det translator: mid-record SYNC slack refused in layout check mode (sibling of #5694)
メンテナーはふだん 1 日以内に返信
まだ誰も着手していません。
評価
調査の方向性
Start with det/engine_layout.py, det/layout.py, and the referenced #6117 test; compare how each lays out a SYNC item after an odd-length field run. First establish from IBM Enterprise COBOL documentation whether the slack belongs to the containing group before changing layout behavior. Done means census blockers no longer show LayoutDisagrees, the stated translation acceptance is met, burned estates remain unchanged, and the #6117 test reflects the new contract.
索引モデルが issue の本文から書いたものです。
説明
Sibling of #5694 (gap named in #6117's "Sibling gaps"), part of #5649.
What
In check mode (det/engine_layout.py, the default) the translator lays a record out twice: from the engine's layout (which models SYNCHRONIZED slack, #4266) and from its own det/layout.py (which models none). A SYNC binary item inside a record, after an odd-length run of fields, gets one byte of slack from the engine and none from the translator, so the program is refused whole with LayoutDisagrees.
#6117 pinned this as intended behaviour and left translator-side SYNC modelling as a frontier gap because no burned program has a mid-record SYNC item (all burned SYNC items are level-01, with no slack). This issue records the first program that hits it.
Evidence (counts only)
Census survey, 2026-10-28, main 4e739d9, translation only (cics_census.py survey over the 9 census clones, no burned or local corpora):
- 1 of 40 census programs is refused whole by this, and it is the only gap class that program has (
only gap1, non-burned 1). - The difference is pure slack: every elementary item before the SYNC item agrees in offset and size; from the SYNC item's group on, the engine is 1 byte further on (
engine @74, translator @73), and the record is 1 byte longer. - It is the only
LayoutDisagreesin the census; the burned estates have none.
Synthetic shape (made-up names):
01 REC.
05 A PIC X(73).
05 GRP.
10 N PIC S9(4) COMP-5 SYNC.
10 B PIC X(10).
Engine: GRP / N at 74 (halfword boundary). Translator: 73. Check mode refuses REC.
Two ways to close it
- Model SYNC slack in
det/layout.py(halfword / fullword / doubleword per USAGE, POINTER at the storage's width, C9), so check mode agrees. Before modelling, confirm from IBM's Enterprise COBOL documentation whether intra-record slack bytes belong to the group holding the SYNC item or sit before it. The engine currently starts the group after the slack, so its length excludes them. If IBM says otherwise, that is an engine fix and lands first (spec before code, no guessed semantics). - Finish #5649 step 3 / 4: retire the translator's own computation after its one-release check window. Engine mode already takes the engine's slack offsets (#6117 test), so the program would translate without a translator change.
Option 2 is the planned end state. Option 1 is only worth doing if check mode stays longer.
Acceptance
cics_census.py blockerson the census survey: noLayoutDisagreesrow; translated whole 8 -> >= 9 of 40 (non-burned), or the program's remaining holes named.- Burned estates unchanged (det-sweep ratchet green); #6117's test updated to the new contract.
- If the IBM check changes the engine's slack placement: engine fix + register entry in the same PR.
Scope: frontier (no burned program needs it; focus lock applies).
🤖 Generated with Claude Code
- 主要言語
- Python
- スター
- 57
- フォーク
- 11
- 平均マージ
- 1時間 14分
- マージ済み PR(30日)
- 967
環境構築
- Dockerfile・Docker Compose ファイルなし
- プルリクエストのテンプレートあり
- コントリビューションガイドを読む
はじめの一歩
- issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
- 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
- リポジトリをフォークし、ブランチを切って変更します。
- issue 番号を参照したプルリクエストを送ります。
squid-protocol/gitgalaxy のほかの issue
-
scope:sweep testing
難易度 5/5 1週間以上 初心者へのやさしさ 25/100
squid-protocol/gitgalaxy#7011 ·
メンテナーはふだん 1 日以内に返信
-
priority: high scope:core
難易度 4/5 3〜5日 初心者へのやさしさ 35/100
squid-protocol/gitgalaxy#6974 · コメント 1 件 ·
メンテナーはふだん 1 日以内に返信
-
claim:melek enhancement legacy-modernization priority: high scope:core testing
難易度 5/5 1週間以上 初心者へのやさしさ 35/100
squid-protocol/gitgalaxy#6967 · コメント 10 件 ·
メンテナーはふだん 1 日以内に返信
-
bug call-graph core-engine priority: high scope:core
難易度 5/5 1週間以上 初心者へのやさしさ 25/100
squid-protocol/gitgalaxy#6904 ·
メンテナーはふだん 1 日以内に返信
-
Siblings of #5042: cics-crucible programs that do not translate whole (17 programs, 8 hole classes)オープンclaim:nyx enhancement legacy-modernization priority: high scope:frontier
難易度 5/5 1週間以上 初心者へのやさしさ 35/100
squid-protocol/gitgalaxy#6876 · コメント 2 件 ·
メンテナーはふだん 1 日以内に返信
squid-protocol/gitgalaxy の issue をすべて見る
似ている issue
-
難易度 2/5 1〜3時間 初心者へのやさしさ 82/100
メンテナーはふだん 2 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 86/100
-
Validate finite output clamping limits in StepPredictor.prepare_clamping_params to prevent silent NaN/Inf propagation対応中かも @muskan-creates352 が今日担当しました。 オープン
難易度 2/5 1〜3時間 初心者へのやさしさ 84/100
mllam/neural-lam#848 ·
メンテナーはふだん 2 日以内に返信
-
infra
難易度 1/5 1時間未満 初心者へのやさしさ 88/100
fab-geocommuns/RNB-coeur#1166 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
deepspeedai/DeepSpeed#8951 ·
メンテナーはふだん 2 日以内に返信