diff options
Diffstat (limited to 'hypervideo_dl/extractor/archiveorg.py')
| -rw-r--r-- | hypervideo_dl/extractor/archiveorg.py | 244 |
1 files changed, 240 insertions, 4 deletions
diff --git a/hypervideo_dl/extractor/archiveorg.py b/hypervideo_dl/extractor/archiveorg.py index 90dda9f..2541cd6 100644 --- a/hypervideo_dl/extractor/archiveorg.py +++ b/hypervideo_dl/extractor/archiveorg.py @@ -3,12 +3,14 @@ import re import urllib.parse from .common import InfoExtractor +from .naver import NaverBaseIE from .youtube import YoutubeBaseInfoExtractor, YoutubeIE -from ..compat import compat_HTTPError, compat_urllib_parse_unquote +from ..compat import compat_urllib_parse_unquote +from ..networking import HEADRequest +from ..networking.exceptions import HTTPError from ..utils import ( KNOWN_EXTENSIONS, ExtractorError, - HEADRequest, bug_reports_message, clean_html, dict_get, @@ -897,7 +899,7 @@ class YoutubeWebArchiveIE(InfoExtractor): video_id, note='Fetching archived video file url', expected_status=True) except ExtractorError as e: # HTTP Error 404 is expected if the video is not saved. - if isinstance(e.cause, compat_HTTPError) and e.cause.code == 404: + if isinstance(e.cause, HTTPError) and e.cause.status == 404: self.raise_no_formats( 'The requested video is not archived, indexed, or there is an issue with web.archive.org (try again later)', expected=True) else: @@ -924,7 +926,7 @@ class YoutubeWebArchiveIE(InfoExtractor): info['thumbnails'] = self._extract_thumbnails(video_id) if urlh: - url = compat_urllib_parse_unquote(urlh.geturl()) + url = compat_urllib_parse_unquote(urlh.url) video_file_url_qs = parse_qs(url) # Attempt to recover any ext & format info from playback url & response headers format = {'url': url, 'filesize': int_or_none(urlh.headers.get('x-archive-orig-content-length'))} @@ -945,3 +947,237 @@ class YoutubeWebArchiveIE(InfoExtractor): if not info.get('title'): info['title'] = video_id return info + + +class VLiveWebArchiveIE(InfoExtractor): + IE_NAME = 'web.archive:vlive' + IE_DESC = 'web.archive.org saved vlive videos' + _VALID_URL = r'''(?x) + (?:https?://)?web\.archive\.org/ + (?:web/)?(?:(?P<date>[0-9]{14})?[0-9A-Za-z_*]*/)? # /web and the version index is optional + (?:https?(?::|%3[Aa])//)?(?: + (?:(?:www|m)\.)?vlive\.tv(?::(?:80|443))?/(?:video|embed)/(?P<id>[0-9]+) # VLive URL + ) + ''' + _TESTS = [{ + 'url': 'https://web.archive.org/web/20221221144331/http://www.vlive.tv/video/1326', + 'md5': 'cc7314812855ce56de70a06a27314983', + 'info_dict': { + 'id': '1326', + 'ext': 'mp4', + 'title': "Girl's Day's Broadcast", + 'creator': "Girl's Day", + 'view_count': int, + 'uploader_id': 'muploader_a', + 'uploader_url': None, + 'uploader': None, + 'upload_date': '20150817', + 'thumbnail': r're:^https?://.*\.(?:jpg|png)$', + 'timestamp': 1439816449, + 'like_count': int, + 'channel': 'Girl\'s Day', + 'channel_id': 'FDF27', + 'comment_count': int, + 'release_timestamp': 1439818140, + 'release_date': '20150817', + 'duration': 1014, + }, + 'params': { + 'skip_download': True, + }, + }, { + 'url': 'https://web.archive.org/web/20221221182103/http://www.vlive.tv/video/16937', + 'info_dict': { + 'id': '16937', + 'ext': 'mp4', + 'title': '첸백시 걍방', + 'creator': 'EXO', + 'view_count': int, + 'subtitles': 'mincount:12', + 'uploader_id': 'muploader_j', + 'uploader_url': 'http://vlive.tv', + 'uploader': None, + 'upload_date': '20161112', + 'thumbnail': r're:^https?://.*\.(?:jpg|png)$', + 'timestamp': 1478923074, + 'like_count': int, + 'channel': 'EXO', + 'channel_id': 'F94BD', + 'comment_count': int, + 'release_timestamp': 1478924280, + 'release_date': '20161112', + 'duration': 906, + }, + 'params': { + 'skip_download': True, + }, + }, { + 'url': 'https://web.archive.org/web/20221127190050/http://www.vlive.tv/video/101870', + 'info_dict': { + 'id': '101870', + 'ext': 'mp4', + 'title': '[ⓓ xV] “레벨이들 매력에 반해? 