> Although for now, the whole table gets read into memory for every read so this won’t work well for very large database files. There’s also no write support… y

> Although for now, the whole table gets read into memory for every read so this won’t work well for very large database files. There’s also no write support… yet. At such a set of features I would prefer a tool that converts databases to a directory of real csv and jsonl files, at least there are no performance issues to worry about