From 514c9c4e311e08ea43302f36568c2837da88cd04 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Mon, 10 Oct 2022 19:08:41 +0100
Subject: [PATCH 01/18] feat: Add batteries included methods for Node

---
 package-lock.json | 54 ++++++++++++++++++++++++++++++++-------
 package.json      |  6 +++--
 src/node.ts       | 65 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 114 insertions(+), 11 deletions(-)
 create mode 100644 src/node.ts

diff --git a/package-lock.json b/package-lock.json
index 476edae4be..3728ee3ad6 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -13,9 +13,11 @@
         "dom-serializer": "^2.0.0",
         "domhandler": "^5.0.3",
         "domutils": "^3.0.1",
+        "encoding-sniffer": "^0.0.0",
         "htmlparser2": "^8.0.1",
         "parse5": "^7.1.2",
-        "parse5-htmlparser2-tree-adapter": "^7.0.0"
+        "parse5-htmlparser2-tree-adapter": "^7.0.0",
+        "parse5-parser-stream": "^7.1.2"
       },
       "devDependencies": {
         "@imgix/js-core": "^3.7.0",
@@ -2671,6 +2673,18 @@
       "integrity": "sha512-L18DaJsXSUk2+42pv8mLs5jJT2hqFkFE4j21wOmgbUqsZ2hL72NsUU785g9RXgo3s0ZNgVl42TiHp3ZtOv/Vyg==",
       "dev": true
     },
+    "node_modules/encoding-sniffer": {
+      "version": "0.0.0",
+      "resolved": "https://registry.npmjs.org/encoding-sniffer/-/encoding-sniffer-0.0.0.tgz",
+      "integrity": "sha512-5JV4vyWTXIPhI9mnBrzBdwQ0R2ConTevpHznpSYzmEFG44rkMygNjwtixTDQEBbL43u/UwbolEZgv/N7RdYueg==",
+      "dependencies": {
+        "iconv-lite": "^0.6.3",
+        "whatwg-encoding": "^2.0.0"
+      },
+      "funding": {
+        "url": "https://github.com/fb55/encoding-sniffer?sponsor=1"
+      }
+    },
     "node_modules/entities": {
       "version": "4.4.0",
       "resolved": "https://registry.npmjs.org/entities/-/entities-4.4.0.tgz",
@@ -3620,7 +3634,6 @@
       "version": "0.6.3",
       "resolved": "https://registry.npmjs.org/iconv-lite/-/iconv-lite-0.6.3.tgz",
       "integrity": "sha512-4fCk79wshMdzMp2rH06qWrJE4iolqLhCUH+OiuIgU++RB0+94NlDL81atO7GX55uUKueo0txHNtvEyI6D7WdMw==",
-      "dev": true,
       "dependencies": {
         "safer-buffer": ">= 2.1.2 < 3.0.0"
       },
@@ -5889,6 +5902,17 @@
         "url": "https://github.com/inikulin/parse5?sponsor=1"
       }
     },
+    "node_modules/parse5-parser-stream": {
+      "version": "7.1.2",
+      "resolved": "https://registry.npmjs.org/parse5-parser-stream/-/parse5-parser-stream-7.1.2.tgz",
+      "integrity": "sha512-JyeQc9iwFLn5TbvvqACIF/VXG6abODeB3Fwmv/TGdLk2LfbWkaySGY72at4+Ty7EkPZj854u4CrICqNk2qIbow==",
+      "dependencies": {
+        "parse5": "^7.0.0"
+      },
+      "funding": {
+        "url": "https://github.com/inikulin/parse5?sponsor=1"
+      }
+    },
     "node_modules/path-exists": {
       "version": "4.0.0",
       "resolved": "https://registry.npmjs.org/path-exists/-/path-exists-4.0.0.tgz",
@@ -6289,8 +6313,7 @@
     "node_modules/safer-buffer": {
       "version": "2.1.2",
       "resolved": "https://registry.npmjs.org/safer-buffer/-/safer-buffer-2.1.2.tgz",
-      "integrity": "sha512-YZo3K82SD7Riyi0E1EQPojLz7kpepnSQI9IyPbHHg1XXXevb5dJI7tpyN2ADxGcQbHG7vcyRHk0cbwqcQriUtg==",
-      "dev": true
+      "integrity": "sha512-YZo3K82SD7Riyi0E1EQPojLz7kpepnSQI9IyPbHHg1XXXevb5dJI7tpyN2ADxGcQbHG7vcyRHk0cbwqcQriUtg=="
     },
     "node_modules/saxes": {
       "version": "6.0.0",
@@ -7095,7 +7118,6 @@
       "version": "2.0.0",
       "resolved": "https://registry.npmjs.org/whatwg-encoding/-/whatwg-encoding-2.0.0.tgz",
       "integrity": "sha512-p41ogyeMUrw3jWclHWTQg1k05DSVXPLcVxRTYsXUk+ZooOCZLcoYgPZ/HL/D/N+uQPOtcp1me1WhBEaX02mhWg==",
-      "dev": true,
       "dependencies": {
         "iconv-lite": "0.6.3"
       },
@@ -9368,6 +9390,15 @@
       "integrity": "sha512-L18DaJsXSUk2+42pv8mLs5jJT2hqFkFE4j21wOmgbUqsZ2hL72NsUU785g9RXgo3s0ZNgVl42TiHp3ZtOv/Vyg==",
       "dev": true
     },
+    "encoding-sniffer": {
+      "version": "0.0.0",
+      "resolved": "https://registry.npmjs.org/encoding-sniffer/-/encoding-sniffer-0.0.0.tgz",
+      "integrity": "sha512-5JV4vyWTXIPhI9mnBrzBdwQ0R2ConTevpHznpSYzmEFG44rkMygNjwtixTDQEBbL43u/UwbolEZgv/N7RdYueg==",
+      "requires": {
+        "iconv-lite": "^0.6.3",
+        "whatwg-encoding": "^2.0.0"
+      }
+    },
     "entities": {
       "version": "4.4.0",
       "resolved": "https://registry.npmjs.org/entities/-/entities-4.4.0.tgz",
@@ -10049,7 +10080,6 @@
       "version": "0.6.3",
       "resolved": "https://registry.npmjs.org/iconv-lite/-/iconv-lite-0.6.3.tgz",
       "integrity": "sha512-4fCk79wshMdzMp2rH06qWrJE4iolqLhCUH+OiuIgU++RB0+94NlDL81atO7GX55uUKueo0txHNtvEyI6D7WdMw==",
-      "dev": true,
       "requires": {
         "safer-buffer": ">= 2.1.2 < 3.0.0"
       }
@@ -11652,6 +11682,14 @@
         "parse5": "^7.0.0"
       }
     },
+    "parse5-parser-stream": {
+      "version": "7.1.2",
+      "resolved": "https://registry.npmjs.org/parse5-parser-stream/-/parse5-parser-stream-7.1.2.tgz",
+      "integrity": "sha512-JyeQc9iwFLn5TbvvqACIF/VXG6abODeB3Fwmv/TGdLk2LfbWkaySGY72at4+Ty7EkPZj854u4CrICqNk2qIbow==",
+      "requires": {
+        "parse5": "^7.0.0"
+      }
+    },
     "path-exists": {
       "version": "4.0.0",
       "resolved": "https://registry.npmjs.org/path-exists/-/path-exists-4.0.0.tgz",
@@ -11922,8 +11960,7 @@
     "safer-buffer": {
       "version": "2.1.2",
       "resolved": "https://registry.npmjs.org/safer-buffer/-/safer-buffer-2.1.2.tgz",
-      "integrity": "sha512-YZo3K82SD7Riyi0E1EQPojLz7kpepnSQI9IyPbHHg1XXXevb5dJI7tpyN2ADxGcQbHG7vcyRHk0cbwqcQriUtg==",
-      "dev": true
+      "integrity": "sha512-YZo3K82SD7Riyi0E1EQPojLz7kpepnSQI9IyPbHHg1XXXevb5dJI7tpyN2ADxGcQbHG7vcyRHk0cbwqcQriUtg=="
     },
     "saxes": {
       "version": "6.0.0",
@@ -12504,7 +12541,6 @@
       "version": "2.0.0",
       "resolved": "https://registry.npmjs.org/whatwg-encoding/-/whatwg-encoding-2.0.0.tgz",
       "integrity": "sha512-p41ogyeMUrw3jWclHWTQg1k05DSVXPLcVxRTYsXUk+ZooOCZLcoYgPZ/HL/D/N+uQPOtcp1me1WhBEaX02mhWg==",
-      "dev": true,
       "requires": {
         "iconv-lite": "0.6.3"
       }
diff --git a/package.json b/package.json
index c9768fb3ae..bb6a92b23b 100644
--- a/package.json
+++ b/package.json
@@ -52,9 +52,11 @@
     "dom-serializer": "^2.0.0",
     "domhandler": "^5.0.3",
     "domutils": "^3.0.1",
+    "encoding-sniffer": "^0.0.0",
     "htmlparser2": "^8.0.1",
     "parse5": "^7.1.2",
-    "parse5-htmlparser2-tree-adapter": "^7.0.0"
+    "parse5-htmlparser2-tree-adapter": "^7.0.0",
+    "parse5-parser-stream": "^7.1.2"
   },
   "devDependencies": {
     "@imgix/js-core": "^3.7.0",
@@ -112,7 +114,7 @@
   "lint-staged": {
     "*.js": [
       "prettier --write",
-      "npm run test:lint -- --fix"
+      "npm run lint:es -- --fix"
     ],
     "*.{json,md,ts,yml}": [
       "prettier --write"
diff --git a/src/node.ts b/src/node.ts
new file mode 100644
index 0000000000..1c5e44977a
--- /dev/null
+++ b/src/node.ts
@@ -0,0 +1,65 @@
+export * from './index.js';
+
+import type { CheerioAPI, CheerioOptions } from './index.js';
+import { load } from './index.js';
+import { flatten as flattenOptions } from './options.js';
+import { adapter as htmlparser2Adapter } from 'parse5-htmlparser2-tree-adapter';
+
+// eslint-disable-next-line node/file-extension-in-import
+import { WritableStream as Htmlparser2Stream } from 'htmlparser2/lib/WritableStream';
+import DomHandler from 'domhandler';
+import { ParserStream as Parse5Stream } from 'parse5-parser-stream';
+import { DecodeStream } from 'encoding-sniffer';
+import { type Writable, finished } from 'node:stream';
+
+export function stream(
+  options: CheerioOptions,
+  cb: (err: Error | null | undefined, $: CheerioAPI) => void
+): Writable {
+  const opts = flattenOptions(options);
+
+  if (opts && (opts.xmlMode || opts._useHtmlParser2)) {
+    const handler: DomHandler = new DomHandler(
+      (err) => cb(err, load(handler.root)),
+      opts
+    );
+
+    return new Htmlparser2Stream(handler, opts);
+  }
+
+  const stream = new Parse5Stream({ ...opts, treeAdapter: htmlparser2Adapter });
+
+  finished(stream, (err) => cb(err, load(stream.document)));
+
+  return stream;
+}
+
+// Get a document from a URL
+export async function request(
+  url: string | URL,
+  options: CheerioOptions
+): Promise<CheerioAPI> {
+  const opts = flattenOptions(options);
+  const str = stream(opts, (err, $) => {
+    if (err) {
+      throw err;
+    }
+  });
+
+  const res = await fetch(url, {
+    method: 'GET',
+    headers: {
+      'user-agent':
+        'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36',
+    },
+  });
+
+  if (!res.ok) {
+    throw new Error(`${res.status} ${res.statusText}`);
+  }
+
+  str.write(res.body);
+  str.end();
+
+  return stream;
+}

From ff4561bef9d7efd2fc867c33540121c1fa4ed778 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Fri, 14 Oct 2022 10:21:33 +0100
Subject: [PATCH 02/18] Update node.ts

---
 src/node.ts | 62 ++++++++++++++++++++++++++++++++++++-----------------
 1 file changed, 42 insertions(+), 20 deletions(-)

diff --git a/src/node.ts b/src/node.ts
index 1c5e44977a..9f6346a324 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -10,9 +10,10 @@ import { WritableStream as Htmlparser2Stream } from 'htmlparser2/lib/WritableStr
 import DomHandler from 'domhandler';
 import { ParserStream as Parse5Stream } from 'parse5-parser-stream';
 import { DecodeStream } from 'encoding-sniffer';
+import * as undici from 'undici';
 import { type Writable, finished } from 'node:stream';
 
-export function stream(
+export function stringStream(
   options: CheerioOptions,
   cb: (err: Error | null | undefined, $: CheerioAPI) => void
 ): Writable {
@@ -34,32 +35,53 @@ export function stream(
   return stream;
 }
 
+export function decodeStream(
+  options: CheerioOptions,
+  cb: (err: Error | null | undefined, $: CheerioAPI) => void
+): Writable {
+  // TODO: Set the encoding to UTF8 for XML mode
+  const decodeStream = new DecodeStream();
+  const loadStream = stringStream(options, cb);
+
+  decodeStream.pipe(loadStream);
+
+  return decodeStream;
+}
+
 // Get a document from a URL
 export async function request(
+  // eslint-disable-next-line node/no-unsupported-features/node-builtins
   url: string | URL,
   options: CheerioOptions
 ): Promise<CheerioAPI> {
-  const opts = flattenOptions(options);
-  const str = stream(opts, (err, $) => {
-    if (err) {
-      throw err;
-    }
-  });
+  let undiciStream: Promise<undici.Dispatcher.StreamData> | undefined;
 
-  const res = await fetch(url, {
-    method: 'GET',
-    headers: {
-      'user-agent':
-        'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36',
-    },
-  });
+  const promise = new Promise<CheerioAPI>((resolve, reject) => {
+    undiciStream = undici.stream(
+      url,
+      {
+        method: 'GET',
+        headers: {
+          'user-agent':
+            'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36',
+        },
+      },
+      (data) => {
+        // TODO Add support for handling status codes.
+        if (!data.statusCode) {
+          throw new Error(`Received ${data.statusCode}`);
+        }
 
-  if (!res.ok) {
-    throw new Error(`${res.status} ${res.statusText}`);
-  }
+        // TODO: Forward the charset from the header to the decodeStream.
+        return decodeStream(options, (err, $) =>
+          err ? reject(err) : resolve($)
+        );
+      }
+    );
+  });
 
-  str.write(res.body);
-  str.end();
+  // Let's make sure the request is completed before returning the promise.
+  await undiciStream;
 
-  return stream;
+  return promise;
 }

From 2779ed3d6689eabc50428d5df5aaf723eb556f0a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Wed, 2 Nov 2022 14:03:27 +0100
Subject: [PATCH 03/18] Start writing tests

---
 src/node.spec.ts | 87 ++++++++++++++++++++++++++++++++++++++++++++++++
 src/node.ts      | 38 +++++++++++++--------
 2 files changed, 112 insertions(+), 13 deletions(-)
 create mode 100644 src/node.spec.ts

diff --git a/src/node.spec.ts b/src/node.spec.ts
new file mode 100644
index 0000000000..65d64b84cd
--- /dev/null
+++ b/src/node.spec.ts
@@ -0,0 +1,87 @@
+/* eslint-disable jest/no-done-callback */
+import * as cheerio from './node.js';
+import { Writable } from 'node:stream';
+
+function noop() {
+  // Ignore
+}
+
+const TEST_HTML = '<h1>Hello World</h1>';
+const TEST_HTML_UTF16 = Buffer.from(TEST_HTML, 'utf16le');
+const TEST_HTML_UTF16_BOM = Buffer.from([
+  // UTF16-LE BOM
+  0xff,
+  0xfe,
+  ...Array.from(TEST_HTML_UTF16),
+]);
+
+describe('Node API', () => {
+  describe('stringStream', () => {
+    it('should use parse5 by default', (cb) => {
+      const stream = cheerio.stringStream({}, (err, $) => {
+        expect(err).toBeUndefined();
+
+        expect($.html()).toBe(
+          `<html><head></head><body>${TEST_HTML}</body></html>`
+        );
+
+        cb();
+      });
+      expect(stream).toBeInstanceOf(Writable);
+
+      stream.end(TEST_HTML);
+    });
+
+    it('should error from parse5 on buffer', () => {
+      const stream = cheerio.stringStream({}, noop);
+      expect(stream).toBeInstanceOf(Writable);
+
+      expect(() => stream.write(Buffer.from(TEST_HTML))).toThrow(
+        'Parser can work only with string streams.'
+      );
+    });
+
+    it('should use htmlparser2 for XML', (cb) => {
+      const stream = cheerio.stringStream({ xmlMode: true }, (err, $) => {
+        expect(err).toBeNull();
+
+        expect($.html()).toBe(TEST_HTML);
+
+        cb();
+      });
+      expect(stream).toBeInstanceOf(Writable);
+
+      stream.end(TEST_HTML);
+    });
+  });
+
+  describe('decodeStream', () => {
+    it('should use parse5 by default', (cb) => {
+      const stream = cheerio.decodeStream({}, (err, $) => {
+        expect(err).toBeUndefined();
+
+        expect($.html()).toBe(
+          `<html><head></head><body>${TEST_HTML}</body></html>`
+        );
+
+        cb();
+      });
+      expect(stream).toBeInstanceOf(Writable);
+
+      stream.end(TEST_HTML_UTF16_BOM);
+    });
+
+    it('should use htmlparser2 for XML', (cb) => {
+      const stream = cheerio.decodeStream({ xmlMode: true }, (err, $) => {
+        expect(err).toBeNull();
+
+        expect($.html()).toBe(TEST_HTML);
+
+        cb();
+      });
+      expect(stream).toBeInstanceOf(Writable);
+
+      stream.end(TEST_HTML_UTF16_BOM);
+    });
+  });
+});
diff --git a/src/node.ts b/src/node.ts
index 9f6346a324..3f27ef987f 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -2,46 +2,58 @@ export * from './index.js';
 
 import type { CheerioAPI, CheerioOptions } from './index.js';
 import { load } from './index.js';
-import { flatten as flattenOptions } from './options.js';
+import { flatten as flattenOptions, InternalOptions } from './options.js';
 import { adapter as htmlparser2Adapter } from 'parse5-htmlparser2-tree-adapter';
 
 // eslint-disable-next-line node/file-extension-in-import
 import { WritableStream as Htmlparser2Stream } from 'htmlparser2/lib/WritableStream';
 import DomHandler from 'domhandler';
 import { ParserStream as Parse5Stream } from 'parse5-parser-stream';
-import { DecodeStream } from 'encoding-sniffer';
+import { DecodeStream, type SnifferOptions } from 'encoding-sniffer';
 import * as undici from 'undici';
 import { type Writable, finished } from 'node:stream';
 
-export function stringStream(
-  options: CheerioOptions,
+function _stringStream(
+  options: InternalOptions | undefined,
   cb: (err: Error | null | undefined, $: CheerioAPI) => void
 ): Writable {
-  const opts = flattenOptions(options);
-
-  if (opts && (opts.xmlMode || opts._useHtmlParser2)) {
+  if (options && (options.xmlMode || options._useHtmlParser2)) {
     const handler: DomHandler = new DomHandler(
       (err) => cb(err, load(handler.root)),
-      opts
+      options
     );
 
-    return new Htmlparser2Stream(handler, opts);
+    return new Htmlparser2Stream(handler, options);
   }
 
-  const stream = new Parse5Stream({ ...opts, treeAdapter: htmlparser2Adapter });
+  const stream = new Parse5Stream({
+    ...options,
+    treeAdapter: htmlparser2Adapter,
+  });
 
   finished(stream, (err) => cb(err, load(stream.document)));
 
   return stream;
 }
 
+export function stringStream(
+  options: CheerioOptions,
+  cb: (err: Error | null | undefined, $: CheerioAPI) => void
+): Writable {
+  return _stringStream(flattenOptions(options), cb);
+}
+
 export function decodeStream(
   options: CheerioOptions,
   cb: (err: Error | null | undefined, $: CheerioAPI) => void
 ): Writable {
-  // TODO: Set the encoding to UTF8 for XML mode
-  const decodeStream = new DecodeStream();
-  const loadStream = stringStream(options, cb);
+  const opts = flattenOptions(options);
+  const snifferOpts: SnifferOptions = {
+    // Set the encoding to UTF8 for XML mode
+    defaultEncoding: opts?.xmlMode ? 'utf8' : 'windows-1252',
+  };
+  const decodeStream = new DecodeStream(snifferOpts);
+  const loadStream = _stringStream(opts, cb);
 
   decodeStream.pipe(loadStream);
 

From 5c4bc880646dd4cbf6a2a554d262b03a7264fe9f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Fri, 4 Nov 2022 15:36:49 +0000
Subject: [PATCH 04/18] Bump encoding-sniffer

---
 package-lock.json | 14 +++++++-------
 package.json      |  2 +-
 2 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/package-lock.json b/package-lock.json
index 3728ee3ad6..da6d2f1063 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -13,7 +13,7 @@
         "dom-serializer": "^2.0.0",
         "domhandler": "^5.0.3",
         "domutils": "^3.0.1",
-        "encoding-sniffer": "^0.0.0",
+        "encoding-sniffer": "^0.0.1",
         "htmlparser2": "^8.0.1",
         "parse5": "^7.1.2",
         "parse5-htmlparser2-tree-adapter": "^7.0.0",
@@ -2674,9 +2674,9 @@
       "dev": true
     },
     "node_modules/encoding-sniffer": {
-      "version": "0.0.0",
-      "resolved": "https://registry.npmjs.org/encoding-sniffer/-/encoding-sniffer-0.0.0.tgz",
-      "integrity": "sha512-5JV4vyWTXIPhI9mnBrzBdwQ0R2ConTevpHznpSYzmEFG44rkMygNjwtixTDQEBbL43u/UwbolEZgv/N7RdYueg==",
+      "version": "0.0.1",
+      "resolved": "https://registry.npmjs.org/encoding-sniffer/-/encoding-sniffer-0.0.1.tgz",
+      "integrity": "sha512-8BhuvYsCaMxDgyQVZgnoXyoSD/jS8Ipj2Fxti+lhA0fAT14aU7Y7ofwc8npIz0x49ikH188Mdi6K7MlUbZOqmg==",
       "dependencies": {
         "iconv-lite": "^0.6.3",
         "whatwg-encoding": "^2.0.0"
@@ -9391,9 +9391,9 @@
       "dev": true
     },
     "encoding-sniffer": {
-      "version": "0.0.0",
-      "resolved": "https://registry.npmjs.org/encoding-sniffer/-/encoding-sniffer-0.0.0.tgz",
-      "integrity": "sha512-5JV4vyWTXIPhI9mnBrzBdwQ0R2ConTevpHznpSYzmEFG44rkMygNjwtixTDQEBbL43u/UwbolEZgv/N7RdYueg==",
+      "version": "0.0.1",
+      "resolved": "https://registry.npmjs.org/encoding-sniffer/-/encoding-sniffer-0.0.1.tgz",
+      "integrity": "sha512-8BhuvYsCaMxDgyQVZgnoXyoSD/jS8Ipj2Fxti+lhA0fAT14aU7Y7ofwc8npIz0x49ikH188Mdi6K7MlUbZOqmg==",
       "requires": {
         "iconv-lite": "^0.6.3",
         "whatwg-encoding": "^2.0.0"
diff --git a/package.json b/package.json
index bb6a92b23b..dcd303cfd6 100644
--- a/package.json
+++ b/package.json
@@ -52,7 +52,7 @@
     "dom-serializer": "^2.0.0",
     "domhandler": "^5.0.3",
     "domutils": "^3.0.1",
-    "encoding-sniffer": "^0.0.0",
+    "encoding-sniffer": "^0.0.1",
     "htmlparser2": "^8.0.1",
     "parse5": "^7.1.2",
     "parse5-htmlparser2-tree-adapter": "^7.0.0",

From 00499b42e21c7f59dd93a90b16cc4073f92024c6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Mon, 7 Nov 2022 16:45:29 +0000
Subject: [PATCH 05/18] Handle mime types, work on getting streams going

---
 package-lock.json | 20 +++++++++---
 package.json      |  4 ++-
 src/node.ts       | 80 +++++++++++++++++++++++++++++++----------------
 tsconfig.json     |  2 +-
 4 files changed, 73 insertions(+), 33 deletions(-)

diff --git a/package-lock.json b/package-lock.json
index da6d2f1063..1401d10a5f 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -17,7 +17,8 @@
         "htmlparser2": "^8.0.1",
         "parse5": "^7.1.2",
         "parse5-htmlparser2-tree-adapter": "^7.0.0",
-        "parse5-parser-stream": "^7.1.2"
+        "parse5-parser-stream": "^7.1.2",
+        "whatwg-mimetype": "^3.0.0"
       },
       "devDependencies": {
         "@imgix/js-core": "^3.7.0",
@@ -25,6 +26,7 @@
         "@types/benchmark": "^2.1.2",
         "@types/jest": "^29.2.3",
         "@types/node": "^18.11.9",
+        "@types/whatwg-mimetype": "^3.0.0",
         "@typescript-eslint/eslint-plugin": "^5.44.0",
         "@typescript-eslint/parser": "^5.44.0",
         "benchmark": "^2.1.4",
@@ -1479,6 +1481,12 @@
       "integrity": "sha512-PBjIUxZHOuj0R15/xuwJYjFi+KZdNFrehocChv4g5hu6aFroHue8m0lBP0POdK2nKzbw0cgV1mws8+V/JAcEkQ==",
       "dev": true
     },
+    "node_modules/@types/whatwg-mimetype": {
+      "version": "3.0.0",
+      "resolved": "https://registry.npmjs.org/@types/whatwg-mimetype/-/whatwg-mimetype-3.0.0.tgz",
+      "integrity": "sha512-xHFOhd41VpUR6Y0k8ZinlyFv5cyhC/r2zghJgWWN8oNxqNo45Nf0qCBInJsFeifLeoHcIF4voEfap4A2GYHWkw==",
+      "dev": true
+    },
     "node_modules/@types/yargs": {
       "version": "17.0.13",
       "resolved": "https://registry.npmjs.org/@types/yargs/-/yargs-17.0.13.tgz",
@@ -7129,7 +7137,6 @@
       "version": "3.0.0",
       "resolved": "https://registry.npmjs.org/whatwg-mimetype/-/whatwg-mimetype-3.0.0.tgz",
       "integrity": "sha512-nt+N2dzIutVRxARx1nghPKGv1xHikU7HKdfafKkLNLindmPU/ch3U31NOCGGA/dmPcmb1VlofO0vnKAcsm0o/Q==",
-      "dev": true,
       "engines": {
         "node": ">=12"
       }
@@ -8535,6 +8542,12 @@
       "integrity": "sha512-PBjIUxZHOuj0R15/xuwJYjFi+KZdNFrehocChv4g5hu6aFroHue8m0lBP0POdK2nKzbw0cgV1mws8+V/JAcEkQ==",
       "dev": true
     },
+    "@types/whatwg-mimetype": {
+      "version": "3.0.0",
+      "resolved": "https://registry.npmjs.org/@types/whatwg-mimetype/-/whatwg-mimetype-3.0.0.tgz",
+      "integrity": "sha512-xHFOhd41VpUR6Y0k8ZinlyFv5cyhC/r2zghJgWWN8oNxqNo45Nf0qCBInJsFeifLeoHcIF4voEfap4A2GYHWkw==",
+      "dev": true
+    },
     "@types/yargs": {
       "version": "17.0.13",
       "resolved": "https://registry.npmjs.org/@types/yargs/-/yargs-17.0.13.tgz",
@@ -12548,8 +12561,7 @@
     "whatwg-mimetype": {
       "version": "3.0.0",
       "resolved": "https://registry.npmjs.org/whatwg-mimetype/-/whatwg-mimetype-3.0.0.tgz",
-      "integrity": "sha512-nt+N2dzIutVRxARx1nghPKGv1xHikU7HKdfafKkLNLindmPU/ch3U31NOCGGA/dmPcmb1VlofO0vnKAcsm0o/Q==",
-      "dev": true
+      "integrity": "sha512-nt+N2dzIutVRxARx1nghPKGv1xHikU7HKdfafKkLNLindmPU/ch3U31NOCGGA/dmPcmb1VlofO0vnKAcsm0o/Q=="
     },
     "whatwg-url": {
       "version": "11.0.0",
diff --git a/package.json b/package.json
index dcd303cfd6..797ebfc993 100644
--- a/package.json
+++ b/package.json
@@ -56,7 +56,8 @@
     "htmlparser2": "^8.0.1",
     "parse5": "^7.1.2",
     "parse5-htmlparser2-tree-adapter": "^7.0.0",
-    "parse5-parser-stream": "^7.1.2"
+    "parse5-parser-stream": "^7.1.2",
+    "whatwg-mimetype": "^3.0.0"
   },
   "devDependencies": {
     "@imgix/js-core": "^3.7.0",
@@ -64,6 +65,7 @@
     "@types/benchmark": "^2.1.2",
     "@types/jest": "^29.2.3",
     "@types/node": "^18.11.9",
+    "@types/whatwg-mimetype": "^3.0.0",
     "@typescript-eslint/eslint-plugin": "^5.44.0",
     "@typescript-eslint/parser": "^5.44.0",
     "benchmark": "^2.1.4",
diff --git a/src/node.ts b/src/node.ts
index 3f27ef987f..9af7392a97 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -11,6 +11,7 @@ import DomHandler from 'domhandler';
 import { ParserStream as Parse5Stream } from 'parse5-parser-stream';
 import { DecodeStream, type SnifferOptions } from 'encoding-sniffer';
 import * as undici from 'undici';
+import MIMEType from 'whatwg-mimetype';
 import { type Writable, finished } from 'node:stream';
 
 function _stringStream(
@@ -43,16 +44,21 @@ export function stringStream(
   return _stringStream(flattenOptions(options), cb);
 }
 
+export interface DecodeStreamOptions extends CheerioOptions {
+  encoding?: SnifferOptions;
+}
+
 export function decodeStream(
-  options: CheerioOptions,
+  options: DecodeStreamOptions,
   cb: (err: Error | null | undefined, $: CheerioAPI) => void
 ): Writable {
-  const opts = flattenOptions(options);
-  const snifferOpts: SnifferOptions = {
-    // Set the encoding to UTF8 for XML mode
-    defaultEncoding: opts?.xmlMode ? 'utf8' : 'windows-1252',
-  };
-  const decodeStream = new DecodeStream(snifferOpts);
+  const { encoding = {}, ...cheerioOptions } = options;
+  const opts = flattenOptions(cheerioOptions);
+
+  // Set the encoding to UTF8 for XML mode
+  encoding.defaultEncoding ??= opts?.xmlMode ? 'utf8' : 'windows-1252';
+
+  const decodeStream = new DecodeStream(encoding);
   const loadStream = _stringStream(opts, cb);
 
   decodeStream.pipe(loadStream);
@@ -60,36 +66,56 @@ export function decodeStream(
   return decodeStream;
 }
 
+interface CheerioRequestOptions extends DecodeStreamOptions {
+  requestOptions?: Parameters<typeof undici.stream>[1];
+}
+
 // Get a document from a URL
 export async function request(
   // eslint-disable-next-line node/no-unsupported-features/node-builtins
   url: string | URL,
-  options: CheerioOptions
+  options: CheerioRequestOptions = {}
 ): Promise<CheerioAPI> {
+  const {
+    requestOptions = { method: 'GET' },
+    encoding = {},
+    ...cheerioOptions
+  } = options;
   let undiciStream: Promise<undici.Dispatcher.StreamData> | undefined;
 
+  requestOptions.method ??= 'GET';
+
   const promise = new Promise<CheerioAPI>((resolve, reject) => {
-    undiciStream = undici.stream(
-      url,
-      {
-        method: 'GET',
-        headers: {
-          'user-agent':
-            'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36',
-        },
-      },
-      (data) => {
-        // TODO Add support for handling status codes.
-        if (!data.statusCode) {
-          throw new Error(`Received ${data.statusCode}`);
-        }
-
-        // TODO: Forward the charset from the header to the decodeStream.
-        return decodeStream(options, (err, $) =>
-          err ? reject(err) : resolve($)
+    undiciStream = undici.stream(url, requestOptions, (res) => {
+      // TODO Add support for handling status codes, such as redirects.
+      if (!res.statusCode) {
+        throw new Error(`Received ${res.statusCode}`);
+      }
+
+      const contentType = res.headers['content-type'];
+      const mimeType = new MIMEType(contentType ?? 'text/html');
+
+      if (!mimeType.isHTML() && !mimeType.isXML()) {
+        throw new RangeError(
+          `The content-type "${contentType}" is neither HTML nor XML.`
         );
       }
-    );
+
+      // Forward the charset from the header to the decodeStream.
+      encoding.transportLayerEncodingLabel ??=
+        mimeType.parameters.get('charset');
+
+      const opts = {
+        ...flattenOptions(cheerioOptions),
+        encoding,
+        // Set XML mode based on the MIME type.
+        xmlMode: mimeType.isXML(),
+        // TODO: Set the baseURL based on the final URL.
+        baseURL: (res.context as any)?.url,
+      };
+
+      return decodeStream(opts, (err, $) => (err ? reject(err) : resolve($)));
+    });
   });
 
   // Let's make sure the request is completed before returning the promise.
diff --git a/tsconfig.json b/tsconfig.json
index 42a229c78b..3c601600f0 100644
--- a/tsconfig.json
+++ b/tsconfig.json
@@ -25,7 +25,7 @@
 
     /* Module Resolution Options */
     "esModuleInterop": true,
-    "moduleResolution": "node",
+    "moduleResolution": "node16",
     "resolveJsonModule": true
   },
   "include": ["src"],

From 944b201ad519be7f59a5eb976193d43e80467248 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Mon, 7 Nov 2022 23:16:58 +0000
Subject: [PATCH 06/18] Set the `baseURL`

---
 src/node.ts | 35 ++++++++++++++++++++++-------------
 1 file changed, 22 insertions(+), 13 deletions(-)

diff --git a/src/node.ts b/src/node.ts
index 9af7392a97..5cf1141279 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -55,7 +55,7 @@ export function decodeStream(
   const { encoding = {}, ...cheerioOptions } = options;
   const opts = flattenOptions(cheerioOptions);
 
-  // Set the encoding to UTF8 for XML mode
+  // Set the default encoding to UTF-8 for XML mode
   encoding.defaultEncoding ??= opts?.xmlMode ? 'utf8' : 'windows-1252';
 
   const decodeStream = new DecodeStream(encoding);
@@ -66,10 +66,19 @@ export function decodeStream(
   return decodeStream;
 }
 
-interface CheerioRequestOptions extends DecodeStreamOptions {
-  requestOptions?: Parameters<typeof undici.stream>[1];
+type UndiciStreamOptions = Parameters<typeof undici.stream>[1];
+
+export interface CheerioRequestOptions extends DecodeStreamOptions {
+  /** The options passed to `undici`'s `stream` method. */
+  requestOptions?: UndiciStreamOptions;
 }
 
+const defaultRequestOptions: UndiciStreamOptions = {
+  method: 'GET',
+  maxRedirections: 5,
+  throwOnError: true,
+};
+
 // Get a document from a URL
 export async function request(
   // eslint-disable-next-line node/no-unsupported-features/node-builtins
@@ -77,7 +86,7 @@ export async function request(
   options: CheerioRequestOptions = {}
 ): Promise<CheerioAPI> {
   const {
-    requestOptions = { method: 'GET' },
+    requestOptions = defaultRequestOptions,
     encoding = {},
     ...cheerioOptions
   } = options;
@@ -87,11 +96,6 @@ export async function request(
 
   const promise = new Promise<CheerioAPI>((resolve, reject) => {
     undiciStream = undici.stream(url, requestOptions, (res) => {
-      // TODO Add support for handling status codes, such as redirects.
-      if (!res.statusCode) {
-        throw new Error(`Received ${res.statusCode}`);
-      }
-
       const contentType = res.headers['content-type'];
       const mimeType = new MIMEType(contentType ?? 'text/html');
 
@@ -102,16 +106,21 @@ export async function request(
       }
 
       // Forward the charset from the header to the decodeStream.
-      encoding.transportLayerEncodingLabel ??=
-        mimeType.parameters.get('charset');
+      encoding.transportLayerEncodingLabel = mimeType.parameters.get('charset');
+
+      /*
+       * If we allow redirects, we will have entries in the history.
+       * The last entry will be the final URL.
+       */
+      const history = (res.context as any)?.history;
 
       const opts = {
         ...flattenOptions(cheerioOptions),
         encoding,
         // Set XML mode based on the MIME type.
         xmlMode: mimeType.isXML(),
-        // TODO: Set the baseURL based on the final URL.
-        baseURL: (res.context as any)?.url,
+        // Set the `baseURL` to the final URL.
+        baseURL: history ? history[history.length - 1] : url,
       };
 
       return decodeStream(opts, (err, $) => (err ? reject(err) : resolve($)));

From 8b4a3eca058d52972a80fb0c6d128d1e5b092c05 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Mon, 7 Nov 2022 23:23:47 +0000
Subject: [PATCH 07/18] Rename `request` to `fromURL`

---
 src/node.ts | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/node.ts b/src/node.ts
index 5cf1141279..507e9cbc1a 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -80,7 +80,7 @@ const defaultRequestOptions: UndiciStreamOptions = {
 };
 
 // Get a document from a URL
-export async function request(
+export async function fromURL(
   // eslint-disable-next-line node/no-unsupported-features/node-builtins
   url: string | URL,
   options: CheerioRequestOptions = {}

From 61ba1a4a561cf02c7c2c4c41e7ecef39309a405e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Tue, 8 Nov 2022 10:42:30 +0000
Subject: [PATCH 08/18] Add `fromURL` tests

---
 src/node.spec.ts | 143 ++++++++++++++++++++++++++++++++---------------
 1 file changed, 97 insertions(+), 46 deletions(-)

diff --git a/src/node.spec.ts b/src/node.spec.ts
index 65d64b84cd..b59fb35673 100644
--- a/src/node.spec.ts
+++ b/src/node.spec.ts
@@ -1,6 +1,7 @@
 /* eslint-disable jest/no-done-callback */
 import * as cheerio from './node.js';
 import { Writable } from 'node:stream';
+import { createServer } from 'node:http';
 
 function noop() {
   // Ignore
@@ -15,73 +16,123 @@ const TEST_HTML_UTF16_BOM = Buffer.from([
   ...Array.from(TEST_HTML_UTF16),
 ]);
 
-describe('Node API', () => {
-  describe('stringStream', () => {
-    it('should use parse5 by default', (cb) => {
-      const stream = cheerio.stringStream({}, (err, $) => {
-        expect(err).toBeUndefined();
+describe('stringStream', () => {
+  it('should use parse5 by default', (cb) => {
+    const stream = cheerio.stringStream({}, (err, $) => {
+      expect(err).toBeUndefined();
 
-        expect($.html()).toBe(
-          `<html><head></head><body>${TEST_HTML}</body></html>`
-        );
+      expect($.html()).toBe(
+        `<html><head></head><body>${TEST_HTML}</body></html>`
+      );
+
+      cb();
+    });
+    expect(stream).toBeInstanceOf(Writable);
 
-        cb();
-      });
-      expect(stream).toBeInstanceOf(Writable);
+    stream.end(TEST_HTML);
+  });
+
+  it('should error from parse5 on buffer', () => {
+    const stream = cheerio.stringStream({}, noop);
+    expect(stream).toBeInstanceOf(Writable);
+
+    expect(() => stream.write(Buffer.from(TEST_HTML))).toThrow(
+      'Parser can work only with string streams.'
+    );
+  });
 
-      stream.end(TEST_HTML);
+  it('should use htmlparser2 for XML', (cb) => {
+    const stream = cheerio.stringStream({ xmlMode: true }, (err, $) => {
+      expect(err).toBeNull();
+
+      expect($.html()).toBe(TEST_HTML);
+
+      cb();
     });
+    expect(stream).toBeInstanceOf(Writable);
+
+    stream.end(TEST_HTML);
+  });
+});
 
-    it('should error from parse5 on buffer', () => {
-      const stream = cheerio.stringStream({}, noop);
-      expect(stream).toBeInstanceOf(Writable);
+describe('decodeStream', () => {
+  it('should use parse5 by default', (cb) => {
+    const stream = cheerio.decodeStream({}, (err, $) => {
+      expect(err).toBeUndefined();
 
-      expect(() => stream.write(Buffer.from(TEST_HTML))).toThrow(
-        'Parser can work only with string streams.'
+      expect($.html()).toBe(
+        `<html><head></head><body>${TEST_HTML}</body></html>`
       );
+
+      cb();
     });
+    expect(stream).toBeInstanceOf(Writable);
 
-    it('should use htmlparser2 for XML', (cb) => {
-      const stream = cheerio.stringStream({ xmlMode: true }, (err, $) => {
-        expect(err).toBeNull();
+    stream.end(TEST_HTML_UTF16_BOM);
+  });
 
-        expect($.html()).toBe(TEST_HTML);
+  it('should use htmlparser2 for XML', (cb) => {
+    const stream = cheerio.decodeStream({ xmlMode: true }, (err, $) => {
+      expect(err).toBeNull();
 
-        cb();
-      });
-      expect(stream).toBeInstanceOf(Writable);
+      expect($.html()).toBe(TEST_HTML);
 
-      stream.end(TEST_HTML);
+      cb();
     });
-  });
+    expect(stream).toBeInstanceOf(Writable);
 
-  describe('decodeStream', () => {
-    it('should use parse5 by default', (cb) => {
-      const stream = cheerio.decodeStream({}, (err, $) => {
-        expect(err).toBeUndefined();
+    stream.end(TEST_HTML_UTF16_BOM);
+  });
+});
 
-        expect($.html()).toBe(
-          `<html><head></head><body>${TEST_HTML}</body></html>`
-        );
+function createTestServer(
+  contentType: string,
+  body: string | Buffer
+): Promise<number> {
+  return new Promise((resolve, reject) => {
+    const server = createServer((_req, res) => {
+      res.writeHead(200, { 'Content-Type': contentType });
+      res.end(body);
+    });
 
-        cb();
-      });
-      expect(stream).toBeInstanceOf(Writable);
+    server.listen(0, () => {
+      const address = server.address();
 
-      stream.end(TEST_HTML_UTF16_BOM);
+      if (typeof address === 'string' || address === null) {
+        reject(new Error('Failed to get port'));
+      } else {
+        resolve(address.port);
+      }
     });
+  });
+}
 
-    it('should use htmlparser2 for XML', (cb) => {
-      const stream = cheerio.decodeStream({ xmlMode: true }, (err, $) => {
-        expect(err).toBeNull();
+describe('fromURL', () => {
+  it('should fetch UTF-8 HTML', async () => {
+    const port = await createTestServer('text/html', TEST_HTML);
+    const $ = await cheerio.fromURL(`http://localhost:${port}`);
 
-        expect($.html()).toBe(TEST_HTML);
+    expect($.html()).toBe(
+      `<html><head></head><body>${TEST_HTML}</body></html>`
+    );
+  });
 
-        cb();
-      });
-      expect(stream).toBeInstanceOf(Writable);
+  it('should fetch UTF-16 HTML', async () => {
+    const port = await createTestServer(
+      'text/html; charset=utf-16le',
+      TEST_HTML_UTF16
+    );
+    const $ = await cheerio.fromURL(`http://localhost:${port}`);
 
-      stream.end(TEST_HTML_UTF16_BOM);
-    });
+    expect($.html()).toBe(
+      `<html><head></head><body>${TEST_HTML}</body></html>`
+    );
+  });
+
+  it('should parse XML based on Content-Type', async () => {
+    const port = await createTestServer('text/xml', TEST_HTML);
+    const $ = await cheerio.fromURL(`http://localhost:${port}`);
+
+    expect($.html()).toBe(TEST_HTML);
   });
 });

From 357cf706f7bcb78e7095a3a1d2514a93a13a5f4a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Wed, 16 Nov 2022 14:10:29 +0000
Subject: [PATCH 09/18] Add `accept` header

---
 src/node.ts | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/src/node.ts b/src/node.ts
index 507e9cbc1a..1765e8d3c8 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -75,8 +75,14 @@ export interface CheerioRequestOptions extends DecodeStreamOptions {
 
 const defaultRequestOptions: UndiciStreamOptions = {
   method: 'GET',
+  // Allow redirects by default
   maxRedirections: 5,
+  // NOTE: `throwOnError` currently doesn't work https://github.com/nodejs/undici/issues/1753
   throwOnError: true,
+  // Set an Accept header
+  headers: {
+    accept: 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
+  },
 };
 
 // Get a document from a URL
@@ -92,7 +98,8 @@ export async function fromURL(
   } = options;
   let undiciStream: Promise<undici.Dispatcher.StreamData> | undefined;
 
-  requestOptions.method ??= 'GET';
+  // Add headers if none were supplied.
+  requestOptions.headers ??= defaultRequestOptions.headers;
 
   const promise = new Promise<CheerioAPI>((resolve, reject) => {
     undiciStream = undici.stream(url, requestOptions, (res) => {
@@ -115,7 +122,7 @@ export async function fromURL(
       const history = (res.context as any)?.history;
 
       const opts = {
-        ...flattenOptions(cheerioOptions),
+        ...cheerioOptions,
         encoding,
         // Set XML mode based on the MIME type.
         xmlMode: mimeType.isXML(),

From e136833617f6dcc69bb4416c01654906923d6359 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Mon, 21 Nov 2022 18:17:58 +0000
Subject: [PATCH 10/18] Add `loadBuffer`

---
 src/node.spec.ts | 18 ++++++++++++++++++
 src/node.ts      | 27 ++++++++++++++++++++++++++-
 2 files changed, 44 insertions(+), 1 deletion(-)

diff --git a/src/node.spec.ts b/src/node.spec.ts
index b59fb35673..8e65553a01 100644
--- a/src/node.spec.ts
+++ b/src/node.spec.ts
@@ -16,6 +16,24 @@ const TEST_HTML_UTF16_BOM = Buffer.from([
   ...Array.from(TEST_HTML_UTF16),
 ]);
 
+describe('loadBuffer', () => {
+  it('should parse UTF-8 HTML', () => {
+    const $ = cheerio.loadBuffer(Buffer.from(TEST_HTML));
+
+    expect($.html()).toBe(
+      `<html><head></head><body>${TEST_HTML}</body></html>`
+    );
+  });
+
+  it('should parse UTF-16 HTML', () => {
+    const $ = cheerio.loadBuffer(TEST_HTML_UTF16_BOM);
+
+    expect($.html()).toBe(
+      `<html><head></head><body>${TEST_HTML}</body></html>`
+    );
+  });
+});
+
 describe('stringStream', () => {
   it('should use parse5 by default', (cb) => {
     const stream = cheerio.stringStream({}, (err, $) => {
diff --git a/src/node.ts b/src/node.ts
index 1765e8d3c8..0d8b941857 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -9,11 +9,36 @@ import { adapter as htmlparser2Adapter } from 'parse5-htmlparser2-tree-adapter';
 import { WritableStream as Htmlparser2Stream } from 'htmlparser2/lib/WritableStream';
 import DomHandler from 'domhandler';
 import { ParserStream as Parse5Stream } from 'parse5-parser-stream';
-import { DecodeStream, type SnifferOptions } from 'encoding-sniffer';
+import {
+  decodeBuffer,
+  DecodeStream,
+  type SnifferOptions,
+} from 'encoding-sniffer';
 import * as undici from 'undici';
 import MIMEType from 'whatwg-mimetype';
 import { type Writable, finished } from 'node:stream';
 
+/**
+ * Sniffs the encoding of a buffer, then creates a querying function bound to a
+ * document created from the buffer.
+ *
+ * @param buffer - The buffer to sniff the encoding of.
+ * @param options - The options to pass to Cheerio.
+ * @returns The loaded document.
+ */
+export function loadBuffer(
+  buffer: Buffer,
+  options: DecodeStreamOptions = {}
+): CheerioAPI {
+  const opts = flattenOptions(options);
+  const str = decodeBuffer(buffer, {
+    defaultEncoding: opts?.xmlMode ? 'utf8' : 'windows-1252',
+    ...options.encoding,
+  });
+
+  return load(str, opts);
+}
+
 function _stringStream(
   options: InternalOptions | undefined,
   cb: (err: Error | null | undefined, $: CheerioAPI) => void

From 81ed6aee2eb3d874c28c3562b4fdf117376385d3 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Mon, 21 Nov 2022 19:29:46 +0000
Subject: [PATCH 11/18] Add Node exports by default for Node ESM

---
 package.json | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/package.json b/package.json
index 797ebfc993..a7db4d83c3 100644
--- a/package.json
+++ b/package.json
@@ -30,7 +30,10 @@
   "exports": {
     ".": {
       "require": "./lib/index.js",
-      "import": "./lib/esm/index.js"
+      "import": {
+        "node": "./lib/esm/node.js",
+        "default": "./lib/esm/index.js"
+      }
     },
     "./lib/slim": {
       "require": "./lib/slim.js",
@@ -39,6 +42,10 @@
     "./lib/utils": {
       "require": "./lib/utils.js",
       "import": "./lib/esm/utils.js"
+    },
+    "./lib/node": {
+      "require": "./lib/node.js",
+      "import": "./lib/esm/node.js"
     }
   },
   "files": [

From 58cd3a1bc99a9e3c85be017f9a6db09a77c1bf1e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Mon, 21 Nov 2022 19:34:58 +0000
Subject: [PATCH 12/18] Close servers after tests

---
 src/node.spec.ts | 47 +++++++++++++++++++++++++++--------------------
 1 file changed, 27 insertions(+), 20 deletions(-)

diff --git a/src/node.spec.ts b/src/node.spec.ts
index 8e65553a01..f56c6599c2 100644
--- a/src/node.spec.ts
+++ b/src/node.spec.ts
@@ -1,7 +1,7 @@
 /* eslint-disable jest/no-done-callback */
 import * as cheerio from './node.js';
 import { Writable } from 'node:stream';
-import { createServer } from 'node:http';
+import { createServer, type Server } from 'node:http';
 
 function noop() {
   // Ignore
@@ -103,29 +103,36 @@ describe('decodeStream', () => {
   });
 });
 
-function createTestServer(
-  contentType: string,
-  body: string | Buffer
-): Promise<number> {
-  return new Promise((resolve, reject) => {
-    const server = createServer((_req, res) => {
-      res.writeHead(200, { 'Content-Type': contentType });
-      res.end(body);
+describe('fromURL', () => {
+  let server: Server | undefined;
+
+  function createTestServer(
+    contentType: string,
+    body: string | Buffer
+  ): Promise<number> {
+    return new Promise((resolve, reject) => {
+      server = createServer((_req, res) => {
+        res.writeHead(200, { 'Content-Type': contentType });
+        res.end(body);
+      });
+
+      server.listen(0, () => {
+        const address = server?.address();
+
+        if (typeof address === 'string' || address == null) {
+          reject(new Error('Failed to get port'));
+        } else {
+          resolve(address.port);
+        }
+      });
     });
+  }
 
-    server.listen(0, () => {
-      const address = server.address();
-
-      if (typeof address === 'string' || address === null) {
-        reject(new Error('Failed to get port'));
-      } else {
-        resolve(address.port);
-      }
-    });
+  afterEach((cb) => {
+    server?.close(cb);
+    server = undefined;
   });
-}
 
-describe('fromURL', () => {
   it('should fetch UTF-8 HTML', async () => {
     const port = await createTestServer('text/html', TEST_HTML);
     const $ = await cheerio.fromURL(`http://localhost:${port}`);

From 1b81b163316e151d8d6988496b56e895dd3a509f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Tue, 22 Nov 2022 11:30:23 +0000
Subject: [PATCH 13/18] Allow inferred request options to be overwritten

---
 src/node.ts | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/src/node.ts b/src/node.ts
index 0d8b941857..b34bc9620c 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -1,11 +1,12 @@
 export * from './index.js';
 
+/* eslint-disable node/no-unsupported-features/node-builtins, node/file-extension-in-import */
+
 import type { CheerioAPI, CheerioOptions } from './index.js';
 import { load } from './index.js';
 import { flatten as flattenOptions, InternalOptions } from './options.js';
 import { adapter as htmlparser2Adapter } from 'parse5-htmlparser2-tree-adapter';
 
-// eslint-disable-next-line node/file-extension-in-import
 import { WritableStream as Htmlparser2Stream } from 'htmlparser2/lib/WritableStream';
 import DomHandler from 'domhandler';
 import { ParserStream as Parse5Stream } from 'parse5-parser-stream';
@@ -112,7 +113,6 @@ const defaultRequestOptions: UndiciStreamOptions = {
 
 // Get a document from a URL
 export async function fromURL(
-  // eslint-disable-next-line node/no-unsupported-features/node-builtins
   url: string | URL,
   options: CheerioRequestOptions = {}
 ): Promise<CheerioAPI> {
@@ -144,15 +144,21 @@ export async function fromURL(
        * If we allow redirects, we will have entries in the history.
        * The last entry will be the final URL.
        */
-      const history = (res.context as any)?.history;
+      const history = (
+        res.context as
+          | {
+              history?: URL[];
+            }
+          | undefined
+      )?.history;
 
       const opts = {
-        ...cheerioOptions,
         encoding,
         // Set XML mode based on the MIME type.
         xmlMode: mimeType.isXML(),
         // Set the `baseURL` to the final URL.
         baseURL: history ? history[history.length - 1] : url,
+        ...cheerioOptions,
       };
 
       return decodeStream(opts, (err, $) => (err ? reject(err) : resolve($)));

From 5cb5cfdb1f1835e538539dc0c1620f61629c88eb Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Tue, 22 Nov 2022 16:19:51 +0000
Subject: [PATCH 14/18] Add documentation comments

---
 src/node.ts | 30 +++++++++++++++++++++++++++++-
 1 file changed, 29 insertions(+), 1 deletion(-)

diff --git a/src/node.ts b/src/node.ts
index b34bc9620c..ad458a0d96 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -63,6 +63,16 @@ function _stringStream(
   return stream;
 }
 
+/**
+ * Creates a stream that parses a sequence of strings into a document.
+ *
+ * The stream is a `Writable` stream that accepts strings. When the stream is
+ * finished, the callback is called with the loaded document.
+ *
+ * @param options - The options to pass to Cheerio.
+ * @param cb - The callback to call when the stream is finished.
+ * @returns The writable stream.
+ */
 export function stringStream(
   options: CheerioOptions,
   cb: (err: Error | null | undefined, $: CheerioAPI) => void
@@ -74,6 +84,16 @@ export interface DecodeStreamOptions extends CheerioOptions {
   encoding?: SnifferOptions;
 }
 
+/**
+ * Parses a stream of buffers into a document.
+ *
+ * The stream is a `Writable` stream that accepts buffers. When the stream is
+ * finished, the callback is called with the loaded document.
+ *
+ * @param options - The options to pass to Cheerio.
+ * @param cb - The callback to call when the stream is finished.
+ * @returns The writable stream.
+ */
 export function decodeStream(
   options: DecodeStreamOptions,
   cb: (err: Error | null | undefined, $: CheerioAPI) => void
@@ -111,7 +131,15 @@ const defaultRequestOptions: UndiciStreamOptions = {
   },
 };
 
-// Get a document from a URL
+/**
+ * `fromURL` loads a document from a URL.
+ *
+ * By default, redirects are allowed and non-2xx responses are rejected.
+ *
+ * @param url - The URL to load the document from.
+ * @param options - The options to pass to Cheerio.
+ * @returns The loaded document.
+ */
 export async function fromURL(
   url: string | URL,
   options: CheerioRequestOptions = {}

From 73c174f18c6c59b4c738e1d3d0f3a4a8c67f84e5 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Tue, 22 Nov 2022 16:23:00 +0000
Subject: [PATCH 15/18] Add jsdom types for benchmarks

---
 benchmark/suite.ts |   1 -
 package-lock.json  | 334 +++++++++++++++++++++++++++++++++++++++++++++
 package.json       |   1 +
 3 files changed, 335 insertions(+), 1 deletion(-)

diff --git a/benchmark/suite.ts b/benchmark/suite.ts
index 8a8db68caf..31d3c0421b 100644
--- a/benchmark/suite.ts
+++ b/benchmark/suite.ts
@@ -2,7 +2,6 @@ import fs from 'fs';
 import path from 'path';
 
 import { Suite, Event } from 'benchmark';
-// @ts-expect-error `jsdom` types currently collide with `parse5` types.
 import { JSDOM } from 'jsdom';
 import { Script } from 'vm';
 import cheerio from '../lib/index.js';
diff --git a/package-lock.json b/package-lock.json
index 1401d10a5f..27f6beee25 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -25,6 +25,7 @@
         "@octokit/graphql": "^5.0.4",
         "@types/benchmark": "^2.1.2",
         "@types/jest": "^29.2.3",
+        "@types/jsdom": "^20.0.1",
         "@types/node": "^18.11.9",
         "@types/whatwg-mimetype": "^3.0.0",
         "@typescript-eslint/eslint-plugin": "^5.44.0",
@@ -1430,6 +1431,185 @@
         "pretty-format": "^29.0.0"
       }
     },
+    "node_modules/@types/jest/node_modules/@jest/expect-utils": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/@jest/expect-utils/-/expect-utils-29.3.1.tgz",
+      "integrity": "sha512-wlrznINZI5sMjwvUoLVk617ll/UYfGIZNxmbU+Pa7wmkL4vYzhV9R2pwVqUh4NWWuLQWkI8+8mOkxs//prKQ3g==",
+      "dev": true,
+      "dependencies": {
+        "jest-get-type": "^29.2.0"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/@jest/schemas": {
+      "version": "29.0.0",
+      "resolved": "https://registry.npmjs.org/@jest/schemas/-/schemas-29.0.0.tgz",
+      "integrity": "sha512-3Ab5HgYIIAnS0HjqJHQYZS+zXc4tUmTmBH3z83ajI6afXp8X3ZtdLX+nXx+I7LNkJD7uN9LAVhgnjDgZa2z0kA==",
+      "dev": true,
+      "dependencies": {
+        "@sinclair/typebox": "^0.24.1"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/@jest/types": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/@jest/types/-/types-29.3.1.tgz",
+      "integrity": "sha512-d0S0jmmTpjnhCmNpApgX3jrUZgZ22ivKJRvL2lli5hpCRoNnp1f85r2/wpKfXuYu8E7Jjh1hGfhPyup1NM5AmA==",
+      "dev": true,
+      "dependencies": {
+        "@jest/schemas": "^29.0.0",
+        "@types/istanbul-lib-coverage": "^2.0.0",
+        "@types/istanbul-reports": "^3.0.0",
+        "@types/node": "*",
+        "@types/yargs": "^17.0.8",
+        "chalk": "^4.0.0"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/ansi-styles": {
+      "version": "5.2.0",
+      "resolved": "https://registry.npmjs.org/ansi-styles/-/ansi-styles-5.2.0.tgz",
+      "integrity": "sha512-Cxwpt2SfTzTtXcfOlzGEee8O+c+MmUgGrNiBcXnuWxuFJHe6a5Hz7qwhwe5OgaSYI0IJvkLqWX1ASG+cJOkEiA==",
+      "dev": true,
+      "engines": {
+        "node": ">=10"
+      },
+      "funding": {
+        "url": "https://github.com/chalk/ansi-styles?sponsor=1"
+      }
+    },
+    "node_modules/@types/jest/node_modules/diff-sequences": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/diff-sequences/-/diff-sequences-29.3.1.tgz",
+      "integrity": "sha512-hlM3QR272NXCi4pq+N4Kok4kOp6EsgOM3ZSpJI7Da3UAs+Ttsi8MRmB6trM/lhyzUxGfOgnpkHtgqm5Q/CTcfQ==",
+      "dev": true,
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/expect": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/expect/-/expect-29.3.1.tgz",
+      "integrity": "sha512-gGb1yTgU30Q0O/tQq+z30KBWv24ApkMgFUpvKBkyLUBL68Wv8dHdJxTBZFl/iT8K/bqDHvUYRH6IIN3rToopPA==",
+      "dev": true,
+      "dependencies": {
+        "@jest/expect-utils": "^29.3.1",
+        "jest-get-type": "^29.2.0",
+        "jest-matcher-utils": "^29.3.1",
+        "jest-message-util": "^29.3.1",
+        "jest-util": "^29.3.1"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/jest-diff": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/jest-diff/-/jest-diff-29.3.1.tgz",
+      "integrity": "sha512-vU8vyiO7568tmin2lA3r2DP8oRvzhvRcD4DjpXc6uGveQodyk7CKLhQlCSiwgx3g0pFaE88/KLZ0yaTWMc4Uiw==",
+      "dev": true,
+      "dependencies": {
+        "chalk": "^4.0.0",
+        "diff-sequences": "^29.3.1",
+        "jest-get-type": "^29.2.0",
+        "pretty-format": "^29.3.1"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/jest-get-type": {
+      "version": "29.2.0",
+      "resolved": "https://registry.npmjs.org/jest-get-type/-/jest-get-type-29.2.0.tgz",
+      "integrity": "sha512-uXNJlg8hKFEnDgFsrCjznB+sTxdkuqiCL6zMgA75qEbAJjJYTs9XPrvDctrEig2GDow22T/LvHgO57iJhXB/UA==",
+      "dev": true,
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/jest-matcher-utils": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/jest-matcher-utils/-/jest-matcher-utils-29.3.1.tgz",
+      "integrity": "sha512-fkRMZUAScup3txIKfMe3AIZZmPEjWEdsPJFK3AIy5qRohWqQFg1qrmKfYXR9qEkNc7OdAu2N4KPHibEmy4HPeQ==",
+      "dev": true,
+      "dependencies": {
+        "chalk": "^4.0.0",
+        "jest-diff": "^29.3.1",
+        "jest-get-type": "^29.2.0",
+        "pretty-format": "^29.3.1"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/jest-message-util": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/jest-message-util/-/jest-message-util-29.3.1.tgz",
+      "integrity": "sha512-lMJTbgNcDm5z+6KDxWtqOFWlGQxD6XaYwBqHR8kmpkP+WWWG90I35kdtQHY67Ay5CSuydkTBbJG+tH9JShFCyA==",
+      "dev": true,
+      "dependencies": {
+        "@babel/code-frame": "^7.12.13",
+        "@jest/types": "^29.3.1",
+        "@types/stack-utils": "^2.0.0",
+        "chalk": "^4.0.0",
+        "graceful-fs": "^4.2.9",
+        "micromatch": "^4.0.4",
+        "pretty-format": "^29.3.1",
+        "slash": "^3.0.0",
+        "stack-utils": "^2.0.3"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/jest-util": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/jest-util/-/jest-util-29.3.1.tgz",
+      "integrity": "sha512-7YOVZaiX7RJLv76ZfHt4nbNEzzTRiMW/IiOG7ZOKmTXmoGBxUDefgMAxQubu6WPVqP5zSzAdZG0FfLcC7HOIFQ==",
+      "dev": true,
+      "dependencies": {
+        "@jest/types": "^29.3.1",
+        "@types/node": "*",
+        "chalk": "^4.0.0",
+        "ci-info": "^3.2.0",
+        "graceful-fs": "^4.2.9",
+        "picomatch": "^2.2.3"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jest/node_modules/pretty-format": {
+      "version": "29.3.1",
+      "resolved": "https://registry.npmjs.org/pretty-format/-/pretty-format-29.3.1.tgz",
+      "integrity": "sha512-FyLnmb1cYJV8biEIiRyzRFvs2lry7PPIvOqKVe1GCUEYg4YGmlx1qG9EJNMxArYm7piII4qb8UV1Pncq5dxmcg==",
+      "dev": true,
+      "dependencies": {
+        "@jest/schemas": "^29.0.0",
+        "ansi-styles": "^5.0.0",
+        "react-is": "^18.0.0"
+      },
+      "engines": {
+        "node": "^14.15.0 || ^16.10.0 || >=18.0.0"
+      }
+    },
+    "node_modules/@types/jsdom": {
+      "version": "20.0.1",
+      "resolved": "https://registry.npmjs.org/@types/jsdom/-/jsdom-20.0.1.tgz",
+      "integrity": "sha512-d0r18sZPmMQr1eG35u12FZfhIXNrnsPU/g5wvRKCUf/tOGilKKwYMYGqh33BNR6ba+2gkHw1EUiHoN3mn7E5IQ==",
+      "dev": true,
+      "dependencies": {
+        "@types/node": "*",
+        "@types/tough-cookie": "*",
+        "parse5": "^7.0.0"
+      }
+    },
     "node_modules/@types/json-schema": {
       "version": "7.0.11",
       "resolved": "https://registry.npmjs.org/@types/json-schema/-/json-schema-7.0.11.tgz",
@@ -1475,6 +1655,12 @@
       "integrity": "sha512-Hl219/BT5fLAaz6NDkSuhzasy49dwQS/DSdu4MdggFB8zcXv7vflBI3xp7FEmkmdDkBUI2bPUNeMttp2knYdxw==",
       "dev": true
     },
+    "node_modules/@types/tough-cookie": {
+      "version": "4.0.2",
+      "resolved": "https://registry.npmjs.org/@types/tough-cookie/-/tough-cookie-4.0.2.tgz",
+      "integrity": "sha512-Q5vtl1W5ue16D+nIaW8JWebSSraJVlK+EthKn7e7UcD4KWsaSJ8BqGPXNaPghgtcn/fhvrN17Tv8ksUsQpiplw==",
+      "dev": true
+    },
     "node_modules/@types/unist": {
       "version": "2.0.6",
       "resolved": "https://registry.npmjs.org/@types/unist/-/unist-2.0.6.tgz",
@@ -8489,6 +8675,148 @@
       "requires": {
         "expect": "^29.0.0",
         "pretty-format": "^29.0.0"
+      },
+      "dependencies": {
+        "@jest/expect-utils": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/@jest/expect-utils/-/expect-utils-29.3.1.tgz",
+          "integrity": "sha512-wlrznINZI5sMjwvUoLVk617ll/UYfGIZNxmbU+Pa7wmkL4vYzhV9R2pwVqUh4NWWuLQWkI8+8mOkxs//prKQ3g==",
+          "dev": true,
+          "requires": {
+            "jest-get-type": "^29.2.0"
+          }
+        },
+        "@jest/schemas": {
+          "version": "29.0.0",
+          "resolved": "https://registry.npmjs.org/@jest/schemas/-/schemas-29.0.0.tgz",
+          "integrity": "sha512-3Ab5HgYIIAnS0HjqJHQYZS+zXc4tUmTmBH3z83ajI6afXp8X3ZtdLX+nXx+I7LNkJD7uN9LAVhgnjDgZa2z0kA==",
+          "dev": true,
+          "requires": {
+            "@sinclair/typebox": "^0.24.1"
+          }
+        },
+        "@jest/types": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/@jest/types/-/types-29.3.1.tgz",
+          "integrity": "sha512-d0S0jmmTpjnhCmNpApgX3jrUZgZ22ivKJRvL2lli5hpCRoNnp1f85r2/wpKfXuYu8E7Jjh1hGfhPyup1NM5AmA==",
+          "dev": true,
+          "requires": {
+            "@jest/schemas": "^29.0.0",
+            "@types/istanbul-lib-coverage": "^2.0.0",
+            "@types/istanbul-reports": "^3.0.0",
+            "@types/node": "*",
+            "@types/yargs": "^17.0.8",
+            "chalk": "^4.0.0"
+          }
+        },
+        "ansi-styles": {
+          "version": "5.2.0",
+          "resolved": "https://registry.npmjs.org/ansi-styles/-/ansi-styles-5.2.0.tgz",
+          "integrity": "sha512-Cxwpt2SfTzTtXcfOlzGEee8O+c+MmUgGrNiBcXnuWxuFJHe6a5Hz7qwhwe5OgaSYI0IJvkLqWX1ASG+cJOkEiA==",
+          "dev": true
+        },
+        "diff-sequences": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/diff-sequences/-/diff-sequences-29.3.1.tgz",
+          "integrity": "sha512-hlM3QR272NXCi4pq+N4Kok4kOp6EsgOM3ZSpJI7Da3UAs+Ttsi8MRmB6trM/lhyzUxGfOgnpkHtgqm5Q/CTcfQ==",
+          "dev": true
+        },
+        "expect": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/expect/-/expect-29.3.1.tgz",
+          "integrity": "sha512-gGb1yTgU30Q0O/tQq+z30KBWv24ApkMgFUpvKBkyLUBL68Wv8dHdJxTBZFl/iT8K/bqDHvUYRH6IIN3rToopPA==",
+          "dev": true,
+          "requires": {
+            "@jest/expect-utils": "^29.3.1",
+            "jest-get-type": "^29.2.0",
+            "jest-matcher-utils": "^29.3.1",
+            "jest-message-util": "^29.3.1",
+            "jest-util": "^29.3.1"
+          }
+        },
+        "jest-diff": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/jest-diff/-/jest-diff-29.3.1.tgz",
+          "integrity": "sha512-vU8vyiO7568tmin2lA3r2DP8oRvzhvRcD4DjpXc6uGveQodyk7CKLhQlCSiwgx3g0pFaE88/KLZ0yaTWMc4Uiw==",
+          "dev": true,
+          "requires": {
+            "chalk": "^4.0.0",
+            "diff-sequences": "^29.3.1",
+            "jest-get-type": "^29.2.0",
+            "pretty-format": "^29.3.1"
+          }
+        },
+        "jest-get-type": {
+          "version": "29.2.0",
+          "resolved": "https://registry.npmjs.org/jest-get-type/-/jest-get-type-29.2.0.tgz",
+          "integrity": "sha512-uXNJlg8hKFEnDgFsrCjznB+sTxdkuqiCL6zMgA75qEbAJjJYTs9XPrvDctrEig2GDow22T/LvHgO57iJhXB/UA==",
+          "dev": true
+        },
+        "jest-matcher-utils": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/jest-matcher-utils/-/jest-matcher-utils-29.3.1.tgz",
+          "integrity": "sha512-fkRMZUAScup3txIKfMe3AIZZmPEjWEdsPJFK3AIy5qRohWqQFg1qrmKfYXR9qEkNc7OdAu2N4KPHibEmy4HPeQ==",
+          "dev": true,
+          "requires": {
+            "chalk": "^4.0.0",
+            "jest-diff": "^29.3.1",
+            "jest-get-type": "^29.2.0",
+            "pretty-format": "^29.3.1"
+          }
+        },
+        "jest-message-util": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/jest-message-util/-/jest-message-util-29.3.1.tgz",
+          "integrity": "sha512-lMJTbgNcDm5z+6KDxWtqOFWlGQxD6XaYwBqHR8kmpkP+WWWG90I35kdtQHY67Ay5CSuydkTBbJG+tH9JShFCyA==",
+          "dev": true,
+          "requires": {
+            "@babel/code-frame": "^7.12.13",
+            "@jest/types": "^29.3.1",
+            "@types/stack-utils": "^2.0.0",
+            "chalk": "^4.0.0",
+            "graceful-fs": "^4.2.9",
+            "micromatch": "^4.0.4",
+            "pretty-format": "^29.3.1",
+            "slash": "^3.0.0",
+            "stack-utils": "^2.0.3"
+          }
+        },
+        "jest-util": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/jest-util/-/jest-util-29.3.1.tgz",
+          "integrity": "sha512-7YOVZaiX7RJLv76ZfHt4nbNEzzTRiMW/IiOG7ZOKmTXmoGBxUDefgMAxQubu6WPVqP5zSzAdZG0FfLcC7HOIFQ==",
+          "dev": true,
+          "requires": {
+            "@jest/types": "^29.3.1",
+            "@types/node": "*",
+            "chalk": "^4.0.0",
+            "ci-info": "^3.2.0",
+            "graceful-fs": "^4.2.9",
+            "picomatch": "^2.2.3"
+          }
+        },
+        "pretty-format": {
+          "version": "29.3.1",
+          "resolved": "https://registry.npmjs.org/pretty-format/-/pretty-format-29.3.1.tgz",
+          "integrity": "sha512-FyLnmb1cYJV8biEIiRyzRFvs2lry7PPIvOqKVe1GCUEYg4YGmlx1qG9EJNMxArYm7piII4qb8UV1Pncq5dxmcg==",
+          "dev": true,
+          "requires": {
+            "@jest/schemas": "^29.0.0",
+            "ansi-styles": "^5.0.0",
+            "react-is": "^18.0.0"
+          }
+        }
+      }
+    },
+    "@types/jsdom": {
+      "version": "20.0.1",
+      "resolved": "https://registry.npmjs.org/@types/jsdom/-/jsdom-20.0.1.tgz",
+      "integrity": "sha512-d0r18sZPmMQr1eG35u12FZfhIXNrnsPU/g5wvRKCUf/tOGilKKwYMYGqh33BNR6ba+2gkHw1EUiHoN3mn7E5IQ==",
+      "dev": true,
+      "requires": {
+        "@types/node": "*",
+        "@types/tough-cookie": "*",
+        "parse5": "^7.0.0"
       }
     },
     "@types/json-schema": {
@@ -8536,6 +8864,12 @@
       "integrity": "sha512-Hl219/BT5fLAaz6NDkSuhzasy49dwQS/DSdu4MdggFB8zcXv7vflBI3xp7FEmkmdDkBUI2bPUNeMttp2knYdxw==",
       "dev": true
     },
+    "@types/tough-cookie": {
+      "version": "4.0.2",
+      "resolved": "https://registry.npmjs.org/@types/tough-cookie/-/tough-cookie-4.0.2.tgz",
+      "integrity": "sha512-Q5vtl1W5ue16D+nIaW8JWebSSraJVlK+EthKn7e7UcD4KWsaSJ8BqGPXNaPghgtcn/fhvrN17Tv8ksUsQpiplw==",
+      "dev": true
+    },
     "@types/unist": {
       "version": "2.0.6",
       "resolved": "https://registry.npmjs.org/@types/unist/-/unist-2.0.6.tgz",
diff --git a/package.json b/package.json
index a7db4d83c3..899e0712ad 100644
--- a/package.json
+++ b/package.json
@@ -71,6 +71,7 @@
     "@octokit/graphql": "^5.0.4",
     "@types/benchmark": "^2.1.2",
     "@types/jest": "^29.2.3",
+    "@types/jsdom": "^20.0.1",
     "@types/node": "^18.11.9",
     "@types/whatwg-mimetype": "^3.0.0",
     "@typescript-eslint/eslint-plugin": "^5.44.0",

From 6e7dbf4197adfaf276e2425315c7ca98bd982dc9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Tue, 22 Nov 2022 16:26:18 +0000
Subject: [PATCH 16/18] Update suite.ts

---
 benchmark/suite.ts | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/benchmark/suite.ts b/benchmark/suite.ts
index 31d3c0421b..e66e6298a6 100644
--- a/benchmark/suite.ts
+++ b/benchmark/suite.ts
@@ -1,9 +1,9 @@
-import fs from 'fs';
-import path from 'path';
+import fs from 'node:fs';
+import path from 'node:path';
+import { Script } from 'node:vm';
 
 import { Suite, Event } from 'benchmark';
 import { JSDOM } from 'jsdom';
-import { Script } from 'vm';
 import cheerio from '../lib/index.js';
 
 const documentDir = path.join(__dirname, 'documents');
@@ -73,11 +73,9 @@ export default class Suites {
 
     jQueryScript.runInContext(dom.getInternalVMContext());
 
-    const setupData: T = options.setup(dom.window.$);
+    const setupData: T = options.setup(dom.window['$']);
 
-    suite.add('jsdom', () => {
-      testFn(dom.window.$, setupData);
-    });
+    suite.add('jsdom', () => testFn(dom.window['$'], setupData));
     suite.run();
   }
 

From 526f1a527ec40fbac223ac32af37e655c732550a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Tue, 22 Nov 2022 16:56:31 +0000
Subject: [PATCH 17/18] Update eslint-plugin-node to -n

---
 src/node.ts | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/node.ts b/src/node.ts
index ad458a0d96..c6154da815 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -1,6 +1,6 @@
 export * from './index.js';
 
-/* eslint-disable node/no-unsupported-features/node-builtins, node/file-extension-in-import */
+/* eslint-disable n/no-unsupported-features/node-builtins, n/file-extension-in-import */
 
 import type { CheerioAPI, CheerioOptions } from './index.js';
 import { load } from './index.js';

From ce9d283fd5ffa421e6f542b4d1093588289348c3 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Felix=20B=C3=B6hm?= <188768+fb55@users.noreply.github.com>
Date: Tue, 22 Nov 2022 17:40:03 +0000
Subject: [PATCH 18/18] Split comment

---
 src/node.ts | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/node.ts b/src/node.ts
index c6154da815..b7f0c98f74 100644
--- a/src/node.ts
+++ b/src/node.ts
@@ -1,12 +1,13 @@
 export * from './index.js';
 
-/* eslint-disable n/no-unsupported-features/node-builtins, n/file-extension-in-import */
+/* eslint-disable n/no-unsupported-features/node-builtins */
 
 import type { CheerioAPI, CheerioOptions } from './index.js';
 import { load } from './index.js';
 import { flatten as flattenOptions, InternalOptions } from './options.js';
 import { adapter as htmlparser2Adapter } from 'parse5-htmlparser2-tree-adapter';
 
+// eslint-disable-next-line n/file-extension-in-import
 import { WritableStream as Htmlparser2Stream } from 'htmlparser2/lib/WritableStream';
 import DomHandler from 'domhandler';
 import { ParserStream as Parse5Stream } from 'parse5-parser-stream';