This pull request introduces a change in how draft and published records are managed by moving drafts to a separate RDF repository. Previously, both drafts and published records were stored together and differentiated using a draft flag in the database. With this update, drafts and published records are now completely separated into two repositories. This structural change simplifies data handling, potentially improves performance, and enables more efficient querying.

• Clear separation of concerns: Keeps draft and published data isolated, reducing complexity.
• Direct SPARQL endpoint for published records: Eliminates the need to filter out drafts during queries.
• Improved performance: Avoids overhead from merging and filtering mixed data sets.
• Simplified architecture: Easier to maintain and extend in the future.
• Reduced risk of accidental exposure: Ensures drafts are never mixed with published data at the query level.