sam.hadow/disk-usage
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

blake 3 instead of blake 2

Browse Source
This commit is contained in:
Sam Hadow 2025-04-08 22:03:49 +02:00
parent 6966b141da
commit 0b41b34c45
2 changed files with 7 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
Cargo.toml
View File

@ -6,6 +6,7 @@ authors = ["Sam Hadow"]
[dependencies] [dependencies]
blake2 = "0.10.6" blake2 = "0.10.6"
blake3 = "1.8.1"
clap = { version = "4.4.6", features = ["derive"] } clap = { version = "4.4.6", features = ["derive"] }
rayon = "1.8.0" rayon = "1.8.0"

11
src/duplicates.rs
View File

@ -8,7 +8,7 @@ use std::fs;
use std::path::{Path, PathBuf}; use std::path::{Path, PathBuf};
// hash // hash
use blake2::{Blake2b512, Digest}; use blake3::Hasher;
use std::io::{BufReader, Read}; use std::io::{BufReader, Read};
// parallelism // parallelism
@ -18,7 +18,7 @@ use std::sync::{Arc, Mutex};
impl FileTree { impl FileTree {
/// Find duplicates in a directory (including sub-directories). /// Find duplicates in a directory (including sub-directories).
/// ///
/// If path exist in tree, find duplicates using Blake2b512. If 2 (or more) files have the same hash they're duplicates. /// If path exist in tree, find duplicates using Blake3. If 2 (or more) files have the same hash they're duplicates.
/// ///
/// returns a `Vec` containing a `Vec<PathBuf>` for each group of duplicates. /// returns a `Vec` containing a `Vec<PathBuf>` for each group of duplicates.
/// ///
@ -52,7 +52,7 @@ impl FileTree {
// parallelized loop // parallelized loop
intersection.par_iter().for_each(|item| { intersection.par_iter().for_each(|item| {
if let Ok(file) = fs::File::open(item) { if let Ok(file) = fs::File::open(item) {
let mut hasher = Blake2b512::new(); let mut hasher = Hasher::new();
// sizable buffer // sizable buffer
let mut buffer = [0; 8192]; let mut buffer = [0; 8192];
@ -69,8 +69,9 @@ impl FileTree {
hasher.update(&buffer[..count]); hasher.update(&buffer[..count]);
} }
let hash = hasher.finalize(); let mut hash = hasher.finalize_xof();
let hash_bytes: [u8; 32] = hash.as_slice()[..32].try_into().unwrap(); let mut hash_bytes: [u8; 32] = [0; 32];
hash.fill(&mut hash_bytes);
// Use a Mutex to update HashMap in parallel // Use a Mutex to update HashMap in parallel
let mut locked_hashes = hashes.lock().unwrap(); let mut locked_hashes = hashes.lock().unwrap();