从国家统计局采集最新行政区划分

最近一新项目要用到国内行政区划数据，bing了一下，已有网友提供sql版本数据下载，但在本地查看数据不够新，至少我老家所在市2010年改名儿了这数据也看不到。所以说呢还是自己动手丰衣足食。 使用了JSON.NET

1、

using System;
using System.Collections.Generic;
using System.Text;
using System.Net;
using System.IO;

namespace DivisonsOfPRC
{
    /// <summary>
    /// 演示代码，不建议在生产环境使用，请搜HttpClient
    /// </summary>
    public class Http
    {
        public string GET(string url)
        {
            HttpWebRequest hwr = (HttpWebRequest)WebRequest.Create(url);
            hwr.Method = "GET";
            hwr.CookieContainer = new CookieContainer();
            hwr.Accept = "*/*";
            hwr.Headers.Add(HttpRequestHeader.AcceptEncoding, "gzip, deflate");
            hwr.Headers.Add(HttpRequestHeader.AcceptLanguage, "zh-CN");

            hwr.Referer = "http://www.stats.gov.cn/tjsj/tjbz/xzqhdm/";
            hwr.UserAgent = "Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64; Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729; Media Center PC 6.0; .NET4.0C; .NET4.0E)";
            hwr.AutomaticDecompression = DecompressionMethods.Deflate | DecompressionMethods.GZip;

            HttpWebResponse response = (HttpWebResponse)hwr.GetResponse();
            Stream receiveStream = response.GetResponseStream();
            Encoding encode = System.Text.Encoding.UTF8;
            StreamReader readStream = new StreamReader(receiveStream, encode);

            String linesOfHTML = readStream.ReadToEnd();
            //System.Console.WriteLine(linesOfHTML);
            //System.Console.ReadKey();

            receiveStream.Close();
            response.Close();
            readStream.Close();
            return linesOfHTML;
        }
    }
}

生成JSON.NET

using System;
using System.Collections.Generic;
using System.Text;
using System.Net;
using System.IO;

namespace DivisonsOfPRC
{
    /// <summary>
    /// 演示代码，不建议在生产环境使用，请搜HttpClient
    /// </summary>
    public class Http
    {
        public string GET(string url)
        {
            HttpWebRequest hwr = (HttpWebRequest)WebRequest.Create(url);
            hwr.Method = "GET";
            hwr.CookieContainer = new CookieContainer();
            hwr.Accept = "*/*";
            hwr.Headers.Add(HttpRequestHeader.AcceptEncoding, "gzip, deflate");
            hwr.Headers.Add(HttpRequestHeader.AcceptLanguage, "zh-CN");

            hwr.Referer = "http://www.stats.gov.cn/tjsj/tjbz/xzqhdm/";
            hwr.UserAgent = "Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64; Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729; Media Center PC 6.0; .NET4.0C; .NET4.0E)";
            hwr.AutomaticDecompression = DecompressionMethods.Deflate | DecompressionMethods.GZip;

            HttpWebResponse response = (HttpWebResponse)hwr.GetResponse();
            Stream receiveStream = response.GetResponseStream();
            Encoding encode = System.Text.Encoding.UTF8;
            StreamReader readStream = new StreamReader(receiveStream, encode);

            String linesOfHTML = readStream.ReadToEnd();
            //System.Console.WriteLine(linesOfHTML);
            //System.Console.ReadKey();

            receiveStream.Close();
            response.Close();
            readStream.Close();
            return linesOfHTML;
        }
    }
}

测试用例：

using System;
using System.Collections.Generic;
using System.Text;
using Newtonsoft.Json;

namespace DivisonsOfPRC
{
    class Program
    {
        static void Main(string[] args)
        {
            string[] directs = new string[] { "北京市", "天津市", "上海市", "重庆市" };
            Http http = new Http();
            //1.采集html到本地
            string html = http.GET("http://www.stats.gov.cn/tjsj/tjbz/xzqhdm/201401/t20140116_501070.html");
            //System.IO.File.WriteAllText(AppDomain.CurrentDomain.BaseDirectory + "dop.html", html, Encoding.UTF8);

            //2.分析html
            //可以用HtmlAgilePack找到行政区划数据的核心html
            //我这里就不演示HAP怎么使用了，咱简单点儿，就用字符串分析截取
            int startIndex = html.LastIndexOf("TRS_Editor");
            startIndex = html.IndexOf("<p class=\"", startIndex);
            html = html.Substring(startIndex);
            html = html.Substring(0, html.IndexOf("</div>"));

            //真正分析省市区逻辑
            string[] lines = html.Split(new string[] { "</p>" }, StringSplitOptions.RemoveEmptyEntries);
            string code = null, name= null,line = null;

            List<Node> nodes = new List<Node>();
            Node PrevCity = null;
            Node PrevProvince = null;
            for (int i = 0; i < lines.Length; i++)
            {
                Node nod = new Node();
                line = ExtractHtml(lines[i], "align=\"justify\">", "");
                code = line.Substring(0, line.IndexOf("&"));
                name = line.Substring(line.LastIndexOf(";")+1).Trim();
                nod.code = code;
                nod.name = name;

                int timesOfSpaceOccure = CountString(line, "&nbsp;");
                nod.spaces = timesOfSpaceOccure;
                if (timesOfSpaceOccure == 3)
                {
                    nodes.Add(nod);
                    PrevProvince = nod;
                    PrevCity = null;
                }
                else
                {
                    if (timesOfSpaceOccure > PrevProvince.spaces)
                    {
                        //下一级别
                        if (PrevCity != null && timesOfSpaceOccure > PrevCity.spaces)
                        {
                            if (PrevCity.cell == null)
                            {
                                PrevCity.cell = new List<Node>();
                            }
                            PrevCity.cell.Add(nod);
                        }
                        else
                        {
                            //市
                            if (PrevProvince.cell == null)
                            {
                                PrevProvince.cell = new List<Node>();
                            }
                            PrevProvince.cell.Add(nod);
                            PrevCity = nod;
                        }
                    }
                }

            }
            JsonSerializerSettings settings = new JsonSerializerSettings();
            settings.NullValueHandling = NullValueHandling.Ignore;
            string json2 = JsonConvert.SerializeObject(nodes, Newtonsoft.Json.Formatting.None, settings);
            System.IO.File.WriteAllText(AppDomain.CurrentDomain.BaseDirectory + "data.js", json2, Encoding.UTF8);
        }

        static string ExtractHtml(string source, string prefix, string suffix)
        {
            if (string.IsNullOrEmpty(source))
            {
                return null;
            }
            int startIndex = source.IndexOf(prefix);
            if (startIndex == -1)
            {
                return string.Empty;
            }
            startIndex = startIndex + prefix.Length;
            int endIndex = source.Length;
            if (!string.IsNullOrEmpty(suffix))
            {
                endIndex = source.IndexOf(suffix, startIndex);
                if (endIndex == -1)
                {
                    //suffix not found
                    return string.Empty;
                }
            }
            return source.Substring(startIndex, endIndex - startIndex);
            //return null;
        }

        static int CountString(string source, string search)
        {
            int count = 0;
            int startIndex = 0;
            startIndex = source.IndexOf(search);
            while (startIndex != -1)
            {
                startIndex = source.IndexOf(search, startIndex + search.Length);
                count++;
            }
            return count;
        }
    }
}

留下备用