From 504334c7be5a56237df2598d338cd494a42fca4c Mon Sep 17 00:00:00 2001 From: Rito Takeuchi Date: Sat, 27 Jan 2024 00:19:41 +0900 Subject: gh-77749: Fix inconsistent behavior of non-ASCII handling in EmailPolicy.fold() (GH-6986) It now always encodes non-ASCII characters in headers if utf8 is false. Co-authored-by: Serhiy Storchaka --- Lib/email/policy.py | 9 ++++++++- Lib/test/test_email/test_policy.py | 17 +++++++++++++++++ .../2024-01-26-16-46-21.gh-issue-77749.NY_7TS.rst | 2 ++ 3 files changed, 27 insertions(+), 1 deletion(-) create mode 100644 Misc/NEWS.d/next/Library/2024-01-26-16-46-21.gh-issue-77749.NY_7TS.rst diff --git a/Lib/email/policy.py b/Lib/email/policy.py index 611deb5..8816c84 100644 --- a/Lib/email/policy.py +++ b/Lib/email/policy.py @@ -210,8 +210,15 @@ class EmailPolicy(Policy): self.refold_source == 'long' and (lines and len(lines[0])+len(name)+2 > maxlen or any(len(x) > maxlen for x in lines[1:]))) - if refold or refold_binary and _has_surrogates(value): + + if not refold: + if not self.utf8: + refold = not value.isascii() + elif refold_binary: + refold = _has_surrogates(value) + if refold: return self.header_factory(name, ''.join(lines)).fold(policy=self) + return name + ': ' + self.linesep.join(lines) + self.linesep diff --git a/Lib/test/test_email/test_policy.py b/Lib/test/test_email/test_policy.py index e87c275..c6b9c80 100644 --- a/Lib/test/test_email/test_policy.py +++ b/Lib/test/test_email/test_policy.py @@ -135,6 +135,23 @@ class PolicyAPITests(unittest.TestCase): for attr, value in expected.items(): self.assertEqual(getattr(added, attr), value) + def test_fold_utf8(self): + expected_ascii = 'Subject: =?utf-8?q?=C3=A1?=\n' + expected_utf8 = 'Subject: á\n' + + msg = email.message.EmailMessage() + s = 'á' + msg['Subject'] = s + + p_ascii = email.policy.default.clone() + p_utf8 = email.policy.default.clone(utf8=True) + + self.assertEqual(p_ascii.fold('Subject', msg['Subject']), expected_ascii) + self.assertEqual(p_utf8.fold('Subject', msg['Subject']), expected_utf8) + + self.assertEqual(p_ascii.fold('Subject', s), expected_ascii) + self.assertEqual(p_utf8.fold('Subject', s), expected_utf8) + def test_fold_zero_max_line_length(self): expected = 'Subject: =?utf-8?q?=C3=A1?=\n' diff --git a/Misc/NEWS.d/next/Library/2024-01-26-16-46-21.gh-issue-77749.NY_7TS.rst b/Misc/NEWS.d/next/Library/2024-01-26-16-46-21.gh-issue-77749.NY_7TS.rst new file mode 100644 index 0000000..f1c99c0 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-01-26-16-46-21.gh-issue-77749.NY_7TS.rst @@ -0,0 +1,2 @@ +:meth:`email.policy.EmailPolicy.fold` now always encodes non-ASCII characters +in headers if :attr:`~email.policy.EmailPolicy.utf8` is false. -- cgit v0.12