X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/15b1d10671b48df598afd70e17ba21e9e64ac766..0433503f78928c1f0c76a8d66ef236dc35ff23df:/youtube_dl/extractor/channel9.py diff --git a/youtube_dl/extractor/channel9.py b/youtube_dl/extractor/channel9.py index 2a05813..81108e7 100644 --- a/youtube_dl/extractor/channel9.py +++ b/youtube_dl/extractor/channel9.py @@ -3,274 +3,260 @@ from __future__ import unicode_literals import re from .common import InfoExtractor -from ..utils import ExtractorError +from ..utils import ( + clean_html, + ExtractorError, + int_or_none, + parse_iso8601, + qualities, + unescapeHTML, +) 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 - not always possible to do. - ''' IE_DESC = 'Channel 9' IE_NAME = 'channel9' - _VALID_URL = r'https?://(?:www\.)?channel9\.msdn\.com/(?P.+)/?' - - _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': 'http://video.ch9.ms/ch9/9d51/03902f2d-fc97-4d3c-b195-0bfe15a19d51/KOS002_220.jpg', - 'session_code': 'KOS002', - 'session_day': 'Day 1', - 'session_room': 'Arena 1A', - 'session_speakers': ['Ed Blankenship', 'Andrew Coates', 'Brady Gaster', 'Patrick Klug', 'Mads Kristensen'], - }, + _VALID_URL = r'https?://(?:www\.)?(?:channel9\.msdn\.com|s\.ch9\.ms)/(?P.+?)(?P/RSS)?/?(?:[?#&]|$)' + + _TESTS = [{ + 'url': 'http://channel9.msdn.com/Events/TechEd/Australia/2013/KOS002', + 'md5': '32083d4eaf1946db6d454313f44510ca', + 'info_dict': { + 'id': '6c413323-383a-49dc-88f9-a22800cab024', + 'ext': 'wmv', + 'title': 'Developer Kick-Off Session: Stuff We Love', + 'description': 'md5:b80bf9355a503c193aff7ec6cd5a7731', + 'duration': 4576, + 'thumbnail': r're:https?://.*\.jpg', + 'timestamp': 1377717420, + 'upload_date': '20130828', + 'session_code': 'KOS002', + 'session_room': 'Arena 1A', + 'session_speakers': ['Andrew Coates', 'Brady Gaster', 'Mads Kristensen', 'Ed Blankenship', 'Patrick Klug'], + }, + }, { + 'url': 'http://channel9.msdn.com/posts/Self-service-BI-with-Power-BI-nuclear-testing', + 'md5': 'dcf983ee6acd2088e7188c3cf79b46bc', + 'info_dict': { + 'id': 'fe8e435f-bb93-4e01-8e97-a28c01887024', + 'ext': 'wmv', + 'title': 'Self-service BI with Power BI - nuclear testing', + 'description': 'md5:2d17fec927fc91e9e17783b3ecc88f54', + 'duration': 1540, + 'thumbnail': r're:https?://.*\.jpg', + 'timestamp': 1386381991, + 'upload_date': '20131207', + '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': '33ad69d2-6a4e-4172-83a1-a523013dec76', + 'ext': 'mp4', + 'title': 'Ranges for the Standard Library', + 'description': 'md5:9895e0a9fd80822d2f01c454b8f4a372', + 'duration': 5646, + 'thumbnail': r're:https?://.*\.jpg', + 'upload_date': '20150930', + 'timestamp': 1443640735, + }, + 'params': { + 'skip_download': True, + }, + }, { + 'url': 'https://channel9.msdn.com/Niners/Splendid22/Queue/76acff796e8f411184b008028e0d492b/RSS', + 'info_dict': { + 'id': 'Niners/Splendid22/Queue/76acff796e8f411184b008028e0d492b', + 'title': 'Channel 9', }, - { - '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': 'http://video.ch9.ms/ch9/87e1/0300391f-a455-4c72-bec3-4422f19287e1/selfservicenuk_512.jpg', - 'authors': ['Mike Wilmot'], - }, - } - ] + 'playlist_mincount': 100, + }, { + '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' - # Sorted by quality - _known_formats = ['MP3', 'MP4', 'Mid Quality WMV', 'Mid Quality MP4', 'High Quality WMV', 'High Quality MP4'] - - def _restore_bytes(self, formatted_size): - if not formatted_size: - return 0 - m = re.match(r'^(?P\d+(?:\.\d+)?)\s+(?P[a-zA-Z]+)', formatted_size) - if not m: - return 0 - units = m.group('units') - try: - exponent = ['B', 'KB', 'MB', 'GB', 'TB', 'PB', 'EB', 'ZB', 'YB'].index(units.upper()) - except ValueError: - return 0 - size = float(m.group('size')) - return int(size * (1024 ** exponent)) - - def _formats_from_html(self, html): - FORMAT_REGEX = r''' - (?x) - [^"]+)">(?P[^<]+)\s* - \((?P[^\)]+)\)\s* - (?:\s* -

File\s+size

