CastReader 如何破解 Kindle 字体加密以实现文本转语音

📄 中文摘要

CastReader 通过拦截亚马逊的字体子集数据,解码 Kindle Cloud Reader 的文本,并利用本地 OCR 校准将混淆的字形代码映射回真实字符。这是唯一能够实现这一功能的 Chrome 扩展程序,其他文本转语音工具由于亚马逊的自定义字体使得 DOM 文本不可读,导致读取的内容毫无意义。Kindle Cloud Reader 并不像普通网站那样渲染文本,页面以预渲染的图像形式呈现,缺乏可读的文本标签和元素,这使得文本转语音工具面临巨大挑战。

📄 English Summary

How CastReader Cracks Kindle's Font Encryption for Text-to-Speech

CastReader decodes text from the Kindle Cloud Reader by intercepting Amazon's font subset data and mapping scrambled glyph codes back to actual characters using local OCR calibration. It is the only Chrome extension capable of this, as all other TTS tools produce gibberish due to Amazon's custom fonts rendering the DOM text unreadable. The Kindle Cloud Reader does not render text like a typical website; there are no readable <p> tags or <span> elements containing words. Instead, the entire page is delivered as a pre-rendered blob image, presenting significant challenges for TTS tools.

Powered by Cloudflare Workers + Payload CMS + Claude 3.5

数据源: OpenAI, Google AI, DeepMind, AWS ML Blog, HuggingFace 等