Added MD2 hash function #7
Conversation
| @@ -0,0 +1,19 @@ | |||
| [package] | |||
| name = "md2" | |||
| version = "0.4.0" | |||
There was a problem hiding this comment.
It's better to start from 0.1.0, as it's not required to synchronize versions across crates.
| digest-buffer = "0.2" | ||
| generic-array = "0.6" | ||
| byte-tools = "0.1" | ||
| fake-simd = "0.1" |
There was a problem hiding this comment.
You don't use neither byte-tools, nor fake-simd, so you can remove them from here.
| @@ -0,0 +1,25 @@ | |||
| Copyright (c) 2016 bacher09, Artyom Pavlov | |||
There was a problem hiding this comment.
It's your work, so you can insert your own credits instead of mine and bacher09. ;)
|
|
||
| #![no_std] | ||
| extern crate generic_array; | ||
| extern crate byte_tools; |
|
|
||
| fn process_block(&mut self, input: &Block) { | ||
| // Update state | ||
| for j in 0..16usize { |
There was a problem hiding this comment.
usize is not necessary here. Rust can infer type in this situation.
| self.finalize(); | ||
|
|
||
| let mut out = GenericArray::default(); | ||
| for (x, y) in self.state.x[0..16].iter().zip(out.iter_mut()) { |
There was a problem hiding this comment.
The simple loop:
for i in 0..16 {
out[i] = self.state.x[i];
}
will be more readable and should have the same efficiency. (compiler can remove bound checks in this case) Or you can use copy_memory from byte_tools.
| } | ||
| } | ||
|
|
||
| impl Default for Md2 { |
There was a problem hiding this comment.
Actually instead of implementing Default like this you can simply derive it for Md2 and Md2State. After that you will not need new() for Md2State and you will be able to simply use Default::default() in Md2::new(). (well some argue that there is no need for new without any arguments and that Default trait would be sufficient, maybe in the future I will remove new() altogether)
| break; | ||
| } | ||
| } | ||
| print_result(&sh.result(), name); |
There was a problem hiding this comment.
Now we can write simply println!("{:x}\t{}", sh.result(), name), so print_result function is not necessary.
There was a problem hiding this comment.
I tried doing that, but the compiler complains that <D as md2::Digest>::OutputSize (which is generic_array::typenum::U16) does not implement std::ops::Add.
Any ideas?
There was a problem hiding this comment.
Hm, implementation of LowerHex and UpperHex requires additional trait bounds to create array with double length internally. The easiest way will be to remove unnecessary generalization over Digest and simply create Md2 instance inside process:
fn process<R: Read>(reader: &mut R, name: &str) {
let mut sh = Md2::new();
...
println!("{:x}\t{}", &sh.result(), name);
}
|
|
||
| // Update checksum | ||
| let mut l = self.checksum[15]; | ||
| for j in 0..16usize { |
| } | ||
|
|
||
| // #[test] | ||
| // fn md2_1million_a() { |
There was a problem hiding this comment.
I guess it's better to add 1million_a test, e.g. by generating it by another tool, or remove it altogether.
There was a problem hiding this comment.
At this point I guess I'll just remove it. Later I'll try to create the 1million_a test file.
|
Anything else? |
|
Everything looks good! Thank you! |
I added the MD2 hash function and 3 basic tests (the ones on the Wikipedia page). I based the structure on the MD4 hash, but I'm not sure if the example code is correct.
If there are any problems, please let me know. :)