DotnetSpider (一) 架构的理解、应用、搭建

2018/04/24 16:48
阅读数 64

** 温馨提示:如需转载本文,请注明内容出处。**

本文链接:http://www.cnblogs.com/grom/p/8931650.html 

 

本文Demo:https://github.com/Grom-Li/DataSpider 适用版本: 2.5.0、2.5.1

 

最近在做爬虫,之前一直在使用 HttpWebRequest 和 WebClient ,很方便快捷,也很适合新手,但随着抓取任务的增多,多任务,多库等情况的出现,使用一个优秀的爬虫框架是十分必要的。于是开始接触dotnetspider。

 

  

借鉴一下框架的设计图,在引入dotnetspider的NuGet包后,我基本也是按照这个进行了分层

  

 

Data.Spider - 存放前台页面(Winform、控制台)和实体爬虫(EntitySpider)等,相当于发起请求的起点。

Spider.Downloader - 封装请求等信息,可实现自定义cookie等,非必须。

Spider.Processor - 处理器,继承 IPageProcessor 实现对抓取内容的处理

Spider.Pipe - 管道,我将它理解为经过了 Processor 处理后的一个回调,将处理好的数据存储(文件、数据库等)

Spider.Entity - 数据实体类,继承 SpiderEntity

Spider.Command - 一些常用的公用命令,我这目前存放着转数据格式类,后台执行JS类,SqlHelper(因架构自带数据库管道,暂时没用)等

这样的分层也是参考了源码的示例

  

随着这几天的尝试,真的发现这个框架真的非常灵活,以凹凸租车的爬虫为例,上代码

  实体类

[EntityTable("CarWinsSpider", "AtzucheCar", EntityTable.Today)]
[EntitySelector(Expression = "$.data.content[*]", Type = SelectorType.JsonPath)]

public class AtzucheModel : SpiderEntity
{
  /// <summary>
  /// 车辆编号
  /// </summary>
  [PropertyDefine(Expression = "$.carNo", Type = SelectorType.JsonPath)]
  public int carNo { get; set; }
  /// <summary>
  /// 品牌
  /// </summary>
  //[ReplaceFormatter(NewValue = "", OldValue = "\r")]
  //[ReplaceFormatter(NewValue = "", OldValue = "\t")]
  //[ReplaceFormatter(NewValue = "", OldValue = "&nbsp;")]
  //[ReplaceFormatter(NewValue = "", OldValue = "\n")]
  //[ReplaceFormatter(NewValue = "", OldValue = "\"")]
  //[ReplaceFormatter(NewValue = "", OldValue = " ")]
  [PropertyDefine(Expression = "$.brand", Type = SelectorType.JsonPath)]
  public string brand { get; set; }
  /// <summary>
  /// 地址
  /// </summary>
  [PropertyDefine(Expression = "$.carAddr", Type = SelectorType.JsonPath)]
  public string carAddr { get; set; }
  /// <summary>
  /// 车系
  /// </summary>
  [PropertyDefine(Expression = "$.type", Type = SelectorType.JsonPath)]
  public string type { get; set; }
  /// <summary>
  /// 排量
  /// </summary>
  [PropertyDefine(Expression = "$.sweptVolum", Type = SelectorType.JsonPath)]
  public string sweptVolum { get; set; }
  /// <summary>
  /// 图片
  /// </summary>
  [PropertyDefine(Expression = "$.coverPic", Type = SelectorType.JsonPath)]
  public string coverPic { get; set; }
  /// <summary>
  /// 日租金
  /// </summary>
  [PropertyDefine(Expression = "$.dayPrice", Type = SelectorType.JsonPath)]
  public int dayPrice { get; set; }
  /// <summary>
  /// 公里数
  /// </summary>
  [PropertyDefine(Expression = "$.distance", Type = SelectorType.JsonPath)]
  public string distance { get; set; }
  /// <summary>
  /// 评分
  /// </summary>
  [PropertyDefine(Expression = "$.evalScore", Type = SelectorType.JsonPath)]
  public string evalScore { get; set; }
  [PropertyDefine(Expression = "$.gbType", Type = SelectorType.JsonPath)]
  public string gbType { get; set; }
  /// <summary>
  /// 车牌
  /// </summary>
  [PropertyDefine(Expression = "$.plateNum", Type = SelectorType.JsonPath)]
  public string plateNum { get; set; }
  [PropertyDefine(Expression = "$.replyTag", Type = SelectorType.JsonPath)]
  public string replyTag { get; set; }
  [PropertyDefine(Expression = "$.transCount", Type = SelectorType.JsonPath)]
  public string transCount { get; set; }
  /// <summary>
  /// 年款
  /// </summary>
  [PropertyDefine(Expression = "$.year", Type = SelectorType.JsonPath)]
  public int year { get; set; }
  [PropertyDefine(Expression = "$.isPrivilege", Type = SelectorType.JsonPath)]
  public int isPrivilege { get; set; }
  [PropertyDefine(Expression = "$.isRecommend", Type = SelectorType.JsonPath)]
  public int isRecommend { get; set; }
  [PropertyDefine(Expression = "$.isUpgrade", Type = SelectorType.JsonPath)]
  public int isUpgrade { get; set; }
  [PropertyDefine(Expression = "$.lat", Type = SelectorType.JsonPath)]
  public string lat { get; set; }
  [PropertyDefine(Expression = "$.lon", Type = SelectorType.JsonPath)]
  public string lon { get; set; }
  [PropertyDefine(Expression = "$.queryId", Type = SelectorType.JsonPath)]
  public string queryId { get; set; }
  [PropertyDefine(Expression = "$.supplyCarService", Type = SelectorType.JsonPath)]
  public int supplyCarService { get; set; }
  [PropertyDefine(Expression = "$.freeCarService", Type = SelectorType.JsonPath)]
  public int freeCarService { get; set; }
  [PropertyDefine(Expression = "$.isShenMaCar", Type = SelectorType.JsonPath)]
  public int isShenMaCar { get; set; }
  [PropertyDefine(Expression = "$.supportGetReturn", Type = SelectorType.JsonPath)]
  public int supportGetReturn { get; set; }
  [PropertyDefine(Expression = "$.confirmation", Type = SelectorType.JsonPath)]
  public int confirmation { get; set; }
}

