BlogJava-cuiyi's blog（崔毅 crazycy）-随笔分类-DBMS

非常有意思的sql排错

crazycy — Tue, 08 Apr 2008 10:30:00 GMT

1) 你看出区别在哪里了么？结果会一样么？
SELECT
(select first 1 longname from ivctrallocpurch a join jcivloc l on a.ivlocid=l.ivlocid and a.ctrid=A.ctrid) as mill
FROM CTR A
JOIN CTRSMRY CS ON (A.CTRID=CS.CTRID)
WHERE A.CTRID=3161

和

SELECT
(select first 1 longname from ivctrallocpurch a join jcivloc l on a.ivlocid=l.ivlocid and a.ctrid=3161) as mill
FROM CTR A
JOIN CTRSMRY CS ON (A.CTRID=CS.CTRID)
WHERE A.CTRID=3161

结果为什么不一样呢？

2） Integer overflow. The result of an integer operation caused the most significant bit of the result to carry.

select sum(rcpt.artothome * rate) as amt
from cashreceipt rcpt
join homerate rate on rcpt.currencyid=rate.currencyid

分析：

artothome @ arheader is : numeric(18,4)
rate @ .. is : numeric(10,10)

改进1 (not work)

select sum(cast(rcpt.artothome * rate as numeric(18,4))) as amt
from cashreceipt rcpt
join homerate rate on rcpt.currencyid=rate.currencyid

改进2 (works)

select sum(cast(cast(rcpt.artothome as numeric(18,4)) * cast(rate as numeric(18,4)) as numeric(18,4))) as amt
from cashreceipt rcpt
join homerate rate on rcpt.currencyid=rate.currencyid

reason, forward from http://www.firebirdfaq.org/faq207/

Integer overflow. The result of an integer operation caused the most significant bit of the result to carry.

Short explanation:
If you use fixed precision datatypes (smallint, integer, bigint, decimal and numeric), it is possible that the result of calculation doesn't fit the datatype. Try casting the values in complex expressions as double precision and see whether the error goes away. If it works and you don't care about being too precise, you can leave it at that. Otherwise you need to check every operation and calculate the result.

Details:
Here's an example: if you multiply 9.12 with 8.11 (both numeric(18,2)) you would get 73.9632. If Firebird would store that into numeric(18,2) datatype, we would lose 0.0032. Doesn't look much, but when you have complex calculations, you can easily loose thousands (dollars or euros). Therefore, the result is stored in numeric(18,4).

Problems are rarely seen with such low precision as 2. Let's use some bigger precision. For example, numeric(18,6) times numeric(18,6) yields numeric(18,12) result, meaning that maximal value it can store is 9223372.036854775807. If (for example) you wish to keep only 6 digits of precision, you could use something like:

cast(value1 as numeric(18,3)) * cast(value2 as numeric(18,3))

which would yield numeric(18,6) result, but it is quite possible that you would get more accurate result by casting to double:

cast(cast(value1 as double precision) * cast(value2 as double precision) as numeric(18,6))

Also, if you have mixed multiplications and divisions it helps to change the order of operations, so that the overflow doesn't happen.

String sql = "select mbrid from jcmbr where reference4=?";
假如没有记录
DynaBean aBean = CxcDataModule.getInstance().getRow(sql, new Object[]{reference4});
如果直接用aBean.get("reference4") 出错；
错误的原因是aBean 为null

Object obj = {spring jdbctemplate}.queryForObject(sql, new Object[]{reference4}, Integer.class);
if (obj instanceof Integer)
         System.out.println("(Integer)obj>>>>>>>>>>" + (Integer)obj);
else if (obj instanceof Map)
            System.out.println("(Integer)obj>>>>>>>>>>" + ((Map)obj).get("REFERENCE4"));
真是的流程是 if 分支
如果有数据，这是对的
如果没有数据，直接出错，因为有个假定有Integer值存在

