jsdomain-parser 1.0.0

package/package.json

@@ -0,0 +1,33 @@
+{
+  "name": "jsdomain-parser",
+  "version": "1.0.0",
+  "description": "Parse URLs to extract TLDs, domain, protocols and more",
+  "main": "index.js",
+  "type": "module",
+  "scripts": {
+    "build": "rollup -c"
+  },
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/ToolsOverflow/jsdomain-parser"
+  },
+  "keywords": [
+    "url",
+    "parser",
+    "domain",
+    "parser",
+    "tld",
+    "extractor",
+    "extract",
+    "tld"
+  ],
+  "author": "ToolsOverflow",
+  "license": "MIT",
+  "devDependencies": {
+    "@rollup/plugin-json": "^6.1.0",
+    "rollup": "^2.79.2",
+    "rollup-plugin-commonjs": "^10.1.0",
+    "rollup-plugin-node-resolve": "^5.2.0",
+    "rollup-plugin-terser": "^7.0.2"
+  }
+}

package/rollup.config.js

@@ -0,0 +1,40 @@
+import resolve from "rollup-plugin-node-resolve";
+import commonjs from "rollup-plugin-commonjs";
+import { terser } from "rollup-plugin-terser";
+import json from "@rollup/plugin-json";
+
+export default {
+  input: "src/index.js",
+  output: [
+    {
+      file: "dist/jsdomain-parser.cjs.js",
+      format: "cjs",
+    },
+    {
+      file: "dist/jsdomain-parser.cjs.min.js",
+      format: "cjs",
+      plugins: [terser()],
+    },
+    {
+      file: "dist/jsdomain-parser.esm.js",
+      format: "esm",
+    },
+    {
+      file: "dist/jsdomain-parser.esm.min.js",
+      format: "esm",
+      plugins: [terser()],
+    },
+    {
+      file: "dist/jsdomain-parser.min.js",
+      format: "iife",
+      name: "DomainParser",
+      plugins: [terser()],
+    },
+    {
+      file: "dist/jsdomain-parser.js",
+      format: "iife",
+      name: "DomainParser",
+    },
+  ],
+  plugins: [resolve(), commonjs(), json()],
+};

package/src/index.js

@@ -0,0 +1,46 @@
+import { parseTld } from "./parseTld.js";
+
+function parse(url, options = {}) {
+  try {
+    if (!url.startsWith("http")) url = `http://${url}`;
+    const urlObject = new URL(url);
+
+    const tldData = parseTld(urlObject.hostname, options);
+
+    let domain = urlObject.hostname;
+    const hostnameParts = urlObject.hostname.split(".");
+    for (let i = hostnameParts.length - 1; i >= 0; i--) {
+      const extended = hostnameParts.slice(i);
+
+      if (extended.join(".") === tldData.name) {
+        domain = hostnameParts[i - 1] + "." + extended;
+        break;
+      }
+    }
+
+    const query = {};
+    for (const [key, value] of urlObject.searchParams.entries()) {
+      query[key] = value;
+    }
+
+    return {
+      tld: tldData,
+      url: {
+        domain: domain,
+        origin: urlObject.origin,
+        protocol: urlObject.protocol,
+        host: urlObject.host,
+        hostname: urlObject.hostname,
+        port: urlObject.port,
+        pathname: urlObject.pathname,
+        search: urlObject.search,
+        hash: urlObject.hash,
+        query,
+      },
+    };
+  } catch (e) {
+    throw new Error(`Invalid URL: ${e}`);
+  }
+}
+
+export { parse, parseTld };

package/src/parseTld.js

@@ -0,0 +1,56 @@
+import tlds from "./tlds.json";
+
+const parseTld = (hostname, options = {}) => {
+  const {
+    allowUnknown = false,
+    allowPrivate = true,
+    extendedTlds = [],
+  } = options;
+
+  if (!Array.isArray(extendedTlds)) {
+    throw new Error("customTlds must be an array");
+  }
+
+  const parts = hostname.split(".");
+
+  const customTlds = extendedTlds.map((tld) => [tld, tld.split(".").length]);
+  const icann = { ...tlds.icann, ...Object.fromEntries(customTlds) };
+  const privateTlds = tlds.private;
+
+  let detected = [];
+
+  for (let p = parts.length - 1; p >= 0; p--) {
+    const extended = parts.slice(p);
+    const prevPart = parts[p - 1];
+
+    if (icann[extended.join(".")] && prevPart !== undefined) {
+      detected = extended;
+    }
+
+    if (allowPrivate) {
+      if (privateTlds[extended.join(".")] && prevPart !== undefined) {
+        detected = extended;
+      }
+    }
+  }
+
+  if (allowUnknown) {
+    if (detected.length == 0) {
+      detected = parts.slice(-1);
+    }
+  }
+
+  if (detected.length == 0) {
+    throw new Error(
+      "Could not detect TLD. You can set allowUnknown to true for allowing unknown TLDs."
+    );
+  }
+
+  return {
+    name: detected.join("."),
+    length: detected.length,
+    parts: detected,
+  };
+};
+
+export { parseTld };