 

起始:

  

/// <summary>
/// 应用程序的主入口点。
/// </summary>
[STAThread]
static void Main()
{
  var site = new Site
  {
    CycleRetryTimes = 1,
    SleepTime = 200,
    Headers = new Dictionary<string, string>()
    {
      {"Accept","application/json, text/javascript, */*; q=0.01" },
      {"Accept-Encoding","gzip, deflate" },
      {"gzip, deflate","zh-CN,zh;q=0.9" },
      {"X-Requested-With","XMLHttpRequest" },
      { "Referer", "http://www.atzuche.com/hz/car/search"},
      { "Connection","keep-alive" },
      { "Content-Type","application/json;charset=UTF-8" },
      { "Host","www.atzuche.com"},
      { "User-Agent","Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/65.0.3325.181 Safari/537.36"}
    }
  };
  List<Request> resList = new List<Request>();

  Request res = new Request();
  //res.PostBody = $"id=7&j=%7B%22createMan%22%3A%2218273159100%22%2C%22createTime%22%3A1518433690000%2C%22row%22%3A5%2C%22siteUserActivityListId%22%3A8553%2C%22siteUserPageRowModuleId%22%3A84959%2C%22topids%22%3A%22%22%2C%22wherePhase%22%3A%221%22%2C%22wherePreferential%22%3A%220%22%2C%22whereUsertype%22%3A%220%22%7D&page={i}&shopid=83106681";//据说是post请求需要
  res.Url = "http://www.atzuche.com/car/searchListMap/2?cityCode=330100&sceneCode=U002&filterCondition%5Blon%5D=120.219294&filterCondition%5Blat%5D=30.259258&filterCondition%5Bseq%5D=4&pageNum=1&pageSize=0";
  res.Method = System.Net.Http.HttpMethod.Get;

  resList.Add(res);

  var spider = DotnetSpider.Core.Spider.Create(site, new QueueDuplicateRemovedScheduler(), new AtzucheProcessor())
    .AddStartRequests(resList.ToArray())//页面抓取整理
    .AddPipeline(new AtzuchePipe());//数据回调

  //----------------------------------
  spider.Monitor = new DotnetSpider.Core.Monitor.NLogMonitor();
  spider.Downloader = new AtzucheDownloader(); //new DotnetSpider.Core.Downloader.HttpClientDownloader();
  spider.ClearSchedulerAfterComplete = false;//爬虫结束后不取消调度器
  //----------------------------------

  spider.ThreadNum = 1;
  spider.Run();

  Console.WriteLine("Press any key to continue...");
  Console.Read();

}

 

这里也可将整个抓取方法当做一个Spider实例单独放置 -> EntitySpider

 

  