int value = JcDataModuleUtils.getJdbcTemplate().queryForInt(sql, new Object[]{reference4});
出错原因是假定一定会有一个int值返回
15:09:00,438 ERROR [STDERR] Caused by: org.springframework.dao.IncorrectResultSizeDataAccessException: Incorrect result size:
expected 1, actual 0
15:09:00,438 ERROR [STDERR]     at org.springframework.dao.support.DataAccessUtils.requiredUniqueResult(DataAccessUtils.java:
66)
15:09:00,469 ERROR [STDERR]     at org.springframework.jdbc.core.JdbcTemplate.queryForObject(JdbcTemplate.java:620)
15:09:00,469 ERROR [STDERR]     at org.springframework.jdbc.core.JdbcTemplate.queryForObject(JdbcTemplate.java:629)
15:09:00,469 ERROR [STDERR]     at org.springframework.jdbc.core.JdbcTemplate.queryForInt(JdbcTemplate.java:656)

crazycy 2008-04-08 18:30 发表评论

mysql乱码编码的解决之道(不含分析)

crazycy — Sun, 23 Dec 2007 10:07:00 GMT

mysql的中文乱码,大多是因为mysql的编码选择造成的。
以mysql的两个相同结构的表间进行数据迁移为例，做了三种情况的测试，并提出解决乱码的解决方法如下:

Migration between MySQLs

1) Migration between character set : utf8

Amend my.ini (2 positions)

default-character-set=latin1

default-character-set=utf8

2) Migration between character set : origin is utf8 ==> the destination is latin1

Alter destination:

alter database [db_name] default character set utf8;
alter table [table_name] default character set utf8;
alter table [table_name] change [field_name] [field_name] [filed type] varchar(45) character set utf8;

example

users(id, user_name)

alter table users change user_name user_name varchar(45) character set utf8;

3) Migration between character set : origin is latin1 ==> the destination is utf8

If the origin is latin1, cannot insert Chinese Character.

crazycy 2007-12-23 18:07 发表评论

firebird技巧和错误分析与总结

crazycy — Sat, 03 Nov 2007 09:06:00 GMT

1）
Cmd >> isql localhost/3050:path/cxcdata.fdb -user sysdba -pass masterkey
Error >>
Statement failed, SQLCODE = -902

Unable to complete network request to host "localhost".
-Failed to establish a connection.
-No connection could be made because the target machine actively refused it.

Use CONNECT or CREATE DATABASE to specify a database
Solution>>
check whether service or application of firebird start or not

2)
Scenario>>Using EMS Data Export for Interbase/Firebird
Error>>

Solution>>windows\system32\drivers\etc目录下的services文件中添加
                gds_db 3050/tcp

3) windows下firebird对db路径要求：与firebird直接交互的fdb文件必须增加前缀：localhost/3050:[driver]/{[path]/}[.fdb file]
    无论是isql 还是 gbak
    比如 test.fdb 在d:\dbfile\firebird\test.fdb
    connect db: cmd>> isql localhost/3050:d:/dbfile/firebird/test.fdb -user [username] -pass [userpass]
    backup db: cmd>> gbak -B localhost/3050:d:/dbfile/firebird/test.fdb test.fbk -user [username] -pass [userpass]
    restore db: cmd>> gbak -C test.fbk localhost/3050:d:/dbfile/firebird/test.fdb -user [username] -pass [userpass]
    NOTE: only the fdb file should add the localhost constraint


4) db installed in d:/dbfile/firebird/test.fdb
    connect way 1>> isql localhost/3050:d:/dbfile/firebird/test.fdb -user [username] -pass [userpass]
    connect way 2>> isql localhost/3050:db -user [username] -pass [userpass]
                               add this line : db = d:/dbfile/firebird/test.fdb in aliases.conf, no need restart firebird service

5）about amention of db's path for different customer
we offten configure db path in apps using jdbc:firebird:localhost:[path]/[db file]   like jdbc:firebird:localhost: d:/dbfile/firebird/test.fdb
As you know, in most time we install db in different folder for different customers; so we must keep re-compile apps or modify war/jar/ear every time.

