Support foreign keys during rebase

Currently if someone uses foreign keys between tables, things may not work correctly.

For example: table "tree_species" (with ID, latin name, common name columns) and table "trees" (with ID, location (X,Y), tree species ID). If two users both add a new tree species and a new tree:

• user1: tree_species: new row with ID 5 named "abc"
• user1: trees: new row with ID 10, referencing ID 5 of tree_species
• user2: tree_species: new row with ID 5 named "def"
• user2: trees: new row with ID 10, referencing ID 5 of tree_species

During rebase, tree_species of one user will be updated from ID 5 to 6, but the entry in trees table will still point to ID 5 (and does not get updated to ID 6).

For this, we should scan all foreign keys and make sure to update also update ID in other tables.

Start by tracing the rebase logic and how it resolves conflicting table IDs, then inspect how foreign-key relationships are exposed or discovered. Reproduce the tree_species and trees example and verify that when a referenced ID changes, corresponding IDs in other tables are updated consistently.