問題文
A team calls the generic load method of the reader with a path and no format argument inside a session and the call succeeds on one directory but fails on a directory of text files. What explains the difference?
選択肢
- The generic reader falls back on a scan of the first file and that scan picks the format from the extension of the name it finds.
- The generic reader falls back on the format of the last successful call so the second call inherits that choice.
- The generic reader falls back on the default source of the settings and that default is the columnar format for every call of the deployment.
- The generic reader falls back on the row based format because that format has no schema for the widest range of inputs.