So basis on this issue, we can change to configure db using firebird self file called aliases.conf
like this form (basis my local configuration):
test.fdb = d:/dbfile/firebird/test.fdb

so now using this form jdbc:firebird:localhost:test.fdb in apps
I tried both windows and linux, it works;

crazycy 2007-11-03 17:06 发表评论

得到PrepareStatement最终执行的sql语句的方法

crazycy — Sat, 22 Jul 2006 13:53:00 GMT

在CSDN的JAVA基础版,常常有人问及如何得到PreparedStatement最终执行的SQL语句;或者如何在控制台输出占位符的真实值.....

原因就是PreparedStatement执行的sql语句有大量的占位符?....

问题诸如JDBC中:

如何得到 conn.prepareStatement 最终执行的sql语句。
sql="update table1 set a=?,b=?"
stmt = con.prepareStatement(sql);
stmt.setObjec t(1,"a");
stmt.setObjec t(2,"b");

希望可以通过stmt或者conn 得到：
update table1 set a='a',b='b'

亦或Hibernate中

如我执行：find("select * from t_table where id = ?",new Integer(5));
在控制台显示SQL时只显示：select * from t_table where id = ?
如何才能做到将控制台显示的占位符用其真实的值来替换？
即控制台输出时显示：select * from t_table where id = 5

无它，无论JDBC还是Hiberante都不提供默认解决方案，但是参数是设置进去的，我们在设置的过程中可以有充分的理由来截取并获得自己想要的东西，类似于AOP理论。

共享我在工程中的使用方法：
插入操作：

/**
     * 执行插入数据库的语句
     * @param sql
     * @param params
     * @return 返回生成的主键
     */
    public int executeInsert(String sql, Object[] params) {
        Connection conn = null;
        PreparedStatement pstmt = null;
        ResultSet rs = null;
        try {
            //1 获得连接
            conn = MyDBConnection.getInstance().getConnection();
            //2 设置提交方式为程序控制
            conn.setAutoCommit(false);
            //3 获得语句对象
            pstmt = conn.prepareStatement(sql, Statement.RETURN_GENERATED_KEYS);
            //4 设置SQL语句的参数
            if (null != params && 0 < params.length) {
                setParams(pstmt, params);
            }
            //5 打印SQL语句
            if (MyDBConstants.showSQL) {
                getPreparedSQL(sql, params);
            }
            //6 执行语句
            pstmt.executeUpdate();
            //7 程序提交
            conn.commit();
            //8 返回生成的主键
            rs = pstmt.getGeneratedKeys();
            int generatedKey = 0;
            if (rs.next()) {
                generatedKey = rs.getInt(1);
            }
            if (0 < generatedKey)
                throw new MySQLException("插入记录时出错");
            return generatedKey;
        } catch (SQLException e) {
            //回滚
            MyDBUtil.rollBack(conn);
            throw new MySQLException(e);
        } finally {
            //关闭打开的操作
            MyDBUtil.close(conn, pstmt, rs);
        }
    }

更新查找操作：

/**
     * 执行更新或者删除数据库的语句
     * @param sql
     * @param params
     * @return 返回执行成功与否
     */
    public boolean executeUpdateDel(String sql, Object[] params) {
        boolean isSuccess = false;
        Connection conn = null;
        PreparedStatement pstmt = null;
        try {
            //1 获得连接
            conn = MyDBConnection.getInstance().getConnection();
            //2 设置提交方式为程序控制
            conn.setAutoCommit(false);
            //3 获得语句对象
            pstmt = conn.prepareStatement(sql);
            //4 设置SQL语句的参数
            if (null != params && 0 < params.length) {
                setParams(pstmt, params);
            }
            //5 打印SQL语句
            if (MyDBConstants.showSQL) {
                getPreparedSQL(sql, params);
            }
            //6 执行语句
            pstmt.executeUpdate();
            //7 程序提交
            conn.commit();
            //8 设置语句执行的标记
            isSuccess = true;
        } catch (SQLException e) {
            //回滚
            MyDBUtil.rollBack(conn);
            throw new MySQLException(e);
        } finally {
            //关闭打开的操作
            MyDBUtil.close(conn, pstmt);
        }
        return isSuccess;
    }

