From e8ce2375e0851e65c4882002297404825fe1045e Mon Sep 17 00:00:00 2001 From: Sergey M? Date: Sun, 8 Nov 2015 06:54:27 +0600 Subject: [PATCH] [viidea] Improve and cleanup (Closes #7390) * Optimize requests for multipart videos * Fix cfg regex * Improve titles and identifiers --- youtube_dl/extractor/viidea.py | 99 ++++++++++++++++++++++++---------- 1 file changed, 72 insertions(+), 27 deletions(-) diff --git a/youtube_dl/extractor/viidea.py b/youtube_dl/extractor/viidea.py index 2541a36eda..525e303d43 100644 --- a/youtube_dl/extractor/viidea.py +++ b/youtube_dl/extractor/viidea.py @@ -35,35 +35,42 @@ class ViideaIE(InfoExtractor): _TESTS = [{ 'url': 'http://videolectures.net/promogram_igor_mekjavic_eng/', 'info_dict': { - 'id': '20171_part1', + 'id': '20171', + 'display_id': 'promogram_igor_mekjavic_eng', 'ext': 'mp4', 'title': 'Automatics, robotics and biocybernetics', 'description': 'md5:815fc1deb6b3a2bff99de2d5325be482', + 'thumbnail': 're:http://.*\.jpg', + 'timestamp': 1372349289, 'upload_date': '20130627', 'duration': 565, - 'thumbnail': 're:http://.*\.jpg', }, }, { # video with invalid direct format links (HTTP 403) 'url': 'http://videolectures.net/russir2010_filippova_nlp/', 'info_dict': { - 'id': '14891_part1', + 'id': '14891', + 'display_id': 'russir2010_filippova_nlp', 'ext': 'flv', 'title': 'NLP at Google', 'description': 'md5:fc7a6d9bf0302d7cc0e53f7ca23747b3', - 'duration': 5352, 'thumbnail': 're:http://.*\.jpg', + 'timestamp': 1284375600, + 'upload_date': '20100913', + 'duration': 5352, }, 'params': { # rtmp download 'skip_download': True, }, }, { + # event playlist 'url': 'http://videolectures.net/deeplearning2015_montreal/', 'info_dict': { 'id': '23181', 'title': 'Deep Learning Summer School, Montreal 2015', 'description': 'md5:0533a85e4bd918df52a01f0e1ebe87b7', + 'thumbnail': 're:http://.*\.jpg', 'timestamp': 1438560000, }, 'playlist_count': 30, @@ -72,37 +79,54 @@ class ViideaIE(InfoExtractor): 'url': 'http://videolectures.net/mlss09uk_bishop_ibi/', 'info_dict': { 'id': '9737', + 'display_id': 'mlss09uk_bishop_ibi', 'title': 'Introduction To Bayesian Inference', + 'thumbnail': 're:http://.*\.jpg', 'timestamp': 1251622800, }, 'playlist': [{ 'info_dict': { 'id': '9737_part1', + 'display_id': 'mlss09uk_bishop_ibi_part1', 'ext': 'wmv', - 'title': 'Introduction To Bayesian Inference', + 'title': 'Introduction To Bayesian Inference (Part 1)', + 'thumbnail': 're:http://.*\.jpg', + 'duration': 4622, + 'timestamp': 1251622800, + 'upload_date': '20090830', }, }, { 'info_dict': { 'id': '9737_part2', + 'display_id': 'mlss09uk_bishop_ibi_part2', 'ext': 'wmv', - 'title': 'Introduction To Bayesian Inference', + 'title': 'Introduction To Bayesian Inference (Part 2)', + 'thumbnail': 're:http://.*\.jpg', + 'duration': 5641, + 'timestamp': 1251622800, + 'upload_date': '20090830', }, }], 'playlist_count': 2, }] def _real_extract(self, url): - lecture_slug, part = re.match(self._VALID_URL, url).groups() + lecture_slug, explicit_part_id = re.match(self._VALID_URL, url).groups() webpage = self._download_webpage(url, lecture_slug) - cfg = self._parse_json(self._search_regex([r'cfg\s*:\s*({.+?