mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2025-07-27 07:38:30 +00:00
diff be gone
This commit is contained in:
parent
d3d29be050
commit
8f26a9d016
53
changed.diff
53
changed.diff
@ -1,53 +0,0 @@
|
|||||||
diff --git a/yt_dlp/extractor/common.py b/yt_dlp/extractor/common.py
|
|
||||||
index b2af2a67b..e2b94c89a 100644
|
|
||||||
--- a/yt_dlp/extractor/common.py
|
|
||||||
+++ b/yt_dlp/extractor/common.py
|
|
||||||
@@ -1807,9 +1807,8 @@ def flatten(flight_data):
|
|
||||||
flatten(f)
|
|
||||||
|
|
||||||
flight_text = ''
|
|
||||||
- # The flight segments regex pattern can afford to be (and should be) strict
|
|
||||||
- # Ref: https://github.com/vercel/next.js/commit/5a4a08fdce91a038f2ed3a70568d3ed040403150
|
|
||||||
- # /packages/next/src/server/app-render/use-flight-response.tsx
|
|
||||||
+ # The non JSON part is written as string in the next.js source and should be matched strictly
|
|
||||||
+ # Ref: https://github.com/vercel/next.js/blob/5a4a08fdce91a038f2ed3a70568d3ed040403150/packages/next/src/server/app-render/use-flight-response.tsx#L189
|
|
||||||
for flight_segment in re.findall(r'<script[^>]*>self\.__next_f\.push\((\[.+?\])\)</script>', webpage):
|
|
||||||
segment = self._parse_json(flight_segment, video_id, fatal=fatal, errnote=None if fatal else False)
|
|
||||||
# Some earlier versions of next.js "optimized" away this array structure; this is unsupported
|
|
||||||
@@ -1818,27 +1817,26 @@ def flatten(flight_data):
|
|
||||||
self.write_debug(
|
|
||||||
f'{video_id}: Unsupported next.js flight data structure detected', only_once=True)
|
|
||||||
continue
|
|
||||||
+ # Ignore useless payload types (1: data, 2: base64)
|
|
||||||
+ # Ref: https://github.com/vercel/next.js/blob/5a4a08fdce91a038f2ed3a70568d3ed040403150/packages/next/src/server/app-render/use-flight-response.tsx#L11-#L14
|
|
||||||
payload_type, chunk = segment
|
|
||||||
- if payload_type == 3:
|
|
||||||
+ if payload_type == 1:
|
|
||||||
+ flight_text += chunk
|
|
||||||
+ elif payload_type == 3:
|
|
||||||
try:
|
|
||||||
- chunk = base64.b64decode(chunk).decode()
|
|
||||||
+ flight_text += base64.b64decode(chunk).decode()
|
|
||||||
except (ValueError, binascii.Error):
|
|
||||||
msg = 'Unable to parse next.js data: unable to decode flight data'
|
|
||||||
if not fatal:
|
|
||||||
self.report_warning(msg, video_id=video_id, only_once=True)
|
|
||||||
continue
|
|
||||||
raise ExtractorError(msg)
|
|
||||||
- elif payload_type != 1:
|
|
||||||
- # Ignore useless payload types (0: bootstrap, 2: form state)
|
|
||||||
- continue
|
|
||||||
- flight_text += chunk
|
|
||||||
|
|
||||||
for f in flight_text.splitlines():
|
|
||||||
prefix, _, body = f.partition(':')
|
|
||||||
- if not (body.startswith('[') and body.endswith(']') and re.fullmatch(r'[0-9a-f]{1,3}', prefix)):
|
|
||||||
- continue
|
|
||||||
- # The body isn't necessarily valid JSON; this should always be non-fatal
|
|
||||||
- flatten(self._parse_json(body, video_id, fatal=False, errnote=False))
|
|
||||||
+ if body.startswith('[') and body.endswith(']') and re.fullmatch(r'[0-9a-f]{1,3}', prefix):
|
|
||||||
+ # The body isn't necessarily valid JSON; this should always be non-fatal
|
|
||||||
+ flatten(self._parse_json(body, video_id, fatal=False, errnote=False))
|
|
||||||
|
|
||||||
return nextjs_data
|
|
||||||
|
|
Loading…
Reference in New Issue
Block a user