Coder/deepseek-harness
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(snapshot): preserve volatile id correlations

Browse Source
This commit is contained in:
Dudu-0223
2026-07-27 15:25:50 +08:00
parent 82637524a3
commit 5eca318cdc
8 changed files with 249 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
.agents/notes/implemented/bug-fix/2026-07-27-stable-snapshot-refresh-volatiles.i18n.yaml
View File

@@ -2,5 +2,5 @@
# side as of the last confirmed-consistent state. Both languages carry equal authority;
# after editing either side, bring the other along and re-record with:
# pnpm run verify-translation-pairing --write .agents/notes/implemented/bug-fix/2026-07-27-stable-snapshot-refresh-volatiles.md
2026-07-27-stable-snapshot-refresh-volatiles.md: e6bff6ef4d20b431cee86d16df863a7c7b138e02
2026-07-27-stable-snapshot-refresh-volatiles.zh.md: 33acbbc0504ca1b61495cbb95a0f6dec28dfa0df
2026-07-27-stable-snapshot-refresh-volatiles.md: 0fd38cabd103657410a468b49e30ab17aad9affb
2026-07-27-stable-snapshot-refresh-volatiles.zh.md: aa2b129f4e1cc5286ff17067fe910dfc969838a1

8
.agents/notes/implemented/bug-fix/2026-07-27-stable-snapshot-refresh-volatiles.md
View File

@@ -12,7 +12,9 @@ ACP snapshot comparison normalizes generated UUIDs, cwd aliases, spill locators,
Refresh write-back uses `normalizeSessionLog` as its sole volatile-value authority. After existing record alignment, it recursively compares fresh and existing leaves through their normalized records: normalized-equivalent leaves retain the existing raw value, while normalized-distinct leaves retain the fresh semantic value.
Object fields align by key. Array elements align only when all corresponding arrays have the same length; otherwise the fresh array wins. Records must retain the same type, and strings remain atomic leaves. Existing packed-chunk timing alignment and inserted-title handling remain separate because they align logical events rather than values inside one record.
Before reuse, the complete logical-record layout must align, apart from the existing packed-chunk and inserted-title equivalences. Normalized-equivalent changed strings form a log-wide bijection: one fresh string maps to exactly one existing string and vice versa, so repeated IDs remain correlated across records. An unexplained record mismatch or conflicting mapping disables normalized string reuse for that log.
Object fields align by key. Array elements align only when all corresponding arrays have the same length; otherwise the fresh array wins. Strings remain atomic leaves. Existing packed-chunk timing alignment and inserted-title handling remain separate because they align logical events rather than values inside one record.
## Alternatives considered
@@ -24,6 +26,6 @@ Object fields align by key. Array elements align only when all corresponding arr
## Consequences
Repeated refreshes no longer rewrite aligned fixture values solely because the normalizer classifies them as volatile, and new volatile categories added to the normalizer automatically inherit the write-back behavior. Structural ambiguity remains conservative: changed record types, resized arrays, and strings containing both semantic and volatile changes use fresh values rather than risk reusing misaligned data.
Repeated refreshes no longer rewrite aligned fixture values solely because the normalizer classifies them as volatile, and new volatile categories added to the normalizer automatically inherit the write-back behavior. Structural ambiguity remains conservative: unmatched records, conflicting string mappings, resized arrays, and strings containing both semantic and volatile changes use fresh values rather than risk reusing misaligned data.
Focused unit coverage pins recursive object/array behavior, volatile strings, and fresh semantic fields. Keyless refresh coverage proves approval UUIDs, cwd aliases, spill paths, and event-read volatility leave their committed fixtures byte-identical.
Focused unit coverage pins recursive object/array behavior, correlated IDs, ambiguous-layout fallback, conflicting mappings, volatile strings, and fresh semantic fields. Keyless refresh coverage proves approval UUIDs, cwd aliases, spill paths, and event-read volatility leave their committed fixtures byte-identical.

8
.agents/notes/implemented/bug-fix/2026-07-27-stable-snapshot-refresh-volatiles.zh.md
View File

@@ -12,7 +12,9 @@ ACP（Agent Client Protocol）快照比较会归一化生成的 UUID、cwd 别
刷新写回以 `normalizeSessionLog` 作为易变值的唯一判定依据。现有记录完成对齐后，系统基于归一化后的记录，递归比较本次生成记录与现有记录的叶节点：归一化后等价的叶节点保留现有原始值，归一化后不同的叶节点则保留本次生成的语义值。
对象字段按键对齐。只有所有对应数组长度相同时，才对齐其元素；否则以本次生成的数组为准。记录必须保持同一类型，字符串始终作为不可拆分的叶节点。现有的打包分片计时对齐与插入标题处理仍保持独立，因为它们对齐的是逻辑事件，而非单条记录内的值。
复用前必须确保完整逻辑记录布局对齐，现有的打包分片与插入标题等价情形除外。归一化后等价但发生变化的字符串在整份日志范围内形成双射：一个本次生成的字符串只映射到一个现有字符串，反向亦然，因此跨记录重复出现的 ID 仍保持关联。出现无法解释的记录不匹配或映射冲突时，该日志会停用规范化字符串复用。
对象字段按键对齐。只有所有对应数组长度相同时，才对齐其元素；否则以本次生成的数组为准。字符串始终作为不可拆分的叶节点。现有的打包分片计时对齐与插入标题处理仍保持独立，因为它们对齐的是逻辑事件，而非单条记录内的值。
## 考虑过的替代方案
@@ -24,6 +26,6 @@ ACP（Agent Client Protocol）快照比较会归一化生成的 UUID、cwd 别
## 后果
重复刷新不再仅仅因为规范化器将已对齐的 fixture 值归类为易变值，就改写这些值；以后加入规范化器的新易变值类别也会自动继承该写回行为。结构有歧义时仍采取保守策略：记录类型发生变化、数组尺寸发生变化，或字符串同时包含语义变化与易变变化时，均使用本次生成的值，避免冒险复用未对齐的数据。
重复刷新不再仅仅因为规范化器将已对齐的 fixture 值归类为易变值，就改写这些值；以后加入规范化器的新易变值类别也会自动继承该写回行为。结构有歧义时仍采取保守策略：记录无法匹配、字符串映射冲突、数组尺寸发生变化，或字符串同时包含语义变化与易变变化时，均使用本次生成的值，避免冒险复用未对齐的数据。
聚焦的单元测试固定了递归处理对象与数组的行为、易变字符串以及本次生成的语义字段。无密钥刷新测试证明，审批 UUID、cwd 别名、spill 路径和事件读取中的易变值不会改变已提交 fixture 的任何字节。
聚焦的单元测试固定了递归处理对象与数组的行为、关联 ID、有歧义布局时的回退、映射冲突、易变字符串以及本次生成的语义字段。无密钥刷新测试证明，审批 UUID、cwd 别名、spill 路径和事件读取中的易变值不会改变已提交 fixture 的任何字节。