WIP: Extract msgpack ReScript implementation as a seperated package

benchmarks/array.cjs

@@ -0,0 +1,38 @@
+const { Benchmark } = require('benchmark');
+
+const n = 10000;
+
+function Person(name, age) {
+  this.name = name;
+  this.age = age;
+}
+
+new Benchmark.Suite()
+.add('init by constructor', () => {
+  const arr = new Array(n);
+  for (let i = 0; i < n; i++) {
+    arr[i] = new Person(i.toString(), i);
+  }
+})
+.add('init by apply', () => {
+  const arr = Array(n);
+  for (let i = 0; i < n; i++) {
+    arr[i] = new Person(i.toString(), i);
+  }
+})
+.add('init by literal + push', () => {
+  const arr = [];
+  for (let i = 0; i < n; i++) {
+    arr.push(new Person(i.toString(), i));
+  }
+})
+.add('init by literal + assign', () => {
+  const arr = [];
+  for (let i = 0; i < n; i++) {
+    arr[i] = new Person(i.toString(), i);
+  }
+})
+.on('cycle', event => {
+  console.log(event.target.toString());
+})
+.run();

benchmarks/utf8-decode.js → benchmarks/utf8-decode.cjs

@@ -2,31 +2,31 @@ const { Benchmark } = require('benchmark');
 
 const textDecoder = new TextDecoder();
 
-const utf8DecodeJs = (function (bytes) {
+const utf8DecodeJs = (bytes) => {
   let offset = 0;
-  const end = bytes.length;
+  let end = bytes.length;
 
-  const units = [];
+  let units = [];
   let result = "";
   while (offset < end) {
-    const byte1 = bytes[offset++];
+    let byte1 = bytes[offset++];
     if ((byte1 & 0x80) === 0) {
       // 1 byte
       units.push(byte1);
     } else if ((byte1 & 0xe0) === 0xc0) {
       // 2 bytes
-      const byte2 = bytes[offset++] & 0x3f;
+      let byte2 = bytes[offset++] & 0x3f;
       units.push(((byte1 & 0x1f) << 6) | byte2);
     } else if ((byte1 & 0xf0) === 0xe0) {
       // 3 bytes
-      const byte2 = bytes[offset++] & 0x3f;
-      const byte3 = bytes[offset++] & 0x3f;
+      let byte2 = bytes[offset++] & 0x3f;
+      let byte3 = bytes[offset++] & 0x3f;
       units.push(((byte1 & 0x1f) << 12) | (byte2 << 6) | byte3);
     } else if ((byte1 & 0xf8) === 0xf0) {
       // 4 bytes
-      const byte2 = bytes[offset++] & 0x3f;
-      const byte3 = bytes[offset++] & 0x3f;
-      const byte4 = bytes[offset++] & 0x3f;
+      let byte2 = bytes[offset++] & 0x3f;
+      let byte3 = bytes[offset++] & 0x3f;
+      let byte4 = bytes[offset++] & 0x3f;
       let unit = ((byte1 & 0x07) << 0x12) | (byte2 << 0x0c) | (byte3 << 0x06) | byte4;
       if (unit > 0xffff) {
         unit -= 0x10000;
@@ -49,7 +49,7 @@ const utf8DecodeJs = (function (bytes) {
   }
 
   return result;
-});
+};
 
 const textEncoded = new Uint8Array([0xac, 0x48, 0x65, 0x6c, 0x6c, 0x6f, 0x20, 0x57, 0x6f, 0x72, 0x6c, 0x64, 0x21]);
 

benchmarks/utf8-encode.cjs

@@ -0,0 +1,66 @@
+const { Benchmark } = require('benchmark');
+
+const textEncoder = new TextEncoder();
+
+const utf8EncodeJs = (str) => {
+  let strLength = str.length;
+  let output = [];
+  let offset = 0;
+  let pos = 0;
+  while (pos < strLength) {
+    let value = str.charCodeAt(pos++);
+
+    if ((value & 0xffffff80) === 0) {
+      // 1-byte
+      output[offset++] = value;
+      continue;
+    } else if ((value & 0xfffff800) === 0) {
+      // 2-bytes
+      output[offset++] = ((value >> 6) & 0x1f) | 0xc0;
+    } else {
+      // handle surrogate pair
+      if (value >= 0xd800 && value <= 0xdbff) {
+        // high surrogate
+        if (pos < strLength) {
+          let extra = str.charCodeAt(pos);
+          if ((extra & 0xfc00) === 0xdc00) {
+            ++pos;
+            value = ((value & 0x3ff) << 10) + (extra & 0x3ff) + 0x10000;
+          }
+        }
+      }
+
+      if ((value & 0xffff0000) === 0) {
+        // 3-byte
+        output[offset++] = ((value >> 12) & 0x0f) | 0xe0;
+        output[offset++] = ((value >> 6) & 0x3f) | 0x80;
+      } else {
+        // 4-byte
+        output[offset++] = ((value >> 18) & 0x07) | 0xf0;
+        output[offset++] = ((value >> 12) & 0x3f) | 0x80;
+        output[offset++] = ((value >> 6) & 0x3f) | 0x80;
+      }
+    }
+
+    output[offset++] = (value & 0x3f) | 0x80;
+  }
+
+  return new Uint8Array(output);
+}
+
+const text = '안녕, 세상!';
+
+console.log(textEncoder.encode(text));
+console.log(utf8EncodeJs(text));
+
+new Benchmark.Suite()
+.add('TextEncoder', () => {
+  textEncoder.encode(text);
+})
+.add('utf8EncodeJs', () => {
+  utf8EncodeJs(text);
+})
+.on('cycle', event => {
+  console.log(event.target.toString());
+})
+.run();

packages/msgpack/.gitignore

@@ -1 +1,5 @@
+.merlin
+.bsb.lock
+*.bs.js
+
 /lib/

packages/msgpack/bsconfig.json

@@ -0,0 +1,21 @@
+{
+  "name": "@urlpack/msgpack",
+  "sources": [
+    {
+      "dir": "src",
+      "subdirs": true
+    }
+  ],
+  "package-specs": [
+    {
+      "module": "es6",
+      "in-source": true
+    }
+  ],
+  "suffix": ".bs.js",
+  "bs-dependencies": [
+    "rescript-msgpack"
+  ],
+  "bs-dev-dependencies": [
+  ]
+}

packages/msgpack/package.json

@@ -26,20 +26,25 @@
     "main": "./lib/index.mjs"
   },
   "files": [
-    "src",
-    "lib"
+    "src/**/*.ts",
+    "src/*.res",
+    "lib/*.mjs",
+    "lib/*.cjs",
+    "lib/*.d.ts",
+    "lib/*.map"
   ],
   "scripts": {
     "prepack": "yarn build",
-    "build": "nanobundle build --standalone",
+    "build": "rescript build && nanobundle build --standalone",
     "test": "uvu -r tsm",
     "test:watch": "yarn test || true && watchlist src tests -- yarn test"
   },
   "devDependencies": {
-    "@rescript/std": "^9.1.3",
     "concurrently": "^6.5.1",
     "esbuild": "^0.14.9",
     "nanobundle": "^0.0.21",
+    "rescript": "^9.1.4",
+    "rescript-msgpack": "workspace:^0.1.0",
     "tsm": "^2.2.1",
     "typescript": "^4.5.4",
     "uvu": "^0.5.2",