etc/yt-dlp
0
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-12-22 06:00:00 +00:00
Code Issues Projects Releases Wiki Activity

[ie/iq.com] Fix extraction and subtitles (#8260)

Browse source 
Closes #7734, Closes #8123
Authored by: AS6939
This commit is contained in:
AS6939 2023-10-07 06:56:12 +08:00 committed by GitHub
parent 2ad3873f0d
commit 35d9cbaf96
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 13 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
yt_dlp/extractor/iqiyi.py
View file

@ -499,9 +499,10 @@ class IqIE(InfoExtractor):
'tm': tm, 'tm': tm,
'qdy': 'a', 'qdy': 'a',
'qds': 0, 'qds': 0,
'k_ft1': 141287244169348, 'k_ft1': '143486267424900',
'k_ft4': 34359746564, 'k_ft4': '1572868',
'k_ft5': 1, 'k_ft7': '4',
'k_ft5': '1',
'bop': JSON.stringify({ 'bop': JSON.stringify({
'version': '10.0', 'version': '10.0',
'dfp': dfp 'dfp': dfp
@ -529,14 +530,22 @@ def _extract_vms_player_js(self, webpage, video_id):
webpack_js_url = self._proto_relative_url(self._search_regex( webpack_js_url = self._proto_relative_url(self._search_regex(
r'<script src="((?:https?:)?//stc\.iqiyipic\.com/_next/static/chunks/webpack-\w+\.js)"', webpage, 'webpack URL')) r'<script src="((?:https?:)?//stc\.iqiyipic\.com/_next/static/chunks/webpack-\w+\.js)"', webpage, 'webpack URL'))
webpack_js = self._download_webpage(webpack_js_url, video_id, note='Downloading webpack JS', errnote='Unable to download webpack JS') webpack_js = self._download_webpage(webpack_js_url, video_id, note='Downloading webpack JS', errnote='Unable to download webpack JS')
webpack_map = self._search_json( webpack_map = self._search_json(
r'["\']\s*\+\s*', webpack_js, 'JS locations', video_id, r'["\']\s*\+\s*', webpack_js, 'JS locations', video_id,
contains_pattern=r'{\s*(?:\d+\s*:\s*["\'][\da-f]+["\']\s*,?\s*)+}', contains_pattern=r'{\s*(?:\d+\s*:\s*["\'][\da-f]+["\']\s*,?\s*)+}',
end_pattern=r'\[\w+\]\+["\']\.js', transform_source=js_to_json) end_pattern=r'\[\w+\]\+["\']\.js', transform_source=js_to_json)
replacement_map = self._search_json(
r'["\']\s*\+\(\s*', webpack_js, 'replacement map', video_id,
contains_pattern=r'{\s*(?:\d+\s*:\s*["\'][\w.-]+["\']\s*,?\s*)+}',
end_pattern=r'\[\w+\]\|\|\w+\)\+["\']\.', transform_source=js_to_json,
fatal=False) or {}
for module_index in reversed(webpack_map): for module_index in reversed(webpack_map):
real_module = replacement_map.get(module_index) or module_index
module_js = self._download_webpage( module_js = self._download_webpage(
f'https://stc.iqiyipic.com/_next/static/chunks/{module_index}.{webpack_map[module_index]}.js', f'https://stc.iqiyipic.com/_next/static/chunks/{real_module}.{webpack_map[module_index]}.js',
video_id, note=f'Downloading #{module_index} module JS', errnote='Unable to download module JS', fatal=False) or '' video_id, note=f'Downloading #{module_index} module JS', errnote='Unable to download module JS', fatal=False) or ''
if 'vms request' in module_js: if 'vms request' in module_js:
self.cache.store('iq', 'player_js', module_js) self.cache.store('iq', 'player_js', module_js)