最近lz在写抓工商公式系统(http://www.gsxt.gov.cn/index.html)的爬虫，其中的难点就是在怎么过极验验证码，搞的我不要不要的！如下：
简直是各种坑，被搞的死去活来以后还是解决了。现在回到主题！
我们不是要抓工商公式系统的数据吗？所以我们先建两个实体BaseInfo（基本信息）和LegInfo（股东信息）
public partial class BaseInfo
{
public BaseInfo()
{ }
public BaseInfo(string html)
{ }
#region Model
/// <summary>
///
/// </summary>
public int Id
{
get;
set;
}
/// <summary>
/// 成立日期
/// </summary>
public string ApprDate
{
get;
set;
}
/// <summary>
/// 公司全称
/// </summary>
public string EntName
{
get;
set;
}
/// <summary>
/// 公司类型
/// </summary>
public string EntType
{
get;
set;
}
/// <summary>
/// 住所
/// </summary>
public string Dom
{
get;
set;
}
/// <summary>
/// 核准日期
/// </summary>
public string EstDate
{
get;
set;
}
/// <summary>
/// 法人
/// </summary>
public string Lerep
{
get;
set;
}
/// <summary>
/// 营业期限自
/// </summary>
public string OpFrom
{
get;
set;
}
/// <summary>
/// 营业期限至
/// </summary>
public string OpTo
{
get;
set;
}
/// <summary>
/// 经营范围
/// </summary>
public string OpScope
{
get;
set;
}
/// <summary>
/// 注册号
/// </summary>
public string RegNo
{
get;
set;
}
/// <summary>
/// 登记机关
/// </summary>
public string RegOrg
{
get;
set;
}
/// <summary>
/// 登记状态
/// </summary>
public string RegState
{
get;
set;
}
/// <summary>
/// 注册资本
/// </summary>
public string RegCap
{
set;
get;
}
/// <summary>
/// 行业领域
/// </summary>
public string IndcodeNameLv2 { get; set; }
/// <summary>
/// 省
/// </summary>
public string Province { get; set; }
/// <summary>
/// 市
/// </summary>
public string City { get; set; }
/// <summary>
/// 网址
/// </summary>
public string Weburl { get; set; }
/// <summary>
/// 评级
/// </summary>
public string Rating { get; set; }
/// <summary>
///
/// </summary>
public int CompanyInfoId
{
get;
set;
}
#endregion Model
#region 导航属性
public virtual CompanyInfo CompanyInfo { get; set; }
#endregion
}
public partial class LegInfo : SpiderModel
{
public LegInfo()
{ }
#region Model
/// <summary>
///
/// </summary>
public int Id
{
get;
set;
}
/// <summary>
///
/// </summary>
public string BlicNo
{
get;
set;
}
/// <summary>
///
/// </summary>
public string BlicType
{
get;
set;
}
/// <summary>
///
/// </summary>
public string ItemId
{
get;
set;
}
/// <summary>
///
/// </summary>
public string Inv
{
get;
set;
}
/// <summary>
///
/// </summary>
public string InvType
{
get;
set;
}
/// <summary>
///
/// </summary>
public int CompanyInfoId
{
get;
set;
}
public string CreateTimeStr { get; set; }
public string MoneyRange { get; set; }
public string Renjiao { get; set; }
#endregion Model
#region 导航属性
/// <summary>
/// 导航属性，公司。
/// </summary>
public virtual CompanyInfo CompanyInfo { get; set; }
#endregion
}
先破解验证码，获取需要查询的公司的URL，然后抓取公司详情也的HTML（过程略）；关键代码有两个方法GetBaseInfo和GetLegInfoes
如下:
/// <summary>
/// 获取工商基本数据
/// </summary>
/// <param name="url"></param>
/// <param name="companyInfo"></param>
public static BaseInfo GetGsxtInfo(string url, out string html)
{
HttpItem item = new HttpItem()
{
URL = url,//URL 必需项
Method = "get",
Referer = "http://www.gsxt.gov.cn/corp-query-homepage.html",
Timeout =
};
