HybridScanMetadata

public class HybridScanMetadata implements AutoCloseable

A parsed Parquet file footer for the experimental hybrid-scan reader.

This class is the Java binding for cudf::io::parquet::experimental::hybrid_scan_metadata. Parsing the footer apart from constructing a HybridScanReader lets a caller inspect the footer before deciding whether to build a reader at all, then build one with HybridScanReader.HybridScanReader(HybridScanMetadata, ParquetOptions) without parsing the footer again. Native code can read the parsed footer in place through getParquetMetadataAddress().

The APIs in this file are experimental and subject to change.

Constructor Summary

HybridScanMetadata(HostMemoryBuffer footerBuffer, ParquetOptions opts)

Parse the bytes of a Parquet file footer.

Method Summary

close()
getParquetMetadataAddress()	Get the address of the parsed native cudf::io::parquet::FileMetaData, for native code to read in place.

Constructors

public HybridScanMetadata(

HostMemoryBuffer footerBuffer,

ParquetOptions opts

)

Parse the bytes of a Parquet file footer.

Parameters:

• footerBuffer – host-resident footer bytes (must remain valid until this constructor returns; the JNI reads the bytes synchronously)

• opts – Parquet reader options. ParquetOptions.DEFAULT by default.

Methods

public long getParquetMetadataAddress()

Get the address of the parsed native cudf::io::parquet::FileMetaData, for native code to read in place.

The memory is owned by this object: the address is valid until close(), and must not be written through or freed. A HybridScanReader that calls HybridScanReader.setupPageIndex(HostMemoryBuffer) fills in the page index of this same footer.

Returns:

the address of the FileMetaData

public void close()