[开源 .NET 跨平台 Crawler 数据采集 爬虫框架: DotnetSpider] [三] 配置式爬虫

[DotnetSpider 系列目录]

• 一、初衷与架构设计
• 二、基本使用
• 三、配置式爬虫
• 四、JSON数据解析与配置系统
• 五、如何做全站采集

上一篇介绍的基本的使用方式，自由度很高，但是编写的代码相对就多了。而我所在的行业其实大部分都是定题爬虫, 只需要采集指定的页面并结构化数据。为了提高开发效率, 我实现了利用实体配置的方式来实现爬虫

创建 Console 项目

利用NUGET添加包

DotnetSpider2.Extension

定义配置式数据对象

• 数据对象必须继承 ISpiderEntity
• Schema 定义数据名称、表名及表名后缀
• Indexes 定义数据表的主键、唯一索引、索引
• EntitySelector 定义从页面数据中抽取数据对象的规则

定义一个原始的数据对象类

public class Product : ISpiderEntity
{
}

使用Chrome打开京东商品页 http://list.jd.com/list.html?cat=9987,653,655&page=2&JL=6_0_0&ms=5#J_main

1. 使用快捷键F12打开开发者工具
2. 选中一个商品，并观察Html结构

          

我们发现每个商品都在class为gl-i-wrap j-sku-item的DIV下面，因此添加EntitySelector到数据对象Product的类名上面。（ XPath的写法不是唯一的，不熟悉的可以去W3CSCHOLL学习一下, 框架也支持使用Css甚至正则来选择出正确的Html片段）。 

　　　 [EntitySelector(Expression = "//li[@class='gl-item']/div[contains(@class,'j-sku-item')]")] public class Product : ISpiderEntity 

3. 添加数据库及索引信息

   [Schema("test", "sku", TableSuffix.Today)]
   [EntitySelector(Expression = "//li[@class='gl-item']/div[contains(@class,'j-sku-item')]")]
   [Indexes(Index = new[] { "category" }, Unique = new[] { "category,sku", "sku" })]
   public class Product : ISpiderEntity

4. 假设你需要采集SKU信息，观察HTML结构，计算出相对的XPath, 为什么是相对XPath？因为EntitySelector已经把HTML截成片段了，内部的Html元素查询都是相对于EntitySelector查询出来的元素。最后再加上数据库中列的信息

   [Schema("test", "sku", TableSuffix.Today)]
   [EntitySelector(Expression = "//li[@class='gl-item']/div[contains(@class,'j-sku-item')]")]
   [Indexes(Index = new[] { "category" }, Unique = new[] { "category,sku", "sku" })]
   public class Product : ISpiderEntity
   {
        [StoredAs("sku", DataType.String, 25)]
        [PropertySelector(Expression = "./@data-sku")]
        public string Sku { get; set; }
    }

5. 爬虫内部，链接是通过Request对象来存储信息的，构造Request对象时可以添加额外的属性值，这时候允许数据对象从Request的额外属性值中查询数据

   [StoredAs("category", DataType.String, 20)]
   [PropertySelector(Expression = "name", Type = SelectorType.Enviroment)]
   public string CategoryName { get; set; }

配置爬虫（继承EntitySpiderBuilder）

    protected override EntitySpider GetEntitySpider()
    {
        EntitySpider context = new EntitySpider(new Site
        {
            //HttpProxyPool = new HttpProxyPool(new KuaidailiProxySupplier("快代理API"))
        })
        {
            UserId = "DotnetSpider",
            TaskGroup = "JdSkuSampleSpider"
        };
        context.SetThreadNum(1);
        context.SetIdentity("JD_sku_store_test_" + DateTime.Now.ToString("yyyy_MM_dd_hhmmss"));
        context.AddEntityPipeline(new MySqlEntityPipeline("Database='test';Data Source=localhost;User ID=root;Password=1qazZAQ!;Port=3306"));
        context.AddStartUrl("http://list.jd.com/list.html?cat=9987,653,655&page=2&JL=6_0_0&ms=5#J_main", new Dictionary<string, object> { { "name", "手机" }, { "cat3", "655" } });
        context.AddEntityType(typeof(Product), new TargetUrlExtractor
        {
            Region = new BaseSelector { Type = SelectorType.XPath, Expression = "//span[@class="p-num"]" },
            Patterns = new List<string> { @"&page=[0-9]+&" }
        });
        return context;
    }

1. 其中AddStartUrl第二个参数Dictionary<string, object>就是用于Enviroment查询的数据

2. 配置Scheduler: 默认是使用内存Queue做Url调度，如果想使用多台机器分布式采集则需要配置为RedisScheduler

   context.SetScheduler(new RedisScheduler
    {
        Host = "",
        Password = "",
        Port = 6379
    });

