Trait ChunksWriter

pub trait ChunksWriter: Sized

Write chunks to a byte destination. Then write each chunk with writer.write_chunk(chunk).

Required Methods

fn total_chunks_count(&self) -> usize

The total number of chunks that the complete file will contain.

fn write_chunk(&mut self, index_in_header_increasing_y: usize, chunk: Chunk) -> UnitResult

Any more calls will result in an error and have no effect. If writing results in an error, the file and the writer may remain in an invalid state and should not be used further. Errors when the chunk at this index was already written.

Provided Methods

fn on_progress<F>(&mut self, on_progress: F) -> OnProgressChunkWriter<'_, Self, F>
where
    F: FnMut(f64),

Obtain a new writer that calls the specified closure for each block that is written to this writer.

fn sequential_blocks_compressor<'w>(&'w mut self, meta: &'w MetaData) -> SequentialBlocksCompressor<'w, Self>

Obtain a new writer that can compress blocks to chunks, which are then passed to this writer.

fn parallel_blocks_compressor<'w>(&'w mut self, meta: &'w MetaData) -> Option<ParallelBlocksCompressor<'w, Self>>

Obtain a new writer that can compress blocks to chunks on multiple threads, which are then passed to this writer. Returns none if the sequential compressor should be used instead (thread pool creation failure or too large performance overhead).

fn compress_all_blocks_sequential(self, meta: &MetaData, blocks: impl Iterator<Item = (usize, UncompressedBlock)>) -> UnitResult

Compresses all blocks to the file. The index of the block must be in increasing line order within the header. Obtain iterator with MetaData::collect_ordered_blocks(...) or similar methods.

fn compress_all_blocks_parallel(self, meta: &MetaData, blocks: impl Iterator<Item = (usize, UncompressedBlock)>) -> UnitResult

Compresses all blocks to the file. The index of the block must be in increasing line order within the header. Obtain iterator with MetaData::collect_ordered_blocks(...) or similar methods. Will fallback to sequential processing where threads are not available, or where it would not speed up the process.

Implementors