diff options
Diffstat (limited to 'youtube_dl')
| -rwxr-xr-x | youtube_dl/InfoExtractors.py | 133 | ||||
| -rw-r--r-- | youtube_dl/extractor/justintv.py | 144 |
2 files changed, 145 insertions, 132 deletions
diff --git a/youtube_dl/InfoExtractors.py b/youtube_dl/InfoExtractors.py index a168e9952..951fe64c0 100755 --- a/youtube_dl/InfoExtractors.py +++ b/youtube_dl/InfoExtractors.py @@ -33,6 +33,7 @@ from .extractor.generic import GenericIE from .extractor.googleplus import GooglePlusIE from .extractor.googlesearch import GoogleSearchIE from .extractor.infoq import InfoQIE +from .extractor.justintv import JustinTVIE from .extractor.metacafe import MetacafeIE from .extractor.mixcloud import MixcloudIE from .extractor.mtv import MTVIE @@ -65,138 +66,6 @@ from .extractor.zdf import ZDFIE -class JustinTVIE(InfoExtractor): - """Information extractor for justin.tv and twitch.tv""" - # TODO: One broadcast may be split into multiple videos. The key - # 'broadcast_id' is the same for all parts, and 'broadcast_part' - # starts at 1 and increases. Can we treat all parts as one video? - - _VALID_URL = r"""(?x)^(?:http://)?(?:www\.)?(?:twitch|justin)\.tv/ - (?: - (?P<channelid>[^/]+)| - (?:(?:[^/]+)/b/(?P<videoid>[^/]+))| - (?:(?:[^/]+)/c/(?P<chapterid>[^/]+)) - ) - /?(?:\#.*)?$ - """ - _JUSTIN_PAGE_LIMIT = 100 - IE_NAME = u'justin.tv' - - def report_download_page(self, channel, offset): - """Report attempt to download a single page of videos.""" - self.to_screen(u'%s: Downloading video information from %d to %d' % - (channel, offset, offset + self._JUSTIN_PAGE_LIMIT)) - - # Return count of items, list of *valid* items - def _parse_page(self, url, video_id): - webpage = self._download_webpage(url, video_id, - u'Downloading video info JSON', - u'unable to download video info JSON') - - response = json.loads(webpage) - if type(response) != list: - error_text = response.get('error', 'unknown error') - raise ExtractorError(u'Justin.tv API: %s' % error_text) - info = [] - for clip in response: - video_url = clip['video_file_url'] - if video_url: - video_extension = os.path.splitext(video_url)[1][1:] - video_date = re.sub('-', '', clip['start_time'][:10]) - video_uploader_id = clip.get('user_id', clip.get('channel_id')) - video_id = clip['id'] - video_title = clip.get('title', video_id) - info.append({ - 'id': video_id, - 'url': video_url, - 'title': video_title, - 'uploader': clip.get('channel_name', video_uploader_id), - 'uploader_id': video_uploader_id, - 'upload_date': video_date, - 'ext': video_extension, - }) - return (len(response), info) - - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - if mobj is None: - raise ExtractorError(u'invalid URL: %s' % url) - - api_base = 'http://api.justin.tv' - paged = False - if mobj.group('channelid'): - paged = True - video_id = mobj.group('channelid') - api = api_base + '/channel/archives/%s.json' % video_id - elif mobj.group('chapterid'): - chapter_id = mobj.group('chapterid') - - webpage = self._download_webpage(url, chapter_id) - m = re.search(r'PP\.archive_id = "([0-9]+)";', webpage) - if not m: - raise ExtractorError(u'Cannot find archive of a chapter') - archive_id = m.group(1) - - api = api_base + '/broadcast/by_chapter/%s.xml' % chapter_id - chapter_info_xml = self._download_webpage(api, chapter_id, - note=u'Downloading chapter information', - errnote=u'Chapter information download failed') - doc = xml.etree.ElementTree.fromstring(chapter_info_xml) - for a in doc.findall('.//archive'): - if archive_id == a.find('./id').text: - break - else: - raise ExtractorError(u'Could not find chapter in chapter information') - - video_url = a.find('./video_file_url').text - video_ext = video_url.rpartition('.')[2] or u'flv' - - chapter_api_url = u'https://api.twitch.tv/kraken/videos/c' + chapter_id - chapter_info_json = self._download_webpage(chapter_api_url, u'c' + chapter_id, - note='Downloading chapter metadata', - errnote='Download of chapter metadata failed') - chapter_info = json.loads(chapter_info_json) - - bracket_start = int(doc.find('.//bracket_start').text) - bracket_end = int(doc.find('.//bracket_end').text) - - # TODO determine start (and probably fix up file) - # youtube-dl -v http://www.twitch.tv/firmbelief/c/1757457 - #video_url += u'?start=' + TODO:start_timestamp - # bracket_start is 13290, but we want 51670615 - self._downloader.report_warning(u'Chapter detected, but we can just download the whole file. ' - u'Chapter starts at %s and ends at %s' % (formatSeconds(bracket_start), formatSeconds(bracket_end))) - - info = { - 'id': u'c' + chapter_id, - 'url': video_url, - 'ext': video_ext, - 'title': chapter_info['title'], - 'thumbnail': chapter_info['preview'], - 'description': chapter_info['description'], - 'uploader': chapter_info['channel']['display_name'], - 'uploader_id': chapter_info['channel']['name'], - } - return [info] - else: - video_id = mobj.group('videoid') - api = api_base + '/broadcast/by_archive/%s.json' % video_id - - self.report_extraction(video_id) - - info = [] - offset = 0 - limit = self._JUSTIN_PAGE_LIMIT - while True: - if paged: - self.report_download_page(video_id, offset) - page_url = api + ('?offset=%d&limit=%d' % (offset, limit)) - page_count, page_info = self._parse_page(page_url, video_id) - info.extend(page_info) - if not paged or page_count != limit: - break - offset += limit - return info diff --git a/youtube_dl/extractor/justintv.py b/youtube_dl/extractor/justintv.py new file mode 100644 index 000000000..9c46baed5 --- /dev/null +++ b/youtube_dl/extractor/justintv.py @@ -0,0 +1,144 @@ +import json +import os +import re +import xml.etree.ElementTree + +from .common import InfoExtractor +from ..utils import ( + ExtractorError, + formatSeconds, +) + + +class JustinTVIE(InfoExtractor): + """Information extractor for justin.tv and twitch.tv""" + # TODO: One broadcast may be split into multiple videos. The key + # 'broadcast_id' is the same for all parts, and 'broadcast_part' + # starts at 1 and increases. Can we treat all parts as one video? + + _VALID_URL = r"""(?x)^(?:http://)?(?:www\.)?(?:twitch|justin)\.tv/ + (?: + (?P<channelid>[^/]+)| + (?:(?:[^/]+)/b/(?P<videoid>[^/]+))| + (?:(?:[^/]+)/c/(?P<chapterid>[^/]+)) + ) + /?(?:\#.*)?$ + """ + _JUSTIN_PAGE_LIMIT = 100 + IE_NAME = u'justin.tv' + + def report_download_page(self, channel, offset): + """Report attempt to download a single page of videos.""" + self.to_screen(u'%s: Downloading video information from %d to %d' % + (channel, offset, offset + self._JUSTIN_PAGE_LIMIT)) + + # Return count of items, list of *valid* items + def _parse_page(self, url, video_id): + webpage = self._download_webpage(url, video_id, + u'Downloading video info JSON', + u'unable to download video info JSON') + + response = json.loads(webpage) + if type(response) != list: + error_text = response.get('error', 'unknown error') + raise ExtractorError(u'Justin.tv API: %s' % error_text) + info = [] + for clip in response: + video_url = clip['video_file_url'] + if video_url: + video_extension = os.path.splitext(video_url)[1][1:] + video_date = re.sub('-', '', clip['start_time'][:10]) + video_uploader_id = clip.get('user_id', clip.get('channel_id')) + video_id = clip['id'] + video_title = clip.get('title', video_id) + info.append({ + 'id': video_id, + 'url': video_url, + 'title': video_title, + 'uploader': clip.get('channel_name', video_uploader_id), + 'uploader_id': video_uploader_id, + 'upload_date': video_date, + 'ext': video_extension, + }) + return (len(response), info) + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + if mobj is None: + raise ExtractorError(u'invalid URL: %s' % url) + + api_base = 'http://api.justin.tv' + paged = False + if mobj.group('channelid'): + paged = True + video_id = mobj.group('channelid') + api = api_base + '/channel/archives/%s.json' % video_id + elif mobj.group('chapterid'): + chapter_id = mobj.group('chapterid') + + webpage = self._download_webpage(url, chapter_id) + m = re.search(r'PP\.archive_id = "([0-9]+)";', webpage) + if not m: + raise ExtractorError(u'Cannot find archive of a chapter') + archive_id = m.group(1) + + api = api_base + '/broadcast/by_chapter/%s.xml' % chapter_id + chapter_info_xml = self._download_webpage(api, chapter_id, + note=u'Downloading chapter information', + errnote=u'Chapter information download failed') + doc = xml.etree.ElementTree.fromstring(chapter_info_xml) + for a in doc.findall('.//archive'): + if archive_id == a.find('./id').text: + break + else: + raise ExtractorError(u'Could not find chapter in chapter information') + + video_url = a.find('./video_file_url').text + video_ext = video_url.rpartition('.')[2] or u'flv' + + chapter_api_url = u'https://api.twitch.tv/kraken/videos/c' + chapter_id + chapter_info_json = self._download_webpage(chapter_api_url, u'c' + chapter_id, + note='Downloading chapter metadata', + errnote='Download of chapter metadata failed') + chapter_info = json.loads(chapter_info_json) + + bracket_start = int(doc.find('.//bracket_start').text) + bracket_end = int(doc.find('.//bracket_end').text) + + # TODO determine start (and probably fix up file) + # youtube-dl -v http://www.twitch.tv/firmbelief/c/1757457 + #video_url += u'?start=' + TODO:start_timestamp + # bracket_start is 13290, but we want 51670615 + self._downloader.report_warning(u'Chapter detected, but we can just download the whole file. ' + u'Chapter starts at %s and ends at %s' % (formatSeconds(bracket_start), formatSeconds(bracket_end))) + + info = { + 'id': u'c' + chapter_id, + 'url': video_url, + 'ext': video_ext, + 'title': chapter_info['title'], + 'thumbnail': chapter_info['preview'], + 'description': chapter_info['description'], + 'uploader': chapter_info['channel']['display_name'], + 'uploader_id': chapter_info['channel']['name'], + } + return [info] + else: + video_id = mobj.group('videoid') + api = api_base + '/broadcast/by_archive/%s.json' % video_id + + self.report_extraction(video_id) + + info = [] + offset = 0 + limit = self._JUSTIN_PAGE_LIMIT + while True: + if paged: + self.report_download_page(video_id, offset) + page_url = api + ('?offset=%d&limit=%d' % (offset, limit)) + page_count, page_info = self._parse_page(page_url, video_id) + info.extend(page_info) + if not paged or page_count != limit: + break + offset += limit + return info |