Struct DictIndexDecoder

pub struct DictIndexDecoder {
    decoder: RleDecoder,
    index_buf: Box<[i32; 1024]>,
    index_buf_len: usize,
    index_offset: usize,
    max_remaining_values: usize,
}

Decoder for Encoding::RLE_DICTIONARY indices

Fields

decoder: RleDecoder

Decoder for the dictionary offsets array

index_buf: Box<[i32; 1024]>

We want to decode the offsets in chunks so we will maintain an internal buffer of decoded offsets

index_buf_len: usize

Current length of index_buf

index_offset: usize

Current offset into index_buf. If index_buf_offset == index_buf_len then we’ve consumed the entire buffer and need to decode another chunk of offsets.

max_remaining_values: usize

This is a maximum as the null count is not always known, e.g. value data from a v1 data page

Implementations

impl DictIndexDecoder

pub fn new(data: Bytes, num_levels: usize, num_values: Option<usize>) -> Self

Create a new DictIndexDecoder with the provided data page, the number of levels associated with this data page, and the number of non-null values (if known)

pub fn read<F: FnMut(&[i32]) -> Result<()>>( &mut self, len: usize, f: F, ) -> Result<usize>

Read up to len values, returning the number of values read and calling f with each decoded dictionary index

Will short-circuit and return on error

pub fn skip(&mut self, to_skip: usize) -> Result<usize>

Skip up to to_skip values, returning the number of values skipped