I only get wiki-18.jsonl after running download_search_data.py, but the server code need corpus.json how to build corpus.json from jsonl file. Appending all items to list?