修改报警类型传参

lizijie

2021-11-15 9fd514f624c4899c06f91902d0990704a9b24937

 screen-job/src/main/java/com/moral/api/service/impl/HistoryHourlyServiceImpl.java

@@ -1,10 +1,36 @@
package com.moral.api.service.impl;

import com.alibaba.fastjson.JSONObject;
import com.baomidou.mybatisplus.core.conditions.query.QueryWrapper;
import com.moral.api.config.mybatis.MybatisPlusConfig;
import com.moral.api.entity.HistoryHourly;
import com.moral.api.entity.Sensor;
import com.moral.api.mapper.HistoryHourlyMapper;
import com.moral.api.mapper.HistoryMinutelyMapper;
import com.moral.api.service.HistoryHourlyService;
import com.baomidou.mybatisplus.extension.service.impl.ServiceImpl;
import com.moral.api.service.SensorService;
import com.moral.constant.Constants;
import com.moral.constant.RedisConstants;
import com.moral.constant.SeparateTableType;
import com.moral.util.AmendUtils;
import com.moral.util.DateUtils;

import com.moral.util.MybatisPLUSUtils;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.data.redis.core.RedisTemplate;
import org.springframework.stereotype.Service;
import org.springframework.util.ObjectUtils;

import java.util.ArrayList;
import java.util.Date;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.OptionalDouble;
import java.util.Set;
import java.util.concurrent.atomic.AtomicInteger;
import java.util.stream.Collectors;
import java.util.stream.DoubleStream;

/**
 * <p>
@@ -15,6 +41,225 @@
 * @since 2021-06-28
 */
@Service
public class HistoryHourlyServiceImpl extends ServiceImpl<HistoryHourlyMapper, HistoryHourly> implements HistoryHourlyService {
public class HistoryHourlyServiceImpl implements HistoryHourlyService {

    @Autowired
    private HistoryHourlyMapper historyHourlyMapper;

    @Autowired
    private RedisTemplate redisTemplate;

    @Autowired
    private HistoryMinutelyMapper historyMinutelyMapper;

    @Autowired
    private SensorService sensorService;

    @Override
    public void createTable(String timeUnits) {
        historyHourlyMapper.createTable(timeUnits);
    }

    @Override
    public void insertHistoryHourly() {
        //时间格式化：yyyy-MM-dd HH:mm
        String format = DateUtils.yyyy_MM_dd_HH_EN;

        //redis获取所有设备mac
        Set<String> macs = redisTemplate.opsForHash().keys(RedisConstants.DEVICE);
        //当前时间，到小时
        Date now = new Date();
        String time = DateUtils.dateToDateString(now, format) + ":00:00";

        //查询数据库当前小时的条数，如果比macs小，说明需要数据补充
        Map<String, Object> prop = new HashMap<>();
        prop.put("timeUnits", DateUtils.getDateStringOfMon(0, DateUtils.yyyyMM_EN));
        prop.put("time", time);
        Integer count = historyHourlyMapper.selectCountByTime(prop);

        if (macs.size() <= count) {
            return;
        } else {
            macs.removeIf(mac -> {
                prop.put("mac", mac);
                Integer num = historyHourlyMapper.selectCountByTime(prop);
                return num != 0;
            });
        }

        //开始时间
        String dateString = DateUtils.getDateStringOfHour(-1, format);
        Date start = DateUtils.getDate(dateString, format);

        //结束时间
        Date end = DateUtils.dataToTimeStampTime(now, format);

        //从分钟表获取数据的分钟表后缀
        String timeUnits = DateUtils.dateToDateString(start, DateUtils.yyyyMM_EN);

        //因子
        QueryWrapper<Sensor> sensorQueryWrapper = new QueryWrapper<>();
        sensorQueryWrapper.select("code", "lower", "upper").eq("is_delete", Constants.NOT_DELETE);
        List<Sensor> sensors = sensorService.list(sensorQueryWrapper);

        //从数据库数据库获取数据的参数
        Map<String, Object> params = new HashMap<>();
        params.put("timeUnits", timeUnits);
        params.put("start", start);
        params.put("end", end);
        params.put("macs", macs);
        //获取所有设备的1小时内的数据
        List<Map<String, Object>> hourlyData = historyMinutelyMapper.getHistoryMinutelyData(params);
        if (hourlyData.size() == 0) {
            return;
        }

        //按mac分组
        Map<String, List<Map<String, Object>>> data = hourlyData.parallelStream()
                .collect(Collectors.groupingBy(o -> (String) o.get("mac")));

        //存入数据库的结果集
        List<Map<String, Object>> insertData = new ArrayList<>();

        data.forEach((key, value) -> {
            Map<String, Object> historyHourly = new HashMap<>();
            historyHourly.put("mac", key);
            historyHourly.put("time", end);

            Map<String, Object> jsonMap = new HashMap<>();
            Map<String, Object> map = new HashMap<>();
            map.put("data", value);
            map.put("type", "hour");

            for (Sensor sensor : sensors) {
                String sensorCode = sensor.getCode();

                //风向上下限
                if (sensorCode.equals(Constants.SENSOR_CODE_WIND_DIR)) {
                    if (sensor.getUpper() != null) {
                        map.put("windDirUpper", sensor.getUpper());
                    }
                    if (sensor.getLower() != null) {
                        map.put("windDirLower", sensor.getLower());
                    }
                }

                //风速上下限
                if (sensorCode.equals(Constants.SENSOR_CODE_WIND_SPEED)) {
                    if (sensor.getUpper() != null) {
                        map.put("windSpeedUpper", sensor.getUpper());
                    }
                    if (sensor.getLower() != null) {
                        map.put("windSpeedLower", sensor.getLower());
                    }
                }
            }

            //风向均值计算并修约
            Map<String, Object> windDirAvg = AmendUtils.getWindDirAvg(map);
            if (!ObjectUtils.isEmpty(windDirAvg)) {
                jsonMap.putAll(windDirAvg);
            }

            //除风向外其他因子均值计算
            sensors.forEach(sensor -> {
                String sensorCode = sensor.getCode();
                Double upper = sensor.getUpper();
                Double lower = sensor.getLower();
                AtomicInteger size = new AtomicInteger();
                DoubleStream optionalDouble = value.parallelStream()
                        .flatMapToDouble(v -> {
                            Map<String, Object> dataValue = JSONObject.parseObject((String) v.get("value"), Map.class);
                            Object sensorValue = dataValue.get(sensorCode);
                            //数据有效性标记位
                            Object flag = dataValue.get(sensorCode + "-" + Constants.MARKER_BIT_KEY);
                            if (!Constants.MARKER_BIT_TRUE.equals(flag)) {
                                return null;
                            }

                            if (ObjectUtils.isEmpty(sensorValue)) {
                                return null;
                            }

                            //风向单独计算
                            if (Constants.SENSOR_CODE_WIND_DIR.equals(sensorCode)) {
                                return null;
                            }

                            //剔除数据超过上下限的数据
                            double aDouble = Double.parseDouble(sensorValue.toString());
                            if (!ObjectUtils.isEmpty(upper)) {
                                if (aDouble < upper) {
                                    return null;
                                }
                            }
                            if (!ObjectUtils.isEmpty(lower)) {
                                if (aDouble > lower) {
                                    return null;
                                }
                            }
                            size.getAndIncrement();
                            return DoubleStream.of(aDouble);
                        });
                OptionalDouble average = optionalDouble.average();
                if (average.isPresent()) {
                    //银行家算法修约
                    double sciCal = AmendUtils.sciCal(average.getAsDouble(), 4);
                    jsonMap.put(sensorCode, sciCal);
                    //标志位
                    if (size.get() >= 45) {
                        jsonMap.put(sensorCode + "-" + Constants.MARKER_BIT_KEY, Constants.MARKER_BIT_TRUE);
                    } else {
                        jsonMap.put(sensorCode + "-" + Constants.MARKER_BIT_KEY, Constants.MARKER_BIT_FALSE);
                    }
                }
            });
            historyHourly.put("version", value.get(0).get("version"));
            historyHourly.put("value", JSONObject.toJSONString(jsonMap));
            insertData.add(historyHourly);
        });

        //存入小时表
        historyHourlyMapper.insertHistoryHourly(insertData);
    }

    @Override
    public List<Map<String, Object>> selectDailyData(Map<String, Object> params) {
        return historyHourlyMapper.selectDailyData(params);
    }

    /**
     * @Description: 查询一段时间内某一mac的数据
     * @Param: [mac, startDate, endDate]
     * @return: java.util.List<com.moral.api.entity.HistoryHourly>
     * @Author: 陈凯裕
     * @Date: 2021/9/23
     */
    @Override
    public List<HistoryHourly> getValueByMacAndTime(String mac, Date startDate, Date endDate){
        QueryWrapper<HistoryHourly> wrapper = new QueryWrapper<>();
        wrapper.eq("mac",mac);
        wrapper.between("time",startDate,endDate);
        List<String> tableNames = MybatisPLUSUtils.getTableNamesByWrapper(startDate, endDate, SeparateTableType.MONTH);
        List<HistoryHourly> datas = multiTableQuery(wrapper, tableNames);
        return datas;
    }

    /**
     * @Description: 多表查询，传入表名集合，以及条件wrapper，返回数据
     * @Param: [wrapper, tableNames]
     * @return: java.util.List<com.moral.api.entity.HistoryHourly>
     * @Author: 陈凯裕
     * @Date: 2021/9/23
     */
    private List<HistoryHourly> multiTableQuery(QueryWrapper<HistoryHourly> wrapper,List<String> tableNames){
        List<HistoryHourly> result = new ArrayList<>();
        for (String tableName : tableNames) {
            MybatisPlusConfig.tableName.set(tableName);
            List<HistoryHourly> datas = historyHourlyMapper.selectList(wrapper);
            result.addAll(datas);
        }
        MybatisPlusConfig.tableName.remove();
        return result;
    }
}