안 반해?” 움직이는 HD 포토 (레드벨벳:Red Velvet)', + 'creator': 'Dispatch', + 'view_count': int, + 'subtitles': 'mincount:6', + 'uploader_id': 'V__FRA08071', + 'uploader_url': 'http://vlive.tv', + 'uploader': None, + 'upload_date': '20181130', + 'thumbnail': r're:^https?://.*\.(?:jpg|png)$', + 'timestamp': 1543601327, + 'like_count': int, + 'channel': 'Dispatch', + 'channel_id': 'C796F3', + 'comment_count': int, + 'release_timestamp': 1543601040, + 'release_date': '20181130', + 'duration': 279, + }, + 'params': { + 'skip_download': True, + }, + }] + + # The wayback machine has special timestamp and "mode" values: + # timestamp: + # 1 = the first capture + # 2 = the last capture + # mode: + # id_ = Identity - perform no alterations of the original resource, return it as it was archived. + _WAYBACK_BASE_URL = 'https://web.archive.org/web/2id_/' + + def _download_archived_page(self, url, video_id, *, timestamp='2', **kwargs): + for retry in self.RetryManager(): + try: + return self._download_webpage(f'https://web.archive.org/web/{timestamp}id_/{url}', video_id, **kwargs) + except ExtractorError as e: + if isinstance(e.cause, HTTPError) and e.cause.status == 404: + raise ExtractorError('Page was not archived', expected=True) + retry.error = e + continue + + def _download_archived_json(self, url, video_id, **kwargs): + page = self._download_archived_page(url, video_id, **kwargs) + if not page: + raise ExtractorError('Page was not archived', expected=True) + else: + return self._parse_json(page, video_id) + + def _extract_formats_from_m3u8(self, m3u8_url, params, video_id): + m3u8_doc = self._download_archived_page(m3u8_url, video_id, note='Downloading m3u8', query=params, fatal=False) + if not m3u8_doc: + return + + # M3U8 document should be changed to archive domain + m3u8_doc = m3u8_doc.splitlines() + url_base = m3u8_url.rsplit('/', 1)[0] + first_segment = None + for i, line in enumerate(m3u8_doc): + if not line.startswith('#'): + m3u8_doc[i] = f'{self._WAYBACK_BASE_URL}{url_base}/{line}?{urllib.parse.urlencode(params)}' + first_segment = first_segment or m3u8_doc[i] + + # Segments may not have been archived. See https://web.archive.org/web/20221127190050/http://www.vlive.tv/video/101870 + urlh = self._request_webpage(HEADRequest(first_segment), video_id, errnote=False, + fatal=False, note='Check first segment availablity') + if urlh: + formats, subtitles = self._parse_m3u8_formats_and_subtitles('\n'.join(m3u8_doc), ext='mp4', video_id=video_id) + if subtitles: + self._report_ignoring_subs('m3u8') + return formats + + # Closely follows the logic of the ArchiveTeam grab script + # See: https://github.com/ArchiveTeam/vlive-grab/blob/master/vlive.lua + def _real_extract(self, url): + video_id, url_date = self._match_valid_url(url).group('id', 'date') + + webpage = self._download_archived_page(f'https://www.vlive.tv/video/{video_id}', video_id, timestamp=url_date) + + player_info = self._search_json(r'__PRELOADED_STATE__\s*=', webpage, 'player info', video_id) + user_country = traverse_obj(player_info, ('common', 'userCountry')) + + main_script_url = self._search_regex(r'<script\s+src="([^"]+/js/main\.