pdf2htmlex/pdf2htmlex ? reverse-engineered prompt

Build me a tool that converts PDF files into HTML while keeping the text, fonts, layout, and images as accurate as possible. I want the output to look as close to the original PDF as it can, including links, bookmarks, printing support, and complicated pages like papers, magazines, and documents with formulas or mixed languages.

It should support two ways of viewing the result, either one single HTML file or page by page loading with JavaScript. If some text is hidden or partly covered, handle that correctly so it does not show through in the wrong place. Please make the output practical for real use, not just a rough preview.

If you need to check current build or dependency details, look up the latest docs online. Also make sure it works well with modern browsers and produces clean output that is not unnecessarily huge.