Package org.apache.parquet.column

Interface ColumnReader

  • All Known Implementing Classes:
    ColumnReaderImpl
    public interface ColumnReader
    Reader for (repetition level, definition level, values) triplets. At any given point in time, a ColumnReader points to a single (r, d, v) triplet. In order to move to the next triplet, call consume().

    Depending on the type and the encoding of the column only a subset of the get* methods are implemented. Dictionary specific methods enable the upper layers to read the dictionary IDs without decoding the data. In particular the Converter will decode the strings in the dictionary only once and iterate on the dictionary IDs instead of the values.

    Each iteration looks at the current definition level and value as well as the next repetition level:

    • The current definition level defines if the value is null.
    • If the value is defined we can read it with the correct get*() method.
    • Looking ahead to the next repetition determines what is the next column to read for in the FSA.

    • Method Detail

      • getTotalValueCount

        @Deprecated
long getTotalValueCount()
        Deprecated.
        will be removed in 2.0.0; Total values might not be able to be counted before reading the values (e.g. in case of column index based filtering)
        Returns:
        the totalCount of values to be consumed

      • consume

        void consume()
        Consume the current triplet, moving to the next value.

      • getCurrentRepetitionLevel

        int getCurrentRepetitionLevel()
        must return 0 when isFullyConsumed() == true
        Returns:
        the repetition level for the current value

      • getCurrentDefinitionLevel

        int getCurrentDefinitionLevel()
        Returns:
        the definition level for the current value

      • writeCurrentValueToConverter

        void writeCurrentValueToConverter()
        writes the current value to the converter

      • skip

        void skip()
        Skip the current value

      • getCurrentValueDictionaryID

        int getCurrentValueDictionaryID()
        available when the underlying encoding is dictionary based
        Returns:
        the dictionary id for the current value

      • getInteger

        int getInteger()
        Returns:
        the current value

      • getBoolean

        boolean getBoolean()
        Returns:
        the current value

      • getLong

        long getLong()
        Returns:
        the current value

      • getBinary

        Binary getBinary()
        Returns:
        the current value

      • getFloat

        float getFloat()
        Returns:
        the current value

      • getDouble

        double getDouble()
        Returns:
        the current value

      • getDescriptor

        ColumnDescriptor getDescriptor()
        Returns:
        Descriptor of the column.