Recently, when converting a mysql project to an oracle project, I encountered a situation where the mysql paging limit cannot be used.
, in the end, only the paging form of oracle can be used, and the rectification of the statements one by one
Mysql pagination uses the limt keyword
select * from t_order limit 5,10; #Return the data of lines 6-15 The first parameter refers to the place to start, and the second parameter refers to how many pieces of data are displayed on each page; Note: The first page is represented by 0 .
select * from t_order limit 5; #return the first 5 rows
select * from t_order limit 0,5; #return the first 5 rows
Mssql 2000 paging uses the top keyword (versions above 20005 also support the keyword rownum)
Select top 10 * from t_order where id not in (select id from t_order where id>5 ) ; //return data from rows 6 to 15
10 means fetching 10 records, 5 means fetching from the 5th record
Oracle paging
①Use rownum keyword ( three levels of nesting )
SELECT * FROM ( SELECT A.*, ROWNUM num FROM ( SELECT * FROM t_order ) A WHERE ROWNUM <=15 ) WHERE num >=5; - - Return rows 5-15
②Using the row_number parsing function for paging ( more efficient )
SELECT xx.* FROM( SELECT t.*,row_number() over(ORDER BY o_id)AS num FROM t_order t )xx WHERE num BETWEEN 5 AND 15; --返回第5-15行数据
Formats available for parsing functions
function() over(pertion by field order by field);
Pertion partitions by a field
Order Sort by fields
Paging query
The database table structure and records are as follows:
1. According to rowid:
16:31:48 SQL> select * from passvehicleinfo p where rowid in ( select rid from (select rownum rn,rid from (select p.rowid rid,p.passvehicleid from passvehicleinfo p order by p.passvehicleid desc) view1 where rownum<10000) view2 where rn >9980)order by p.passvehicleid asc;
2. According to the analysis function
17:02:42 SQL> select * from (select p.*,row_number() over (order by p.passvehicleid desc ) rk from passvehicleinfo p) where rk>9980 and rk<10000;
3. According to rownum
17:07:38 SQL> select * from (select view1.*,rownum rn from (select p.* from passvehicleinfo p order by p.passvehicleid desc) view1 where rownum<10000) view2 where rn>9980;
The following mainly introduces the third type: according to rownum
1. rownum paging
SELECT * FROM emp;
2. Display rownum[oracle allocated]
SELECT e.*, ROWNUM rn FROM (SELECT * FROM emp) e;
rn is equivalent to the ID number of the row assigned by Oracle
3. Pick out 6-10 records
First find 1-10 records
SELECT e.*, ROWNUM rn FROM (SELECT * FROM emp) e WHERE ROWNUM <= 10;
If it is not possible to add rownum>=6 after it,
4. Then find 6-10 records
SELECT * FROM (SELECT e.*, ROWNUM rn FROM (SELECT * FROM emp) e WHERE ROWNUM <= 10) WHERE rn >= 6;
5. Several query changes
a. Specify the query column, only need to modify the innermost subquery
Query only employee number and salary
SELECT * FROM (SELECT e.*, ROWNUM rn FROM (SELECT ename, sal FROM emp)
e WHERE ROWNUM <= 10) WHERE rn >= 6;
