From 9af0f299bfaaa78c4353e2e38f15db67781c55c5 Mon Sep 17 00:00:00 2001 From: Zenon Mousmoulas Date: Thu, 11 Nov 2021 08:21:41 +0200 Subject: [PATCH 1/3] Add Ant1NewsGr IEs * Add new IEs * Ant1NewsGrBaseIE: Base IE class * Ant1NewsGrWatchIE: Extract videos from TV VOD pages * Ant1NewsGrArticleIE: Extract videos from news articles * Ant1NewsGrEmbedIE: Extract iframe-embeddable ant1news.gr videos * NB: There is a different platform at vod.antenna.gr, which is not covered here * The Generic extractor can also be used to extract videos from news article pages (through Ant1NewsGrEmbed._extract_urls), however a specific IE is used to extract JSON-LD @type='NewsArticle' metadata * However that does not work currently, as _json_ld does not support @graph nesting * Ant1NewsGrArticleIE defers to Ant1NewsGrEmbedIE, either as a playlist or a single video * Ant1NewsGrWatchIE and Ant1NewsGrEmbedIE query an API endpoint to extract metadata, get the respective stream/source URLs and detect video formats * The endpoint HTTP path varies per IE * Ant1NewsGrEmbedIE first resolves any redirects for its' own URL, to derive the properl base URL for the API query --- youtube_dl/extractor/ant1newsgr.py | 190 +++++++++++++++++++++++++++++ youtube_dl/extractor/extractors.py | 5 + youtube_dl/extractor/generic.py | 8 ++ 3 files changed, 203 insertions(+) create mode 100644 youtube_dl/extractor/ant1newsgr.py diff --git a/youtube_dl/extractor/ant1newsgr.py b/youtube_dl/extractor/ant1newsgr.py new file mode 100644 index 000000000..ac50accdc --- /dev/null +++ b/youtube_dl/extractor/ant1newsgr.py @@ -0,0 +1,190 @@ +# coding: utf-8 +from __future__ import unicode_literals + +import re + +from .common import InfoExtractor +from ..compat import ( + compat_parse_qs, + compat_urllib_parse_urlparse, + compat_urlparse, +) +from ..utils import ( + HEADRequest, + ExtractorError, + determine_ext, + smuggle_url, + unsmuggle_url, + unescapeHTML, +) + + +class Ant1NewsGrBaseIE(InfoExtractor): + @staticmethod + def _smuggle_parent_info(url, **info_dict): + return smuggle_url(url, {'parent_info': info_dict}) + + @staticmethod + def _unsmuggle_parent_info(url): + unsmuggled_url, data = unsmuggle_url(url, default={'parent_info': {}}) + return unsmuggled_url, data['parent_info'] + + def _download_api_data(self, netloc, cid, scheme='https'): + url_parts = (scheme, netloc, self._API_PATH, None, None, None) + url = compat_urlparse.urlunparse(url_parts) + query = {'cid': cid} + return self._download_json( + url, cid, + 'Downloading JSON', + 'Unable to download JSON', + query=query) + + def _download_and_extract_api_data(self, video_id, *args, **kwargs): + info = self._download_api_data(*args, **kwargs) + try: + source = info['url'] + except KeyError: + raise ExtractorError('no source found for %s' % video_id) + formats = self._extract_m3u8_formats(source, video_id, 'mp4') \ + if determine_ext(source) == 'm3u8' else [source] + self._sort_formats(formats) + return { + 'id': video_id, + 'title': info['title'], + 'thumbnail': info['thumb'], + 'formats': formats, + } + + +class Ant1NewsGrWatchIE(Ant1NewsGrBaseIE): + IE_NAME = 'ant1newsgr:watch' + IE_DESC = 'ant1news.gr videos' + _VALID_URL = r'https?://(?:www\.)