[JarryShaw]

Please follow the guide below

• You will be asked some questions, please read them carefully and answer honestly

• Put an x into all the boxes [ ] relevant to your pull request (like that [x])

• Use Preview tab to see how your pull request will actually look like

• Searched for similar pull requests

• Followed the coding style (make pylint, make mypy, make isort)

• make test passes, and a test case covers the change. Narrow runs only, listed below.

• Added a changelog entry under docs/source/changelog/ and regenerated CHANGELOG.md, if the change is user-visible: this PR is the changelog pass

What is the purpose of your pull request?

Tick the commit type your subject line carries.

• fix — corrects a defect
• feat — adds a feature
• perf — changes performance, not behaviour
• refactor — changes neither behaviour nor performance
• test — tests only
• docs — documentation only
• ci — workflows or build tooling
• release — bumps the version or rolls up a distribution
• chore — anything else

---

Third whole-repo pass over the changelog for #1103.

• 1.5.0: adds entries for the merged fixes the file lacked, from the PadN change to the Conda packaging (fix(protocols): two octets of option padding are written as two Pad1 options, not one PadN #1793 to fix(conda): pin cryptography 48.0.1 on Intel macOS, which has no newer wheel (#1827) #1828, plus fix(protocols): export the L2TP data model from pcapkit.protocols.data.link (#1797) #1806), and drops tallies of codebase items. Measurements that justify a choice, and code constants, stay.
• Released files, checked against the tags:
  • the Extractor unsupported-call raise and the Reassembly not working since 0.16.3 #155 fix move from 1.2.0 to 1.1.1 (31282487e is in v1.1.1);
  • the reasm_strict and register_dumper renames are recorded under 1.1.1's post-release (3e0a581e5 is in v1.1.1.post2, not v1.1.1);
  • 1.3.0 names Engine, not _Engine;
  • 0.16.0's module path and 1.0.0's diffstat are corrected;
  • 1.4.0 lists the TCP construction fixes from 78e28d811.
• Markup: a bare TCP:80/8080 no longer autolinks; section signs and an issue role are fixed. CHANGELOG.md is regenerated.

Tests: tests/project passed.

Refs #1103

[JarryShaw]

Verdict: NEEDS CHANGES at be78b27c2 (merged as 34a494b1f). A Sonnet cross-review (author: Opus) confirmed every tag attribution, the new 1.5.0 entries, and that CHANGELOG.md is exact generator output. It found defects in the count rewrite:

• 1.5.0.rst:2389: a continuation line lost its indent, which splits the HIP bullet and adds two docutils warnings (reproduced in a clean Sphinx build);
• garbled rewrites at lines ~140, ~1087 and ~2276, and an overclaim at ~2536 (ENCAP_TYPE_TO_LINKTYPE excludes unmapped encapsulations);
• the preamble stops at b12, but b13 is out;
• "broken since 0.16.3" for Reassembly not working since 0.16.3 #155 should read 1.0.0, both in 1.1.1.rst and in the index;
• counts remain in 1.5.0.rst.

A worker is preparing a docs-only follow-up commit for main.

[JarryShaw]

Fixed on main in e5f7fb66b. Changes:

• the HIP bullet indent is restored, so docutils reports no warnings;
• the garbled rewrites are repaired;
• ENCAP_TYPE_TO_LINKTYPE is scoped to encapsulations whose DLT has a LinkType member;
• the preamble runs to b13 (2026-10-16), and Reassembly not working since 0.16.3 #155 dates the breakage to 1.0.0;
• the remaining counts are removed, keeping only measurements and spec quantities.

A Sonnet re-review found one new overclaim: "any other built-in is still refused" is false for the dumper registrars. I reproduced it, since register_dumper('json', XML) is accepted with a warning. The entry is now scoped to the engine, reassembly and traceflow registrars. changelog_md.py --check passes.