package aa;
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
/**
* 全国省市县镇村数据爬取
* @author liushaofeng
* @date 2015-10-11 上午12:19:39
* @version 1.0.0
*/
public class JsoupTest
{
private static Map<Integer, String> cssMap = new HashMap<Integer, String>();
static
{
cssMap.put(1, "provincetr");// 省
cssMap.put(2, "citytr");// 市
cssMap.put(3, "countytr");// 县
cssMap.put(4, "towntr");// 镇
cssMap.put(5, "villagetr");// 村
}
public static void main(String[] args) throws Exception
{
int level = 1;
// 获取全国各个省级信息
Document connect = connect("http://www.stats.gov.cn/tjsj/tjbz/tjyqhdmhcxhfdm/2016/");
Elements rowProvince = connect.select("tr." + cssMap.get(level));
for (Element provinceElement : rowProvince)// 遍历每一行的省份城市
{
Elements select = provinceElement.select("a");
for (Element province : select)// 每一个省份(四川省)
{
parseNextLevel(province, level + 1);
}
}
}
private static void parseNextLevel(Element parentElement, int level) throws Exception
{
try
{
Thread.sleep(2);//睡眠一下,否则可能出现各种错误状态码
} catch (InterruptedException e)
{
e.printStackTrace();
}
Document doc = connect(parentElement.attr("abs:href"));
if (doc != null)
{
Elements newsHeadlines = doc.select("tr." + cssMap.get(level));//
// 获取表格的一行数据
for (Element element : newsHeadlines)
{
printInfo(element, level + 1);
Elements select = element.select("a");// 在递归调用的时候,这里是判断是否是村一级的数据,村一级的数据没有a标签
if (select.size() != 0)
{
parseNextLevel(select.last(), level + 1);
}
}
}
}
/**
* 写一行数据到数据文件中去
* @param element 爬取到的数据元素
* @param level 城市级别
* @throws Exception
*/
private static void printInfo(Element element, int level) throws Exception
{
String value=element.select("td").last().text();
String code=element.select("td").first().text();
System.out.println(code+","+value+","+level);
}
private static Document connect(String url)
{
if (url == null || url.isEmpty())
{
throw new IllegalArgumentException("The input url('" + url + "') is invalid!");
}
try
{
return Jsoup.connect(url).timeout(100 * 1000).get();
} catch (IOException e)
{
e.printStackTrace();
return null;
}
}
}
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
/**
* 全国省市县镇村数据爬取
* @author liushaofeng
* @date 2015-10-11 上午12:19:39
* @version 1.0.0
*/
public class JsoupTest
{
private static Map<Integer, String> cssMap = new HashMap<Integer, String>();
static
{
cssMap.put(1, "provincetr");// 省
cssMap.put(2, "citytr");// 市
cssMap.put(3, "countytr");// 县
cssMap.put(4, "towntr");// 镇
cssMap.put(5, "villagetr");// 村
}
public static void main(String[] args) throws Exception
{
int level = 1;
// 获取全国各个省级信息
Document connect = connect("http://www.stats.gov.cn/tjsj/tjbz/tjyqhdmhcxhfdm/2016/");
Elements rowProvince = connect.select("tr." + cssMap.get(level));
for (Element provinceElement : rowProvince)// 遍历每一行的省份城市
{
Elements select = provinceElement.select("a");
for (Element province : select)// 每一个省份(四川省)
{
parseNextLevel(province, level + 1);
}
}
}
private static void parseNextLevel(Element parentElement, int level) throws Exception
{
try
{
Thread.sleep(2);//睡眠一下,否则可能出现各种错误状态码
} catch (InterruptedException e)
{
e.printStackTrace();
}
Document doc = connect(parentElement.attr("abs:href"));
if (doc != null)
{
Elements newsHeadlines = doc.select("tr." + cssMap.get(level));//
// 获取表格的一行数据
for (Element element : newsHeadlines)
{
printInfo(element, level + 1);
Elements select = element.select("a");// 在递归调用的时候,这里是判断是否是村一级的数据,村一级的数据没有a标签
if (select.size() != 0)
{
parseNextLevel(select.last(), level + 1);
}
}
}
}
/**
* 写一行数据到数据文件中去
* @param element 爬取到的数据元素
* @param level 城市级别
* @throws Exception
*/
private static void printInfo(Element element, int level) throws Exception
{
String value=element.select("td").last().text();
String code=element.select("td").first().text();
System.out.println(code+","+value+","+level);
}
private static Document connect(String url)
{
if (url == null || url.isEmpty())
{
throw new IllegalArgumentException("The input url('" + url + "') is invalid!");
}
try
{
return Jsoup.connect(url).timeout(100 * 1000).get();
} catch (IOException e)
{
e.printStackTrace();
return null;
}
}
}