?ant1news\.gr/watch/(?P\d+)/' + _API_PATH = '/templates/data/player' + + _TEST = { + 'url': 'https://www.ant1news.gr/watch/1506168/ant1-news-09112021-stis-18-45', + 'md5': '60a984da5ffc98c9924e6d9dd46c6f04', + 'info_dict': { + 'id': '1506168', + 'ext': 'mp4', + 'title': 'md5:0ad00fa66ecf8aa233d26ab0dba7514a', + 'description': 'md5:18665af715a6dcfeac1d6153a44f16b0', + }, + } + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + scheme, netloc, _, _, _, _ = compat_urllib_parse_urlparse(url) + info = self._download_and_extract_api_data( + video_id, netloc, video_id, scheme=scheme) + info['description'] = self._og_search_description(webpage) + return info + + +class Ant1NewsGrArticleIE(Ant1NewsGrBaseIE): + IE_NAME = 'ant1newsgr:article' + IE_DESC = 'ant1news.gr articles' + _VALID_URL = r'https?://(?:www\.)?ant1news\.gr/[^/]+/article/(?P\d+)/' + + _TESTS = [{ + 'url': 'https://www.ant1news.gr/afieromata/article/549468/o-tzeims-mpont-sta-meteora-oi-apeiles-kai-o-xesikomos-ton-kalogeron', + 'md5': 'eb635a194c15272c2611a751766b0200', + 'info_dict': { + 'id': '_xvg/m_cmbatw=', + 'ext': 'mp4', + 'title': 'md5:a93e8ecf2e4073bfdffcb38f59945411', + }, + 'expected_warnings': [r'^[Uu]nable to extract JSON-LD'], + }, { + 'url': 'https://ant1news.gr/Society/article/620286/symmoria-anilikon-dikigoros-thymaton-ithelan-na-toys-apoteleiosoyn', + 'info_dict': { + 'id': '620286', + 'title': 'md5:91fe569e952e4d146485740ae927662b', + }, + 'expected_warnings': [r'^[Uu]nable to extract JSON-LD'], + 'playlist_mincount': 2, + 'params': { + 'skip_download': True, + }, + }] + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + info = self._search_json_ld(webpage, video_id, + expected_type='NewsArticle', + fatal=False) + # workaround as _json_ld does not recognize @graph nesting + if not info: + info['title'] = self._og_search_title(webpage) + embed_urls = list(Ant1NewsGrEmbedIE._extract_urls(webpage, url, **info)) + if not embed_urls: + raise ExtractorError('no videos found for %s' % video_id) + if len(embed_urls) == 1: + return self.url_result(embed_urls[0], ie=Ant1NewsGrEmbedIE.ie_key(), + video_title=info['title']) + return self.playlist_from_matches( + embed_urls, video_id, info['title'], ie=Ant1NewsGrEmbedIE.ie_key()) + + +class Ant1NewsGrEmbedIE(Ant1NewsGrBaseIE): + IE_NAME = 'ant1newsgr:embed' + IE_DESC = 'ant1news.gr embedded videos' + _VALID_URL = r'''(?x)https?://(?:[a-zA-Z0-9\-]+\.)? + (?:antenna|ant1news)\.gr/templates/pages/player + \?(?:(?:cid=(?P[^&#]+)|[^&=#]+=[^&#]+)&?)+''' + _API_PATH = '/news/templates/data/jsonPlayer' + + _TEST = { + 'url': 'https://www.antenna.gr/templates/pages/player?cid=3f_li_c_az_jw_y_u=&w=670&h=377', + 'md5': '12872b12af18b5dbf76528786728de8c', + 'info_dict': { + 'id': '3f_li_c_az_jw_y_u=', + 'ext': 'mp4', + 'title': 'md5:a30c93332455f53e1e84ae0724f0adf7', + }, + } + + @classmethod + def _extract_urls(cls, webpage, origin_url=None, **parent_info): + # make the scheme in _VALID_URL optional + _URL_RE = r'(?:https?:)?//' + cls._VALID_URL.split('://', 1)[1] + # simplify the query string part of _VALID_URL; after extracting iframe + # src, the URL will be matched again + _URL_RE = _URL_RE.split(r'\?', 1)[0] + r'\?(?:(?!(?P=_q1)).)+' + EMBED_RE = r'''(?x) + ]+?src=(?P<_q1>%(quot_re)s)(?P%(url_re)s)(?P=_q1) + ''' % {'quot_re': r'[\"\']', 'url_re': _URL_RE} + for mobj in re.finditer(EMBED_RE, webpage): + url = unescapeHTML(mobj.group('url')) + if url.startswith('//'): + scheme = compat_urllib_parse_urlparse(origin_url).scheme \ + if origin_url else 'https' + url = '%s:%s' % (scheme, url) + if not cls.suitable(url): + continue + if isinstance(parent_info, dict): + url = cls._smuggle_parent_info(url, **parent_info) + yield url + + def _real_extract(self, url): + url, parent_info = type(self)._unsmuggle_parent_info(url) + video_id = self._match_id(url) + + # resolve any redirects, to derive the proper base URL for the API query + canonical_url = self._request_webpage( + HEADRequest(url), video_id, + note='Resolve