用ruby写了一个搜索下载歌曲的工具

[url=http://fuliang.iteye.com/blog/176323]前几天用java写了一个GUI的搜索下载工具[/url],主要利用baidu mp3搜索的结果。[url=http://hi.baidu.com/david_jlu/blog/item/d76622a1f1a6878f46106446.html]david同学用perl写了命令行的类似的下载工具[/url],为了练练ruby,我又写了ruby版的。
Fetcher类：
根据url来Fetch到页面，供Parser分析之用

 require "net/http" 

 class Fetcher

  def fetch(url)
    host = url.scan(/\/\/(.*?)\//m)[0][0]
    path = url.split(/#{host}\//)[1]
   # print "host: ",host,"\n"
   # print "path: ",path,"\n"
    h = Net::HTTP.new(host,80)
    resp = h.get("/#{path}",nil)

    if resp.message == "OK"
     # puts "建立连接成功..." 
      return resp.body     
    end 
    return ""
  end

end

Parser类：
提取出可供下载的链接，并通过ping，来选取速度最快的连接，供Download之用：

class Parser
public
  def initialize()
    @fetcher = Fetcher.new
  end

  def parse_mp3(html)
    urls = html.scan(/<a href="(.*?)"/m)
    download_hosts_urls = {}
    parse_threads = []
    for url in urls do
        if url[0] =~ /.*?\.mp3,,.*?/
           parse_threads << Thread.new(url) do |url|
              song_url = url[0].gsub(" ","%20")
              download_url = parse_download_url(song_url)
              if download_url
              	host =  download_url.scan(/\/\/(.*?)\//m)[0][0] 
              	#We only want to find the best download url,so we needn't care duplicate key
              	download_hosts_urls[host] = download_url
              end 
           end
        end
    end
    parse_threads.each{|t| t.join}
    puts "已经搜索到#{download_hosts_urls.size}个链接可以下载..."
    exit(1) if download_hosts_urls.size == 0
    puts "正在选择速度最快的链接..."
    host = select_best_host(download_hosts_urls.keys)
    download_hosts_urls[host]
  end

private
  def select_best_host(hosts)
    times_hosts = {}
    threads = []
    hosts.each do |host|
      threads << Thread.new(host) do |host|
           response = `ping -c 1 -W 30 #{host}` #use`ping -n 1 -w 30 #{host}` in windows
           r_t = response.scan(/time=(\d+)/m) #only get integer part
           times_hosts[r_t[0][0]] = host unless r_t.empty? #duplicate key no problem 
      end
    end

    threads.each{|t| t.join}

    times = times_hosts.keys
    min = times.min
    times_hosts[min]
  end

  def parse_download_url(song_url)
     html = @fetcher.fetch(song_url)
     urls = html.scan(/<a href="(.*?)"/m)
     return nil if urls.empty? || urls[0][0] =~ /.*?\.html/
     return urls[0][0]      
  end
end

Download类：

 require "open-uri"
require "parser"
require "fetcher"

class Download
public
  def initialize(song_name)
    @song_name = song_name
    @search_url = "http://mp3.baidu.com/m?f=ms&tn=baidump3&ct=134217728&lf=&rn=&word=#@song_name&lm=0"
    @parser = Parser.new
    @fetcher = Fetcher.new
  end

  def download
    puts "正在建立连接..."
    html = @fetcher.fetch(@search_url)
    puts "正在获取搜索结果..."
    url = @parser.parse_mp3(html)
    puts "已经获得最快的下载连接:#{url}.\n开始下载..."
    doDownload(url)    
    puts "下载完毕..."
  end
private
  def doDownload(url)
    open(url) do |fin|
  	size = fin.size
  	download_size = 0
  	puts "大小: #{size / 1024}KB"
  	filename = url[url.rindex('/')+1, url.length-1]
  	puts "歌曲名: #{filename}"
  	open(File.basename("./#{filename}"),"wb") do |fout|
     	    while buf = fin.read(1024) do
       		fout.write buf
       		download_size += buf.size
                print "已经下载： #{download_size * 100 / size}%\r"
                STDOUT.flush 
           end
       end
    end 
    puts
  end
end

download = Download.new(ARGV[0])
download.download

[quote]
fuliang@fuliang-desktop:~/program/ruby/mp3download$ ruby download.rb pretty body
正在建立连接...
正在获取搜索结果...
已经搜索到25个链接可以下载...
正在选择速度最快的链接...
已经获得最快的下载连接:http://www.jxggzp.com/muisc/20051122185348.mp3.
开始下载...
大小: 6570KB
歌曲名: 20051122185348.mp3
已经下载： 100%
下载完毕...
[/quote]
基本上可以使用。现在还存在一些问题，下载链接中有中文，往往会失败，主要是没有进行编码，知道ruby有个Iconv.conv来转换编码，不知道如何直接对中文进行编码：不知道没有像encode("gb2312","大海")之类的方法。另一个是下载问题：进度条有问题，主要open-uri使用open貌似就把文件下载到本地了，造成open很长时间，fin.read,fout.write是本地操作则非常快，结果下载进度从开始出现到下载完成瞬间就完成。希望各位达人可以帮助修正两个问题。