If the Content-Transfer-Encoding header field of a message part has trailing whitespace, for example "base64 ", get_payload(decode=True) does not return the properly decoded payload.

Here is a minimal code example. Sample message file attached.

import email
from email import policy

with open('msg.txt', 'rb') as f:
    msg = email.message_from_binary_file(f, policy=policy.default)

parts = list(msg.walk())
parts[1].get_payload(decode=True)
> b'SGVsbG8uIFRlc3Rpbmc=\n'

The parsed content-transfer-encoding header "cte" value is truncated, but it's string value is not.

>>> header = parts[1].get('content-transfer-encoding')
>>> header.cte
'base64'
>>> str(header)
'base64 '

Which is what appears to be used in the decode attempt
https://github.com/python/cpython/blob/main/Lib/email/message.py#L289

• CPython versions tested on: 3.9.13, 3.10.7
• Operating system and architecture: macOS 12.6 Intel

msg.txt

Linked PRs

• gh-98188: Fix EmailMessage.get_payload to decode data #127547
• [3.13] gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) #128528
• [3.12] gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) #128529