canonical player URL', + errnote='Could not resolve canonical player URL').geturl() + scheme, netloc, _, _, query, _ = compat_urllib_parse_urlparse( + canonical_url) + query = compat_parse_qs(query) + cid = query['cid'][0] + + info = self._download_and_extract_api_data( + video_id, netloc, cid, scheme=scheme) + if 'timestamp' not in info and 'timestamp' in parent_info: + info['timestamp'] = parent_info['timestamp'] + return info diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py index 6e8fc3961..21d8047a5 100644 --- a/youtube_dl/extractor/extractors.py +++ b/youtube_dl/extractor/extractors.py @@ -1038,6 +1038,11 @@ from .rutube import ( RutubePersonIE, RutubePlaylistIE, ) +from .ant1newsgr import ( + Ant1NewsGrWatchIE, + Ant1NewsGrArticleIE, + Ant1NewsGrEmbedIE, +) from .rutv import RUTVIE from .ruutu import RuutuIE from .ruv import RuvIE diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py index a9c064105..36c90dcc4 100644 --- a/youtube_dl/extractor/generic.py +++ b/youtube_dl/extractor/generic.py @@ -100,6 +100,7 @@ from .ustream import UstreamIE from .arte import ArteTVEmbedIE from .videopress import VideoPressIE from .rutube import RutubeIE +from .ant1newsgr import Ant1NewsGrEmbedIE from .limelight import LimelightBaseIE from .anvato import AnvatoIE from .washingtonpost import WashingtonPostIE @@ -3199,6 +3200,13 @@ class GenericIE(InfoExtractor): return self.playlist_from_matches( rutube_urls, video_id, video_title, ie=RutubeIE.ie_key()) + # Look for ant1news.gr embeds + ant1newsgr_urls = list(Ant1NewsGrEmbedIE._extract_urls(webpage, url, + title=video_title)) + if ant1newsgr_urls: + return self.playlist_from_matches( + ant1newsgr_urls, video_id, video_title, ie=Ant1NewsGrEmbedIE.ie_key()) + # Look for WashingtonPost embeds wapo_urls = WashingtonPostIE._extract_urls(webpage) if wapo_urls: From 264544f90e5fccd4efb70a4babf6db4fc35bf70e Mon Sep 17 00:00:00 2001 From: Zenon Mousmoulas Date: Sat, 13 Nov 2021 08:45:46 +0200 Subject: [PATCH 2/3] Remove unnecessary quote escape --- youtube_dl/extractor/ant1newsgr.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/youtube_dl/extractor/ant1newsgr.py b/youtube_dl/extractor/ant1newsgr.py index ac50accdc..794e241f7 100644 --- a/youtube_dl/extractor/ant1newsgr.py +++ b/youtube_dl/extractor/ant1newsgr.py @@ -156,7 +156,7 @@ class Ant1NewsGrEmbedIE(Ant1NewsGrBaseIE): _URL_RE = _URL_RE.split(r'\?', 1)[0] + r'\?(?:(?!(?P=_q1)).)+' EMBED_RE = r'''(?x) ]+?src=(?P<_q1>%(quot_re)s)(?P%(url_re)s)(?P=_q1) - ''' % {'quot_re': r'[\"\']', 'url_re': _URL_RE} + ''' % {'quot_re': r'["\']', 'url_re': _URL_RE} for mobj in re.finditer(EMBED_RE, webpage): url = unescapeHTML(mobj.group('url')) if url.startswith('//'): From 3872619ed5210e8a60f68cff7dbb64164bdc7b5c Mon Sep 17 00:00:00 2001 From: Zenon Mousmoulas Date: Sat, 13 Nov 2021 11:55:12 +0200 Subject: [PATCH 3/3] Ant1NewsGrEmbedIE._extract_urls: Simplify redundant statement --- youtube_dl/extractor/ant1newsgr.py | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/youtube_dl/extractor/ant1newsgr.py b/youtube_dl/extractor/ant1newsgr.py index 794e241f7..e1d3bb0a9 100644 --- a/youtube_dl/extractor/ant1newsgr.py +++ b/youtube_dl/extractor/ant1newsgr.py @@ -165,9 +165,7 @@ class Ant1NewsGrEmbedIE(Ant1NewsGrBaseIE): url = '%s:%s' % (scheme, url) if not cls.suitable(url): continue - if isinstance(parent_info, dict): - url = cls._smuggle_parent_info(url, **parent_info) - yield url + yield cls._smuggle_parent_info(url, **parent_info) def _real_extract(self, url): url, parent_info = type(self)._unsmuggle_parent_info(url)