\s*(?P.*?)\s* - )? # File size part may be missing - ''' - # Extract known formats - formats = [{ - 'url': x.group('url'), - 'format_id': x.group('quality'), - 'format_note': x.group('note'), - 'format': '%s (%s)' % (x.group('quality'), x.group('note')), - 'filesize': self._restore_bytes(x.group('filesize')), # File size is approximate - 'preference': self._known_formats.index(x.group('quality')), - 'vcodec': 'none' if x.group('note') == 'Audio only' else None, - } for x in list(re.finditer(FORMAT_REGEX, html)) if x.group('quality') in self._known_formats] - - self._sort_formats(formats) - - return formats - - def _extract_title(self, html): - title = self._html_search_meta('title', html, 'title') - if title is None: - title = self._og_search_title(html) - 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) - \s* - \s* - (?P.+?)\s* - \s* - - ''' - m = re.search(DESCRIPTION_REGEX, html) - if m is not None: - return m.group('description') - return self._html_search_meta('description', html, 'description') - - def _extract_duration(self, html): - m = re.search(r'"length": *"(?P\d{2}):(?P\d{2}):(?P\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'Slides', html) - return m.group('slidesurl') if m is not None else None - - def _extract_zip(self, html): - m = re.search(r'Zip', html) - return m.group('zipurl') if m is not None else None - - def _extract_avg_rating(self, html): - m = re.search(r'

Avg Rating: (?P[^<]+)

', html) - return float(m.group('avgrating')) if m is not None else 0 - - def _extract_rating_count(self, html): - m = re.search(r'
\((?P[^<]+)\)
', 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'
  • \s*(?P[^<]+) Views\s*
    • ', 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'
  • \s*\s*(?P[^<]+) Comments\s*\s*
    • ', 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)
  • (.*?)
    • ', html) - if m is None: - return None - return re.findall(r'([^<]+)', m.group(1)) - - def _extract_session_code(self, html): - m = re.search(r'
  • \s*(?P.+?)\s*
    • ', html) - return m.group('code') if m is not None else None - - def _extract_session_day(self, html): - m = re.search(r'
  • \s*(?P[^<]+)\s*
    • ', html) - return m.group('day') if m is not None else None + @staticmethod + def _extract_urls(webpage): + return re.findall( + r']+src=["\'](https?://channel9\.msdn\.com/(?:[^/]+/)+)player\b', + webpage) - def _extract_session_room(self, html): - m = re.search(r'
  • \s*(?P.+?)\s*
    • ', html) - return m.group('room') if m is not None else None - - def _extract_session_speakers(self, html): - return re.findall(r'([^<]+)', html) - - def _extract_content(self, html, content_path): - # 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: - 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) - - 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() - d.update({'title': title + '-Slides', 'url': slides}) - result.append(d) - - if zip_ is not None: - d = common.copy() - d.update({'title': title + '-Zip', 'url': zip_}) - result.append(d) - - if len(formats) > 0: - d = common.copy() - 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 - - authors = self._extract_authors(html) - - for content in contents: - content['authors'] = authors - - return contents - - 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 contents - - def _extract_list(self, content_path): - rss = self._download_xml(self._RSS_URL % content_path, content_path, 'Downloading RSS') + 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') 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, content_path, 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') - - webpage = self._download_webpage(url, content_path, 'Downloading web page') - - page_type_m = re.search(r'', webpage) - if page_type_m is not None: - page_type = page_type_m.group('pagetype') - 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) + content_path, rss = re.match(self._VALID_URL, url).groups() + + if rss: + return self._extract_list(content_path, url) + + webpage = self._download_webpage( + url, content_path, 'Downloading web page') + + episode_data = self._search_regex( + r"data-episode='([^']+)'", webpage, 'episode data', default=None) + if episode_data: + episode_data = self._parse_json(unescapeHTML( + episode_data), content_path) + content_id = episode_data['contentId'] + is_session = '/Sessions(' in episode_data['api'] + content_url = 'https://channel9.msdn.com/odata' + episode_data['api'] + if is_session: + content_url += '?