[Oracle SQL] NLS_SORT, NLS_COMP, NLSSORT, etc

假设我有如下一张表 test_sort_comp,   注意列NAME上有一个索引。

SQL> select * from test_sort_comp;

NAME
--------------------------------------------------
Frank
frank
frAnk
Bob
BOB
Candy
candy

7 rows selected.

SQL> select index_name, table_name,column_name from user_ind_columns where table_name = 'TEST_SORT_COMP';

INDEX_NAME                     TABLE_NAME                     COLUMN_NAME
------------------------------ ------------------------------ ------------------------------
INDX_TEST_SORT_COMP            TEST_SORT_COMP                 NAME

SQL>

 现在如果我需要查找Name为frank的记录，很简单，如下 （注意，设置了autotrace，方便查看执行计划）

SQL> set autotrace on
SQL> select * from test_sort_comp where name = 'frank';

NAME
--------------------------------------------------
frank


Execution Plan
----------------------------------------------------------
Plan hash value: 3593911111

----------------------------------------------------------------------------------------
| Id  | Operation        | Name                | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT |                     |     1 |    27 |     1   (0)| 00:00:01 |
|*  1 |  INDEX RANGE SCAN| INDX_TEST_SORT_COMP |     1 |    27 |     1   (0)| 00:00:01 |
----------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - access("NAME"='frank')

Note
-----
   - dynamic sampling used for this statement


Statistics
----------------------------------------------------------
          0  recursive calls
          0  db block gets
          1  consistent gets
          0  physical reads
          0  redo size
        410  bytes sent via SQL*Net to client
        396  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
          1  rows processed

SQL>

 很显然，查询结果返回了一条记录，而且采用了index scan的方式。 

如果现在我想得到name为frank的记录，但是不管字母的大小写，只要是f,r,a,n,k这几个字母按照这个顺序拼出来就可以，很显然表test_sort_comp中有3条记录满足条件 -- frank, Frank,  frAnk. 

很容易想到一种方法，通过upper或lower函数，把name转成要么是大写要么是小写，但是由于在name上的index是个普通索引(不是函数索引），因此会采用全表扫描的方式，如下，

SQL> select * from test_sort_comp where upper(name) = 'FRANK';

NAME
--------------------------------------------------
Frank
frank
frAnk


Execution Plan
----------------------------------------------------------
Plan hash value: 796376470

------------------------------------------------------------------------------------
| Id  | Operation         | Name           | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT  |                |     3 |    81 |     3   (0)| 00:00:01 |
|*  1 |  TABLE ACCESS FULL| TEST_SORT_COMP |     3 |    81 |     3   (0)| 00:00:01 |
------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - filter(UPPER("NAME")='FRANK')

Note
-----
   - dynamic sampling used for this statement


Statistics
----------------------------------------------------------
          5  recursive calls
          0  db block gets
         17  consistent gets
          0  physical reads
          0  redo size
        463  bytes sent via SQL*Net to client
        396  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
          3  rows processed

 那么可不可以通过改变参数NLS_SORT, NLS_COMP来做呢？ 可以尝试一下....

 但是如果单单设置NLS_SORT或NLS_COMP的话，是不起作用的，(NLS_COMP依赖于NLS_SORT， NLS_SORT作用于排序，

SQL> alter session set NLS_SORT = BINARY_CI;

Session altered.

SQL> select * from test_sort_comp where name = 'frank';

NAME
--------------------------------------------------
frank


Execution Plan
----------------------------------------------------------
Plan hash value: 3593911111

----------------------------------------------------------------------------------------
| Id  | Operation        | Name                | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT |                     |     1 |    27 |     1   (0)| 00:00:01 |
|*  1 |  INDEX RANGE SCAN| INDX_TEST_SORT_COMP |     1 |    27 |     1   (0)| 00:00:01 |
----------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - access("NAME"='frank')

Note
-----
   - dynamic sampling used for this statement


Statistics
----------------------------------------------------------
          0  recursive calls
          0  db block gets
          1  consistent gets
          0  physical reads
          0  redo size
        410  bytes sent via SQL*Net to client
        396  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
          1  rows processed

SQL>

接着设置下参数NLS_COMP为ANSI，表示在比较的时候参照NLS_SORT的值，

SQL> alter session set nls_comp = ansi;

Session altered.

