From 02fb9804513ce1bfe28ec7c285526db7989e5844 Mon Sep 17 00:00:00 2001 From: remitamine Date: Sun, 1 Nov 2015 02:08:19 +0100 Subject: [PATCH 1/5] [flickr] extract more info and formats --- youtube_dl/extractor/flickr.py | 110 ++++++++++++++++++--------------- 1 file changed, 60 insertions(+), 50 deletions(-) diff --git a/youtube_dl/extractor/flickr.py b/youtube_dl/extractor/flickr.py index 2fe76d661..5ca754105 100644 --- a/youtube_dl/extractor/flickr.py +++ b/youtube_dl/extractor/flickr.py @@ -1,67 +1,77 @@ from __future__ import unicode_literals -import re - from .common import InfoExtractor -from ..compat import compat_urllib_request +from ..compat import compat_urllib_parse from ..utils import ( - ExtractorError, - find_xpath_attr, + int_or_none, + qualities, ) class FlickrIE(InfoExtractor): - _VALID_URL = r'https?://(?:www\.|secure\.)?flickr\.com/photos/(?P[\w\-_@]+)/(?P\d+).*' + _VALID_URL = r'https?://(?:www\.|secure\.)?flickr\.com/photos/[\w\-_@]+/(?P\d+)' _TEST = { 'url': 'http://www.flickr.com/photos/forestwander-nature-pictures/5645318632/in/photostream/', - 'md5': '6fdc01adbc89d72fc9c4f15b4a4ba87b', + 'md5': '164fe3fa6c22e18d448d4d5af2330f31', 'info_dict': { 'id': '5645318632', - 'ext': 'mp4', - "description": "Waterfalls in the Springtime at Dark Hollow Waterfalls. These are located just off of Skyline Drive in Virginia. They are only about 6/10 of a mile hike but it is a pretty steep hill and a good climb back up.", - "uploader_id": "forestwander-nature-pictures", - "title": "Dark Hollow Waterfalls" + 'ext': 'mpg', + 'description': 'Waterfalls in the Springtime at Dark Hollow Waterfalls. These are located just off of Skyline Drive in Virginia. They are only about 6/10 of a mile hike but it is a pretty steep hill and a good climb back up.', + 'uploader_id': 'forestwander-nature-pictures', + 'title': 'Dark Hollow Waterfalls', + 'duration': 19, + 'timestamp': 1303528740, + 'upload_date': '20110423', + 'uploader_id': '10922353@N03', + 'uploader': 'Forest Wander', + 'comment_count': int, } } - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) + _API_BASE_URL = 'https://api.flickr.com/services/rest?' + _API_KEY = '61b16865f916058e63580a912d9143be' - video_id = mobj.group('id') - video_uploader_id = mobj.group('uploader_id') - webpage_url = 'http://www.flickr.com/photos/' + video_uploader_id + '/' + video_id - req = compat_urllib_request.Request(webpage_url) - req.add_header( - 'User-Agent', - # it needs a more recent version - 'Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20150101 Firefox/38.0 (Chrome)') - webpage = self._download_webpage(req, video_id) - - secret = self._search_regex(r'secret"\s*:\s*"(\w+)"', webpage, 'secret') - - first_url = 'https://secure.flickr.com/apps/video/video_mtl_xml.gne?v=x&photo_id=' + video_id + '&secret=' + secret + '&bitrate=700&target=_self' - first_xml = self._download_xml(first_url, video_id, 'Downloading first data webpage') - - node_id = find_xpath_attr( - first_xml, './/{http://video.yahoo.com/YEP/1.0/}Item', 'id', - 'id').text - - second_url = 'https://secure.flickr.com/video_playlist.gne?node_id=' + node_id + '&tech=flash&mode=playlist&bitrate=700&secret=' + secret + '&rd=video.yahoo.com&noad=1' - second_xml = self._download_xml(second_url, video_id, 'Downloading second data webpage') - - self.report_extraction(video_id) - - stream = second_xml.find('.