oralce分類統計

@Transactional
@Component("statisDAO")
@Scope("prototype")
public class StatisDAO extends AbstractDAO {
    private static final String SQL_STATIS_CHANNELTYPE = "INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) " +
    		"SELECT DBMS_RANDOM.STRING('A', 32) AS uuid, count(*) AS counter, channeltype, accountarea,sysdate AS statistime,to_char(?)||'/'||to_char(?) AS STATISRANGE " +
    		"FROM ACTIVITY WHERE txtime>=? AND txtime<=?  GROUP BY ROLLUP (channeltype,accountarea)";

    @Transactional(propagation = Propagation.REQUIRED, rollbackFor = Exception.class)
    public boolean statisChanneltype(String startDateStr,String endDateStr) {
        try {
            DateFormat format = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");
            Date startDate=format.parse(startDateStr+" 00:00:00");
            Date endDate=format.parse(endDateStr+" 23:59:59");
            
            Object[] values = {startDateStr,endDateStr,startDate,endDate};
            this.getSmartDAO().update("", SQL_STATIS_CHANNELTYPE, values);
            return true;
        } catch (Exception e) {
            e.printStackTrace();
            return false;
        }
    }
public static void main(String[] args) {
        ApplicationContext applicationContext = new FileSystemXmlApplicationContext("file:E:\\CommonContext.xml");
        StatisDAO statisDAO = (StatisDAO) applicationContext.getBean("statisDAO");
        System.out.println(statisDAO.statisChanneltype("2011-04-01", "2011-08-02"));
    }

}

按日期時長統計：

統計每個時間段得條目數，txtime是date類型，sql如下：

按小時---select to_char(txtime, 'YYYY-MM-dd hh24') a,count(*) from mytable group by to_char(txtime,'YYYY-MM-dd hh24');

按天---select to_char(txtime, 'YYYY-MM-dd') a,count(*) from mytable group by to_char(txtime,'YYYY-MM-dd');

按月---select to_char(txtime, 'YYYY-MM') a,count(*) from mytable group by to_char(txtime,'YYYY-MM');

按年---select to_char(txtime, 'YYYY') a,count(*) from mytable group by to_char(txtime,'YYYY');

按季度---select to_char(txtime, 'YYYY-q') a,count(*) from mytable group by to_char(txtime,'YYYY-q');

按周---select to_char(txtime, 'ww') a,count(*) from mytable group by to_char(txtime,'ww');

參考http://hi.baidu.com/qq5910225/blog/item/4a8c91d7ef0ec514a08bb74e.html

同字符串類型統計

select
SUM(CASE
WHEN city = '海口市' THEN
1
ELSE
0
END) haikou_num
,SUM(CASE
WHEN city = '廣州市' THEN
1
ELSE
0
END) guangzhou_num
FROM ACTIVITY_HIS

decode方式

SQL> select id,num from test1;

ID NUM
---------- ----------
1 3
1 4
2 2
2 5
3 1
3 8

6 rows selected

SQL> select decode(grouping(id),1, '總計 ',id) id,sum(num) num
2 from test1
3 group by rollup(id);

ID NUM
---------------------------------------- ----------
1 7
2 7
3 9
總計 23

最終大神：select count(*), province, city from ACTIVITY_HIS where to_char(txtime, 'YYYY-MM-dd')='2011-06-01' group by rollup (province,city);

rollup是數據挖掘中的上卷操作，運行效果截圖

另外，將取出來得數據直接插入表中：

select ...into 用在存儲過程裏面的,保存變量
insert ...select 就是插入語句,插入的部分是表中的數據

舉例來說：

insert 表 select * from 表的方法主要有兩種：

1、若兩張表完全相同：
insert into table1
select * from table2
where condition(條件)

2、若兩張表字段有不同的：
insert into table1(字段1，字段2，字段3....)
select 字段1，字段2，字段3....
from table2
where condition(條件)

上述兩種方法均不需要寫values.

綜上，我的sql是：

INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange)

SELECT DBMS_RANDOM.STRING('A', 32) AS uuid, count(*) AS counter, channeltype, accountarea,sysdate AS statistime,TO_CHAR(?,'yyyy-mm-dd') AS STATISRANGE

FROM ACTIVITY

WHERE where to_char(txtime, 'yyyy-mm-dd')>'?' AND TO_CHAR(txtime, 'yyyy-mm-dd')<'?'

GROUP BY ROLLUP (channeltype,accountarea)

channeltypestatis數據字典：

CHANNELTYPESTATIS UUID VARCHAR2
CHANNELTYPESTATIS COUNTER NUMBER
CHANNELTYPESTATIS CHANNELTYPE VARCHAR2
CHANNELTYPESTATIS ACCOUNTAREA VARCHAR2
CHANNELTYPESTATIS STATISTIME DATE
CHANNELTYPESTATIS STATISRANGE VARCHAR2

在編寫過程中還有問題：

代碼片段

@Transactional
@Component("statisDAO")
@Scope("prototype")
public class StatisDAO extends AbstractDAO {
    private static final String SQL_STATIS_CHANNELTYPE = "INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) SELECT DBMS_RANDOM.STRING('A', 32) AS uuid, count(*) AS counter, channeltype, accountarea,sysdate AS statistime,TO_CHAR(?/?) AS STATISRANGE FROM ACTIVITY WHERE txtime>=to_date(?, 'yyyy-mm-dd') AND txtime<=to_date(?, 'yyyy-mm-dd')  GROUP BY ROLLUP (channeltype,accountarea)";

    @Transactional(propagation = Propagation.REQUIRED, rollbackFor = Exception.class)
    public boolean statisChanneltype(String startDate,String endDate) {
        try {
            Object[] values = {startDate,endDate,startDate,endDate};
            this.getSmartDAO().update("", SQL_STATIS_CHANNELTYPE, values);
            return true;
        } catch (Exception e) {
            e.printStackTrace();
            return false;
        }
    }

    public static void main(String[] args) {
        ApplicationContext applicationContext = new FileSystemXmlApplicationContext("file:E:\\CommonContext.xml");
        StatisDAO statisDAO = (StatisDAO) applicationContext.getBean("statisDAO");
        System.out.println(statisDAO.statisChanneltype("2011-04-01", "2011-08-02"));
    }
}

如果報無效的列索引，原因是在sql語句中？不能加' 而應如上所示

但上述代碼依舊報錯，錯誤的數字格式，因爲是我傳 2011-08-02 參數的時候，解析sql時，會產生錯誤，最終改成了如下格式

@Transactional
@Component("statisDAO")
@Scope("prototype")
public class StatisDAO extends AbstractDAO {
    private static final String SQL_STATIS_CHANNELTYPE = "INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) SELECT DBMS_RANDOM.STRING('A', 32) AS uuid, count(*) AS counter, channeltype, accountarea,sysdate AS statistime,TO_CHAR(?-?) AS STATISRANGE FROM ACTIVITY WHERE txtime>=to_date(?, 'yyyymmdd') AND txtime<=to_date(?, 'yyyymmdd')  GROUP BY ROLLUP (channeltype,accountarea)";

    @Transactional(propagation = Propagation.REQUIRED, rollbackFor = Exception.class)
    public boolean statisChanneltype(String startDate,String endDate) {
        try {
            Object[] values = {startDate,endDate,startDate,endDate};
            this.getSmartDAO().update("", SQL_STATIS_CHANNELTYPE, values);
            return true;
        } catch (Exception e) {
            e.printStackTrace();
            return false;
        }
    }

    public static void main(String[] args) {
        ApplicationContext applicationContext = new FileSystemXmlApplicationContext("file:E:\\CommonContext.xml");
        StatisDAO statisDAO = (StatisDAO) applicationContext.getBean("statisDAO");
        System.out.println(statisDAO.statisChanneltype("20110401", "20110802"));
    }
}

但是TO_CHAR(?-?) 這個函數給我解析成數學符號後，全給我相減了NND,應該用oracle中的字符串拼接。其他常用oracle函數見下一篇

帶拼接的，還應注意時間，前面的不能滿足當天查詢

@Transactional
@Component("statisDAO")
@Scope("prototype")
public class StatisDAO extends AbstractDAO {
    private static final String SQL_STATIS_CHANNELTYPE = "INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) " +
    		"SELECT DBMS_RANDOM.STRING('A', 32) AS uuid, count(*) AS counter, channeltype, accountarea,sysdate AS statistime,to_char(?)||'/'||to_char(?) AS STATISRANGE " +
    		"FROM ACTIVITY WHERE txtime>=? AND txtime<=?  GROUP BY ROLLUP (channeltype,accountarea)";

    @Transactional(propagation = Propagation.REQUIRED, rollbackFor = Exception.class)
    public boolean statisChanneltype(String startDateStr,String endDateStr) {
        try {
            DateFormat format = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");
            Date startDate=format.parse(startDateStr+" 00:00:00");
            Date endDate=format.parse(endDateStr+" 23:59:59");
            
            Object[] values = {startDateStr,endDateStr,startDate,endDate};
            this.getSmartDAO().update("", SQL_STATIS_CHANNELTYPE, values);
            return true;
        } catch (Exception e) {
            e.printStackTrace();
            return false;
        }
    }

public static void main(String[] args) {
        ApplicationContext applicationContext = new FileSystemXmlApplicationContext("file:E:\\CommonContext.xml");
        StatisDAO statisDAO = (StatisDAO) applicationContext.getBean("statisDAO");
        System.out.println(statisDAO.statisChanneltype("2011-08-01", "2011-08-01"));
    }

}

但是上述sql語句是採用SELECT DBMS_RANDOM.STRING('A', 32)隨機數的，這個是可能相同的，因此要求使用sequence，但是sequence和group by一塊使用會報錯：ORA-02287: 此處不允許序號，

INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) 
SELECT SEQ_PK.nextVal as uuid , count(*) AS counter, channeltype, accountarea,sysdate AS statistime,to_char('2011-08-01')||'/'||to_char('2011-08-01') AS STATISRANGE FROM ACTIVITY 
WHERE txtime>=to_date('2011-08-01 00:00:00','yyyy-mm-dd hh24:mi:ss') AND txtime<=to_date('2011-08-01 23:59:59','yyyy-mm-dd hh24:mi:ss')  GROUP BY ROLLUP (channeltype,accountarea)

報錯：oracle sequence ORA-02287: 此處不允許序號

採用

INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) 
SELECT SEQ_PK.nextVal as uuid , P.* from 
(select count(*) AS counter, channeltype, accountarea,sysdate AS statistime,to_char('2011-08-01')||'/'||to_char('2011-08-01') AS STATISRANGE 
FROM ACTIVITY WHERE txtime>=to_date('2011-08-01 00:00:00','yyyy-mm-dd hh24:mi:ss') AND txtime<=to_date('2011-08-01 23:59:59','yyyy-mm-dd hh24:mi:ss')  GROUP BY ROLLUP (channeltype,accountarea)) P

進行使用即可

最終代碼爲：

@Transactional
@Component("statisDAO")
@Scope("prototype")
public class StatisDAO extends AbstractDAO {
    private static final String SQL_STATIS_CHANNELTYPE ="INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) " +
    		"SELECT SEQ_PK.nextVal as uuid , P.* from " +
    		"(select count(*) AS counter, channeltype, accountarea,sysdate AS statistime,to_char(?)||'/'||to_char(?) AS STATISRANGE " +
    		"FROM ACTIVITY WHERE txtime>=? AND txtime<=? " +
    		"GROUP BY ROLLUP (channeltype,accountarea)) P";

    @Transactional(propagation = Propagation.REQUIRED, rollbackFor = Exception.class)
    public boolean statisChanneltype(String startDateStr,String endDateStr) {
        try {
            DateFormat format = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");
            Date startDate=format.parse(startDateStr+" 00:00:00");
            Date endDate=format.parse(endDateStr+" 23:59:59");
            
            Object[] values = {startDateStr,endDateStr,startDate,endDate};
            this.getSmartDAO().update("", SQL_STATIS_CHANNELTYPE, values);
            return true;
        } catch (Exception e) {
            e.printStackTrace();
            return false;
        }
    }
    
    
}

哎，又有新問題出現了，當用rollup進行統計時，如果原有數據中有null，在所有時它統計也過也是null於是就杯具了，如同第一個圖，經過搜索

將上述sql更改爲

INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) 
SELECT SEQ_PK_CHANNELTYPESTATIS.nextVal as uuid , P.* from    
(select count(*) AS counter, Decode(Grouping(channeltype),1,'所有渠道',channeltype) channeltype,Decode(Grouping(accountarea),1,'所有地區',accountarea) accountarea,sysdate AS statistime,to_char('2011-08-01')||'/'||to_char('2011-08-01') AS STATISRANGE    
FROM ACTIVITY WHERE txtime>=to_date('2011-08-01 00:00:00','yyyy-mm-dd hh24:mi:ss') AND txtime<=to_date('2011-08-01 23:59:59','yyyy-mm-dd hh24:mi:ss')  GROUP BY ROLLUP (channeltype,accountarea)) P

即可，其中搜索出的結果如圖2所示

問題還在：就是accountarea區域不能就合計，就是不能求得北京的all值，因此sql再次修改爲

INSERT INTO channeltypestatis(uuid,counter,channeltype,accountarea,statistime,statisrange) 

SELECT SEQ_PK_CHANNELTYPESTATIS.nextVal AS uuid , P.* FROM
(SELECT count(*) AS counter, Decode(Grouping(channeltype),1,'ALL',channeltype) channeltype,Decode(Grouping(accountarea),1,'ALL',accountarea) accountarea,sysdate AS statistime,to_char('2011-08-01')||'/'||to_char('2011-08-01') AS STATISRANGE 
FROM activity WHERE txtime>=to_date('2011-04-01','yyyy-mm-dd') AND txtime<=to_date('2011-08-01','yyyy-mm-dd')
GROUP BY CUBE (channeltype,accountarea)) P

注意將rollup換成了cube即可,關於cube和rollup的區別詳見：

http://space.itpub.net/519536/viewspace-610997

oralce分類統計

如何使用 JS 判斷用戶是否處於活躍狀態

Mono 支持LoongArch架構

lightdb秒級增加列和刪除列（not null帶默認值）

lightdb數據庫超時相關控制參數

通過HPA+CronHPA組合應對業務複雜彈性伸縮場景

❤️‍🔥 Solon Cloud Event 新的事務特性與應用

lightdb mysql 8.0兼容之不可見主鍵

使用 JS 實現在瀏覽器控制檯打印圖片 console.image()

基於Ubuntu-22.04安裝K8s-v1.28.2實驗（四）使用域名訪問網站應用

java shutdownhook

併發集合類的那些事

java基本類實現二級緩存

oralce分類統計

向AIX上傳.bin文件時最好用binary方式完成上傳

https://yachay.unat.edu.pe/blog/index.php?comment_area=format_blog&comment_component=blog&comment_co

linux以太網驅動總結