kumi/yt-dlp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

[outtmpl] Smarter replacing of unsupported characters

Browse source 
Closes #1330
This commit is contained in:
pukkandan 2022-08-04 20:19:32 +05:30
parent 05e2243e80
commit 989a01c261
No known key found for this signature in database
GPG key ID: 7EEE9E1E817D0A39
1 changed files with 6 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
yt_dlp/utils.py
View file

@ -40,6 +40,7 @@ import tempfile
import time import time
import traceback import traceback
import types import types
import unicodedata
import urllib.error import urllib.error
import urllib.parse import urllib.parse
import urllib.request import urllib.request
@ -647,6 +648,9 @@ def sanitize_filename(s, restricted=False, is_id=NO_DEFAULT):
return ACCENT_CHARS[char] return ACCENT_CHARS[char]
elif not restricted and char == '\n': elif not restricted and char == '\n':
return '\0 ' return '\0 '
elif is_id is NO_DEFAULT and not restricted and char in '"*:<>?|/\\':
# Replace with their full-width unicode counterparts
return {'/': '\u29F8', '\\': '\u29f9'}.get(char, chr(ord(char) + 0xfee0))
elif char == '?' or ord(char) < 32 or ord(char) == 127: elif char == '?' or ord(char) < 32 or ord(char) == 127:
return '' return ''
elif char == '"': elif char == '"':
@ -659,6 +663,8 @@ def sanitize_filename(s, restricted=False, is_id=NO_DEFAULT):
return '\0_' return '\0_'
return char return char
if restricted and is_id is NO_DEFAULT:
s = unicodedata.normalize('NFKC', s)
s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s) # Handle timestamps s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s) # Handle timestamps
result = ''.join(map(replace_insane, s)) result = ''.join(map(replace_insane, s))
if is_id is NO_DEFAULT: if is_id is NO_DEFAULT: