快捷導(dǎo)航

Java實(shí)現(xiàn)獲取行政區(qū)劃的示例代碼

更新時(shí)間：2023年03月13日 08:34:13 作者：歸去來(lái) 兮

這篇文章主要為大家詳細(xì)介紹了如何利用Java語(yǔ)言實(shí)現(xiàn)獲取行政區(qū)劃的功能，文中的示例代碼講解詳細(xì)，感興趣的小伙伴可以跟隨小編一起學(xué)習(xí)游戲

背景

公司的行政區(qū)劃代碼有問題，有的沒有街道信息，有的關(guān)聯(lián)信息有誤，然后找到了國(guó)家的網(wǎng)站國(guó)家統(tǒng)計(jì)局-行政區(qū)劃，這個(gè)里面是包含了所有的行政信息，但是全是html頁(yè)面，這個(gè)就沒法自動(dòng)獲取了，只能去爬取這個(gè)數(shù)據(jù)了，java語(yǔ)音有三方類庫(kù)Jsoup，他是一個(gè)仿瀏覽器的三方庫(kù)，可以通過(guò)他來(lái)獲取頁(yè)面信息。

一、導(dǎo)入jar包

下面是筆者用到的全部jar包

	<dependency>
        <groupId>org.apache.poi</groupId>
        <artifactId>poi-ooxml</artifactId>
        <version>3.9</version>
    </dependency>
    <dependency>
        <groupId>com.google.guava</groupId>
        <artifactId>guava</artifactId>
        <version>30.1.1-jre</version>
    </dependency>
    <dependency>
        <groupId>cn.hutool</groupId>
        <artifactId>hutool-json</artifactId>
        <version>5.4.0</version>
    </dependency>
    <dependency>
        <groupId>com.alibaba</groupId>
        <artifactId>fastjson</artifactId>
        <version>1.2.44</version>
    </dependency>

    <dependency>
        <groupId>org.jsoup</groupId>
        <artifactId>jsoup</artifactId>
        <version>1.14.3</version>
    </dependency>

    <dependency>
        <groupId>org.apache.httpcomponents</groupId>
        <artifactId>httpclient</artifactId>
        <version>4.5.5</version>
    </dependency>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-web</artifactId>
        <version>2.5.4</version>
    </dependency>

這里說(shuō)下，maven配置要從阿里云下載jar，若是從中央倉(cāng)庫(kù)將會(huì)非常的慢。

二、代碼展示

這里是代碼的展示，筆者是網(wǎng)上搜的代碼改造的，不然網(wǎng)站有反爬，大概爬取2000條左右就會(huì)中斷，筆者加了延時(shí)這樣就避開了反爬（可能還有別的規(guī)避措施）。這里爬取的是4級(jí)行政區(qū)劃：省、市、區(qū)縣、街道

package com.cheng.controller;
import org.apache.poi.ss.usermodel.Row;
import org.apache.poi.xssf.streaming.SXSSFSheet;
import org.apache.poi.xssf.streaming.SXSSFWorkbook;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.select.Elements;

import java.io.FileOutputStream;
import java.io.IOException;
import java.net.ConnectException;
import java.net.SocketTimeoutException;
import java.util.*;

/**

@author pcc

@version 1.0.0

@className JsoupTest

@date 2023-03-02 10:39
*/
public class JsoupTestPluMdm {
static int i = 1;

static String url1 = “http://www.stats.gov.cn/tjsj/tjbz/tjyqhdmhcxhfdm/2022”;
static String url2 = “”;
public static void main(String[] args) throws IOException {

 try{
     List<Map<String,String>> listMap = new ArrayList<>();

     Document document = Jsoup.connect(url1)
             .header("Cookie", "wzws_sessionid=oGQAAyWBMmNlMWZkgjdlZDJkMIAyMjEuMjM4LjEzMi41MA==; SF_cookie_1=15502425; wzws_cid=6e8cdc0aea81349b05c8a0b6c05cd7204b6e0f10e5a48d462175473d23abcb4891edf1ceb73464398cb1ce7e6f53999f7545dd0014a15b1fb4eec5c6cf37421f0c2b08528de36f728ec4c676ed264c7d")
             .get();

     //獲取他所有的省
     Elements elements = document.select("body > table:nth-child(3) > tbody > tr:nth-child(1) > td > table > tbody > tr:nth-child(2) > td > table > tbody > tr > td > table > tbody");
     //解析省的超鏈接
     Elements elements1 = elements.select("tbody > tr > td > a");
     for(int j=0;j<elements1.size();j++){

// Thread.sleep(100);
String s = elements1.get(j).select(“a”).attr(“href”);
String provinceCode = s.replaceAll(“.html”,“”)+“0000”;
System.out.println(“省代碼:” + provinceCode);
String provinceName = elements1.get(j).text();
System.out.println(“省名稱:” + provinceName);
Map<String,String> map = new HashMap<>();
map.put(provinceCode,provinceName);
listMap.add(map);
}

        for (int i1 = 0; i1 <31; i1++) {
            System.out.println("**********************i********************:"+i);
            if(i%1000==0){
                Thread.sleep(1000*60*10);
            }
            Map<String, String> stringStringMap = listMap.get(i1);
            Iterator<Map.Entry<String, String>> iterator = stringStringMap.entrySet().iterator();
            while(iterator.hasNext()){
                Map.Entry<String,String> entry = iterator.next();
                String provinceCode = entry.getKey();
                String provinceName = entry.getValue();
                String index = provinceCode.substring(0,2)+".html";



                SXSSFWorkbook wb = new SXSSFWorkbook(100);
                SXSSFSheet sheet = (SXSSFSheet) wb.createSheet();
                // TODO 這里改成自己的地址即可，也可以存放到一個(gè)文件里
                String enterFileName = "C:\\Users\\pcc\\Desktop\\xingzhengquhua\\"+provinceName+".xlsx";
                FileOutputStream fileOut = new FileOutputStream(enterFileName);
                Row row = sheet.createRow(0);




                sheet.createRow(i).createCell(0).setCellValue(provinceCode);// id
                sheet.getRow(i).createCell(1).setCellValue(provinceName);// name
                sheet.getRow(i).createCell(2).setCellValue(""); // pid
                sheet.getRow(i).createCell(3).setCellValue("1"); // type
                i++;
                try {
                    jsoupList2(url1 + "/" + index, provinceName, provinceCode, sheet);
                } catch (SocketTimeoutException e) {
                    e.printStackTrace();
                    jsoupList2(url1 + "/" + index, provinceName, provinceCode, sheet);
                } catch (ConnectException e) {
                    e.printStackTrace();
                    jsoupList2(url1 + "/" + index, provinceName, provinceCode, sheet);
                }



                row.createCell(0).setCellValue("id");
                row.createCell(1).setCellValue("district_name");
                row.createCell(2).setCellValue("pid");
                row.createCell(3).setCellValue("type");
                wb.write(fileOut);
                fileOut.close();
            }


        }

    }catch (Exception e){
        e.printStackTrace();
    }finally {

    }

}



//市級(jí)頁(yè)面
public static void jsoupList2(String url,String provinceName,String provinceCode,SXSSFSheet sheet) throws Exception {
    String cityName = "";
    String cityCode = "";
    url2 = url.replace(".html","");
    Document document = Jsoup.connect(url).get();
    Elements elements = document.select("body > table:nth-child(3) > tbody > tr:nth-child(1) > td > table > tbody > tr:nth-child(2) > td > table > tbody > tr > td > table > tbody");
    Elements elements1 = elements.select("tbody > tr > td");
    //j從2開始是因?yàn)樗袀€(gè)表頭 統(tǒng)計(jì)用區(qū)劃代碼 名稱
    for (int j = 2; j < elements1.size(); j++) {
        System.out.println("**********************i********************:"+i);
        if(i%1000==0){
            Thread.sleep(1000*60*10);
        }
// Thread.sleep(500);
//判斷是否是超鏈接，不是超鏈接也要獲取數(shù)據(jù)
if(elements1.get(j).select(“td > a”).toString().equals(“”)){
String text = elements1.get(j).text();
if (j % 2 == 0) {
System.out.println(“市代碼:” + text);
sheet.createRow(i).createCell(0).setCellValue(text);
} else {
System.out.println(“市名稱:” + text);
sheet.getRow(i).createCell(1).setCellValue(text);
sheet.getRow(i).createCell(2).setCellValue(provinceCode);
sheet.getRow(i).createCell(3).setCellValue(“3”);
i++;
}
}else {
Elements elements2 = elements1.get(j).select(“td > a”);
for (int j1 = 0; j1 < elements2.size(); j1++) {
String text = elements2.get(j1).text();
if (j % 2 == 0) {
System.out.println(“市代碼:” + text);
cityCode = text;
sheet.createRow(i).createCell(0).setCellValue(text);
} else {
System.out.println(“市名稱:” + text);
cityName = text;
sheet.getRow(i).createCell(1).setCellValue(text);
sheet.getRow(i).createCell(2).setCellValue(provinceCode);
sheet.getRow(i).createCell(3).setCellValue(“2”);
i++;
String s = elements2.get(j1).select(“a”).attr(“href”);
//TODO 這里排除了海南的幾個(gè)市區(qū),更改為不排除任何市區(qū)
if(true) {
try {
jsoupList3(url1 + “/” + s,cityName,cityCode,provinceName,provinceCode, sheet);
} catch (SocketTimeoutException e) {
e.printStackTrace();
jsoupList3(url1 + “/” + s,cityName,cityCode,provinceName,provinceCode, sheet);
} catch (ConnectException e) {
e.printStackTrace();
jsoupList3(url1 + “/” + s,cityName,cityCode,provinceName,provinceCode, sheet);
}
}
}
}
}
}
}
//縣級(jí)頁(yè)面
public static void jsoupList3(String url,String cityName,String cityCode,String provinceName,String provinceCode,SXSSFSheet sheet) throws Exception {
Document document = Jsoup.connect(url).get();
Elements elements = document.select(“body > table:nth-child(3) > tbody > tr:nth-child(1) > td > table > tbody > tr:nth-child(2) > td > table > tbody > tr > td > table > tbody”);
Elements elements1 = elements.select(“tbody > tr > td”);
String xianName = “”;
String xianCode = “”;
//j從2開始是因?yàn)樗袀€(gè)表頭 統(tǒng)計(jì)用區(qū)劃代碼 名稱
for (int j = 2; j < elements1.size(); j++) {
System.out.println(“i:"+i);
if(i%1000==0){
Thread.sleep(10006010);
}
// Thread.sleep(500);
//判斷是否是超鏈接，不是超鏈接也要獲取數(shù)據(jù)
if(elements1.get(j).select(“td > a”).toString().equals(”“)){
String text = elements1.get(j).text();
if (j % 2 == 0) {
System.out.println(“縣代碼:” + text);
sheet.createRow(i).createCell(0).setCellValue(text);
} else {
System.out.println(“縣名稱:” + text);
sheet.getRow(i).createCell(1).setCellValue(text);
sheet.getRow(i).createCell(2).setCellValue(cityCode);
sheet.getRow(i).createCell(3).setCellValue(“3”);
i++;
}
}else {
Elements elements2 = elements1.get(j).select(“td > a”);
for (int j1 = 0; j1 < elements2.size(); j1++) {
String text = elements2.get(j1).text();
xianName = text;
if (j % 2 == 0) {
xianCode = text;
System.out.println(“縣代碼:” + xianCode);
sheet.createRow(i).createCell(0).setCellValue(text);
} else {
System.out.println(“縣名稱:” + text);
sheet.getRow(i).createCell(1).setCellValue(text);
sheet.getRow(i).createCell(2).setCellValue(cityCode);
sheet.getRow(i).createCell(3).setCellValue(“3”);
i++;
String s = elements2.get(j1).select(“a”).attr(“href”);
try {
jsoupList4(url1 + “/” +provinceCode.substring(0,2)+”/“+ s,xianName,xianCode,cityName,cityCode,provinceName,provinceCode, sheet);
} catch (SocketTimeoutException e) {
e.printStackTrace();
jsoupList4(url1 + “/” +provinceCode.substring(0,2)+”/“+ s,xianName,xianCode,cityName,cityCode,provinceName,provinceCode, sheet);
} catch (ConnectException e) {
e.printStackTrace();
jsoupList4(url1 + “/” +provinceCode.substring(0,2)+”/"+ s,xianName,xianCode,cityName,cityCode,provinceName,provinceCode, sheet);
}
}
}
}
}
}

//街道頁(yè)面
public static void jsoupList4(String url,String xianName,String xianCode,String cityName,String cityCode,String provinceName,String provinceCode,SXSSFSheet sheet) throws Exception {
    Document document = Jsoup.connect(url).get();
    Elements elements = document.select("body > table:nth-child(3) > tbody > tr:nth-child(1) > td > table > tbody > tr:nth-child(2) > td > table > tbody > tr > td > table > tbody");
    Elements elements1 = elements.select("tbody > tr > td");
    //j從2開始是因?yàn)樗袀€(gè)表頭 統(tǒng)計(jì)用區(qū)劃代碼 名稱
    for (int j = 2; j < elements1.size(); j++) {
        System.out.println("**********************i********************:"+i);
        if(i%1000==0){
            Thread.sleep(1000*60*10);
        }

// Thread.sleep(500);
//判斷是否是超鏈接，不是超鏈接也要獲取數(shù)據(jù)
if(elements1.get(j).select(“td > a”).toString().equals(“”)){
String text = elements1.get(j).text();
if (j % 2 == 0) {
System.out.println(“街道代碼:” + text);
sheet.createRow(i).createCell(0).setCellValue(text);
} else {
System.out.println(“街道名稱:” + text);
sheet.getRow(i).createCell(1).setCellValue(text);
sheet.getRow(i).createCell(2).setCellValue(xianCode);
sheet.getRow(i).createCell(3).setCellValue(“4”);
i++;
}
}else {
Elements elements2 = elements1.get(j).select(“td > a”);
for (int j1 = 0; j1 < elements2.size(); j1++) {
String text = elements2.get(j1).text();
if (j % 2 == 0) {
System.out.println(“街道代碼:” + text);// TODO 這里不能截取，不然街道界別數(shù)據(jù)截不全
sheet.createRow(i).createCell(0).setCellValue(text);
} else {
System.out.println(“街道名稱:” + text);
sheet.getRow(i).createCell(1).setCellValue(text);
sheet.getRow(i).createCell(2).setCellValue(xianCode);
sheet.getRow(i).createCell(3).setCellValue(“4”);
i++;
}
}
}
}
}
}

到此這篇關(guān)于Java實(shí)現(xiàn)獲取行政區(qū)劃的示例代碼的文章就介紹到這了,更多相關(guān)Java獲取行政區(qū)劃內(nèi)容請(qǐng)搜索腳本之家以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持腳本之家！

您可能感興趣的文章: