Fix the V22 phase 1 migrations to use new tables for mapping rather than editing existing tables

The ALTER TABLE operations previously used were causing the DB to die when run on the production TagManifest table which has 7 million rows. We instead now use new mapping tables, which is less nice, but these are temporary anyway, so hopefully we only have to deal with their ugliness for a short duration.

endpoints/api/test/test_repository.py

@@ -128,6 +128,8 @@ def test_create_repository(repo_name, expected_status, client):
 def test_get_repo(has_tag_manifest, client, initialized_db):
   with client_with_identity('devtable', client) as cl:
     if not has_tag_manifest:
+      database.TagManifestLabelMap.delete().execute()
+      database.TagManifestToManifest.delete().execute()
       database.TagManifestLabel.delete().execute()
       database.TagManifest.delete().execute()