执行查询

1 /**
2      * 执行查询数据库的语句;
9      *
10      * @return
11      */
12     public Object executeQuery(String sql, Object[] params) {
13         Connection conn = null;
14         PreparedStatement pstmt = null;
15         ResultSet rs = null;
16         try {
17             //1 获得连接
18             conn = MyDBConnection.getInstance().getConnection();
19             //2 设置提交方式为程序控制
20             conn.setAutoCommit(false);
21             //3 获得语句对象
22             pstmt = conn.prepareStatement(sql, ResultSet.TYPE_SCROLL_SENSITIVE, ResultSet.CONCUR_UPDATABLE);
23             //4 设置SQL语句的参数
24             if (null != params && 0 < params.length) {
25                 setParams(pstmt, params);
26             }
27             //5 打印SQL语句
28             if (MyDBConstants.showSQL) {
29                 getPreparedSQL(sql, params);
30             }
31             //6 执行语句
32             rs = pstmt.executeQuery();
33
34             //9 程序提交
35             conn.commit();
36
37             //10 获得记录
38             Object vo = new Object();
39             if (null != rs && rs.next()) {
40                 vo = rs2vo(rs);
41             }
42 //            return results;
43             return vo;
44         } catch (SQLException e) {
45             //回滚
46             MyDBUtil.rollBack(conn);
47             throw new MySQLException(e);
48         } finally {
49             //关闭打开的操作
50             MyDBUtil.close(conn, pstmt, rs);
51         }
52     }

看到getPreparedSQL(sql, params)了么？这个地方就是要实现我们预期效果的地方：

得到PrepareStatement最终执行的sql语句的方法

然后轻松核实你的控制台或者日志文件吧......

crazycy 2006-07-22 21:53 发表评论

An internal error occurred during: "Generating Artifacts"

crazycy — Thu, 22 Jun 2006 07:41:00 GMT

    今天是用Eclipse3.1.2+MyEclipse4.1.1+MySQL5.0.12，用Database Explorer创建Hibernate3所需的HBM映射文件和VO对象时，居然报An internal error occurred during: "Generating Artifacts" 这个错误。

    google了一把，看到Blog（软道，url：http://blog.csdn.net/langtaojin/archive/2006/03/28/641722.aspx）介绍如下：
今天在使用myEclipse 的　Database Explorer 视图中Create Hibernate Mapping时，出现了“An internal error occurred during: "Generating Artifacts".”错误提示。费了好大劲，终于解决了。据myEclipse官方网站上的介绍，发生该错误主要有两种情况，一是你eclipse上安装了jboss公司的hibernate tools plugin for eclipse,这些插件与myEclipse发生冲突，解决方法是删除这些插件。二是你在没有卸载以前myEclipse版本的情况下安装了新的 myEclipse版本，解决方法是把myEclipse卸载后重新安装，我遇到的情况就属于每二种情况，重装myEclipse后问题解决了。

    第一个问题不存在，然后咣咣的卸载、安装了MyEclipse4.1.1。问题依然。

    手工建立测试表，却没有这个问题。突然想到了可能是数据表建立的格式有问题：我用PD直接把物理实体模型导入到Mysql中。就它了。通过MyCC（当然也可以通过命令行 mysql -h localhost -u root -p）把PD生成的SQL语句重新运行，OK，问题解决。

    结合上面的blog，可以看出出现这个问题具有三种情况了。

    当然或许你用Oracle，也可能遇到这个问题，如果以上三种情况都没有解决了An internal error occurred during: "Generating Artifacts" 这个错误。你注意两点：1PLSQL工具的问题，建议直接命令行。2Orcle的class12驱动的版本，建议换新版本。

crazycy 2006-06-22 15:41 发表评论