html = GetHtml(item);
string companyName = GetXpathNode(html, "//h1[@class=\"fullName\"]");
string companyNo = GetXpathNode(html, "//*[@class=\"nameBoxColor\"]");
if (companyNo != "")
{
//CompanyInfo companyInfo = new CompanyInfo();
//companyInfo.CompanyName = companyName;
//companyInfo.CompanyNo = companyNo;
//companyInfo.State = 1;
//companyInfo.AddTime = DateTime.Now;
//companyInfo.NextTime = DateTime.Now;
//companyInfo.BaseInfos = new List<BaseInfo>();
var baseInfo = new BaseInfo();
baseInfo.EntName = companyName;
baseInfo.RegNo = companyNo;
baseInfo.ApprDate = GetXpathNode(html, "//*[@class=\"companyDetail clearfix\"]/span[4]/span[1]");
baseInfo.RegState = GetXpathNode(html, "//*[@class=\"companyStatus\"]");
baseInfo.EntType = GetXpathNode(html, "//div[@class=\"overview\"]/dl[3]/dd[1]");
baseInfo.Lerep = GetXpathNode(html, "//div[@class=\"overview\"]/dl[4]/dd[1]");
baseInfo.RegCap = GetXpathNode(html, "//div[@class=\"overview\"]/dl[5]/dd[1]");
baseInfo.OpFrom = GetXpathNode(html, "//div[@class=\"overview\"]/dl[7]/dd[1]");
baseInfo.OpTo = GetXpathNode(html, "//div[@class=\"overview\"]/dl[8]/dd[1]");
baseInfo.RegOrg = GetXpathNode(html, "//*[@class=\"companyDetail clearfix\"]/span[3]/span[1]");
baseInfo.EstDate = GetXpathNode(html, "//div[@class=\"overview\"]/dl[10]/dd[1]");
baseInfo.Dom = GetXpathNode(html, "//div[@class=\"overview\"]/dl[12]/dd[1]");
baseInfo.OpScope = GetXpathNode(html, "//div[@class=\"overview\"]/dl[13]/dd");
return baseInfo;
}
else
{
return null;
}
}
/// <summary>
/// 股东信息
/// </summary>
/// <param name="html"></param>
/// <param name="companyInfo"></param>
/// <param name="draw"></param>
/// <param name="start"></param>
, )
{
string url = string.Format("http://www.gsxt.gov.cn{0}", GetFirstInnerText(html, "var shareholderUrl = \"", "\""));
//HttpHelper http = new HttpHelper();
HttpItem item = new HttpItem()
{
URL = url,//URL 必需项
Method = "post",
Referer = "http://www.gsxt.gov.cn/corp-query-search-1.html",
Postdata = string.Format("draw={0}&start={1}&length=5", draw, start),
ContentType = "application/x-www-form-urlencoded",
Timeout =
};
string rhtml = GetHtml(item); if (rhtml.Equals("")) { return; }
var legInfoesListPage = JObject.Parse(rhtml);
var legInfoesListList = legInfoesListPage["data"].ToList();
//删除数据库中的数据
//if (draw == 1)
//{
// if (legInfoesListList.Count > 0)
// {
// foreach (var leginfo in companyInfo.LegInfos.ToList())
// {
// companyInfo.LegInfos.Remove(leginfo);
// }
// }
//}
//add
foreach (var legInfoes in legInfoesListList)
{
reflegInfos.Add(new LegInfo { BlicNo = GetText(legInfoes["bLicNo"].ToString().Replace("\"", "")), BlicType = legInfoes["blicType_CN"].ToString().Replace("\"", ""), ItemId = legInfoes["invId"].ToString().Replace("\"", ""), Inv = legInfoes["inv"].ToString().Replace("\"", ""), InvType = GetText(legInfoes["invType_CN"].ToString().Replace("\"", "")) });
}
///下页
if (int.Parse(legInfoesListPage["totalPage"].ToString()) > draw) //获取下一页的数据
{
draw++;
start += ;
Console.WriteLine(string.Format("查询股东信息第{0}页", draw));
