[JarryShaw]

• Searched for similar pull requests
• Followed the coding style
• make test passes, and a test case covers the change
• Changelog entry: N/A — centralised in docs(changelog): shared 1.5.0 changelog — long-lived, merges last (#610, #616, #617, #618, #620) #657

What is the purpose of your pull request?

• fix — bug fix

---

Description

Closes #2298.

• Problem: http.HTTP's version, data and version-guess errors, httpv1.HTTP._make_start_line's errors, and httpv2.HTTP._make_http_settings's rejection all broke the house <prefix>: <lowercase description>: … {value!r} form — missing or malformed prefixes, capitalised with a full stop, and some without the rejected values.
• Fix: reword as HTTP: invalid version: 3, HTTP: invalid data: …, HTTP: unknown version, HTTP: invalid start line: … (naming the method and status it got), and HTTP/2: [Type 4] invalid settings: 42 (dropping the stray space before the colon and adding the rejected value).
• Test: test_http_message_prefix_2298_unit pins each message, including the settings one; the existing pins in test_http_unit, test_http_guess_http2_1141_unit and test_http_from_data_1154_unit follow the new text.

[JarryShaw]

Review: NEEDS CHANGES at 28ec3824c (Sonnet). Everything touched is correct and well pinned; one scope gap from the issue.

1. Missed site: httpv2.HTTP._make_http_settings's else: branch still raises HTTP/2 : [Type 4] invalid settings (space before the colon, no value shown). fix(http): HTTP version, data and start-line messages lack the HTTP: prefix #2298 names this site explicitly. Fix + pin needed.
2. Stale docstring tense: test_http_unit.py::test_guess_version_identifies_the_http2_connection_preface still says "would raise unknown HTTP version" — the old text, missed among the sibling updates.

Confirmed correct: every other touched site (version/data/version-guess in http.py, start-line in httpv1.py) matches the stated form exactly, live-triggered on both trees; no other missed site anywhere else in http.py/httpv1.py; tests pass, fail-without-fix gives 21 failures on origin/main. Fix in progress.

[JarryShaw]

Rebased and amended 28ec3824c → 336ea4aa1 (6 commits onto current origin/main, clean): added the missed httpv2._make_http_settings fix (HTTP/2: [Type 4] invalid settings: 42, dropping the stray space and adding the value), a pinning test (TestHTTPv2SettingsMessage), and fixed the stale docstring tense in test_http_unit.py. Title/body widened to mention the settings category. 7/7 message-prefix tests pass. Fresh review round dispatched.

[JarryShaw]

Review: GOOD TO GO at 336ea4aa1 (Sonnet, round 2). Independently re-verified: HTTP/2: [Type 4] invalid settings: 42 triggers correctly live (9 more value types checked), fails without the fix (11 assertion sites), 8 mutation variants each caught by exactly the right subtest, docstring tense fix confirmed accurate, and a repo-wide AST walk of all 35 HTTP raise/warn sites found none missing the house form. Passes on Python 3.9-3.12 and 3.14. Added labels.

All findings across both rounds now addressed.

[github-actions]

Coverage: 91.58% (unit tier, Python 3.15, 336ea4aa1, Unit Tests run success)

Package	Statements	Missed	Branches	Partial	Cover
pcapkit (top level)	104	4	20	2	95.16%
pcapkit/const	19601	881	2620	867	92.12%
pcapkit/corekit	2596	74	818	37	96.51%
pcapkit/dumpkit	282	3	102	3	98.44%
pcapkit/foundation	3029	23	1050	25	98.82%
pcapkit/interface	138	4	52	3	96.32%
pcapkit/protocols	17455	218	4598	185	98.15%
pcapkit/toolkit	914	36	334	10	95.51%
pcapkit/utilities	502	7	144	5	98.14%
pcapkit/vendor	4432	1988	1018	167	51.87%

Per-file detail: the coverage-html artifact of this run.

[JarryShaw]

Ready to merge.

Review: GOOD TO GO (Sonnet, round 2) at 336ea4aa1.
CI: Required checks passed, 0 failures (only the non-blocking experimental GraalPy leg still pending).
Mergeable, 0 unresolved threads.