With big stores, update CLEAR ALL queries or pyOxigraph Store.clear() commands take very long (>2 hours). Deleting the store folder, however, is done swiftly.

Yet, if you're running oxigraph as a docker service, deleting all files isn't possible, and thus an OS error will prevent populating/loading the store again. Stopping the service before removing files works, but this is inconvenient. I wonder if pyOxigraph might get a function to remove files directly or to close() a Store so that the RocksDB is not preventing OS file operations.

My current deletion loops for the Store folder is

                for filename in os.listdir(endpoint):
                    file_path = os.path.join(endpoint, filename)
                    if os.path.isfile(file_path):
                        os.chmod(file_path, stat.S_IWRITE)
                        os.unlink(file_path)