Exp导出的数据dmp文件中的第2，3字节表示导出数据时的OS字符集ID(NLS_LANG环境变量保存)

下图中：

①表示目标OS的字符集

②表示目标DB的字符集

③表示目标DB的NCHAR字符集

④表示源OS的字符集

⑤表示源DB的NCHAR字符集

下图例子中，导出时OS字符集为ZHT16MSWIN950，目标OS字符集US7ASCII，导入后出现乱码。

将目标OS字符集①改为”AMERICAN_AMERICA.UTF8”或”TRADITIONAL CHINESE_TAIWAN.ZHT16MSWIN950”导入后正常。

此例中导出的数据文件在使用exp导出时，没有将OS字符集设定为和DB一致(最好在exp时将OS字符集NLS_LANG设定为和DB一致)，但” ZHT16MSWIN950”字符集正常解析了该DB中的所有字符，imp导入时要保证目标OS能正确解析数据文件中的所有字符，方法是设定目标OS的字符集与数据文件字符集(源OS字符集)一致，或为其超集。

-------------------------------例子------------------------------------------

[oracle@hi0-ibmsv508 u01]$ imp

Import: Release 10.2.0.1.0 - Production on Mon Apr 7 14:59:45 2008

Copyright (c) 1982, 2005, Oracle. All rights reserved.

Username: kevin

Password:

Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bit Production

With the Partitioning, OLAP and Data Mining options

Import file: expdat.dmp > CCIGPM.DMP

Enter insert buffer size (minimum is 8192) 30720>

Export file created by EXPORT:V10.01.00 via conventional path

Warning: the objects were exported by CCIGPM, not by you

import done in ①US7ASCIIcharacter set and③AL16UTF16NCHAR character set

import server uses ②AL32UTF8character set (possible charset conversion)

export client uses ④ZHT16MSWIN950character set (possible charset conversion)

export server uses ⑤UTF8NCHAR character set (possible ncharset conversion)

List contents of import file only (yes/no): no >

Ignore create error due to object existence (yes/no): no >

Import grants (yes/no): yes >

Import table data (yes/no): yes >------------------------------------------------------------------------------------------------------------