GetLegInfoes(html, ref reflegInfos, draw, start);
}
}
到这里为了完成任务写的代码，如果需要对代码让它更加优美，就需要用IOC的模式是重构它
先创建父类：
public class SpiderModel
{
public SpiderModel()
{
}
public SpiderModel(JToken token)
{
ToObje(token);
}
public virtual void ToObje(JToken token)
{
}
public virtual SpiderModel ToObje(string html)
{
return new SpiderModel();
}
/// <summary>
/// Xpath获取值
/// </summary>
/// <param name="html"></param>
/// <param name="xpath"></param>
/// <returns></returns>
public static string GetXpathNode(string html, string xpath)
{
string result = string.Empty;
#region Xpath提取
try
{
HtmlDocument htmlDoc = new HtmlDocument();
htmlDoc.LoadHtml(html);
HtmlNode node = htmlDoc.DocumentNode.SelectSingleNode(xpath);
if (node != null)
{
result = node.InnerHtml;
result = new Regex("\\t", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(result, string.Empty);
result = TextRemover.RemoveHTML(result);//去除HTML标签
result = TextRemover.RemoveWhiteSpace(result).Trim();//去空白字符
}
}
catch (Exception)
{
}
return result;
#endregion
}
public static string GetText(string result)
{
result = new Regex(@"<(p|br)[^<]*>", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(result, "[$1]");
result = new Regex("\\[p]", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(result, "\r\n\r\n");
result = new Regex("\\[br]", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(result, "\r\n");
result = new Regex("\\t", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(result, " ");
result = TextRemover.RemoveHTML(result);//去除HTML标签
result = result.Replace("+", "");
result = result.Trim();
result = RegexHelper.RegexFilter(result.ToString().Replace("\"", ""), "([a-zA-Z0-9]+)", false, RegexOptions.None);
return result;
}
}
然后实体BaseInfo（基本信息）和LegInfo（股东信息）继承自SpiderModel
然后给BaseInfo（基本信息）和LegInfo（股东信息）重写函数
BaseInfo：
public override SpiderModel ToObje(string html)
{
string companyName = GetXpathNode(html, "//h1[@class=\"fullName\"]");
string companyNo = GetXpathNode(html, "//*[@class=\"nameBoxColor\"]");
if (companyNo != "")
{
EntName = companyName;
RegNo = companyNo;
ApprDate = GetXpathNode(html, "//*[@class=\"companyDetail clearfix\"]/span[4]/span[1]");
RegState = GetXpathNode(html, "//*[@class=\"companyStatus\"]");
EntType = GetXpathNode(html, "//div[@class=\"overview\"]/dl[3]/dd[1]");
Lerep = GetXpathNode(html, "//div[@class=\"overview\"]/dl[4]/dd[1]");
RegCap = GetXpathNode(html, "//div[@class=\"overview\"]/dl[5]/dd[1]");
OpFrom = GetXpathNode(html, "//div[@class=\"overview\"]/dl[7]/dd[1]");
OpTo = GetXpathNode(html, "//div[@class=\"overview\"]/dl[8]/dd[1]");
RegOrg = GetXpathNode(html, "//*[@class=\"companyDetail clearfix\"]/span[3]/span[1]");
EstDate = GetXpathNode(html, "//div[@class=\"overview\"]/dl[10]/dd[1]");
Dom = GetXpathNode(html, "//div[@class=\"overview\"]/dl[12]/dd[1]");
OpScope = GetXpathNode(html, "//div[@class=\"overview\"]/dl[13]/dd");
}
return this;
}
LegInfo：
public override void ToObje(JToken token)
{
BlicNo = SpiderModel.GetText(token["bLicNo"].ToString().Replace("\"", ""));
