Package com.pkgrove.pkgrovekit.storage

Interface RowPartFormat

  • All Implemented Interfaces:
    
public interface RowPartFormat

    How rows become object-part bytes and back (HEL-236 scenarios 1–2). The format is a SEAM: PkgroveKit ships the dependency-free JsonLinesRowFormat; columnar formats (Parquet) are deliberately delegated to engines that already own them — DuckDB reads/writes Parquet on S3 natively (docs/storage.md shows that path) — because a faithful Parquet writer would drag the Hadoop dependency tree into a module whose contract is "lightweight".

    • Nested Class Summary

      Nested Classes 
      Modifier and Type Class Description

    • Field Summary

      Fields 
      Modifier and Type Field Description

    • Constructor Summary

      Constructors 
      Constructor Description

    • Enum Constant Summary

      Enum Constants 
      Enum Constant Description

    • Method Summary

      Modifier and Type Method Description
      abstract Long write(Schema schema, Iterable<Row> rows, OutputStream out) Encode rows to out; returns the row count written.
      abstract Sequence<RowBatch> read(Schema schema, InputStream input, Integer batchRows) Decode a part back into batches of at most batchRows rows.
      abstract String getId() Stable identifier recorded in manifests (a reader must match it).
      abstract String getFileExtension() Suggested key suffix, e.g.

      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait