[Muffi7869]

Add a focused specification test for the decoding order in RFC 6901 section 4: ~01 must become the literal member name ~1, not /.

The implementation is already correct. Reversing the two replacements in unescape() still passes the existing 84 tests, but fails this new test. The test checks decoded parts, a round trip, resolution against distinct ~1 and / members, and assignment that leaves the slash member unchanged. No runtime code changes.

Validation on Linux, Python 3.10.12:

• Full unittest suite: 85 tests passed, also from a fresh clone of the published branch.
• Reversed-decoding mutation: new test fails as expected.
• flake8 and git diff --check: passed.
• Stubtest, Sphinx warnings-as-errors, wheel/sdist build and strict twine checks: passed.

AI assistance was used to prepare and review this test; Claude also reviewed the exact patch. The reported validation above was run locally.

[Copilot]

🟢 Approval recommended

No active review issues remain for the test-only RFC 6901 decoding-order regression coverage.

No open issues 🎉

What changed in this PR

Adds a focused regression test for RFC 6901 decoding order, ensuring ~01 resolves to the literal member name ~1, not /. Runtime code remains unchanged.

Changes:

• Verifies decoded pointer parts and round-trip encoding.
• Checks resolution and assignment against distinct ~1 and / members, ensuring assignment leaves the slash member unchanged.

File	Description
tests.py	Adds regression coverage for escaped-tilde decoding order.

_{Review effort: Balanced}

[coveralls]

coverage: 97.714%. remained the same — Muffi7869:test-rfc6901-escape-order into stefankoegl:master

[stefankoegl]

Thanks!