mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-27 14:14:54 +01:00
[youtube] Improve signature caching
and refactor related functions
This commit is contained in:
pukkandan
parent
2f1a299c50
commit
580ce00782
|
|
@ -2512,20 +2512,17 @@ def _extract_signature_function(self, video_id, player_url, example_sig):
|
|||
assert os.path.basename(func_id) == func_id
|
||||
|
||||
self.write_debug(f'Extracting signature function {func_id}')
|
||||
cache_spec = self.cache.load('youtube-sigfuncs', func_id)
|
||||
if cache_spec is not None:
|
||||
return lambda s: ''.join(s[i] for i in cache_spec)
|
||||
cache_spec, code = self.cache.load('youtube-sigfuncs', func_id), None
|
||||
|
||||
if not cache_spec:
|
||||
code = self._load_player(video_id, player_url)
|
||||
if code:
|
||||
res = self._parse_sig_js(code)
|
||||
|
||||
test_string = ''.join(map(chr, range(len(example_sig))))
|
||||
cache_res = res(test_string)
|
||||
cache_spec = [ord(c) for c in cache_res]
|
||||
|
||||
cache_spec = [ord(c) for c in res(test_string)]
|
||||
self.cache.store('youtube-sigfuncs', func_id, cache_spec)
|
||||
return res
|
||||
|
||||
return lambda s: ''.join(s[i] for i in cache_spec)
|
||||
|
||||
def _print_sig_code(self, func, example_sig):
|
||||
if not self.get_param('youtube_print_sig_code'):
|
||||
|
|
@ -2593,18 +2590,29 @@ def _parse_sig_js(self, jscode):
|
|||
initial_function = jsi.extract_function(funcname)
|
||||
return lambda s: initial_function([s])
|
||||
|
||||
def _cached(self, func, *cache_id):
|
||||
def inner(*args, **kwargs):
|
||||
if cache_id not in self._player_cache:
|
||||
try:
|
||||
self._player_cache[cache_id] = func(*args, **kwargs)
|
||||
except ExtractorError as e:
|
||||
self._player_cache[cache_id] = e
|
||||
except Exception as e:
|
||||
self._player_cache[cache_id] = ExtractorError(traceback.format_exc(), cause=e)
|
||||
|
||||
ret = self._player_cache[cache_id]
|
||||
if isinstance(ret, Exception):
|
||||
raise ret
|
||||
return ret
|
||||
return inner
|
||||
|
||||
def _decrypt_signature(self, s, video_id, player_url):
|
||||
"""Turn the encrypted s field into a working signature"""
|
||||
try:
|
||||
player_id = (player_url, self._signature_cache_id(s))
|
||||
if player_id not in self._player_cache:
|
||||
func = self._extract_signature_function(video_id, player_url, s)
|
||||
self._player_cache[player_id] = func
|
||||
func = self._player_cache[player_id]
|
||||
extract_sig = self._cached(
|
||||
self._extract_signature_function, 'sig', player_url, self._signature_cache_id(s))
|
||||
func = extract_sig(video_id, player_url, s)
|
||||
self._print_sig_code(func, s)
|
||||
return func(s)
|
||||
except Exception as e:
|
||||
raise ExtractorError(traceback.format_exc(), cause=e, video_id=video_id)
|
||||
|
||||
def _decrypt_nsig(self, s, video_id, player_url):
|
||||
"""Turn the encrypted n field into a working signature"""
|
||||
|
|
@ -2612,54 +2620,47 @@ def _decrypt_nsig(self, s, video_id, player_url):
|
|||
raise ExtractorError('Cannot decrypt nsig without player_url')
|
||||
player_url = urljoin('https://www.youtube.com', player_url)
|
||||
|
||||
sig_id = ('nsig_value', s)
|
||||
if sig_id in self._player_cache:
|
||||
return self._player_cache[sig_id]
|
||||
|
||||
try:
|
||||
player_id = ('nsig', player_url)
|
||||
if player_id not in self._player_cache:
|
||||
self._player_cache[player_id] = self._extract_n_function(video_id, player_url)
|
||||
func = self._player_cache[player_id]
|
||||
self._player_cache[sig_id] = func(s)
|
||||
self.write_debug(f'Decrypted nsig {s} => {self._player_cache[sig_id]}')
|
||||
return self._player_cache[sig_id]
|
||||
except Exception as e:
|
||||
raise ExtractorError(traceback.format_exc(), cause=e, video_id=video_id)
|
||||
|
||||
def _extract_n_function_name(self, jscode):
|
||||
nfunc, idx = self._search_regex(
|
||||
r'\.get\("n"\)\)&&\(b=(?P<nfunc>[a-zA-Z0-9$]+)(?:\[(?P<idx>\d+)\])?\([a-zA-Z0-9]\)',
|
||||
jscode, 'Initial JS player n function name', group=('nfunc', 'idx'))
|
||||
if not idx:
