Package org.apache.drill.exec.vector.accessor.writer

Class UnionWriterImpl

java.lang.Object
org.apache.drill.exec.vector.accessor.writer.UnionWriterImpl
All Implemented Interfaces:
ColumnWriter, VariantWriter, WriterEvents, WriterPosition
public class UnionWriterImpl extends Object implements VariantWriter, WriterEvents
Writer to a union vector.

A union vector has three attributes: null flag, type and value. The union vector holds the type: a bundle of other vectors hold the value. The type says which of the other vectors to consult to write the value. If a column is null, then we consult no other vectors. If all columns (thus far) are null, then there are no associated data vectors.

The protocol is to first set the type. Doing so creates the associated data vector, if it does not yet exist. This highlights the poor design of this vector: if we have even one value of a given type, we must have a vector that holds values for all rows, then we ignore the unwanted values.

  • Constructor Details

  • Method Details

    • bindIndex

      public void bindIndex(ColumnWriterIndex index)
      Description copied from interface: WriterEvents
      Bind the writer to a writer index.
      Specified by:
      bindIndex in interface WriterEvents
      Parameters:
      index - the writer index (top level or nested for arrays)

    • bindListener

      public void bindListener(VariantWriter.VariantWriterListener listener)

    • bindListener

      public void bindListener(WriterEvents.ColumnWriterListener listener)
      Description copied from interface: WriterEvents
      Bind a listener to the underlying vector writer. This listener reports on vector events (overflow, growth), and so is called only when the writer is backed by a vector. The listener is ignored (and never called) for dummy (non-projected) columns. If the column is compound (such as for a nullable or repeated column, or for a map), then the writer is bound to the individual components.
      Specified by:
      bindListener in interface WriterEvents
      Parameters:
      listener - the vector event listener to bind

    • state

      public WriterEvents.State state()

    • index

      public ColumnWriterIndex index()

    • listener

      public VariantWriter.VariantWriterListener listener()

    • shim

      public UnionShim shim()

    • elementPosition

      public WriterPosition elementPosition()

    • bindShim

      public void bindShim(UnionShim shim)

    • type

      public ObjectType type()
      Description copied from interface: ColumnWriter
      Return the object (structure) type of this writer.
      Specified by:
      type in interface ColumnWriter
      Returns:
      type indicating if this is a scalar, tuple or array

    • nullable

      public boolean nullable()
      Description copied from interface: ColumnWriter
      Whether this writer allows nulls. This is not as simple as checking for the TypeProtos.DataMode.OPTIONAL type in the schema. List entries are nullable, if they are primitive, but not if they are maps or lists. Unions are nullable, regardless of cardinality.
      Specified by:
      nullable in interface ColumnWriter
      Returns:
      true if a call to ColumnWriter.setNull() is supported, false if not

    • schema

      public ColumnMetadata schema()
      Description copied from interface: ColumnWriter
      Returns the schema of the column associated with this writer.
      Specified by:
      schema in interface ColumnWriter
      Returns:
      schema for this writer's column

    • variantSchema

      public VariantMetadata variantSchema()
      Description copied from interface: VariantWriter
      Metadata description of the variant that includes the set of types, along with extended properties of the types such as expected allocations sizes, expected array cardinality, etc.
      Specified by:
      variantSchema in interface VariantWriter
      Returns:
      metadata for the variant

    • size

      public int size()
      Description copied from interface: VariantWriter
      Returns the number of types in the variant. Some implementations (such as lists) impart special meaning to a variant with a single type.
      Specified by:
      size in interface VariantWriter
      Returns:
      number of types in the variant

    • hasType

      public boolean hasType(TypeProtos.MinorType type)
      Description copied from interface: VariantWriter
      Determine if the union vector has materialized storage for the given type. (The storage will be created as needed during writing.)
      Specified by:
      hasType in interface VariantWriter
      Parameters:
      type - data type
      Returns:
      true if a value of the given type has been written and storage allocated (or storage was allocated implicitly), false otherwise

    • setNull

      public void setNull()
      Description copied from interface: ColumnWriter
      Set the current value to null. Support depends on the underlying implementation: only nullable types support this operation. throws IllegalStateException if called on a non-nullable value.
      Specified by:
      setNull in interface ColumnWriter

    • memberWriter

      public ObjectWriter memberWriter(TypeProtos.MinorType type)
      Description copied from interface: VariantWriter
      Create or retrieve a writer for the given type. Use this form when caching writers. This form does not set the type of the current row; call VariantWriter.setType(MinorType) per row when the writers are cached. This method can be called at any time as it does not depend on an active batch.
      Specified by:
      memberWriter in interface VariantWriter
      Parameters:
      type - the type of the writer to cache
      Returns:
      the writer for that type without setting the type of the current row.

