Example algorithm to textify HTML of tweetish posts

html2text.rb

require 'nokogiri'

def content_text(nodes)
	nodes.map do |el|
		if el.text? || el.attributes['class'].to_s.match(/\b(?:h\-card|vcard|h\-x\-username)\b/) || el.attributes['rel'].to_s.match(/\btag\b/)
			el.text
		elsif el.name == 'a'
			href = el.attributes['href'].to_s
			if el.text.strip == ''
				''
			elsif el.text.match(/[\/\.]/) && href.gsub(/[^\w\-_\/]/, '').include?(el.text.gsub(/[^\w\-_\/]/, ''))
				href
			else
				el.text + " #{href}"
			end
		elsif el.name == 'img'
			el.attributes['src'].to_s
		else
			content_text(el.children)
		end
	end.join
end

content_text(Nokogiri::HTML.fragment(item[:content].to_s).children)