Html to Slack Markdown Converter

slack_markdown_converter.rb

require 'nokogiri'

  class SlackMarkdownConverter
    class << self
      def convert(html, options = {})
        converter = new(options)
        converter.convert(html)
      end
    end

    def initialize(options = {})
      @options = options
      @include_images = options.fetch(:include_images, false)
    end

    def convert(html)
      return '' if html.nil? || html.empty?

      doc = Nokogiri::HTML.fragment(html)
      process_node(doc)
    end

    private

    def process_node(node)
      result = ''

      node.children.each do |child|
        result += process_element(child)
      end

      # Clean up extra whitespace
      result.gsub(/\n{3,}/, "\n\n").strip
    end

    def process_element(element)
      case element.name
      when 'text'
        element.text
      when 'br'
        "\n"
      when 'p'
        process_node(element) + "\n\n"
      when 'strong', 'b', 'em', 'i', 'code'
        process_formatting_element(element)
      when 'pre'
        "```\n#{process_node(element)}\n```"
      when 'blockquote'
        process_blockquote(element)
      when 'a'
        process_link(element)
      when 'ul', 'ol'
        process_list_container(element)
      when 'li'
        process_node(element) + "\n"
      when 'h1', 'h2', 'h3', 'h4', 'h5', 'h6'
        "*#{process_node(element)}*\n\n"
      when 'img'
        @include_images ? process_image(element) : ''
      when 'div', 'span'
        process_node(element)
      when 'hr'
        "---\n"
      else
        # For any other tags, just process their contents
        process_node(element)
      end
    end

    def process_formatting_element(element)
      case element.name
      when 'strong', 'b'
        "*#{process_node(element)}*"
      when 'em', 'i'
        "_#{process_node(element)}_"
      when 'code'
        "`#{process_node(element)}`"
      end
    end

    def process_list_container(element)
      marker = element.name == 'ul' ? '• ' : lambda { |i| "#{i + 1}. " }
      process_list(element, marker)
    end

    def process_blockquote(node)
      content = process_node(node)
      lines = content.split("\n")
      quoted_lines = lines.map { |line| "> #{line}" }
      quoted_lines.join("\n") + "\n\n"
    end

    def process_link(node)
      text = process_node(node)
      href = node['href']

      # If text is the same as href or just a basic placeholder, show href only
      if text.strip == href || text.strip.downcase == 'link'
        "<#{href}>"
      else
        "<#{href}|#{text}>"
      end
    end

    def process_list(node, marker)
      result = "\n"
      node.children.each_with_index do |child, i|
        next unless child.name == 'li'

        prefix = marker.is_a?(Proc) ? marker.call(i) : marker
        content = process_node(child).strip
        result += "#{prefix}#{content}\n"
      end
      result + "\n"
    end

    def process_image(node)
      alt_text = node['alt'] || 'image'
      src = node['src']
      "<#{src}|#{alt_text}>"
    end
  end