htmlの本文のxpathを探すアルゴリズム(その2)

mobilizer.rb

require 'nokogiri'

def mobilizer(page)
	page = Nokogiri::HTML(page.to_s) if page.class == String
	remove_tags = ['script', 'style', 'select', 'a']
	remove_tags.each do |tag|
		page.xpath('//' + tag).each { |n| n.remove }
	end
	
	path = identifier(page, '/html/body')
end

def identifier(page, recent_path)
	sum = 0
	pathes = []
	page.xpath(recent_path + '/node()').each do |node|
		size = node.text.size
		puts size
		sum += size > 100 ? size : 0
		pathes << node.path
	end
	
	least_importance = sum * 0.5r
	pathes.each do |path|
		return identifier(page,path) if page.xpath(path).text.size > least_importance
	end
	recent_path
end