Regenerate streams with missing compacted records

[corylown]

• remove association between compacted upload and its constituent obsolete uploads and set status to active
• delete existing compacted upload
• run generate full dump
• run compact job
• run full dump job and verify compacted upload still has associated marc records
• turm scheduled full dump job back on

When testing on stage:

• ensure re-running full dump job does not liberate the MARC Records from their compacted upload.

Process:

Needed for Princeton, Duke, MIT, Harvard

stream = Organization.find_by(slug: 'stanford').default_stream
stream.uploads.where(status: 'compacted').first.compacted_uploads.update_all(status: 'active', compacted_upload_id: nil)
stream.uploads.where(status: 'compacted').destroy_all
GenerateFullDumpJob.perform_later(stream)
CompactUploadsJob.perform_later(stream)