sis-background/gyee-sample-impala/src/main/java/com/gyee/impala/controller/diagnose/TrainFaultPredictController.java

package com.gyee.impala.controller.diagnose;


import com.alibaba.fastjson.JSONObject;
import com.gyee.impala.common.result.JsonResult;
import com.gyee.impala.common.result.ResultCode;
import com.gyee.impala.common.util.DateUtil;
import com.gyee.impala.model.custom.diagnose.DataInfo;
import com.gyee.impala.model.custom.diagnose.ExecuteInfo;
import com.gyee.impala.model.master.Casefault;
import com.gyee.impala.model.master.diagnose.Diagnosepoint;
import com.gyee.impala.model.master.diagnose.Diagnosetrainhistory;
import com.gyee.impala.service.custom.diagnose.DataPredictService;
import com.gyee.impala.service.custom.diagnose.CmdFaultPredictService;;
import lombok.extern.slf4j.Slf4j;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.scheduling.concurrent.ThreadPoolTaskExecutor;
import org.springframework.web.bind.annotation.*;
import org.springframework.web.multipart.MultipartFile;

import javax.annotation.Resource;
import java.util.Calendar;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;

/**
 * 故障预测模型训练
 */
@Slf4j
@CrossOrigin
@RestController
@RequestMapping("/api/predict")
public class TrainFaultPredictController {


    @Autowired
    private CmdFaultPredictService cmdFaultPredictService;
    /**
     * 参数
     */
    private ExecuteInfo executeInfo;

    /**
     * 数据服务
     */
    @Autowired
    private DataPredictService dataDiagnoseService;

    /**
     * 线程池
     */
    @Resource
    private ThreadPoolTaskExecutor taskExecutor;

    /**
     * 预测生成文件名
     */
    private String fileName;

    private static final Object locker = new Object();

    private String name1;
    private String forecastLabel1;
    private String[] inputLabel1;
    private String host1;
    private MultipartFile file1;


    /**
     * 离线文件模式训练故障预测接口
     * @param forecastLabel
     * @param inputLabel
     * @param host
     * @param file
     * @return
     */
    @PostMapping("/trainfile")
    @ResponseBody
    public JSONObject getTrainfile(String forecastLabel, String[] inputLabel, String host, MultipartFile file) {

        if (!cmdFaultPredictService.isComplete()) {
            return JsonResult.error(4000, "命令正在执行...");
        }

        if (file.isEmpty()) {
            return JsonResult.error(ResultCode.ERROR_FILE_NO);
        }
        try {
            synchronized (locker) {
                forecastLabel1 = forecastLabel;
                inputLabel1 = inputLabel;
                host1 = host;
                file1 = file;
                taskExecutor.submit(this::executeFile);
            }

            return JsonResult.success(ResultCode.SUCCESS);
        } catch (Exception e) {
            return JsonResult.error(ResultCode.ERROR_DATA_FILE);
        }
    }



    /**
     * 调用执行脚本
     */
    private void executeFile() {
        cmdFaultPredictService.exec(forecastLabel1, inputLabel1, host1, file1);
    }


    /** 开始预测 查询 golden 所有原始数据
     *  flag  ture: 所有数据 ；false: 前10条数据
     * **/
    @PostMapping("/pointfaultdata")
    public JSONObject getPointData(@RequestBody JSONObject json){
        if (json == null)
            return JsonResult.error(ResultCode.PARAM_IS_BLANK);


        log.info("进入预测程序");
        boolean flag = json.getBooleanValue("flag");
        List<Diagnosepoint> points = JSONObject.parseArray(json.getJSONArray("points").toString(), Diagnosepoint.class);
        List<Casefault> faults = JSONObject.parseArray(json.getJSONArray("faults").toString(), Casefault.class);

        /** 组装数据  删除添加的故障类型**/
        dataDiagnoseService.formatUniformcode(points.stream().filter(a -> !a.getUniformcode().equals("faulttype")).collect(Collectors.toList()));
        executeInfo = new ExecuteInfo();
        Calendar cal = Calendar.getInstance();
        DataInfo[] dataInfos = new DataInfo[faults.size()];
        for (int i = 0; i < faults.size(); i++){
            DataInfo data = new DataInfo();
            data.setId(Long.valueOf(faults.get(i).getId()));
            data.setStationId(faults.get(i).getStationen());
            data.setThingId(faults.get(i).getWindturbineid());
            data.setModelId(faults.get(i).getModel());
            data.setTag(faults.get(i).getFaultcode());
            data.setFaultTime(faults.get(i).getStarttime());