//STREAM') - if stream is None: - raise ExtractorError('Unable to extract video url') - video_url = stream.attrib['APP'] + stream.attrib['FULLPATH'] - - return { - 'id': video_id, - 'url': video_url, - 'ext': 'mp4', - 'title': self._og_search_title(webpage), - 'description': self._og_search_description(webpage), - 'thumbnail': self._og_search_thumbnail(webpage), - 'uploader_id': video_uploader_id, + def _call_api(self, method, video_id, secret=None): + query = { + 'photo_id': video_id, + 'method': 'flickr.%s' % method, + 'api_key': self._API_KEY, + 'format': 'json', + 'nojsoncallback': 1, } + if secret: + query['secret'] = secret + return self._download_json(self._API_BASE_URL + compat_urllib_parse.urlencode(query), video_id) + + def _real_extract(self, url): + video_id = self._match_id(url) + + video_info = self._call_api('photos.getInfo', video_id)['photo'] + if video_info['media'] == 'video': + streams = self._call_api('video.getStreamInfo', video_id, video_info['secret'])['streams'] + + preference = qualities(['iphone_wifi', '700', 'appletv', 'orig']) + + formats = [] + for stream in streams['stream']: + stream_type = str(stream.get('type')) + formats.append({ + 'format_id': stream_type, + 'url': stream['_content'], + 'preference': preference(stream_type), + }) + self._sort_formats(formats) + + owner = video_info.get('owner', {}) + + return { + 'id': video_id, + 'title': video_info['title']['_content'], + 'description': video_info.get('description', {}).get('_content'), + 'formats': formats, + 'timestamp': int_or_none(video_info.get('dateuploaded')), + 'duration': int_or_none(video_info.get('video', {}).get('duration')), + 'uploader_id': owner.get('nsid'), + 'uploader': owner.get('realname'), + 'comment_count': int_or_none(video_info.get('comments', {}).get('_content')), + } From 146672254e409bf97c82a302095fbfabf2c48928 Mon Sep 17 00:00:00 2001 From: remitamine Date: Sun, 1 Nov 2015 13:23:23 +0100 Subject: [PATCH 2/5] [flickr] extract fresh api key and remove duplication in test --- youtube_dl/extractor/flickr.py | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/youtube_dl/extractor/flickr.py b/youtube_dl/extractor/flickr.py index 5ca754105..0d5d6b0b9 100644 --- a/youtube_dl/extractor/flickr.py +++ b/youtube_dl/extractor/flickr.py @@ -17,7 +17,6 @@ class FlickrIE(InfoExtractor): 'id': '5645318632', 'ext': 'mpg', 'description': 'Waterfalls in the Springtime at Dark Hollow Waterfalls. These are located just off of Skyline Drive in Virginia. They are only about 6/10 of a mile hike but it is a pretty steep hill and a good climb back up.', - 'uploader_id': 'forestwander-nature-pictures', 'title': 'Dark Hollow Waterfalls', 'duration': 19, 'timestamp': 1303528740, @@ -29,26 +28,27 @@ class FlickrIE(InfoExtractor): } _API_BASE_URL = 'https://api.flickr.com/services/rest?' - _API_KEY = '61b16865f916058e63580a912d9143be' - def _call_api(self, method, video_id, secret=None): + def _call_api(self, method, video_id, api_key, note, secret=None): query = { 'photo_id': video_id, 'method': 'flickr.