One pipeline should process different source tables using the same definition. Which design supports this?

1. Rename the source table before each run
2. Parameters and dynamic expressions for table-specific settings
3. Hardcode one table and copy the pipeline for every change
4. Use a report slicer as the source-table variable