3. 在添加数据对象时，可以配置数据链接的合法性验证。用在一个网站采集多种链接时映射到不同的数据对象。同时此验证会抽取当前页面中符合规则的Url加入到Scheduler中继续采集。

   context.AddEntityType(typeof(Product), new TargetUrlExtractor
   {
        Region = new BaseSelector { Type = SelectorType.XPath, Expression = "//span[@class="p-num"]" },
       Patterns = new List<string> { @"&page=[0-9]+&" }
   });

          

4. 添加一个MySql的数据管道，只需要配置好连接字符串即可

   context.AddEntityPipeline(new MySqlEntityPipeline("Database='test';Data Source=localhost;User ID=root;Password=1qazZAQ!;Port=3306"));

完整代码

public class JdSkuSampleSpider : EntitySpiderBuilder
    {
        protected override EntitySpider GetEntitySpider()
        {
            EntitySpider context = new EntitySpider(new Site
            {
                //HttpProxyPool = new HttpProxyPool(new KuaidailiProxySupplier("快代理API"))
            })
            {
                UserId = "DotnetSpider",
                TaskGroup = "JdSkuSampleSpider"
            };
            context.SetThreadNum(1);
            context.SetIdentity("JD_sku_store_test_" + DateTime.Now.ToString("yyyy_MM_dd_hhmmss"));
            context.AddEntityPipeline(new MySqlEntityPipeline("Database='test';Data Source=localhost;User ID=root;Password=1qazZAQ!;Port=3306"));
            context.AddStartUrl("http://list.jd.com/list.html?cat=9987,653,655&page=2&JL=6_0_0&ms=5#J_main", new Dictionary<string, object> { { "name", "手机" }, { "cat3", "655" } });
            context.AddEntityType(typeof(Product), new TargetUrlExtractor
            {
                Region = new BaseSelector { Type = SelectorType.XPath, Expression = "//span[@class="p-num"]" },
                Patterns = new List<string> { @"&page=[0-9]+&" }
            });
            return context;
        }

        [Schema("test", "sku", TableSuffix.Today)]
        [EntitySelector(Expression = "//li[@class='gl-item']/div[contains(@class,'j-sku-item')]")]
        [Indexes(Index = new[] { "category" }, Unique = new[] { "category,sku", "sku" })]
        public class Product : ISpiderEntity
        {
            [StoredAs("sku", DataType.String, 25)]
            [PropertySelector(Expression = "./@data-sku")]
            public string Sku { get; set; }

            [StoredAs("category", DataType.String, 20)]
            [PropertySelector(Expression = "name", Type = SelectorType.Enviroment)]
            public string CategoryName { get; set; }

            [StoredAs("cat3", DataType.String, 20)]
            [PropertySelector(Expression = "cat3", Type = SelectorType.Enviroment)]
            public int CategoryId { get; set; }

            [StoredAs("url", DataType.Text)]
            [PropertySelector(Expression = "./div[1]/a/@href")]
            public string Url { get; set; }

            [StoredAs("commentscount", DataType.String, 32)]
            [PropertySelector(Expression = "./div[5]/strong/a")]
            public long CommentsCount { get; set; }

            [StoredAs("shopname", DataType.String, 100)]
            [PropertySelector(Expression = ".//div[@class='p-shop']/@data-shop_name")]
            public string ShopName { get; set; }

            [StoredAs("name", DataType.String, 50)]
            [PropertySelector(Expression = ".//div[@class='p-name']/a/em")]
            public string Name { get; set; }

            [StoredAs("venderid", DataType.String, 25)]
            [PropertySelector(Expression = "./@venderid")]
            public string VenderId { get; set; }

            [StoredAs("jdzy_shop_id", DataType.String, 25)]
            [PropertySelector(Expression = "./@jdzy_shop_id")]
            public string JdzyShopId { get; set; }

            [StoredAs("run_id", DataType.Date)]
            [PropertySelector(Expression = "Monday", Type = SelectorType.Enviroment)]
            public DateTime RunId { get; set; }

            [PropertySelector(Expression = "Now", Type = SelectorType.Enviroment)]
            [StoredAs("cdate", DataType.Time)]
            public DateTime CDate { get; set; }
        }
    }

运行爬虫

public class Program
{
    public static void Main(string[] args)
    {
        JdSkuSampleSpider spiderBuilder = new JdSkuSampleSpider();
        spiderBuilder.Run("rerun");
    }
}

不到100行代码完成一个爬虫，是不是异常的简单？