org.apache.spark.sql.execution.columnar.encoding
BigDictionaryDecoderNullable
Related Doc:
package encoding
org.apache.spark.sql.execution.columnar.encoding
Random access to the encoded data.
Random access to the encoded data.
Normally not used by decoder but supplied by caller to the methods but can be used if required but needs to be set by caller explicitly.
Normally not used by decoder but supplied by caller to the methods but can be used if required but needs to be set by caller explicitly.
Delta encoder/decoder depend on initialize being final and invoking initializeCursor and initializeNulls as below.
Delta encoder/decoder depend on initialize being final and invoking initializeCursor and initializeNulls as below.
Initialization will fill in the dictionaries as written by the DictionaryEncoder.
Initialization will fill in the dictionaries as written by the DictionaryEncoder. For string maps it reads in the value array written using ByteBufferHashMap by the encoder expecting the size of UTF8 encoded string followed by the string contents. Long and integer dictionaries are still using the old ObjectHashSet which needs to be moved to ByteBufferHashMap once DictionaryEncoder adds support for long/integer dictionary encoding.
Sequential null check for a value which should be invoked for all values of ordinal from 0 until maximum in order.
Sequential null check for a value which should be invoked for all values of ordinal from 0 until maximum in order.
Absolute ordinal null check for random access.
Absolute ordinal null check for random access.
Get the number of null values till given 0-based position (exclusive) for random access.
Get the number of null values till given 0-based position (exclusive) for random access.