Package org.apache.drill.exec.store.hive

Class HiveMetadataProvider.LogicalInputSplit

java.lang.Object
org.apache.drill.exec.store.hive.HiveMetadataProvider.LogicalInputSplit
Enclosing class:
HiveMetadataProvider
public static class HiveMetadataProvider.LogicalInputSplit extends Object
Contains group of input splits along with the partition. For non-partitioned tables, the partition field is null. Input splits can be logically grouped together, for example, in case of when table has header / footer. In this case all splits from the same file should be processed together.

  • Constructor Details

    • LogicalInputSplit

      public LogicalInputSplit(org.apache.hadoop.mapred.InputSplit inputSplit, org.apache.hadoop.hive.metastore.api.Partition partition)

    • LogicalInputSplit

      public LogicalInputSplit(Collection<? extends org.apache.hadoop.mapred.InputSplit> inputSplits, org.apache.hadoop.hive.metastore.api.Partition partition)

  • Method Details

    • getInputSplits

      public Collection<org.apache.hadoop.mapred.InputSplit> getInputSplits()

    • getPartition

      public org.apache.hadoop.hive.metastore.api.Partition getPartition()

    • getLength

      public long getLength() throws IOException
      Returns:
      returns total length of all stored input splits
      Throws:
      IOException

    • getLocations

      public Collection<String> getLocations() throws IOException
      Returns:
      collection of unique locations where input splits are stored
      Throws:
      IOException

    • serialize

      public List<String> serialize() throws IOException
      Serializes each input split to string using Base64 encoding.
      Returns:
      list of serialized input splits
      Throws:
      IOException

    • getType

      public String getType()
      Returns:
      returns input split class name if at least one input split is present, null otherwise.