feat(authoring): ADVISORY-D7 extract-method cyclomatic false-FAIL det…#153
Merged
Yuu6798 merged 4 commits intoJun 12, 2026
Merged
Conversation
…ector (D7 closure) dogfooding_findings_tracker D7 (2026-05-28 real-PR pass FINDING-F2) を candidate path (a)+(b) 両方で解決し、D-class を 8/8 全解決にする (ROADMAP v0.1.0 exit criterion 'D 全解決/waive' 充足)。 complexity_delta.cyclomatic は extractor 可視関数の総和 delta で、各関数は base 1 から始まるため、faithful な extract-method refactor は分岐を完全 保存しても総和が必ず +1/関数 微増する。primary_kind=refactor が cyclomatic ≤ 0 を宣言すると、まさに承認したい refactor で構造的に false-FAIL する。cognitive は抽出で下がる側の metric。 - nested_defs.py に count_visible_defs 追加: extract_python_complexity に 委譲して emission 規則との parity drift を構造的に排除 + VisibleDefGrowth - hazards.py detect_d7: refactor kind × cyclomatic 増加禁止 shape (≤0 / <1 / equals ≤0 / within_range [·, ≤0]) × 可視 def 増加で発火。 ≤3 等の許容幅つき lock は対象外 (false-FAIL trap でないため)。 canonical order は D6 の直後 - target-doctor: _DiffContext に visible_def_growth 追加、blob 1 回取得で nested/visible 両カウントを計算 (_def_counts_at)。out-of-scope 側 0 / syntax error skip の D6 契約をそのまま共有 - docs: guide Hazard 5 新設 (refactor パターン別 metric 選択表、D6/D7 は 相補ペア)、cli_usage / json_schema / exit_codes / schema enum に D7 追記、 tracker D7 行を解決 + 8/8 完了宣言 - tests: 可視 def counter 6 件 + detect_d7 9 件 + CLI git 統合 4 件 (extract-method 発火 / cognitive 沈黙 / 抽出なし沈黙 / D6+D7 同時発火)。 全 1810 件 pass / ruff check + format clean https://claude.ai/code/session_01BtUNrsnRygJY52xAy5bFRw
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1854d5da1e
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Codex review P2 ×2 対応: - net growth: complexity_delta.cyclomatic は extracted state 全体の総和 delta のため、ファイル間の関数移動 (+1/-1) は相殺されて lock は通る。 per-file 正のみで発火すると誤 advisory → aggregate (net) 比較に変更。 pure relocation で沈黙 / 部分 shrinkage 込み net 正で発火をテスト encode - tolerance: evaluator は lt/le/within_range に tolerance を適用する (le: observed <= expected + tol)。'<= 0, tolerance: 1' は +1 を既に 予算化しているので増加禁止 shape ではない → +1 が violate する条件 (le: expected+tol < 1 / lt: <= 1 / within_range: high+tol < 1) に変更。 equals は evaluator が tolerance 非適用のため従来通り expected <= 0。 予算化 3 形 / 不足 tolerance / equals+tolerance をテスト encode https://claude.ai/code/session_01BtUNrsnRygJY52xAy5bFRw
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: e3a3c2cd7e
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Codex review P2 対応: compatibility policy (docs/json_schema.md) は使用中 envelope の enum 値追加に schema version bump を要求する。advisory-1 の advisories[].code enum へ in-place 追加していた ADVISORY-D6 (PR #152、 未リリース main 上) / ADVISORY-D7 を advisory-2 として発行し直す。 producer 定数 / schema const / cli help / docs (json_schema 本文 + version 履歴表に v1→v2 diff 行) / test assert を同時更新 (tests/discipline/test_json_schema_version_sync.py が producer↔docs 同期を enforce)。 https://claude.ai/code/session_01BtUNrsnRygJY52xAy5bFRw
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 85183e27aa
