Native is twice as fast as Wasm

web-tree-sitter (npm install web-tree-sitter) can run in both the browser and natively, node-tree-sitter (npm install tree-sitter) can only run natively but it runs faster.

Parsing Python's entire standard library (about 800,000 lines of code) with tree-sitter-python, we get these results on an M1 (Arm) Macbook Air with 8GB of RAM (lower is better):

node-tree-sitter: 2221ms
web-tree-sitter:  5144ms

a 2.3x speedup.

Here's how to reproduce these results:

mkdir /tmp/tree-sitter-vs-node-tree-sitter
cd /tmp/tree-sitter-vs-node-tree-sitter
npm init -y
# add `"type": "module",` to package.json
npm install tree-sitter tree-sitter-cli web-tree-sitter tree-sitter-python
npx tree-sitter build-wasm node_modules/tree-sitter-python

Create a file to parse as a benchmark, for example all files in Python's standard library concatenated into one file:

git clone https://github.com/python/cpython.git
#cd cpython && checkout f508800 && cd ..
find cpython/Lib/ -type f -name "*.py" -exec cat {} + > sample_python_code.py

Create the two JavaScript files that will parse that file in the two possible ways:

native.js

import Parser from "tree-sitter";
import Python from "tree-sitter-python";
import fs from "fs";

const parser = new Parser();
parser.setLanguage(Python);

const file = fs.readFileSync("./sample_python_code.py", "utf8");

const start = process.hrtime.bigint();
parser.parse(file);
const end = process.hrtime.bigint();
console.log(`node-tree-sitter parse time: ${(end - start) / 1000000n}ms`);

wasm.js

import Parser from "web-tree-sitter";
import fs from "fs";

await Parser.init();
const Python = await Parser.Language.load("./tree-sitter-python.wasm");
const parser = new Parser();
parser.setLanguage(Python);

const file = fs.readFileSync("./sample_python_code.py", "utf8");

const start = process.hrtime.bigint();
parser.parse(file);
const end = process.hrtime.bigint();
console.log(`web-tree-sitter parse time:  ${(end - start) / 1000000n}ms`);

and run them:

node native.js
node wasm.js