flink UTDF函数

代码示例IP解析demo案例

https://help.aliyun.com/zh/flink/developer-reference/udtfs

java 复制代码
package com.xxx.udx;


import com.alibaba.fastjson.JSON;
import com.alibaba.fastjson.JSONObject;
import com.maxmind.db.CHMCache;
import com.maxmind.geoip2.DatabaseReader;
import com.maxmind.geoip2.exception.GeoIp2Exception;
import com.maxmind.geoip2.model.CountryResponse;
import org.apache.flink.table.functions.FunctionContext;
import org.apache.flink.table.functions.TableFunction;
//import org.apache.flink.table.sources.parquet.update.UpdateVectorizedColumnRowInputParquetFormat;
import org.apache.flink.table.types.DataType;
//import org.apache.flink.table.types.DataTypes;
import org.apache.flink.types.Row;
import org.apache.http.HttpHost;
import org.apache.http.client.config.RequestConfig;
import org.apache.http.client.methods.CloseableHttpResponse;
import org.apache.http.client.methods.HttpGet;
import org.apache.http.conn.routing.HttpRoute;
import org.apache.http.impl.client.CloseableHttpClient;
import org.apache.http.impl.client.HttpClientBuilder;
import org.apache.http.impl.client.HttpClients;
import org.apache.http.impl.conn.PoolingHttpClientConnectionManager;
import org.apache.http.util.EntityUtils;

import java.io.IOException;
import java.io.InputStream;
import java.net.InetAddress;
import java.net.UnknownHostException;

/**
 * 使用节点ip解析服务
 */

public class GetResolveIp4Ip6AP extends TableFunction<Row> {
    CloseableHttpClient client = null;
    private static RequestConfig requestConfig = null;
    InputStream path =  null;
    DatabaseReader reader = null;
    @Override
    public void open(FunctionContext context) throws Exception {

        super.open(context);
        requestConfig = RequestConfig.custom()
                .setConnectTimeout(50000) //一、连接超时:connectionTimeout-->指的是连接一个url的连接等待时间
                .setSocketTimeout(5000)  // 二、读取数据超时:SocketTimeout-->指的是连接上一个url,获取response的返回等待时间
                .setConnectionRequestTimeout(5000)
                .setMaxRedirects(0)
                .build();

        client = getConnection();
        //client.getParams().setParameter(HttpMethodParams.HTTP_CONTENT_CHARSET, "UTF-8");

        path=GetResolveIp4Ip6AP.class.getResourceAsStream("/GeoLite2-Country.mmdb");
        reader = new DatabaseReader.Builder(path).withCache(new CHMCache()).build();

    }

    public void eval (String ip) throws Exception{
        String countryCode = "";
        String countryName =""; //todo 新加
        String city = "";
        String stateprov = "";

        //todo
        Row row = new Row(3);

        String res="";
        if(null != ip  && !ip.isEmpty() && ip.split("\\.").length==4 ){ //todo IP4解析
            res = doGet(ip,client);

        }else if(null != ip  && !ip.isEmpty() && !ip.equals("127.0.0.1")){//todo IP6解析

            try {
                InetAddress ipAddress = InetAddress.getByName(ip);
                //-com.maxmind.geoip2.record.Country country=null;

                CountryResponse response = reader.country(ipAddress);
                com.maxmind.geoip2.record.Country country = response.getCountry();
                countryCode= country.getIsoCode();

            }catch (Exception e){
                System.out.println("IP:"+ip);
                e.printStackTrace();
            }

