Crates.io | brotlic |
lib.rs | brotlic |
version | 0.8.2 |
source | src |
created_at | 2022-04-03 17:43:12.948362 |
updated_at | 2024-01-27 13:39:03.363613 |
description | Bindings to the brotli library featuring a low-overhead encoder and decoder, io::Write and io::Read wrappers for compression and decompression at customizable compression qualities and window sizes. |
homepage | |
repository | https://github.com/AronParker/brotlic |
max_upload_size | |
id | 561383 |
size | 154,192 |
Brotlic (or BrotliC) is a thin wrapper around brotli.
A C compiler is required for building brotli with cargo.
Brotlic provides Rust bindings to all compression and decompression APIs. On the fly compression and
decompression is supported for both BufRead
and Write
via CompressorReader<R>
,
CompressorWriter<W>
, DecompressorReader<R>
and DecompressorWriter<W>
. For low-level
instances, see BrotliEncoder
and BrotliDecoder
. These can be configured via
BrotliEncoderOptions
and BrotliDecoderOptions
respectively.
When dealing with BufRead
:
DecompressorReader<R>
- Reads a brotli compressed input stream and decompresses it.CompressorReader<R>
- Reads a stream and compresses it while reading.When dealing with Write
:
CompressorWriter<W>
- Writes brotli compressed data to the underlying writer.DecompressorWriter<W>
- Writes brotli decompressed data to the underlying writer.To simplify this decision, the following table outlines all the differences:
Input | Output | Wraps | |
---|---|---|---|
CompressorReader<R> |
Uncompressed | Compressed | BufRead |
DecompressorReader<R> |
Compressed | Uncompressed | BufRead |
CompressorWriter<W> |
Uncompressed | Compressed | Write |
DecompressorWriter<W> |
Compressed | Uncompressed | Write |
To compress a file with brotli:
use std::fs::File;
use std::io::{self, Write};
use brotlic::CompressorWriter;
let mut input = File::open("test.txt")?; // uncompressed text file
let mut output = File::create("test.brotli")?; // compressed text output file
let mut output_compressed = CompressorWriter::new(output);
output_compressed.write_all(b"test")?;
To decompress that same file:
use std::fs::File;
use std::io::{self, BufReader, Read};
use brotlic::DecompressorReader;
let mut input = BufReader::new(File::open("test.brotli")?); // uncompressed text file
let mut input_decompressed = DecompressorReader::new(input); // requires BufRead
let mut text = String::new();
input_decompressed.read_to_string(&mut text)?;
assert_eq!(text, "test");
To compress and decompress in memory:
use std::io::{self, Read, Write};
use brotlic::{CompressorWriter, DecompressorReader};
let input = vec![0; 1024];
// create a wrapper around Write that supports on the fly brotli compression.
let mut compressor = CompressorWriter::new(Vec::new()); // write to memory
compressor.write_all(input.as_slice());
let encoded_input = compressor.into_inner()?; // read to vec
// create a wrapper around BufRead that supports on the fly brotli decompression.
let mut decompressed_reader = DecompressorReader::new(encoded_input);
let mut decoded_input = Vec::new();
decompressed_reader.read_to_end(&mut decoded_input)?;
assert_eq!(input, decoded_input);
Sometimes it can be desirable to trade run-time costs for an even better compression ratio by customizing compression quality:
use brotlic::{BlockSize, BrotliEncoderOptions, CompressorWriter, Quality, WindowSize};
let encoder = BrotliEncoderOptions::new()
.quality(Quality::best())
.window_size(WindowSize::best())
.block_size(BlockSize::best())
.build()?;
let writer = Vec::new();
let compressed_writer = CompressorWriter::with_encoder(encoder, writer);
It is recommended to not use the encoder directly but instead pass it onto the higher level
abstractions like CompressorWriter<W>
or DecompressorReader<R>
.
Brotlic has comparable performance to the rust-brotli library:
To conduct your own testing, run cargo bench
. This will compare the performance of this library and the rust brotli
library using inputs with different sizes and different amounts of entropy.
Licensed under either of
at your option.
Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.