LonamiWebs/Telethon
1
1
Fork 0
mirror of https://github.com/LonamiWebs/Telethon.git synced 2024-11-10 19:46:36 +03:00
Code Issues Packages Projects Releases Wiki Activity

Fix UnicodeDecodeError with malformed input on unparse text

Browse Source
This commit is contained in:
Lonami Exo 2019-01-01 20:31:39 +01:00
parent 619e4dc2d6
commit 229969192a
1 changed files with 22 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
telethon/extensions/markdown.py
View File

@ -169,6 +169,28 @@ def unparse(text, entities, delimiters=None, url_fmt=None):
elif isinstance(entity, MessageEntityMentionName):
url = 'tg://user?id={}'.format(entity.user_id)
if url:
# It's possible that entities are malformed and end up in the
# middle of some character, like emoji, by using malformed
# clients or bots. Try decoding the current one to check if
# this is the case, and if it is, advance the entity.
while e <= len(text):
try:
del_surrogate(text[s:e])
break
except UnicodeDecodeError:
e += 1
else:
# Out of bounds, no luck going forward
while e > s:
try:
del_surrogate(text[s:e])
break
except UnicodeDecodeError:
e -= 1
else:
# No luck going backwards either, ignore entity
continue
text = (
text[:s] +
add_surrogate(url_fmt.format(text[s:e], url)) +