ORC Schema
An ORC schema describes the column types and nested structure of Hive ORC columnar files, supporting compound types such as union, struct, list, and map.
Purpose
To provide a type description for ORC files so Hive/Spark read and write column data and nested structures correctly.
Mechanism
The schema is a tree of column types: scalars (int/string/...), struct (named fields), list (element type), map (key/value types), and union (list of types). Each column is encoded and indexed independently.
Use cases
- Hive ORC files
- Spark data reading
- Columnar storage schemas