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
…rowth Codex review P2 対応: '+1 が violate するか' の判定では '<= 0, tolerance: 1' (1 helper 分の予算) が 2 helper 抽出 (+2) を suppress して しまうが、評価は observed=+2 で依然 structural false-FAIL する。 _forbids_cyclomatic_increase を _rejects_cyclomatic_delta(constraint, net) に置き換え、net_added を先に計算してから「実際の +net を reject する lock」だけを flag する。equals は net と一致すれば予算化扱いで沈黙。 budget < helper count で発火 / budget が helper count を被覆で沈黙 / equals=net で沈黙 をテスト encode、message と docs も +N 比較の表現に同期。 https://claude.ai/code/session_01BtUNrsnRygJY52xAy5bFRw
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Completion Summary: D7-closure (ADVISORY-D7)
Phase
D-class closure 最終 (
STATUS.md 次の発行順序/ ROADMAP v0.1.0 exit criterion「D 全解決/waive」充足)。docs/dogfooding_findings_tracker.mdD7 = 2026-05-28 real-PR complexity dogfood FINDING-F2。本 PR で D-class 8/8 全解決。What Changed
authoring/nested_defs.pyにcount_visible_defs追加:extract_python_complexityに委譲して extractor の emission 規則との parity drift を構造的に排除 +VisibleDefGrowthrecordhazards.py::detect_d7新設: refactor kind × in-scope diff の可視 def net 増加 +N × 「+N を reject する」complexity_delta.cyclomatic制約(tolerance含め evaluator semantics を逐語 mirror)で発火。canonical order は D6 の直後target-doctor:_DiffContextにvisible_def_growth追加、blob 1 回取得で nested/visible 両カウント(_def_counts_at)。out-of-scope 側 0 / syntax error skip の D6 契約を共有advisory-2に bump(compatibility policy: 使用中 enum への値追加は bump 必須。D6 の in-place 追加分も本 bump に畳む)cli_usage/json_schema(v1→v2 diff 行含む)/exit_codes/ schema enum に D7 追記、tracker D7 行を解決 + 8/8 完了宣言Acceptance Criteria Status
<= 3/tolerance: N/equals Nが増加を被覆する場合は沈黙、budget < helper count は発火)Tests
detect_d716 件(net-zero relocation / tolerance 予算化 / budget<helper count 発火 / equals=net 沈黙 / equals×tolerance 非適用 含む)/ CLI git 統合 4 件(extract-method 発火 / cognitive 沈黙 / 抽出なし沈黙 / D6+D7 同時発火)ruff check .+ruff format --check .cleanDogfooding
processから_keepを module-level 抽出、全分岐保存)を作成し、semantic-ci check --target target.yaml --package-root src --baseline-rev <sha>(cyclomatic ≤ 0)→semantic-ci target-doctor ... --format human→ target をcomplexity_delta.cognitiveに書き換えてcheck再実行target-doctorがADVISORY-D7を発火(evidence:visible_defs_added: 1,src/mod.py 1→2)、guide 推奨どおり cognitive に替えると同一 refactor が PASS(fail+pass 両方実演 + 推奨 metric の実証)Files Changed
初版 15 files +566 −65、レビュー対応 3 commit(net growth + tolerance / advisory-2 bump / net-aware allowance)
Deviations from Brief
事前 brief なし(session 内で user が継続タスクを委任、tracker candidate paths (a)+(b) を逐語実装)。
Open Questions / Deferred
e3a3c2c、advisory-1→advisory-2 schema bump =85183e2、allowance vs 実 net helper 数比較 =3341d62Next Handoff
merge 後: STATUS.md の Phase / 次の発行順序 から D-class closure を sweep(wrap-up 時)。v0.1.0 exit criteria の D 項目は完了、残る軸は Phase X-2 実行(別 session)と H-5 申し送り P3。
https://claude.ai/code/session_01BtUNrsnRygJY52xAy5bFRw