b. Sort query, only need to modify the innermost subquery
Query 6-10 pieces of data after salary sorting
SELECT * FROM (SELECT e.*, ROWNUM rn FROM (SELECT ename, sal FROM emp ORDER
by sal) e WHERE ROWNUM <= 10) WHERE rn >= 6;
Paging is often used when displaying record entries. A common method is to use the positioning interface that comes with various databases to rewrite the original query statement, so that only certain records in a specific range are retrieved. Different databases have different query and positioning interfaces. The following is a summary:
| database |
Paging query statement |
illustrate |
| MySql |
"QUERY_SQL limit ?,?" | Using the limit keyword, the first "?" is the starting line number, and the second "?" is the number of returned items |
| Oracle |
SELECT * FROM |
结合rownum关键字，利用嵌套三层select 语句实现。第一个"?"表示终止行号， 第二个"?"表示其实行号 |
| Sql Server |
尚无通用语句 | 可使用top n来返回前n条记录或使用存储过程 |
| DB2 |
假设查询语句：select t1.* from t1 order by t1.id； 分页语句可为： "select * from ( select rownumber() over (order by t1.id) as row_, t1.* from t1 order by t1.id) as temp_ where row_ between ?+1 and ?" |
返回两个"?"之间的记录 |
| InterBase |
“QUERY_SQL row ? to ?” | 返回两个"?"之间的记录 |
| PostgreSQL | “QUERY_SQL limit ? offset ?” | 第一个"?"为起始行号，第二个"?"代表 返回记录数 |
SQL Server
关于分页 SQL 的资料许多，有的使用存储过程，有的使用游标。本人不喜欢使用游标，我觉得它耗资、效率低；使用存储过程是个不错的选择，因为存储过程是颠末预编译的，执行效率高，也更灵活。先看看单条 SQL 语句的分页 SQL 吧。
方法1：
适用于 SQL Server 2000/2005
SELECT TOP 页大小 * FROM table1 WHERE id NOT IN ( SELECT TOP 页大小*(页数-1) id FROM table1 ORDER BY id ) ORDER BY id
方法2：
适用于 SQL Server 2000/2005
SELECT TOP 页大小 * FROM table1 WHERE id > ( SELECT ISNULL(MAX(id),0) FROM ( SELECT TOP 页大小*(页数-1) id FROM table1 ORDER BY id ) A ) ORDER BY id
方法3：
适用于 SQL Server 2005
SELECT TOP 页大小 * FROM ( SELECT ROW_NUMBER() OVER (ORDER BY id) AS RowNumber,* FROM table1 ) A WHERE RowNumber > 页大小*(页数-1)
说明，页大小：每页的行数；页数：第几页。使用时，请把“页大小”以及“页大小*(页数-1)”替换成数码。
MYSQL
SELECT * FROM TT LIMIT 1,20
SELECT * FROM TT LIMIT 21,30
/*
如果你是几千上万数据，就直接使用mysql自带的函数 limit的普通用法就ok了，如果是100万以上的数据，可能就要讲方法了，下面我们来做个百万级数据的分页查询语句.
mysql> select * from news where id>=(select id from news limit 490000,1) limit 10; //0.18 sec //很 明显,这 种方式胜出 .
mysql> select * from news limit 490000,10 //0.22 sec;
*/
以下的文章主要介绍的是MySQL分页的实际操作方案，其实关于实现MySQL分页的最简单的方法就是利用利用mysql数据库的LIMIT函数,LIMIT [offset,] rows可以从MySQL数据库表中第M条记录开始检索N条记录的语句为：
- SELECT * FROM 表名称 LIMIT M,N
例如从表Sys_option（主键为sys_id)中从第10条记录开始检索20条记录，语句如下：
- select * from sys_option limit 10,20
- select * from table [查询条件] order by id limit ?,?
Oracle
Oracle的分页查询语句基本上可以按照这篇了，下一篇文章会通过例子来申述。下面简单讨论一下多表联合的情况。对最多见的等值表连接查询，CBO 一般可能会采用两种连接方式NESTED LOOP以及HASH JOIN（MERGE JOIN效率比HASH JOIN效率低，一般CBO不会考虑）。在这里，由于使用了分页，因此指定了一个归回的最大记载数，NESTED LOOP在归回记载数跨越最大值时可以顿时遏制并将结果归回给中心层，而HASH JOIN必需处理完所有成集（MERGE JOIN也是）。那么在大部分的情况下，对分页查询选择NESTED LOOP作为查询的连接方法具有较高的效率（分页查询的时候绝大部分的情况是查询前几页的数据，越靠后面的页数访问概率越小）。
因此，如果不介意在体系中使用HINT的话，可以将分页的查询语句改写为：
SELECT /*+ FIRST_ROWS */ * FROM
(
SELECT A.*, ROWNUM RN
FROM (SELECT * FROM TABLE_NAME) A
WHERE ROWNUM <= 40
)
WHERE RN >= 21