Use pg_export_snapshot for consistent parallel reads

[nafg]

Summary

• Exports a PostgreSQL snapshot from a coordinator JDBC connection in DbCopier.run() before any table copies begin
• Each CopyAction imports the snapshot via SET TRANSACTION SNAPSHOT within a REPEATABLE READ transaction on the source, ensuring all parallel table reads see the same consistent point-in-time view
• Prevents FK violations when copying from a live database where new rows can appear between parent and child table copies

Test plan

• Compiles on both Scala 2.13 and Scala 3
• All 7 unit test suites pass
• Integration tests (require Docker)
• Manual test against live PostgreSQL database with concurrent writes

🤖 Generated with Claude Code

[Copilot]

Pull request overview

This PR implements PostgreSQL snapshot-based consistency for parallel table copying to prevent foreign key violations when copying from a live database. It exports a snapshot from a coordinator connection before copying begins and imports that snapshot into each parallel worker transaction.

Changes:

• Exports a PostgreSQL snapshot at the start of DbCopier.run() via pg_export_snapshot()
• Passes the snapshot ID through copyTablesByLevel and TableCopier.run to CopyAction
• Each CopyAction imports the snapshot using SET TRANSACTION SNAPSHOT within a REPEATABLE READ transaction

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
simple-anonymizer/src/scala/simpleanonymizer/DbCopier.scala	Creates coordinator connection, exports snapshot, passes snapshot ID to table copiers, closes connection via andThen
simple-anonymizer/src/scala/simpleanonymizer/TableCopier.scala	Adds optional snapshotId parameter and passes it to CopyAction; minor whitespace change
simple-anonymizer/src/scala/simpleanonymizer/CopyAction.scala	Imports snapshot via SET TRANSACTION SNAPSHOT before executing source SELECT query

Comments suppressed due to low confidence (1)

simple-anonymizer/src/scala/simpleanonymizer/TableCopier.scala:27

• The snapshotId parameter has a default value of None, which allows TableCopier to be used independently without snapshot consistency. However, when called from DbCopier.copyTablesByLevel, it always passes Some(snapshotId). Document this behavior in the method's scaladoc to clarify that the default None is for standalone usage while DbCopier will always provide a snapshot ID for consistent parallel reads.

          target.db

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 5 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.