UnquoteMail

TypeScript port of unquotemail - parse HTML/text emails and extract only the new message content, removing quoted replies. This fork adds html cleaning, markdown output support, getQuote method and performance optimizations.

Installation

npm install unquotemail

Usage

import { Unquote } from 'unquotemail';

// Create instance - parsing is lazy (happens on first getter call)
const unquote = new Unquote(htmlContent, textContent);

// Primary content (reply history stripped)
unquote.getHtml();              // Cleaned HTML (default)
unquote.getHtml({ raw: true }); // Raw HTML (original structure)
unquote.getText();              // Plain text
unquote.getMarkdown();          // Markdown

// The stripped quote block
unquote.getQuote();             // Quote HTML (null if none)

Standalone Converters

import { htmlToText, htmlToMarkdown } from 'unquotemail';

const text = htmlToText(html);
const markdown = htmlToMarkdown(html);

Markdown Converter Features

The markdown converter (used by both getMarkdown() and htmlToMarkdown()) is optimized for email HTML:

• Flattens layout tables (common in email templates) instead of rendering them as markdown tables
• Ignores data URI images, scripts, and styles
• Handles non-standard HTML from various email clients
• Preserves nested blockquotes with proper > syntax

How it works

1. First tries to identify and remove known quote markup (.gmail_quote, .protonmail_quote, etc.)
2. Falls back to regex patterns to identify "On DATE, NAME wrote:" style headers
3. Removes everything from the quote marker onwards

Credits

• Original Python implementation by Cyril Nicodeme
• Regex patterns from Talon (Mailgun) and Email Reply Parser (Crisp)
• Markdown conversion by node-html-markdown

License

MIT