Read and Comment Color-Palette from https://github.com/alyssa/color-palette

color_palette.rb

require 'open-uri'
require 'nokogiri'                  # Nokogiri 是 Ruby 上的一個 HTML, XML, SAX 的 parser

class ColorPalette

  @@site_name                       # 类变量

  def initialize(url)               # 构造函数
    url = check_url(url)
    @@site_name = URI.parse(url).host   # 从url中取得网址
    @color_map = Hash.new           # 类实例变量，哈希表用来保存出现的颜色
    urls = get_stylesheet_urls(url) # 取得相应的样式表，因为里面记录了颜色文件
    build_color_palette(urls)       # 生成颜色列表
    sort_palette                    # 进行排序
  end

  def check_url(url)
    url = "http://" + url if !url.start_with?("http", "https", "ftp")
    url = url.chomp('/') if url.end_with? '/'
    return url
  end

  def get_stylesheet_urls(url)
    urls = Array.new                                          # 用来保存可能获取颜色信息的网页位置
    begin
      page = Nokogiri::HTML(open(url))
    rescue                                                    # rescue 块用来处理异常
      abort("Not a valid link. Try another one.")
    end
    page.css('head').css('link[rel=stylesheet]').each {|stylesheet|
      if !stylesheet['href'].start_with? '/'
        css_url = stylesheet['href']
      else
        css_url = url +  stylesheet['href'] # local files, must add to url for uri parsing
      end
      urls << css_url                                         # 数组这样也可以追加元素
    }
    urls
  end

  def build_color_palette(urls)
    urls.each{ |css_url|
      #    puts css_url
      begin
        page_source = Nokogiri::HTML(open(css_url)).text
      rescue
        puts "Did not pull any colors from badly formed url:\n"+css_url
        next
      end
      page_source = Nokogiri::HTML(open(css_url)).text
      # |color| 5个元素的数组 (5个正则模式去匹配)
      # 种类 1: hex
      # 种类 2,3,4: rgb respectively notes: handles rgba but ignores opacity
      # 种类 5: 颜色名
      color_array = page_source.scan(/color\s*:\s*(#[0-9A-Fa-f]{3,6}+)\s*|rgba?\s*\(\s*(\d{1,3})\s*,\s*(\d{1,3})\s*,\s*(\d{1,3})\s*|color\s*:\s*(white|aqua|black|blue|fuchsia|gray|green|lime|maroon|navy|olive|orange|purple)/)
      color_array.each{|color|
        #puts color.inspect
        if color[0] != nil
          color = color[0].downcase
          color = expand_hex(color) if color.length == 4 # usually length 3 but including '#' so use length 4
        elsif color[1] != nil
          color = "#" + color[1].to_i.to_s(16) + color[2].to_i.to_s(16) + color[3].to_i.to_s(16)
          # below is the case that you have r: 0 g: 0 b: 0 => #000 .... need to expand
          color = expand_hex(color) if color.length == 4 # usually length 3 but including '#' so use length 4
        else
          color = color[4].downcase
          hex_color = get_hex(color)
          color = hex_color if hex_color != nil # try to assign a hex value, if not, will remain as is.
        end
        # 颜色哈希表中，对应颜色的频率加1
        if @color_map[color] == nil
          @color_map[color] = 1
        else
          @color_map[color] += 1
        end
      }
    }
  end

  # 有些颜色使用名称指定的，那么需要转化为rgb码
  def get_hex(color)                               # 注意 case 语句的返回值
    hex = case color
            when "white"    then "#ffffff"
            when "black"    then "#000000"
            when "blue"     then "#0000ff"
            when "fuchsia"  then "#ff00ff"
            when "gray"     then "#808080"
            when "green"    then "#008000"
            when "lime"     then "#00ff00"
            when "maroon"   then "#800000"
            when "olive"    then "#808000"
            when "orange"   then "#ffA500"
            when "purple"   then "#800080"
          end
  end

  # 类似于 #765 这种rgb码，需要转化
  def expand_hex(color)
    new_hex = color[0]
    for i in 1..3
      new_hex += color[i]+color[i]
    end
    new_hex
  end

  # 对哈希表排序，返回仅含颜色名的数组
  def sort_palette
    @color_palette = Array.new
    @color_map = @color_map.sort_by {|k,v| v}.reverse
    @color_map.each{|key, value|
      @color_palette << key
    }
    return @color_palette
  end

  def print_palette_html
    file = File.new("#{@@site_name}.html", "w+")        # File puts 直接写入文件中
    file.puts "<html>"
    file.puts "<title>#{@@site_name} Color Page</title>"
    file.puts "<body>"
    file.puts "<table>"
    file.puts "<tr>"
    file.puts "<th> Color </th>"
    file.puts "<th> Hex </th>"
    file.puts "<th> Frequency </th>"
    file.puts "</tr>"
    @color_map.each{|key, value|
      file.puts "<tr>"
      file.puts "<td style='width:50px; height:50px; background-color: #{key}'></td>"
      file.puts "<td> #{key} </td>"
      file.puts "<td> #{value} </td>"
      file.puts "</tr>"
    }
    file.puts "</body>"
    file.puts "</html>"
  end

  def print_color_palette
    @color_palette.each{|c| puts c}
  end

  def print_palette_with_freq
    puts "number of unique colors: " + @color_map.size.to_s
    printf("--------|-----------\n")
    printf("%.8s\t|%s\n", "Color", "Frequency")
    printf("--------|-----------\n")
    @color_map.each{|key, value|
      printf("%.8s |%d\n", key, value)
    }

  end
end

if ARGV.length == 0
  abort("must provide a url")       # 终止程序，并输出提示信息
elsif ARGV.length > 1
  abort("Too much info, buddy. I just need one url.")
elsif ARGV.length == 1
  param_url = ARGV[0]               # ARGV是一个参数数组变量
end

cp = ColorPalette.new("#{param_url}")     # Perl只需要 $param_url ，这里略显复杂
puts "Palette for " + param_url
cp.print_palette_with_freq          # 控制台，输出对应网页使用颜色的频率列表
cp.print_palette_html               # 记录颜色列表信息，并生成相应网页

result

Palette for http://www.cityu.edu.hk/
number of unique colors: 27
--------|-----------
Color  |Frequency
--------|-----------
#ffffff |31
#555555 |10
#666666 |7
#333333 |5
#398600 |5
#000000 |4
#444444 |3
#dddddd |2
#f3f3ed |2
#cccccc |2
#323232 |2
#0072aa |2
#222222 |2
#888888 |1
#acd5ea |1
#025d8c |1
#323f41 |1
#ff0000 |1
#367e00 |1
#dd0000 |1
#667172 |1
#2094c3 |1
#d7e1e1 |1
#9ec990 |1
#0074a3 |1
#999999 |1
#767775 |1