}),[\da-zA-Z_]:\(?function', r'cfg\s*:\s*({[^}]+})'], webpage, 'cfg'), lecture_slug, js_to_json) + cfg = self._parse_json(self._search_regex( + [r'cfg\s*:\s*({.+?})\s*,\s*[\da-zA-Z_]+\s*:\s*\(?\s*function', + r'cfg\s*:\s*({[^}]+})'], + webpage, 'cfg'), lecture_slug, js_to_json) lecture_id = compat_str(cfg['obj_id']) base_url = self._proto_relative_url(cfg['livepipe'], 'http:') - lecture_data = self._download_json('%s/site/api/lecture/%s?format=json' % (base_url, lecture_id), lecture_id)['lecture'][0] + lecture_data = self._download_json( + '%s/site/api/lecture/%s?format=json' % (base_url, lecture_id), + lecture_id)['lecture'][0] lecture_info = { 'id': lecture_id, @@ -113,31 +137,52 @@ def _real_extract(self, url): 'thumbnail': lecture_data.get('thumb'), } - entries = [] - parts = cfg.get('videos') + playlist_entries = [] + lecture_type = lecture_data.get('type') + parts = [compat_str(video) for video in cfg.get('videos', [])] if parts: - if len(parts) == 1: - part = compat_str(parts[0]) - if part: - smil_url = '%s/%s/video/%s/smil.xml' % (base_url, lecture_slug, part) + multipart = len(parts) > 1 + + def extract_part(part_id): + smil_url = '%s/%s/video/%s/smil.xml' % (base_url, lecture_slug, part_id) smil = self._download_smil(smil_url, lecture_id) info = self._parse_smil(smil, smil_url, lecture_id) - info['id'] = '%s_part%s' % (lecture_id, part) + info['id'] = lecture_id if not multipart else '%s_part%s' % (lecture_id, part_id) + info['display_id'] = lecture_slug if not multipart else '%s_part%s' % (lecture_slug, part_id) + if multipart: + info['title'] += ' (Part %s)' % part_id switch = smil.find('.//switch') if switch is not None: info['duration'] = parse_duration(switch.attrib.get('dur')) - return info + item_info = lecture_info.copy() + item_info.update(info) + return item_info + + if explicit_part_id or not multipart: + result = extract_part(explicit_part_id or parts[0]) else: - for part in parts: - entries.append(self.url_result('%s/%s/video/%s' % (base_url, lecture_slug, part), 'Viidea')) - lecture_info['_type'] = 'multi_video' - if not parts or lecture_data.get('type') == 'evt': - # Probably a playlist - playlist_webpage = self._download_webpage('%s/site/ajax/drilldown/?id=%s' % (base_url, lecture_id), lecture_id) + result = { + '_type': 'multi_video', + 'entries': [extract_part(part) for part in parts], + } + result.update(lecture_info) + + # Immediately return explicitly requested part or non event item + if explicit_part_id or lecture_type != 'evt': + return result + + playlist_entries.append(result) + + # It's probably a playlist + if not parts or lecture_type == 'evt': + playlist_webpage = self._download_webpage( + '%s/site/ajax/drilldown/?id=%s' % (base_url, lecture_id), lecture_id) entries = [ self.url_result(compat_urlparse.urljoin(url, video_url), 'Viidea') - for _, video_url in re.findall(r']+href=(["\'])(.+?)\1[^>]+id=["\']lec=\d+', playlist_webpage)] - lecture_info['_type'] = 'playlist' + for _, video_url in re.findall( + r']+href=(["\'])(.+?)\1[^>]+id=["\']lec=\d+', playlist_webpage)] + playlist_entries.extend(entries) - lecture_info['entries'] = entries - return lecture_info + playlist = self.playlist_result(playlist_entries, lecture_id) + playlist.update(lecture_info) + return playlist