SQL> select * from test_sort_comp where name = 'frank';

NAME
--------------------------------------------------
Frank
frank
frAnk


Execution Plan
----------------------------------------------------------
Plan hash value: 796376470

------------------------------------------------------------------------------------
| Id  | Operation         | Name           | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT  |                |     1 |    27 |     3   (0)| 00:00:01 |
|*  1 |  TABLE ACCESS FULL| TEST_SORT_COMP |     1 |    27 |     3   (0)| 00:00:01 |
------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - filter(NLSSORT("NAME",'nls_sort=''BINARY_CI''')=HEXTORAW('6672616E6B00') )

Note
-----
   - dynamic sampling used for this statement


Statistics
----------------------------------------------------------
          0  recursive calls
          0  db block gets
          8  consistent gets
          0  physical reads
          0  redo size
        463  bytes sent via SQL*Net to client
        396  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
          3  rows processed

 虽然结果返回了三条记录，但是这条SQL语句的执行计划显示进行的是Full Table Scan. 虽然没有改变SQL语句的写法，但是ORACLE在执行SQL语句的时候，根据刚才设置的参数NLS_SORT, NLS_COMP对SQL语句进行了改写，

  1 - filter(NLSSORT("NAME",'nls_sort=''BINARY_CI''')=HEXTORAW('6672616E6B00') )

其实相当于我们自己显示地用NLSSORT函数来查询，

首先清楚当前session的参数设置，只需要从新connect一下就OK了，

SQL> conn hr/hr;
Connected.
SQL> set autotrace on;

SQL> select * from test_sort_comp where NLSSORT(NAME,'NLS_SORT=BINARY_CI') = NLSSORT('frank', 'NLS_SORT=BINARY_CI');

NAME
--------------------------------------------------
Frank
frank
frAnk


Execution Plan
----------------------------------------------------------
Plan hash value: 796376470

------------------------------------------------------------------------------------
| Id  | Operation         | Name           | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT  |                |     1 |    27 |     3   (0)| 00:00:01 |
|*  1 |  TABLE ACCESS FULL| TEST_SORT_COMP |     1 |    27 |     3   (0)| 00:00:01 |
------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - filter(NLSSORT("NAME",'nls_sort=''BINARY_CI''')=HEXTORAW('6672616E6B00') )

Note
-----
   - dynamic sampling used for this statement


Statistics
----------------------------------------------------------
          4  recursive calls
          0  db block gets
         17  consistent gets
          0  physical reads
          0  redo size
        463  bytes sent via SQL*Net to client
        396  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
          3  rows processed

SQL>

因此通过设置NLS_SORT, NLS_COMP只是可以使得我们在不改变原有SQL的基础上来得到我们想要的结果，但是要注意SQL的执行计划会发生变化，有可能会对性能产生影响。

[Update on 2011-1-5]

本来打算写一篇关于NLS_SORT, NLS_COMP参数的设置的文章，结果发现之前已经写过了，记性好差:(   

关于上面讲到的是用NLS_SORT/NLS_COMP对查询性能的影响，如果情况允许，可以通过创建一个FBI来避免full table scan的发生，比如

create index fbi_tab on tab_name(nlssort(column_name, 'nls_sort=''binary_ci'''));

关于NLS_SORT 和 NLS_COMP这两个参数的介绍，可以参见oracle的官方文档，如下

NLS_SORT: http://download.oracle.com/docs/cd/B19306_01/server.102/b14237/initparams130.htm#REFRN10127

NLS_COMP:http://download.oracle.com/docs/cd/B19306_01/server.102/b14237/initparams120.htm#REFRN10117

需要注意的是，对NLS_SORT的介绍中有如下一段注释，

Setting NLS_SORT to anything other than BINARY causes a sort to use a full table scan, regardless of the path chosen by the optimizer. BINARY is the exception because indexes are built according to a binary order of keys. Thus the optimizer can use an index to satisfy the ORDER BY clause when NLS_SORT is set to BINARY. If NLS_SORT is set to any linguistic sort, the optimizer must include a full table scan and a full sort in the execution plan.

另外注意NLS_SORT和NLS_COMP 可以起修改的scope都是局限于session的，也就是只可以通过ALTER SESSION来改变这两个参数的设置。

SQL> alter system set nls_sort=binary;
alter system set nls_sort=binary
                 *
ERROR at line 1:
ORA-02096: specified initialization parameter is not modifiable with this option


SQL> alter system set nls_comp=ansi;
alter system set nls_comp=ansi
                 *
ERROR at line 1:
ORA-02096: specified initialization parameter is not modifiable with this option