            cal.setTime(DateUtil.parseStrtoDate(faults.get(i).getStarttime(), DateUtil.YYYY_MM_DD_HH_MM_SS));
            cal.add(Calendar.MINUTE, -30);
            data.setStartTs(cal.getTimeInMillis() + "");
            cal.add(Calendar.MINUTE, 30);
            data.setEndTs(cal.getTimeInMillis() + "");
            dataInfos[i] = data;
        }
        executeInfo.setDataInfos(dataInfos);
        if (flag){
            if (!cmdFaultPredictService.isComplete()) {
                return JsonResult.error(4000, "已有正在预测的故障模型...");
            }
            synchronized (locker) {
                taskExecutor.submit(this::execute);
            }
            return JsonResult.success(ResultCode.SUCCESS);
        }else {
            Map<String, Object> mp = dataDiagnoseService.getFormData(executeInfo);
            return JsonResult.successData(ResultCode.SUCCESS, mp);
        }
    }


    /**
     * 调用执行脚本
     */
    private void execute() {

        fileName = dataDiagnoseService.getFormDataAll(executeInfo);
        cmdFaultPredictService.exec();

    }



    /**
     * py 获取预测数据
     *
     * @return
     */
    @GetMapping("/trainfaultdata")
    public JSONObject getData() {
        Map<String, Object> map = new HashMap<>();
        map.put("info", this.executeInfo);
        map.put("filename", fileName);
        return JsonResult.successData(ResultCode.SUCCESS, map);
    }




    /**
     * 缓存训练预测模型结果
     *
     * @param history
     * @return
     */
    @PostMapping("/putHistory")
    public JSONObject putDiagnosetrainhistory(@RequestBody String history) {
        try {
            log.info("训练模型结果：" + history);
            cmdFaultPredictService.putDiagnosetrainhistory(history);
            return JsonResult.success(ResultCode.SUCCESS);
        } catch (Exception e) {
            log.error("请求错误", e);
            return JsonResult.error(ResultCode.ERROR);
        }
    }


    /**
     * 获取当前训练预测模型结果
     * @return
     */
    @GetMapping("/getHistory")
    public JSONObject getHistory() {
        try {
            Diagnosetrainhistory d = cmdFaultPredictService.getHistoryQueue();
            return JsonResult.successData(ResultCode.SUCCESS, d);
        } catch (Exception e) {
            log.error("请求错误", e);
            return JsonResult.error(ResultCode.ERROR);
        }
    }


    /** 开始预测 查询 golden 所有原始数据
     *  flag  ture: 所有数据 ；false: 前10条数据
     * **/
    @PostMapping("/pointfaultdata")
    public JSONObject getWarningData(@RequestBody JSONObject json){
        if (json == null)
            return JsonResult.error(ResultCode.PARAM_IS_BLANK);


        log.info("进入预测程序");
        boolean flag = json.getBooleanValue("flag");
        List<Diagnosepoint> points = JSONObject.parseArray(json.getJSONArray("points").toString(), Diagnosepoint.class);
        List<Casefault> faults = JSONObject.parseArray(json.getJSONArray("faults").toString(), Casefault.class);

        /** 组装数据  删除添加的故障类型**/
        dataDiagnoseService.formatUniformcode(points.stream().filter(a -> !a.getUniformcode().equals("faulttype")).collect(Collectors.toList()));
        executeInfo = new ExecuteInfo();
        Calendar cal = Calendar.getInstance();
        DataInfo[] dataInfos = new DataInfo[faults.size()];
        for (int i = 0; i < faults.size(); i++){
            DataInfo data = new DataInfo();
            data.setId(Long.valueOf(faults.get(i).getId()));
            data.setStationId(faults.get(i).getStationen());
            data.setThingId(faults.get(i).getWindturbineid());
            data.setModelId(faults.get(i).getModel());
            data.setTag(faults.get(i).getFaultcode());
            data.setFaultTime(faults.get(i).getStarttime());

            cal.setTime(DateUtil.parseStrtoDate(faults.get(i).getStarttime(), DateUtil.YYYY_MM_DD_HH_MM_SS));
            cal.add(Calendar.MINUTE, -30);
            data.setStartTs(cal.getTimeInMillis() + "");
            cal.add(Calendar.MINUTE, 30);
            data.setEndTs(cal.getTimeInMillis() + "");
            dataInfos[i] = data;
        }
        executeInfo.setDataInfos(dataInfos);
        if (flag){
            if (!cmdFaultPredictService.isComplete()) {
                return JsonResult.error(4000, "已有正在预测的故障模型...");
            }
            synchronized (locker) {
                taskExecutor.submit(this::execute);
            }
            return JsonResult.success(ResultCode.SUCCESS);
        }else {
            Map<String, Object> mp = dataDiagnoseService.getFormData(executeInfo);
            return JsonResult.successData(ResultCode.SUCCESS, mp);
        }
    }



}