public class FileDescrip extends Object
Describes one file within a scan and is used to populate implicit columns. Specify the file name and optional selection root. If the selection root is provided, then partitions are defined as the portion of the file name that is not also part of the selection root. That is, if selection root is /a/b and the file path is /a/b/c/d.csv, then dir0 is c.
  • Field Details

    • rowGroupIndex

      protected Integer rowGroupIndex
    • rowGroupStart

      protected Long rowGroupStart
    • rowGroupLength

      protected Long rowGroupLength
    • isEmpty

      protected boolean isEmpty
  • Constructor Details

    • FileDescrip

      public FileDescrip(DrillFileSystem dfs, FileWork fileWork, org.apache.hadoop.fs.Path selectionRoot)
  • Method Details

    • fileSystem

      public DrillFileSystem fileSystem()
      Gives the Drill file system for this operator.
    • filePath

      public org.apache.hadoop.fs.Path filePath()
      Returns Drill's version of the Hadoop file split.
    • split

      public org.apache.hadoop.mapred.FileSplit split()
      Describes the file split (path and block offset) for this scan.
      Hadoop file split object with the file path, block offset, and length.
    • fileWork

      public FileWork fileWork()
    • partition

      public String partition(int index)
    • dirPathLength

      public int dirPathLength()
    • setRowGroupAttribs

      public void setRowGroupAttribs(int index, long start, long length)
    • getModTime

      public String getModTime()
    • setModTime

      public void setModTime(String modTime)
      Explicitly set the cached modification time. For testing only.
    • setCompressible

      public void setCompressible(boolean isCompressed)
    • isCompressible

      public boolean isCompressible()
    • open

      public InputStream open() throws IOException
    • markEmpty

      public void markEmpty()