/// <summary>
/// 应用程序的主入口点。
/// </summary>
[STAThread]
static void Main()
{
  AtzucheEntitySpider dDengEntitySpider = new AtzucheEntitySpider();
  dDengEntitySpider.AddPageProcessor(new AtzucheProcessor());//控制器
  dDengEntitySpider.AddPipeline(new AtzuchePipe());//回调
  dDengEntitySpider.ThreadNum = 1;
  dDengEntitySpider.Run();
  Console.WriteLine("Press any key to continue...");
  Console.Read();
}

 

Downloader

对目标的请求全部包含着这里,可以根据需要自行设置,下篇将进行自定义Request的应用

public class AtzucheDownloader : BaseDownloader
{
  protected override Page DowloadContent(Request request, ISpider spider)
  { 
    return new HttpClientDownloader().Download(request, spider);
  }
  
//v2.5.0+版本   protected override Task<Page> DowloadContent(Request request, ISpider spider)   {     return new HttpClientDownloader().Download(request, spider);   }
}

 

 

新建爬虫实体类 (如在Main里写入实体爬虫,此方法可省略)

public class AtzucheEntitySpider : EntitySpider
{
  protected override void MyInit(params string[] arguments)
  {
    AddPipeline(new SqlServerEntityPipeline("Server=.;Database=AuzucheSpider;uid=sa;pwd=123;MultipleActiveResultSets=true"));//注意连接字符串中数据库不能带 .  亲测报错。。。
    AddStartUrl("http://www.atzuche.com/car/searchListMap/2?cityCode=330100&sceneCode=U002&filterCondition%5Blon%5D=120.219294&filterCondition%5Blat%5D=30.259258&filterCondition%5Bseq%5D=4&pageNum=1&pageSize=0");
    AddEntityType<AtzucheModel>();//如添加此实体类,框架将会根据此实体类上面的特性选择进行匹配,匹配成功后插入数据库,固可以省略Processor和Pipe,或者不使用此句,通过控制器和回调自定义存储方法
  }

public AtzucheEntitySpider() : base("AuzucheSpider", new Site
{
  CycleRetryTimes = 1,
  SleepTime = 200,
  Headers = new Dictionary<string, string>()
  {
    {"Accept","application/json, text/javascript, */*; q=0.01" },
    {"Accept-Encoding","gzip, deflate" },
    {"gzip, deflate","zh-CN,zh;q=0.9" },
    {"X-Requested-With","XMLHttpRequest" },
    { "Referer", "http://www.atzuche.com/hz/car/search"},
    { "Connection","keep-alive" },
    { "Content-Type","application/json;charset=UTF-8" },
    { "Host","www.atzuche.com"},
    { "User-Agent","Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/65.0.3325.181 Safari/537.36"}
  }
  })
  {
  }

}

 

接下来是处理器:

解析抓取的数据封装到"AtzucheList"内,可Pipe内通过此名称获取处理好的数据。

public class AtzucheProcessor : IPageProcessor
{
  public void Process(Page page, ISpider spider)
  {
    List<AtzucheModel> list = new List<AtzucheModel>();
    var html = page.Selectable.JsonPath("$.data.content").GetValue();
    list = JsonConvert.DeserializeObject<List<AtzucheModel>>(html);
    page.AddResultItem("AtzucheList", list);
  }
}

 

最后是回调,可在此加入保存数据的代码,至此结束。

public class AtzuchePipe : BasePipeline
{
  public override void Process(IEnumerable<ResultItems> resultItems, ISpider spider)
  {
    var result = new List<AtzucheModel>();
    foreach (var resultItem in resultItems)
    {
      Console.WriteLine((resultItem.Results["AtzucheList"] as List<AtzucheModel>).Count);
      foreach (var item in (resultItem.Results["AtzucheList"] as List<AtzucheModel>))
      {
        result.Add(new AtzucheModel()
        {
          carNo = item.carNo
        });
        Console.WriteLine($"{item.carNo}:{item.type} ");
      }
    }
  }
}

 

   结果图:

 

总体来说,此框架对新手还是很友好的,灵活写法可以让我们有较多的方式去实现爬虫,因为这个爬虫比较简单,就先写到这里,未来如果可能,会再尝试使用框架内的多线程、代理等功能,如有心得将继续分享,希望能对跟我一样的新手有所帮助,十分感谢。

 

展开阅读全文
打赏
0
1 收藏
分享
加载中
更多评论
打赏
0 评论
1 收藏
0
分享
返回顶部
顶部