方法一：建一个拼音表 t_cosler ，存放每个字母开头的第一个汉字的编号和最后一个汉字的编号。
BatchFile code
+------+--------+-------+
| f_PY | cBegin | cEnd |
+------+--------+-------+
| A | 45217 | 45252 |
| B | 45253 | 45760 |
.....
| Z | 54481 | 55289 |
+------+--------+-------+
然后直接查询就行了。
SQL code
mysql> create table t_cosler(
-> f_PY char primary key,
-> cBegin SMALLINT UNSIGNED not null,
-> cEnd SMALLINT UNSIGNED not null
-> );
Query OK, 0 rows affected (0.09 sec)
mysql> insert into t_cosler values
-> ('A',0xB0A1,0xB0C4),
-> ('B',0xB0C5,0xB2C0),
-> ('C',0xB2C1,0xB4ED),
-> ('D',0xB4EE,0xB6E9),
-> ('E',0xB6EA,0xB7A1),
-> ('F',0xB7A2,0xB8C0),
-> ('G',0xB8C1,0xB9FD),
-> ('H',0xB9FE,0xBBF6),
-> ('J',0xBBF7,0xBFA5),
-> ('K',0xBFA6,0xC0AB),
-> ('L',0xC0AC,0xC2E7),
-> ('M',0xC2E8,0xC4C2),
-> ('N',0xC4C3,0xC5B5),
-> ('O',0xC5B6,0xC5BD),
-> ('P',0xC5BE,0xC6D9),
-> ('Q',0xC6DA,0xC8BA),
-> ('R',0xC8BB,0xC8F5),
-> ('S',0xC8F6,0xCBF9),
-> ('T',0xCBFA,0xCDD9),
-> ('W',0xCDDA,0xCEF3),
-> ('X',0xCEF4,0xD188),
-> ('Y',0xD1B9,0xD4D0),
-> ('Z',0xD4D1,0xD7F9);
Query OK, 23 rows affected (0.16 sec)
Records: 23 Duplicates: 0 Warnings: 0
mysql> select * from o_personnel;
+------+------------+
| A_Id | A_UserName |
+------+------------+
| 1 | 首先 |
| 2 | 检查 |
| 3 | 我们 |
| 4 | 的二 |
| 5 | 进制 |
| 6 | 是否 |
| 7 | 适合 |
| 8 | 你的 |
| 9 | 平台 |
+------+------------+
9 rows in set (0.00 sec)
mysql> select p.*,c.*
-> from o_personnel p , t_cosler c
-> where CONV(HEX(left(A_UserName,1)),16,10) between c.cBegin and c.cEnd;
+------+------------+------+--------+-------+
| A_Id | A_UserName | f_PY |
SQL #1: select * from uchome_feed where uid in('0',501...) order by dateline desc limit 0,50;
SQL #2: select * from uchome_feed where uid in(0,501...) order by dateline desc limit 0,50;
SQL #1 是CSDN 个人空间获取个人动态的SQL，属UCH原版SQL。
SQL #2 是去掉单引号(SQL #1中红色部分)后的SQL，两� ......
今天这么激动又想写文章的原因是MySQL5.1的发布带来了设计超强动力数据库的强有力的武器，任何MySQL的DBA都应该尽快学习并使用它。我觉得如果能很好滴使用这个5.1版带来的新特性，DBA可以使自己管理的VLDB（不知道什么是VLDB？告诉你，是好大好大的数据库的意思，Very Large DB）或数据仓库奇迹般的获得巨大的性能提升 ......