[^"]+\.js)"', webpage, 'main script url') + main_script = self._download_archived_page(main_script_url, video_id, note='Downloading main script') + app_id = self._search_regex(r'appId\s*=\s*"([^"]+)"', main_script, 'app id') + + inkey = self._download_archived_json( + f'https://www.vlive.tv/globalv-web/vam-web/video/v1.0/vod/{video_id}/inkey', video_id, note='Fetching inkey', query={ + 'appId': app_id, + 'platformType': 'PC', + 'gcc': user_country, + 'locale': 'en_US', + }, fatal=False) + + vod_id = traverse_obj(player_info, ('postDetail', 'post', 'officialVideo', 'vodId')) + + vod_data = self._download_archived_json( + f'https://apis.naver.com/rmcnmv/rmcnmv/vod/play/v2.0/{vod_id}', video_id, note='Fetching vod data', query={ + 'key': inkey.get('inkey'), + 'pid': 'rmcPlayer_16692457559726800', # partially unix time and partially random. Fixed value used by archiveteam project + 'sid': '2024', + 'ver': '2.0', + 'devt': 'html5_pc', + 'doct': 'json', + 'ptc': 'https', + 'sptc': 'https', + 'cpt': 'vtt', + 'ctls': '%7B%22visible%22%3A%7B%22fullscreen%22%3Atrue%2C%22logo%22%3Afalse%2C%22playbackRate%22%3Afalse%2C%22scrap%22%3Afalse%2C%22playCount%22%3Atrue%2C%22commentCount%22%3Atrue%2C%22title%22%3Atrue%2C%22writer%22%3Atrue%2C%22expand%22%3Afalse%2C%22subtitles%22%3Atrue%2C%22thumbnails%22%3Atrue%2C%22quality%22%3Atrue%2C%22setting%22%3Atrue%2C%22script%22%3Afalse%2C%22logoDimmed%22%3Atrue%2C%22badge%22%3Atrue%2C%22seekingTime%22%3Atrue%2C%22muted%22%3Atrue%2C%22muteButton%22%3Afalse%2C%22viewerNotice%22%3Afalse%2C%22linkCount%22%3Afalse%2C%22createTime%22%3Afalse%2C%22thumbnail%22%3Atrue%7D%2C%22clicked%22%3A%7B%22expand%22%3Afalse%2C%22subtitles%22%3Afalse%7D%7D', + 'pv': '4.26.9', + 'dr': '1920x1080', + 'cpl': 'en_US', + 'lc': 'en_US', + 'adi': '%5B%7B%22type%22%3A%22pre%22%2C%22exposure%22%3Afalse%2C%22replayExposure%22%3Afalse%7D%5D', + 'adu': '%2F', + 'videoId': vod_id, + 'cc': user_country, + }) + + formats = [] + + streams = traverse_obj(vod_data, ('streams', ...)) + if len(streams) > 1: + self.report_warning('Multiple streams found. Only the first stream will be downloaded.') + stream = streams[0] + + max_stream = max( + stream.get('videos') or [], + key=lambda v: traverse_obj(v, ('bitrate', 'video'), default=0), default=None) + if max_stream is not None: + params = {arg.get('name'): arg.get('value') for arg in stream.get('keys', []) if arg.get('type') == 'param'} + formats = self._extract_formats_from_m3u8(max_stream.get('source'), params, video_id) or [] + + # For parts of the project MP4 files were archived + max_video = max( + traverse_obj(vod_data, ('videos', 'list', ...)), + key=lambda v: traverse_obj(v, ('bitrate', 'video'), default=0), default=None) + if max_video is not None: + video_url = self._WAYBACK_BASE_URL + max_video.get('source') + urlh = self._request_webpage(HEADRequest(video_url), video_id, errnote=False, + fatal=False, note='Check video availablity') + if urlh: + formats.append({'url': video_url}) + + return { + 'id': video_id, + 'formats': formats, + **traverse_obj(player_info, ('postDetail', 'post', { + 'title': ('officialVideo', 'title', {str}), + 'creator': ('author', 'nickname', {str}), + 'channel': ('channel', 'channelName', {str}), + 'channel_id': ('channel', 'channelCode', {str}), + 'duration': ('officialVideo', 'playTime', {int_or_none}), + 'view_count': ('officialVideo', 'playCount', {int_or_none}), + 'like_count': ('officialVideo', 'likeCount', {int_or_none}), + 'comment_count': ('officialVideo', 'commentCount', {int_or_none}), + 'timestamp': ('officialVideo', 'createdAt', {lambda x: int_or_none(x, scale=1000)}), + 'release_timestamp': ('officialVideo', 'willStartAt', {lambda x: int_or_none(x, scale=1000)}), + })), + **traverse_obj(vod_data, ('meta', { + 'uploader_id': ('user', 'id', {str}), + 'uploader': ('user', 'name', {str}), + 'uploader_url': ('user', 'url', {url_or_none}), + 'thumbnail': ('cover', 'source', {url_or_none}), + }), expected_type=lambda x: x or None), + **NaverBaseIE.process_subtitles(vod_data, lambda x: [self._WAYBACK_BASE_URL + x]), + } |