Field names in CSV and JSON are case sensitive.
Unsupported files and compression types are skipped.
Records that contain errors are skipped. Errors can include mismatched quotation marks (" "), embedded line feeds, and so forth.
External files with UTF-8 BOMs (Byte Order Markers) are supported, but UTF-16LE, UTF-16BE, UTF-32LE, UTF-32BE BOMs are not supported.
- File type (JSON, CSV, or Parquet)
- File compression (GZIP for JSON or CSV, SNAPPY for Parquet, or uncompressed)
- Character encoding (LATIN or UTF-8)
- Field delimiter (comma, tab, and so on)
- Record delimiter (line feed)