Struct ChunkDecoder 

pub struct ChunkDecoder { /* private fields */ }

Stateful chunk decoder — holds bzip2 stream parameters.

Implementations

impl ChunkDecoder

pub fn from_header(header: &[u8]) -> Result<Self, BlockError>

Create a decoder from the bzip2 stream header (first 4 bytes: BZhN).

Call this once with the first 4 bytes of the file, then use decode_chunk() repeatedly on subsequent compressed chunks.

pub fn decode_chunk_segments( &self, data: &[u8], is_last: bool, ) -> Result<(Vec<Vec<u8>>, usize), BlockError>

Like decode_chunk, but returns each segment’s output separately.

Avoids the single-threaded assembly of a giant Vec — the caller can send each segment to the writer independently.

Returns (segments, bytes_consumed).

pub fn decode_chunk( &self, data: &[u8], is_last: bool, ) -> Result<(Vec<u8>, usize), BlockError>

Decode all complete bzip2 blocks in data.

data is a raw slice of compressed bzip2 data (may include the 4-byte header on first call, or start mid-stream on subsequent calls).

Returns (decompressed_output, bytes_consumed).

• decompressed_output: concatenated decoded blocks, in order.
• bytes_consumed: how many bytes of data were fully decoded. The caller must carry data[bytes_consumed..] into the next chunk.

If is_last is true, all blocks are decoded (even the last one). Otherwise the last block is skipped (it may be incomplete).