From: dirkf Date: Tue, 6 Aug 2024 19:59:09 +0000 (+0100) Subject: [Youtube] Rework n function extraction pattern X-Git-Url: http://git.oshgnacknak.de/?a=commitdiff_plain;h=refs%2Fheads%2Fmaster;p=youtube-dl [Youtube] Rework n function extraction pattern Now also succeeds with player b12cc44b --- diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 509e374a4..6fe520e9a 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -1661,23 +1661,33 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c) # or: (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c) # or: (PL(a),b=a.j.n||null)&&(b=nfunc[idx](b) - # old: .get("n"))&&(b=nfunc[idx](b) - # older: .get("n"))&&(b=nfunc(b) + # or: (b="nn"[+a.D],vL(a),c=a.j[b]||null)&&(c=narray[idx](c),a.set(b,c),narray.length||nfunc("") + # old: (b=a.get("n"))&&(b=nfunc[idx](b)(?P[a-z])\s*=\s*[a-z]\s* + # older: (b=a.get("n"))&&(b=nfunc(b) r'''(?x) - (?:\((?:[\w$()\s]+,)*?\s*(?P[a-z])\s*=\s*(?: - String\s*\.\s*fromCharCode\s*\(\s*110\s*\)| - "n+"\[\s*\+?s*[\w$.]+\s*]| - (?P(?:[\w$]+\s*\.\s*)+n\b(?:(?!&&).)+\)) - )\s* - (?(b1) - &&\s*\(\s*(?P=b)| - (?: - ,(?P[a-z])\s*=\s*[a-z]\s*)? - \.\s*get\s*\(\s*(?(b)(?P=b)|"n{1,2}")(?:\s*\)){2}\s* - &&\s*\(\s*(?(c)(?P=c)|(?P=b)) - ) - )\s*=\s* - (?P[a-zA-Z_$][\w$]*)(?:\s*\[(?P\d+)\])?\s*\(\s*[\w$]+\s*\) + \((?:[\w$()\s]+,)*?\s* # ( + (?P[a-z])\s*=\s* # b= + (?: + (?: # expect ,c=a.get(b) (etc) + String\s*\.\s*fromCharCode\s*\(\s*110\s*\)| + "n+"\[\s*\+?s*[\w$.]+\s*] + )\s*(?:,[\w$()\s]+(?=,))*| + (?P[\w$]+) # a (old[er]) + )\s* + (?(old) + # b.get("n") + (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*? + (?:\.\s*n|\[\s*"n"\s*]|\.\s*get\s*\(\s*"n"\s*\)) + | # ,c=a.get(b) + ,\s*(?P[a-z])\s*=\s*[a-z]\s* + (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*? + (?:\[\s*(?P=b)\s*]|\.\s*get\s*\(\s*(?P=b)\s*\)) + ) + # interstitial junk + \s*(?:\|\|\s*null\s*)?(?:\)\s*)?&&\s*(?:\(\s*)? + (?(c)(?P=c)|(?P=b))\s*=\s* # [c|b]= + # nfunc|nfunc[idx] + (?P[a-zA-Z_$][\w$]*)(?:\s*\[(?P\d+)\])?\s*\(\s*[\w$]+\s*\) ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'), default=(None, None)) # thx bashonly: yt-dlp/yt-dlp/pull/10611 @@ -1690,7 +1700,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor): \s*\{(?:(?!};).)+?["']enhanced_except_ ''', jscode, 'Initial JS player n function name', group='name') if not idx: - self.report_warning('Falling back to generic n function search') return func_name return self._parse_json(self._search_regex(