diff options
Diffstat (limited to 'tests')
| -rw-r--r-- | tests/test_case.py | 10 | ||||
| -rw-r--r-- | tests/test_cli.py | 7 | ||||
| -rw-r--r-- | tests/test_parse.py | 74 |
3 files changed, 90 insertions, 1 deletions
diff --git a/tests/test_case.py b/tests/test_case.py index c6964a9..ab7eeeb 100644 --- a/tests/test_case.py +++ b/tests/test_case.py @@ -45,6 +45,16 @@ class TestCaseCreation(unittest.TestCase): manifest = json.loads((created.path / "manifest.json").read_text()) self.assertEqual(manifest["format"], case.FORMAT_VERSION) + def test_every_block_is_seeded_present_and_empty(self): + # A created-but-unparsed case must have the same SHAPE as a parsed + # one, so a later reader indexes a block rather than guarding every + # access. report will read headers and would hit a KeyError. + created = case.create(self.root, b"x") + manifest = json.loads((created.path / "manifest.json").read_text()) + for block in ("iocs", "auth", "contacts", "destinations", "headers"): + self.assertIn(block, manifest) + self.assertEqual(manifest[block], []) + def test_two_cases_do_not_collide(self): a = case.create(self.root, b"one") b = case.create(self.root, b"two") diff --git a/tests/test_cli.py b/tests/test_cli.py index 425bce9..b8fffea 100644 --- a/tests/test_cli.py +++ b/tests/test_cli.py @@ -160,7 +160,12 @@ class TestParse(unittest.TestCase): headers = manifest.pop("headers") self.assertNotIn("example.org", json.dumps(manifest)) # And nothing shaped like an address survives in the exception. - self.assertNotIn("@example.org", json.dumps(headers)) + # Both spellings: you%40example.org is not a hypothetical, it is why + # leaky.eml exists, and docs/plans/2026-09-09-contacts.md records a + # From of phish@victim%40example.org.invalid. + blob = json.dumps(headers) + self.assertNotIn("@example.org", blob) + self.assertNotIn("you%40example.org", blob) names = [name for name, _ in headers] for name in ("To", "Cc", "Delivered-To", "X-Original-To"): self.assertNotIn(name, names) diff --git a/tests/test_parse.py b/tests/test_parse.py index fa3526a..6e3b7a4 100644 --- a/tests/test_parse.py +++ b/tests/test_parse.py @@ -336,6 +336,80 @@ class ReportHeaders(unittest.TestCase): # The rest of the hop survives; this is a cut, not a blanking. self.assertIn("203.0.113.42", received[0]) + def test_every_for_clause_shape_loses_the_address(self): + # RFC 5321 4.4 puts For inside Opt-info, so With, ID, Via or a CFWS + # comment may legitimately follow it, and its ABNF is + # 1*( Path / Mailbox ) where Mailbox carries no angle brackets. + # Anchoring on "for" being immediately followed by the clause + # terminator matched only the neatest shape and let four routine + # ones through, each publishing the victim's address. + hop = "from a.invalid (a.invalid [203.0.113.5]) by mx.example.org " + shapes = ( + "for <you@example.org> (envelope-from <b@c.invalid>); Mon, 07 Sep 2026 09:12:40 +0000", + "for you@example.org; Mon, 07 Sep 2026 09:12:40 +0000", + "for <you@example.org> with ESMTP; Mon, 07 Sep 2026 09:12:40 +0000", + "id qq; Mon, 07 Sep 2026 09:12:40 +0000 (for <you@example.org>)", + "for <you@example.org>; Mon, 07 Sep 2026 09:12:40 +0000", + ) + for tail in shapes: + with self.subTest(tail=tail): + stripped = parse._strip_envelope_recipient(hop + tail) + self.assertNotIn("you@example.org", stripped) + # The hop's own evidence survives: this is a cut, not a + # blanking, and a rule that ate the line would pass the + # assertion above while destroying the report. + self.assertIn("203.0.113.5", stripped) + self.assertIn("mx.example.org", stripped) + + def test_stripping_leaves_no_doubled_space_or_stray_separator(self): + # Cosmetic in isolation, but the result is published verbatim to a + # third party, so a mangled line reads as a broken tool. + hop = ("from a.invalid (a.invalid [203.0.113.5]) by mx.example.org" + " for <you@example.org>; Mon, 07 Sep 2026 09:12:40 +0000") + stripped = parse._strip_envelope_recipient(hop) + self.assertNotIn(" ", stripped) + self.assertNotIn(" ;", stripped) + self.assertIn("mx.example.org; Mon", stripped) + + def test_a_comment_holding_only_the_clause_leaves_no_debris(self): + hop = ("from a.invalid (a.invalid [203.0.113.5]) by mx.example.org" + " id qq; Mon, 07 Sep 2026 09:12:40 +0000 (for <you@example.org>)") + stripped = parse._strip_envelope_recipient(hop) + self.assertNotIn("you@example.org", stripped) + self.assertFalse(stripped.endswith("(")) + self.assertTrue(stripped.endswith("+0000")) + + def test_the_envelope_sender_comment_survives_the_cut(self): + # envelope-from is the SENDER, which is what the report is about, so + # cutting the recipient must not take it along. + hop = ("from a.invalid (a.invalid [203.0.113.5]) by mx.example.org" + " for <you@example.org> (envelope-from <bounce@sender.invalid>);" + " Mon, 07 Sep 2026 09:12:40 +0000") + stripped = parse._strip_envelope_recipient(hop) + self.assertNotIn("you@example.org", stripped) + self.assertIn("bounce@sender.invalid", stripped) + + def test_the_whitelist_does_not_filter_attacker_free_text(self): + # A DOCUMENTED LIMIT, not a guarantee. The spec keeps Subject and the + # From display name knowing both are attacker-controlled free text, + # because they are what lets a desk recognise a campaign. An attacker + # who writes the recipient's own address into one, obfuscated or not, + # gets it published: the whitelist governs WHICH headers travel, never + # what is inside one. + # + # This is asserted so the limit is visible and deliberate. Do not + # "fix" it by filtering free text, which is the judgement-shaped + # problem AGENTS.md names as the source of every leak here. The + # sweep over real mail is what covers this class, per AGENTS.md. + raw = (b"Received: from a.invalid (a.invalid [203.0.113.5])" + b" by mx.example.org with ESMTP id X;" + b" Mon, 07 Sep 2026 09:12:40 +0000\r\n" + b"From: <phish@sender.invalid>\r\n" + b"Subject: Verify you%40example.org\r\n\r\nbody\r\n") + headers = parse.report_headers(raw, trusted=["192.0.2.0/24"]) + subject = dict(headers)["Subject"] + self.assertIn("you%40example.org", subject) + def test_a_forged_chain_publishes_no_hop_below_the_boundary(self): # The same job test_a_forged_chain_stops_at_the_first_untrusted_hop # does for sending_ip(), asserted over what actually gets published: |