            String deal_countryCode;
            if(countryCode==null|| countryCode.isEmpty()){
                deal_countryCode="UNKNOWN";
            }else{
                deal_countryCode=countryCode;
            }
            row.setField(0,deal_countryCode);
            //row.setField(1,countryName);
            row.setField(1,"未知");
            row.setField(2,"未知");


            collect(row);
            return;
        }

        if(res ==""){
            row.setField(0,"UNKNOWN");
            //row.setField(1,"未知");//todo 新加
            row.setField(1,"未知");
            row.setField(2,"未知");
            collect(row);
            return;
        }

        try{
            JSONObject json = JSON.parseObject(res);
            countryCode = json.getString("country_code");
            countryName = "";//todo 新加
            stateprov = json.getString("region_name");
            city = json.getString("city");
        }catch (Exception e){
            e.printStackTrace();
        }


        if(countryCode==null || countryCode.isEmpty()){
            countryCode="UNKNOWN";
        }
        if(stateprov==null || stateprov.isEmpty()){
            stateprov="未知";
        }
        if(city==null || city.isEmpty()){
            city="未知";
        }
        row.setField(0,countryCode);
        //row.setField(1,countryName);//todo 新加
        row.setField(1,stateprov);
        row.setField(2,city);

        collect(row);



    }
    @Override
    public void close() throws Exception {
        if(null != client){
            client.close();
            reader.close();
            path.close();
        }
        super.close();
    }

    //todo flink 注掉
//    @Override
//    // 如果返回值是Row,则必须重载实现getResultType方法,显式地声明返回的字段类型。
//    public DataType getResultType(Object[] arguments, Class[] argTypes) {
//        return DataTypes.createRowType(DataTypes.STRING, DataTypes.STRING, DataTypes.STRING);
//    }

    private static CloseableHttpClient getConnection() {
        HttpHost target = new HttpHost("0.0.0.0", 80);



        PoolingHttpClientConnectionManager connectionManager = new PoolingHttpClientConnectionManager();
        connectionManager.setMaxTotal(2000);//客户端总并行链接最大数
        connectionManager.setDefaultMaxPerRoute(2000);//每个主机的最大并行链接数
        connectionManager.setMaxPerRoute(new HttpRoute(target), 2000);

        HttpClientBuilder httpBuilder = HttpClients.custom();
        httpBuilder.setConnectionManager(connectionManager);

        CloseableHttpClient httpClient = httpBuilder.build();
        return httpClient;
    }

    private static String doGet(String ip, CloseableHttpClient client) {
        //简单的对ip地址的合法性做一下验证
        if(null == ip  || ip.isEmpty() || ip.split("\\.").length!=4 ){
            return "";
        }

        if (null == client) {
            client =  getConnection();
        }
        HttpGet getMethod = new HttpGet("http://0.0.0.0:80/json/" + ip);
        getMethod.setConfig(requestConfig);
        String res = "";
        try {
            CloseableHttpResponse response = client.execute(getMethod);


            if (response.getStatusLine().getStatusCode() == 200) {
                res =  EntityUtils.toString(response.getEntity());
//                System.out.println(res);
            }
        } catch (IOException e) {
            e.printStackTrace();
        } finally {
            getMethod.releaseConnection();
        }
        return res;
    }

}
相关推荐
rit843249924 分钟前
Git常用命令的详细指南
大数据·git·elasticsearch
赵谨言1 小时前
基于Python Web的大数据系统监控平台的设计与实现
大数据·开发语言·经验分享·python
南棱笑笑生1 小时前
20251028在Ubuntu20.04.6上编译AIO-3576Q38开发板的Buildroot系统
大数据·linux·服务器·rockchip
武子康1 小时前
大数据-139 ClickHouse MergeTree 最佳实践:Replacing 去重、Summing 求和、分区设计与物化视图替代方案
大数据·后端·nosql
我要升天!1 小时前
Git的原理与使用 -- 分支管理
大数据·git·elasticsearch
培培说证4 小时前
2025年高职大数据技术专业需要什么基础?
大数据
北邮-吴怀玉4 小时前
1.4.2 大数据方法论与实践指南-质量治理(准确性&及时性)
大数据·数据治理
2501_938782096 小时前
《大数据框架选型指南:Hadoop 与 Spark 的性能、成本与扩展性对比》
大数据·hadoop·spark
TMT星球6 小时前
AI重构兴趣内容与营销生态,驱动消费全链路升级
大数据·人工智能·重构
HitpointNetSuite7 小时前
科技行业ERP系统选择指南:Oracle NetSuite的全面解析
大数据·科技·netsuite·企业·erp