MarcoBuster/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-12-25 20:55:52 +01:00
Code Issues Projects Releases Wiki Activity

[teamcoco] Rewrite preload data extraction

Browse Source 
Idea: "puncture" some consecutive fragments and check whether the
b64decode result of a punctured string is a valid JSON or not.

It's a O(N^3) algorithm, but should be fast for a small N (less than 30
fragments in all test cases)
This commit is contained in:
Yen Chi Hsuan 2015-05-15 02:17:22 +08:00
parent 1ae72fb23d
commit 3a105f7b20
1 changed files with 26 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

53
youtube_dl/extractor/teamcoco.py
View File

@ -4,6 +4,7 @@
import base64 import base64
import binascii import binascii
import re import re
import json
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import ( from ..utils import (
@ -68,41 +69,39 @@ def _real_extract(self, url):
video_id = self._html_search_regex( video_id = self._html_search_regex(
self._VIDEO_ID_REGEXES, webpage, 'video id') self._VIDEO_ID_REGEXES, webpage, 'video id')
data = preload = None data = None
preloads = re.findall(r'"preload":\s*"([^"]+)"', webpage)
if preloads:
preload = max([(len(p), p) for p in preloads])[1]
if not preload: preload_codes = self._html_search_regex(
preload = ''.join(re.findall(r'this\.push\("([^"]+)"\);', webpage)) r'(function.+)setTimeout\(function\(\)\{playlist',
webpage, 'preload codes')
base64_fragments = re.findall(r'"([a-zA-z0-9+/=]+)"', preload_codes)
base64_fragments.remove('init')
if not preload: def _check_sequence(cur_fragments):
preload = self._html_search_regex([ if not cur_fragments:
r'player,\[?"([^"]+)"\]?', r'player.init\(\[?"([^"]+)"\]?\)' return
], webpage.replace('","', ''), 'preload data', default=None) for i in range(len(cur_fragments)):
cur_sequence = (''.join(cur_fragments[i:] + cur_fragments[:i])).encode('ascii')
if not preload:
preload_codes = self._html_search_regex(
r'(function.+)setTimeout\(function\(\)\{playlist',
webpage, 'preload codes')
base64_fragments = re.findall(r'"([a-zA-z0-9+/=]+)"', preload_codes)
base64_fragments.remove('init')
for i in range(len(base64_fragments)):
cur_sequence = (''.join(base64_fragments[i:] + base64_fragments[:i])).encode('ascii')
try: try:
raw_data = base64.b64decode(cur_sequence) raw_data = base64.b64decode(cur_sequence)
except (TypeError, binascii.Error): if compat_ord(raw_data[0]) == compat_ord('{'):
return json.loads(raw_data.decode('utf-8'))
except (TypeError, binascii.Error, UnicodeDecodeError, ValueError):
continue continue
if compat_ord(raw_data[0]) == compat_ord('{'):
data = self._parse_json(raw_data.decode('utf-8'), video_id, fatal=False)
if not preload and not data: def _check_data():
raise ExtractorError( for i in range(len(base64_fragments) + 1):
'Preload information could not be extracted', expected=True) for j in range(i, len(base64_fragments) + 1):
data = _check_sequence(base64_fragments[:i] + base64_fragments[j:])
if data:
return data
self.to_screen('Try to compute possible data sequence. This may take some time.')
data = _check_data()
if not data: if not data:
data = self._parse_json( raise ExtractorError(
base64.b64decode(preload.encode('ascii')).decode('utf-8'), video_id) 'Preload information could not be extracted', expected=True)
formats = [] formats = []
get_quality = qualities(['500k', '480p', '1000k', '720p', '1080p']) get_quality = qualities(['500k', '480p', '1000k', '720p', '1080p'])