What is everythingtohtml?

everythingtohtml is a Python library and command-line tool that converts a wide range of file formats into clean, self-contained HTML. You give it a PDF, a Word document, a spreadsheet, a Markdown file, a CSV, an EPUB, an email, a Jupyter notebook, or several other common formats, and it gives back a single HTML file you can open in any browser. No internet connection is required for local files, and no separate program needs to be running. The project describes itself as the inverse of tools that convert rich documents down to Markdown. The argument is that Markdown loses structure: tables get flattened, styling disappears, and slide layouts vanish. HTML preserves that structure, keeping proper table tags, heading hierarchy, sections, and links, while still being something any browser can display without additional software. The output documents also include a small built-in stylesheet and support dark mode. Installing the base package covers plain text, Markdown, CSV, JSON, EPUB, email, Jupyter notebooks, RSS feeds, and a few other formats. Office formats like Word, Excel, and PowerPoint need separate optional installs, as does PDF support. A single command installs everything at once if you want the full set. The tool requires Python 3.10 or newer. Usage is straightforward from both Python code and the command line. In code, you pass a file path, a URL, or an open file stream to a convert method and get back an object containing the full HTML string and a detected document title. On the command line, you point it at a file and optionally name an output file, or pipe the result elsewhere. A shorthand command alias is available for faster typing. The tool also supports merging multiple documents of any supported format into a single HTML page with a table of contents, displaying two documents side by side for comparison, and generating a highlighted line-by-line diff between two documents. New file format support can be added by writing a small Python class following a documented pattern. The project is MIT licensed.