if m is None:
return None
return re.findall(r'<a href="/Niners/[^"]+">([^<]+)</a>', m.group(1))
def _extract_session_code(self, html):
m = re.search(r'<li class="code">\s*(?P<code>.+?)\s*</li>', html)
return m.group('code') if m is not None else None
def _extract_session_day(self, html):
m = re.search(r'<li class="day">\s*<a href="/Events/[^"]+">(?P<day>[^<]+)</a>\s*</li>', html)
2015-10-13 21:18:30 +02:00
return m.group('day').strip() if m is not None else None
def _extract_session_room(self, html):
m = re.search(r'<li class="room">\s*(?P<room>.+?)\s*</li>', html)
return m.group('room') if m is not None else None
def _extract_session_speakers(self, html):
return re.findall(r'<a href="/Events/Speakers/[^"]+">([^<]+)</a>', html)
def _extract_content(self, html, content_path):
2014-11-04 23:20:34 +01:00
# Look for downloadable content
formats = self._formats_from_html(html)
slides = self._extract_slides(html)
zip_ = self._extract_zip(html)
# Nothing to download
if len(formats) == 0 and slides is None and zip_ is None:
2014-01-28 20:00:56 +01:00
self._downloader.report_warning('None of recording, slides or zip are available for %s' % content_path)
return
# Extract meta
title = self._extract_title(html)
description = self._extract_description(html)
thumbnail = self._og_search_thumbnail(html)
duration = self._extract_duration(html)
avg_rating = self._extract_avg_rating(html)
rating_count = self._extract_rating_count(html)
view_count = self._extract_view_count(html)
comment_count = self._extract_comment_count(html)
2014-11-23 22:21:46 +01:00
common = {
'_type': 'video',
'id': content_path,
'description': description,
'thumbnail': thumbnail,
'duration': duration,
'avg_rating': avg_rating,
'rating_count': rating_count,
'view_count': view_count,
'comment_count': comment_count,
}
result = []
if slides is not None:
d = common.copy()
2014-11-23 20:41:03 +01:00
d.update({'title': title + '-Slides', 'url': slides})
result.append(d)
if zip_ is not None:
d = common.copy()
2014-11-23 20:41:03 +01:00
d.update({'title': title + '-Zip', 'url': zip_})
result.append(d)
if len(formats) > 0:
d = common.copy()
2014-11-23 20:41:03 +01:00
d.update({'title': title, 'formats': formats})
result.append(d)
return result
def _extract_entry_item(self, html, content_path):
contents = self._extract_content(html, content_path)
if contents is None:
return contents
if len(contents) > 1:
raise ExtractorError('Got more than one entry')
result = contents[0]
result['authors'] = self._extract_authors(html)
return result
def _extract_session(self, html, content_path):
contents = self._extract_content(html, content_path)
if contents is None:
return contents
session_meta = {
'session_code': self._extract_session_code(html),
'session_day': self._extract_session_day(html),
'session_room': self._extract_session_room(html),
'session_speakers': self._extract_session_speakers(html),
}
for content in contents:
content.update(session_meta)
return self.playlist_result(contents)
def _extract_list(self, video_id, rss_url=None):
if not rss_url:
rss_url = self._RSS_URL % video_id
rss = self._download_xml(rss_url, video_id, 'Downloading RSS')
2013-12-12 09:19:23 +01:00
entries = [self.url_result(session_url.text, 'Channel9')
for session_url in rss.findall('./channel/item/link')]
title_text = rss.find('./channel/title').text
return self.playlist_result(entries, video_id, title_text)
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
content_path = mobj.group('contentpath')
rss = mobj.group('rss')
if rss:
return self._extract_list(content_path, url)
webpage = self._download_webpage(
url, content_path, 'Downloading web page')
page_type = self._search_regex(
r'<meta[^>]+name=(["\'])WT\.entryid\1[^>]+content=(["\'])(?P<pagetype>[^:]+).+?\2',
webpage, 'page type', default=None, group='pagetype')
if page_type:
2014-11-04 23:20:34 +01:00
if page_type == 'Entry': # Any 'item'-like page, may contain downloadable content
return self._extract_entry_item(webpage, content_path)
elif page_type == 'Session': # Event session page, may contain downloadable content
return self._extract_session(webpage, content_path)
elif page_type == 'Event':
return self._extract_list(content_path)
else:
raise ExtractorError('Unexpected WT.entryid %s' % page_type, expected=True)
2014-11-23 20:41:03 +01:00
else: # Assuming list
return self._extract_list(content_path)