    • member

      public ObjectWriter member(TypeProtos.MinorType type)
      Description copied from interface: VariantWriter
      Set the type of the present value and get the writer for that type. Available only when a batch is active. Use this form to declare the type of the current row, and retrieve a writer for that value.
      Specified by:
      member in interface VariantWriter
      Parameters:
      type - type to set for the current row
      Returns:
      writer for the type just set

    • setType

      public void setType(TypeProtos.MinorType type)
      Description copied from interface: VariantWriter
      Explicitly set the type of the present value. Use this when the writers are cached. The writer must already exist.
      Specified by:
      setType in interface VariantWriter
      Parameters:
      type - type to set for the current row

    • addMember

      public ObjectWriter addMember(ColumnMetadata colSchema)
      Specified by:
      addMember in interface VariantWriter

    • addMember

      public ObjectWriter addMember(TypeProtos.MinorType type)
      Specified by:
      addMember in interface VariantWriter

    • addMember

      protected void addMember(AbstractObjectWriter writer)
      Add a column writer to an existing union writer. Used for implementations that support "live" schema evolution: column discovery while writing. The corresponding metadata must already have been added to the schema. Called by the shim's addMember to do writer-level tasks.
      Parameters:
      writer - the column writer to add

    • scalar

      public ScalarWriter scalar(TypeProtos.MinorType type)
      Specified by:
      scalar in interface VariantWriter

    • tuple

      public TupleWriter tuple()
      Specified by:
      tuple in interface VariantWriter

    • array

      public ArrayWriter array()
      Specified by:
      array in interface VariantWriter

    • isProjected

      public boolean isProjected()
      Description copied from interface: ColumnWriter
      Whether this writer is projected (is backed by a materialized vector), or is unprojected (is just a dummy writer.) In most cases, clients can ignore whether the column is projected and just write to the writer. This flag handles those special cases where it is helpful to know if the column is projected or not.
      Specified by:
      isProjected in interface ColumnWriter

    • startWrite

      public void startWrite()
      Description copied from interface: WriterEvents
      Start a write (batch) operation. Performs any vector initialization required at the start of a batch (especially for offset vectors.)
      Specified by:
      startWrite in interface WriterEvents

    • startRow

      public void startRow()
      Description copied from interface: WriterEvents
      Start a new row. To be called only when a row is not active. To restart a row, call WriterEvents.restartRow() instead.
      Specified by:
      startRow in interface WriterEvents

    • endArrayValue

      public void endArrayValue()
      Description copied from interface: WriterEvents
      End a value. Similar to WriterEvents.saveRow(), but the save of a value is conditional on saving the row. This version is primarily of use in tuples nested inside arrays: it saves each tuple within the array, advancing to a new position in the array. The update of the array's offset vector based on the cumulative value saves is done when saving the row.
      Specified by:
      endArrayValue in interface WriterEvents

    • restartRow

      public void restartRow()
      Description copied from interface: WriterEvents
      During a writer to a row, rewind the the current index position to restart the row. Done when abandoning the current row, such as when filtering out a row at read time.
      Specified by:
      restartRow in interface WriterEvents

    • saveRow

      public void saveRow()
      Description copied from interface: WriterEvents
      Saves a row. Commits offset vector locations and advances each to the next position. Can be called only when a row is active.
      Specified by:
      saveRow in interface WriterEvents

    • preRollover

      public void preRollover()
      Description copied from interface: WriterEvents
      The vectors backing this vector are about to roll over. Finish the current batch up to, but not including, the current row.
      Specified by:
      preRollover in interface WriterEvents

    • postRollover

      public void postRollover()
      Description copied from interface: WriterEvents
      The vectors backing this writer rolled over. This means that data for the current row has been rolled over into a new vector. Offsets and indexes should be shifted based on the understanding that data for the current row now resides at the start of a new vector instead of its previous location elsewhere in an old vector.
      Specified by:
      postRollover in interface WriterEvents

    • endWrite

      public void endWrite()
      Description copied from interface: WriterEvents
      End a batch: finalize any vector values.
      Specified by:
      endWrite in interface WriterEvents

    • lastWriteIndex

      public int lastWriteIndex()
      Description copied from interface: WriterPosition
      Return the last write position in the vector. This may be the same as the writer index position (if the vector was written at that point), or an earlier point. In either case, this value points to the last valid value in the vector.
      Specified by:
      lastWriteIndex in interface WriterPosition
      Returns:
      index of the last valid value in the vector

    • rowStartIndex

      public int rowStartIndex()
      Description copied from interface: WriterPosition
      Position within the vector of the first value for the current row. Note that this is always the first value for the row, even for a writer deeply nested within a hierarchy of arrays. (The first position for the current array is not exposed in this API.)
      Specified by:
      rowStartIndex in interface WriterPosition
      Returns:
      the vector offset of the first value for the current row

    • writeIndex

      public int writeIndex()
      Description copied from interface: WriterPosition
      Current write index for the writer. This is the global array location for arrays, same as the row index for top-level columns.
      Specified by:
      writeIndex in interface WriterPosition
      Returns:
      current write index

    • copy

      public void copy(ColumnReader from)
      Description copied from interface: ColumnWriter
      Copy a single value from the given reader, which must be of the same type as this writer.
      Specified by:
      copy in interface ColumnWriter
      Parameters:
      from - reader to provide the data

    • setObject

      public void setObject(Object value)
      Description copied from interface: ColumnWriter
      Generic technique to write data as a generic Java object. The type of the object must match the target writer. Primarily for testing.
      • Scalar: The type of the Java object must match the type of the target vector. String or byte[] can be used for Varchar vectors.
      • Array: Write the array given an array of values. The object must be a Java array. The type of the array must match the type of element in the repeated vector. That is, if the vector is a Repeated Int, provide an int[] array.
      • Tuple (map or row): The Java object must be an array of objects in which the members of the array have a 1:1 correspondence with the members of the tuple in the order defined by the writer metadata. That is, if the map is (Int, Varchar), provide a Object[] array like this: {10, "fred"}.
      • Union: Uses the Java object type to determine the type of the backing vector. Creates a vector of the required type if needed.
      Specified by:
      setObject in interface ColumnWriter
      Parameters:
      value - value to write to the vector. The Java type of the object indicates the Drill storage type

    • dump

      public void dump(HierarchicalFormatter format)
      Specified by:
      dump in interface WriterEvents