%s' % method, - 'api_key': self._API_KEY, + 'api_key': api_key, 'format': 'json', 'nojsoncallback': 1, } if secret: query['secret'] = secret - return self._download_json(self._API_BASE_URL + compat_urllib_parse.urlencode(query), video_id) + return self._download_json(self._API_BASE_URL + compat_urllib_parse.urlencode(query), video_id, note) def _real_extract(self, url): video_id = self._match_id(url) - video_info = self._call_api('photos.getInfo', video_id)['photo'] + api_key = self._download_json('https://www.flickr.com/hermes_error_beacon.gne', video_id, 'Downloading api key',)['site_key'] + + video_info = self._call_api('photos.getInfo', video_id, api_key, 'Downloading video info')['photo'] if video_info['media'] == 'video': - streams = self._call_api('video.getStreamInfo', video_id, video_info['secret'])['streams'] + streams = self._call_api('video.getStreamInfo', video_id, api_key, 'Downloading streams info', video_info['secret'])['streams'] preference = qualities(['iphone_wifi', '700', 'appletv', 'orig']) From f3003531a5622cc01501325b9f35dcb2424cfb70 Mon Sep 17 00:00:00 2001 From: remitamine Date: Sun, 1 Nov 2015 13:38:11 +0100 Subject: [PATCH 3/5] [flickr] handle error message --- youtube_dl/extractor/flickr.py | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/youtube_dl/extractor/flickr.py b/youtube_dl/extractor/flickr.py index 0d5d6b0b9..e97754d36 100644 --- a/youtube_dl/extractor/flickr.py +++ b/youtube_dl/extractor/flickr.py @@ -3,6 +3,7 @@ from __future__ import unicode_literals from .common import InfoExtractor from ..compat import compat_urllib_parse from ..utils import ( + ExtractorError, int_or_none, qualities, ) @@ -39,7 +40,10 @@ class FlickrIE(InfoExtractor): } if secret: query['secret'] = secret - return self._download_json(self._API_BASE_URL + compat_urllib_parse.urlencode(query), video_id, note) + data = self._download_json(self._API_BASE_URL + compat_urllib_parse.urlencode(query), video_id, note) + if data['stat'] != 'ok': + raise ExtractorError(data['message']) + return data def _real_extract(self, url): video_id = self._match_id(url) From 967c9076a31ca2a2b43fb71082ad1a8db88116bd Mon Sep 17 00:00:00 2001 From: remitamine Date: Thu, 5 Nov 2015 18:01:13 +0100 Subject: [PATCH 4/5] raise ExtractorError if the page doesn't contain a video --- youtube_dl/extractor/flickr.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/youtube_dl/extractor/flickr.py b/youtube_dl/extractor/flickr.py index e97754d36..92d2ac553 100644 --- a/youtube_dl/extractor/flickr.py +++ b/youtube_dl/extractor/flickr.py @@ -79,3 +79,5 @@ class FlickrIE(InfoExtractor): 'uploader': owner.get('realname'), 'comment_count': int_or_none(video_info.get('comments', {}).get('_content')), } + else: + raise ExtractorError('not a video', expected=True) From 5b95419ca59e6238ab0041d48a53a9e51485e1cc Mon Sep 17 00:00:00 2001 From: remitamine Date: Mon, 21 Dec 2015 00:20:22 +0100 Subject: [PATCH 5/5] [flickr] extract views_count and tags --- youtube_dl/extractor/flickr.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/youtube_dl/extractor/flickr.py b/youtube_dl/extractor/flickr.py index 92d2ac553..452b27b26 100644 --- a/youtube_dl/extractor/flickr.py +++ b/youtube_dl/extractor/flickr.py @@ -25,6 +25,8 @@ class FlickrIE(InfoExtractor): 'uploader_id': '10922353@N03', 'uploader': 'Forest Wander', 'comment_count': int, + 'view_count': int, + 'tags': list, } } @@ -78,6 +80,8 @@ class FlickrIE(InfoExtractor): 'uploader_id': owner.get('nsid'), 'uploader': owner.get('realname'), 'comment_count': int_or_none(video_info.get('comments', {}).get('_content')), + 'view_count': int_or_none(video_info.get('views')), + 'tags': [tag.get('_content') for tag in video_info.get('tags', {}).get('tag', [])] } else: raise ExtractorError('not a video', expected=True)