Compression (DBTL-1)

[seb-hyland]

This code is heavily WIP at the moment and is untested. Exec.rs likely contains bugs or problems.

The goal of this branch is to create:

1. A function that can arbitrarily run binaries on Windows/Linux (the binary will be written into a variable at compile time, then ran as an anonymous file at runtime)
2. Create the DBTL-1 variant of the compression implementation.

[seb-hyland]

Provides compression and decompression mechanisms for lz4 and ts_zip. Exec.rs has been removed.

Compression takes inpath:PathBuf as input and outputs Result<PathBuf, io::Error> where PathBuf references the output file.
Decompression takes inpath:PathBuf and outpath:PathBuf as input and outputs Result<()>.

[lhao03]

can you add a test with quickcheck, just to quickly verify the decompress and compress work?

dna-software/src-tauri/src/encoder.rs

Lines 75 to 92 in 1519163

	#[quickcheck]
	fn rotation_encode_decode(bytes: Vec<u8>) -> bool {
	let encoder = RotationEncoder {};
	let decoder = RotationDecoder {};
	let bits = BitVec::from_vec(bytes);
	bits == decoder.decode(&encoder.encode(&bits))
	}
	#[quickcheck]
	fn quaternary_encode_decode(bytes: Vec<u8>) -> bool {
	if bytes.len() % 2 == 1 {
	true
	} else {
	let encoder = QuaternaryEncoder {};
	let decoder = QuaternaryDecoder {};
	let bits = BitVec::from_vec(bytes);
	bits == decoder.decode(&encoder.encode(&bits))
	}

[lhao03]

also can you edit main.rs to add the compression during encode_sequence?

dna-software/src-tauri/src/main.rs

Lines 35 to 47 in 1519163

	#[tauri::command]
	fn encode_sequence(encoder_type: &str, file_path: &str) -> Result<Vec<Base>, String> {
	let bytes = fs::read(file_path).map_err(|err| err.to_string())?;
	let bits = BitVec::<_, Msb0>::from_vec(bytes);
	let encoder: Box<dyn Encoder> = match encoder_type {
	"quaternary" => Box::new(QuaternaryEncoder {}),
	"rotation" => Box::new(RotationEncoder {}),
	_ => return Err("Selected encoder does not exist.".to_string()),
	};
	Ok(encoder.encode(&bits).into())
	}

it should follow the old deleted sequence.rs almost exactly

dna-software/src-tauri/src/sequence.rs

Lines 11 to 21 in 2b88993

	fn encode(
	path: impl AsRef<Path>,
	primer: Primer,
	compressor: impl Compressor,
	encoder: impl Encoder,
	) -> io::Result<Vec<Base>> {
	let file = fs::read(path)?;
	let compressed = compressor.compress(file);
	let bit_sequence = BitVec::from_vec(compressed);
	Ok(encoder.encode(bit_sequence))
	}

[lhao03]

To test this, do the below

[lhao03]

The assumption that this directory exists causes an error to be thrown; additionally, there is no handling of the case where the directory does not exist; I pushed a way to fix this, but this assumption is made in several other places in the code; please fix them and add proper handling of the case when the directory doesn't exist (either make it through Rust, or prompt the user to make it themselves).

[seb-hyland]

Hi Lucy, I pushed up a hefty merge with my changes, your changes, and changes in main. This should include:

1. Your changes to compress_lz4 to ensure the directory exists
2. My updates to compressor.rs and the updated encode_sequence function with my changes
3. I have kept your dbg!("here"); macros; do you wish to keep those for now?

Although I haven't gone over everything again since the merge (except for resolving conflicts), the project is compiling and tests are passing (including the quickcheck for (de)compression). Do you wish to make a test for encode_sequence? I'm not sure how we would do that save for a test similar to chaosDNA (no assertion, just use some test sentence and print the output to CLI).

[lhao03]

Hi Lucy, I pushed up a hefty merge with my changes, your changes, and changes in main. This should include:

1. Your changes to compress_lz4 to ensure the directory exists
2. My updates to compressor.rs and the updated encode_sequence function with my changes
3. I have kept your dbg!("here"); macros; do you wish to keep those for now?

Although I haven't gone over everything again since the merge (except for resolving conflicts), the project is compiling and tests are passing (including the quickcheck for (de)compression). Do you wish to make a test for encode_sequence? I'm not sure how we would do that save for a test similar to chaosDNA (no assertion, just use some test sentence and print the output to CLI).

Yeah we can have a test for encode_sequence using quickcheck, I can set that up (but yeah it would be a test sentence or str that we can check compress -> bit -> base -> bit -> decompress)

[lhao03]

Thanks for working on this Sebastian!