|
||||
return nfunc
|
||||
return json.loads(js_to_json(self._search_regex(
|
||||
rf'var {re.escape(nfunc)}\s*=\s*(\[.+?\]);', jscode,
|
||||
f'Initial JS player n function list ({nfunc}.{idx})')))[int(idx)]
|
||||
|
||||
def _extract_n_function(self, video_id, player_url):
|
||||
player_id = self._extract_player_info(player_url)
|
||||
func_code = self.cache.load('youtube-nsig', player_id)
|
||||
|
||||
if func_code:
|
||||
jsi = JSInterpreter(func_code)
|
||||
else:
|
||||
jscode = self._load_player(video_id, player_url)
|
||||
funcname = self._extract_n_function_name(jscode)
|
||||
jsi = JSInterpreter(jscode)
|
||||
func_code = jsi.extract_function_code(funcname)
|
||||
self.cache.store('youtube-nsig', player_id, func_code)
|
||||
|
||||
jsi, player_id, func_code = self._extract_n_function_code(video_id, player_url)
|
||||
if self.get_param('youtube_print_sig_code'):
|
||||
self.to_screen(f'Extracted nsig function from {player_id}:\n{func_code[1]}\n')
|
||||
|
||||
extract_nsig = self._cached(self._extract_n_function_from_code, 'nsig func', player_url)
|
||||
ret = extract_nsig(jsi, func_code)(s)
|
||||
|
||||
self.write_debug(f'Decrypted nsig {s} => {ret}')
|
||||
return ret
|
||||
|
||||
def _extract_n_function_code(self, video_id, player_url):
|
||||
player_id = self._extract_player_info(player_url)
|
||||
func_code = self.cache.load('youtube-nsig', player_id)
|
||||
jscode = func_code or self._load_player(video_id, player_url)
|
||||
jsi = JSInterpreter(jscode)
|
||||
|
||||
if func_code:
|
||||
return jsi, player_id, func_code
|
||||
|
||||
funcname, idx = self._search_regex(
|
||||
r'\.get\("n"\)\)&&\(b=(?P<nfunc>[a-zA-Z0-9$]+)(?:\[(?P<idx>\d+)\])?\([a-zA-Z0-9]\)',
|
||||
jscode, 'Initial JS player n function name', group=('nfunc', 'idx'))
|
||||
if idx:
|
||||
funcname = json.loads(js_to_json(self._search_regex(
|
||||
rf'var {re.escape(funcname)}\s*=\s*(\[.+?\]);', jscode,
|
||||
f'Initial JS player n function list ({funcname}.{idx})')))[int(idx)]
|
||||
|
||||
func_code = jsi.extract_function_code(funcname)
|
||||
self.cache.store('youtube-nsig', player_id, func_code)
|
||||
return jsi, player_id, func_code
|
||||
|
||||
def _extract_n_function_from_code(self, jsi, func_code):
|
||||
func = jsi.extract_function_from_code(*func_code)
|
||||
|
||||
def inner(s):
|
||||
def extract_nsig(s):
|
||||
ret = func([s])
|
||||
if ret.startswith('enhanced_except_'):
|
||||
raise ExtractorError('Signature function returned an exception')
|
||||
return ret
|
||||
return inner
|
||||
|
||||
return extract_nsig
|
||||
|
||||
def _extract_signature_timestamp(self, video_id, player_url, ytcfg=None, fatal=False):
|
||||
"""
|
||||
|
|
@ -3225,7 +3226,8 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, i
|
|||
self._decrypt_signature(encrypted_sig, video_id, player_url)
|
||||
)
|
||||
except ExtractorError as e:
|
||||
self.report_warning('Signature extraction failed: Some formats may be missing', only_once=True)
|
||||
self.report_warning('Signature extraction failed: Some formats may be missing',
|
||||
video_id=video_id, only_once=True)
|
||||
self.write_debug(e, only_once=True)
|
||||
continue
|
||||
|
||||
|
|
@ -3233,12 +3235,14 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, i
|
|||
throttled = False
|
||||
if query.get('n'):
|
||||
try:
|
||||
decrypt_nsig = self._cached(self._decrypt_nsig, 'nsig', query['n'][0])
|
||||
fmt_url = update_url_query(fmt_url, {
|
||||
'n': self._decrypt_nsig(query['n'][0], video_id, player_url)})
|
||||
'n': decrypt_nsig(query['n'][0], video_id, player_url)
|
||||
})
|
||||
except ExtractorError as e:
|
||||
self.report_warning(
|
||||
'nsig extraction failed: You may experience throttling for some formats\n'
|
||||
f'n = {query["n"][0]} ; player = {player_url}', only_once=True)
|
||||
f'n = {query["n"][0]} ; player = {player_url}', video_id=video_id, only_once=True)
|
||||
self.write_debug(e, only_once=True)
|
||||
throttled = True
|
||||
|
||||
|
|
|
|||
Loading…
Reference in New Issue
Block a user