Issue

email.quoprimime.decode() does decoded[:-1] to strip a trailing line ending, but this only removes one character. When eol='\r\n' (2 characters), a stray \r is left in the output.

Reproducer

import email.quoprimime as qp
print(repr(qp.decode('abc', eol='\r\n')))  # 'abc\r' — expected 'abc'

Fix

In Lib/email/quoprimime.py, change:

decoded = decoded[:-1]

to:

decoded = decoded[:-len(eol)]

Tests

Three test cases added to TestQuopri in Lib/test/test_email/test_email.py:

• test_decode_crlf_eol_no_trailing_newline — verifies decode('abc', eol='\r\n') returns 'abc' (direct reproducer for the bug; without the fix, returns 'abc\r').
• test_decode_crlf_eol_multiline_no_trailing_newline — verifies decode('a\r\nb', eol='\r\n') returns 'a\r\nb' (multi-line input without trailing newline).
• test_decode_crlf_eol_with_trailing_newline — verifies decode('abc\r\n', eol='\r\n') returns 'abc\r\n' (trailing newline is legitimate and should be preserved).

• Issue: email.quoprimime.decode() leaves stray \r when eol='\r\n' #145831