Skip to content

Recover legacy Jira attachment metadata during file reads - #697

Merged
witbrock merged 1 commit into
mainfrom
codex/legacy-jira-file-export
Sep 14, 2026
Merged

witbrock merged 1 commit into
mainfrom
codex/legacy-jira-file-export

Conversation

@witbrock

Copy link
Copy Markdown
Member

The paired Jira migration file export stopped with not_found for an existing, authorised attachment. Early imports can retain complete blob attributes without the later attributes.v1 marker or blob text relations. Recover that legacy Jira format when canonical blob text is absent, retaining text precedence and all existing actor checks.

Seven focused retrieval tests pass, including owner read, unrelated-user denial, text precedence and unrelated/incomplete format rejection. A real read in the isolated restored Atlas snapshot retrieved the stopped file: 83,038 bytes with the exact retained SHA-256, without metadata or access changes. An adjacent artifact-ingestion suite was stopped during an unrelated database wait after eight passes; it is not a release gate for this read-path fix.

Von task: #V#task_agent_6bb01d1014e4b888a724b8e7ba2c77fe. Activation is limited to the authorised pinned migration runtime; no public web deployment or writer cutover.

@witbrock
witbrock merged commit 363eb39 into main Sep 14, 2026
2 checks passed
@witbrock
witbrock deleted the codex/legacy-jira-file-export branch September 14, 2026 17:24
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant