From ac0f2292507e280ecc9d60293bf89325e89aec50 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E2=88=9A=28noham=29=C2=B2?=
 <100566912+NohamR@users.noreply.github.com>
Date: Fri, 19 Jun 2026 16:17:04 +0200
Subject: [PATCH] Add local HTML fallback for bot protection

---
 src/index.js | 33 ++++++++++++++++++++++++++++++++-
 1 file changed, 32 insertions(+), 1 deletion(-)

diff --git a/src/index.js b/src/index.js
index f3e081f..b1e46c5 100644
--- a/src/index.js
+++ b/src/index.js
@@ -1,5 +1,7 @@
+import fs from 'node:fs'
 import process from 'node:process'
 import path from 'node:path'
+import readline from 'node:readline/promises'
 import { parseArgs, VALID_THEMES } from './cli.js'
 import { loadPreferences, readCustomCss } from './config.js'
 import logger from './logger.js'
@@ -26,6 +28,16 @@ import {
 
 const EXTENSION_BASE = 'chrome-extension://'
 
+async function promptForInput () {
+  const rl = readline.createInterface({
+    input: process.stdin,
+    output: process.stdout
+  })
+  const answer = await rl.question('> ')
+  rl.close()
+  return answer.trim().replace(/^['"]|['"]$/g, '')
+}
+
 async function main () {
   const { url, opts } = parseArgs()
 
@@ -63,7 +75,26 @@ async function main () {
     logger.info(`Fetching article: ${url}`)
     const articlePage = await browser.newPage()
     await articlePage.goto(url, { waitUntil: 'networkidle0', timeout: 30000 })
-    const article = await extractArticle(articlePage, getReadabilityPath())
+    let article = await extractArticle(articlePage, getReadabilityPath())
+
+    if (article.title === 'Just a moment...' || article.length < 500) {
+      logger.warn('Cloudflare or bot protection detected..')
+      logger.warn('Open the URL in your browser, save the page as complete HTML (File > Save Page As),')
+      logger.warn('then paste the saved file path below and press Enter:')
+
+      const htmlPath = await promptForInput()
+      const resolvedPath = path.resolve(htmlPath)
+
+      if (!fs.existsSync(resolvedPath)) {
+        logger.error(`File not found: ${resolvedPath}`)
+        process.exit(1)
+      }
+
+      logger.info(`Loading article from local file: ${resolvedPath}`)
+      await articlePage.goto(`file://${resolvedPath}`, { waitUntil: 'networkidle0', timeout: 30000 })
+      article = await extractArticle(articlePage, getReadabilityPath())
+    }
+
     await articlePage.close()
 
     const extPage = await browser.newPage()