ContainerMigrateSorting migration reappears due to restored deleted records #644

After upgrading from v11 to v12, we ran the ContainerMigrateSorting migration multiple times and successfully updated the sorting of the records. However, after a few days, the Upgrade Wizard again indicates that the migration needs to be performed. Upon investigation, we found that this is caused by restored records that had been deleted before the migration was initially run. Since the migration only processes records where deleted = 0, these restored records still have the old sorting in the database. As a result, the dry run detects that there are records left to migrate.

To Reproduce

1. Delete some pages with containers and content in v11
2. Perform the upgrade from v11 to v12 and run the migrations
3. Restore records that were deleted before the migration.
4. The Upgrade Wizard shows that the migration needs to be run again, as the restored records still have the old sorting.

Restored records should either be included in the migration (which is not possible by using a datahandler paste command, I guess) or handled in a way that prevents the migration from being repeatedly flagged as incomplete.

The issue occurs because the migration only updates records with deleted = 0, so any records restored after the migration retain the old sorting and trigger the migration check again.