Class PlainValuesReader

java.lang.Object
org.apache.parquet.column.values.ValuesReader
org.apache.parquet.column.values.plain.PlainValuesReader
Direct Known Subclasses:
PlainValuesReader.DoublePlainValuesReader, PlainValuesReader.FloatPlainValuesReader, PlainValuesReader.IntegerPlainValuesReader, PlainValuesReader.LongPlainValuesReader

public abstract class PlainValuesReader extends ValuesReader
Plain encoding for float, double, int, long.

Reads directly from a ByteBuffer with ByteOrder.LITTLE_ENDIAN byte order, bypassing the deprecated LittleEndianDataInputStream wrapper to avoid per-value virtual dispatch overhead. The underlying page data is obtained as a single contiguous ByteBuffer via ByteBufferInputStream.slice(int).

  • Field Details

  • Constructor Details

    • PlainValuesReader

      public PlainValuesReader()
  • Method Details

    • initFromPage

      public void initFromPage(int valueCount, org.apache.parquet.bytes.ByteBufferInputStream stream) throws IOException
      Description copied from class: ValuesReader
      Called to initialize the column reader from a part of a page.

      Implementations must consume all bytes from the input stream, leaving the stream ready to read the next section of data. The underlying implementation knows how much data to read, so a length is not provided.

      Each page may contain several sections:

      • repetition levels column
      • definition levels column
      • data column
      Overrides:
      initFromPage in class ValuesReader
      Parameters:
      valueCount - count of values in this page
      stream - an input stream containing the page data at the correct offset
      Throws:
      IOException - if there is an exception while reading from the input stream
    • skip

      public void skip()
      Description copied from class: ValuesReader
      Skips the next value in the page
      Specified by:
      skip in class ValuesReader