BlicType = token["blicType_CN"].ToString().Replace("\"", "");
ItemId = token["invId"].ToString().Replace("\"", "");
Inv = token["inv"].ToString().Replace("\"", "");
InvType = GetText(token["invType_CN"].ToString().Replace("\"", ""));
}
最后要一个IOC管理类
public class SpiderManage
{
public SpiderManage(HttpItem item)
{
this.Item = item;
}
public SpiderManage(HttpItem item,SpiderModel spiderModel)
{
this.Item = item;
this.SpiderModel = spiderModel;
}
public string Html { get; set; }
public HttpItem Item { get; set; }
public List<SpiderModel> SpiderModelList{ get; set; }
public SpiderModel SpiderModel { get; set; }
public virtual string GetHtml()
{
;
)
{
i--;
HttpHelper http = new HttpHelper();
HttpResult result;
object oj = new object();
lock (oj)
{
Thread.Sleep();
result = http.GetHtml(Item);
}
if (result.StatusCode == System.Net.HttpStatusCode.OK)
{
string rhtml = result.Html;
if (!rhtml.Equals("<script>window.location.href='/index/invalidLink'</script>"))
{
Html = result.Html;
return Html;
}
}
}
Html = "";
return Html;
}
public SpiderModel GetOjb()
{
return SpiderModel.ToObje(Html);
}
public void toList()
{
// SpiderModelList
//SpiderModel
}
}
最后面我就只要 List<SpiderManage> sManageList用于保存对象就可以了
重新改写前面的GetBaseInfo和GetLegInfoes函数。
public static List<SpiderManage> sManageList = new List<SpiderManage>();
//////////////////////////////////////////////
, )
{
string url = string.Format("http://www.gsxt.gov.cn{0}", GetFirstInnerText(html, "var shareholderUrl = \"", "\""));
HttpItem item = new HttpItem()
{
URL = url,//URL 必需项
Method = "post",
Referer = "http://www.gsxt.gov.cn/corp-query-search-1.html",
Postdata = string.Format("draw={0}&start={1}&length=5", draw, start),
ContentType = "application/x-www-form-urlencoded",
Timeout =
};
SpiderManage sManage = new SpiderManage(item);
sManage.SpiderModel = new LegInfo();
sManage.GetHtml();
string rhtml = sManage.Html;
if (rhtml.Equals(""))
{ return; }
var legInfoesListPage = JObject.Parse(rhtml);
sManageList.Add(sManage);
//var legInfoesListList = legInfoesListPage["data"].ToList();
//删除数据库中的数据
//add
//foreach (var legInfoes in legInfoesListList)
// {
// reflegInfos.Add(new LegInfo { BlicNo = GetText(legInfoes["bLicNo"].ToString().Replace("\"", "")), BlicType = legInfoes["blicType_CN"].ToString().Replace("\"", ""), ItemId = legInfoes["invId"].ToString().Replace("\"", ""), Inv = legInfoes["inv"].ToString().Replace("\"", ""), InvType = GetText(legInfoes["invType_CN"].ToString().Replace("\"", "")) });
// }
///下页
while (int.Parse(legInfoesListPage["totalPage"].ToString()) > draw) //获取下一页的数据
{
draw++;
start += ;
sManageList.Add(new SpiderManage(new HttpItem()
{
URL = url,//URL 必需项
Method = "post",
Referer = "http://www.gsxt.gov.cn/corp-query-search-1.html",
Postdata = string.Format("draw={0}&start={1}&length=5", draw, start),
ContentType = "application/x-www-form-urlencoded",
Timeout =
}, new LegInfo()));
//Console.WriteLine(string.Format("查询股东信息第{0}页", draw));
//GetLegInfoes(html, ref reflegInfos, draw, start);
}
}
后面怎么用就不讨论了，只要是把sManageList拿过去调度分配抓取就可以了