$expand=Speakers' else: - raise ExtractorError('Unexpected WT.entryid %s' % page_type, expected=True) - - else: # Assuming list + content_url += '?$expand=Authors' + content_data = self._download_json(content_url, content_id) + title = content_data['Title'] + + QUALITIES = ( + 'mp3', + 'wmv', 'mp4', + 'wmv-low', 'mp4-low', + 'wmv-mid', 'mp4-mid', + 'wmv-high', 'mp4-high', + ) + + quality_key = qualities(QUALITIES) + + def quality(quality_id, format_url): + return (len(QUALITIES) if '_Source.' in format_url + else quality_key(quality_id)) + + formats = [] + urls = set() + + SITE_QUALITIES = { + 'MP3': 'mp3', + 'MP4': 'mp4', + 'Low Quality WMV': 'wmv-low', + 'Low Quality MP4': 'mp4-low', + 'Mid Quality WMV': 'wmv-mid', + 'Mid Quality MP4': 'mp4-mid', + 'High Quality WMV': 'wmv-high', + 'High Quality MP4': 'mp4-high', + } + + formats_select = self._search_regex( + r'(?s)]+name=["\']format[^>]+>(.+?)]+\bvalue=(["\'])(?P(?:(?!\1).)+)\1[^>]*>\s*(?P[^<]+?)\s*<', + formats_select): + format_url = mobj.group('url') + if format_url in urls: + continue + urls.add(format_url) + format_id = mobj.group('format') + quality_id = SITE_QUALITIES.get(format_id, format_id) + formats.append({ + 'url': format_url, + 'format_id': quality_id, + 'quality': quality(quality_id, format_url), + 'vcodec': 'none' if quality_id == 'mp3' else None, + }) + + API_QUALITIES = { + 'VideoMP4Low': 'mp4-low', + 'VideoWMV': 'wmv-mid', + 'VideoMP4Medium': 'mp4-mid', + 'VideoMP4High': 'mp4-high', + 'VideoWMVHQ': 'wmv-hq', + } + + for format_id, q in API_QUALITIES.items(): + q_url = content_data.get(format_id) + if not q_url or q_url in urls: + continue + urls.add(q_url) + formats.append({ + 'url': q_url, + 'format_id': q, + 'quality': quality(q, q_url), + }) + + self._sort_formats(formats) + + slides = content_data.get('Slides') + zip_file = content_data.get('ZipFile') + + if not formats and not slides and not zip_file: + raise ExtractorError( + 'None of recording, slides or zip are available for %s' % content_path) + + subtitles = {} + for caption in content_data.get('Captions', []): + caption_url = caption.get('Url') + if not caption_url: + continue + subtitles.setdefault(caption.get('Language', 'en'), []).append({ + 'url': caption_url, + 'ext': 'vtt', + }) + + common = { + 'id': content_id, + 'title': title, + 'description': clean_html(content_data.get('Description') or content_data.get('Body')), + 'thumbnail': content_data.get('Thumbnail') or content_data.get('VideoPlayerPreviewImage'), + 'duration': int_or_none(content_data.get('MediaLengthInSeconds')), + 'timestamp': parse_iso8601(content_data.get('PublishedDate')), + 'avg_rating': int_or_none(content_data.get('Rating')), + 'rating_count': int_or_none(content_data.get('RatingCount')), + 'view_count': int_or_none(content_data.get('Views')), + 'comment_count': int_or_none(content_data.get('CommentCount')), + 'subtitles': subtitles, + } + if is_session: + speakers = [] + for s in content_data.get('Speakers', []): + speaker_name = s.get('FullName') + if not speaker_name: + continue + speakers.append(speaker_name) + + common.update({ + 'session_code': content_data.get('Code'), + 'session_room': content_data.get('Room'), + 'session_speakers': speakers, + }) + else: + authors = [] + for a in content_data.get('Authors', []): + author_name = a.get('DisplayName') + if not author_name: + continue + authors.append(author_name) + common['authors'] = authors + + contents = [] + + if slides: + d = common.copy() + d.update({'title': title + '-Slides', 'url': slides}) + contents.append(d) + + if zip_file: + d = common.copy() + d.update({'title': title + '-Zip', 'url': zip_file}) + contents.append(d) + + if formats: + d = common.copy() + d.update({'title': title, 'formats': formats}) + contents.append(d) + return self.playlist_result(contents) + else: return self._extract_list(content_path)