A base64 codec using wasm32 SIMD intrinsics.
use based64::{decode, encode};
use wasm_bindgen::{wasm_bindgen, JsValue};
#[wasm_bindgen]
fn main() -> Result<(), JsValue> {
let ascii = b"VGhlIGRvZyBsaWNrZWQgdGhlIG9pbCwgYW5kIGV2ZXJ5Ym9keSBsYXVnaGVkLg==";
let message = decode(ascii)?; // The dog licked the oil, and everybody laughed.
let encoded_to_ascii = encode(&message)?;
assert_eq!(encoded_to_ascii, ascii.to_vec());
Ok(())
}
or
import init, {
encode,
decode,
} from "./pkg/based64.js";
async function run(): boolean {
await init();
let data = "howdy";
let bytes = new TextEncoder().encode(data);
let ascii: Uint8Array = encode(bytes);
let rawString = decode(ascii);
return data === rawString;
}
# make sure to have the wasm32 target installed
rustup target add wasm32-unknown-unknown
RUSTFLAGS=\"-C target-feature=+simd128 cargo test --target=wasm32-unknown-unknown
To run benchmarks, run just bench
. It should lead you to a web page, you can view the console.
The benchmark rules are very simple, it must follow window.btoa
and window.atob
's function header: String
-> String
.
Since certain functions have different function signatures, the work needed to convert into a String
is included in the measurement.
Codecs measured:
base64
withwasm_bindgen
bindingswindow.atob()
,window.btoa()
based64
Uint8Array
->Uint8Array
withwasm_bindgen
bindingsbased64
String
->String
withwasm_bindgen
bindings
- Mcyoung's Designing a SIMD Algorithm from Scratch
- Daniel Lemire's Ridiculously fast base64 encoding and decoding
- core::arch::wasm32::v128