一、mysql中的编码

mysql> show variables like 'collation_%';
+----------------------+-------------------+
| Variable_name | Value |
+----------------------+-------------------+
| collation_connection | latin1_swedish_ci |
| collation_database | latin1_swedish_ci |
| collation_server | latin1_swedish_ci |
+----------------------+-------------------+
3 rows in set (0.00 sec)
mysql> show variables like 'character_set_%';
+--------------------------+----------------------------+

| Variable_name | Value |
+--------------------------+----------------------------+
| character_set_client | latin1 |
| character_set_connection | latin1 |
| character_set_database | latin1 |
| character_set_filesystem | binary |
| character_set_results | latin1 |
| character_set_server | latin1 |
| character_set_system | utf8 |
| character_sets_dir | /usr/share/mysql/charsets/ |
+--------------------------+----------------------------+
8 rows in set (0.00 sec)

可见原来缺省是latin1编码，会导致中文乱码。

可以在mysql中设置编码，单个设置
mysql> alter database name character set utf8;
mysql> set character_set_connection=utf8;

Query OK, 0 rows affected (0.00 sec)
mysql> set character_set_connection=utf8;
Query OK, 0 rows affected (0.00 sec)
mysql> set character_set_results=utf8;
Query OK, 0 rows affected (0.00 sec)
mysql> set character_set_server=utf8;
Query OK, 0 rows affected (0.00 sec)
但重启后会失效。

可以修改配置文件：

[root@Hadoop48 ~]# vi /etc/my.cnf
[mysql]
default-character-set=utf8
[client]
default-character-set=utf8
[mysqld]
default-character-set=utf8
character_set_server=utf8
init_connect='SET NAMES utf8'

重启mysql，这样确保缺省编码是utf8

[root@Hadoop48 ~]# service mysqld restart

查看是否变成utf8：

mysql> \s
--------------
mysql Ver 14.12 Distrib 5.0.95, for redhat-linux-gnu (x86_64) using readline 5.1

Connection id: 2
Current database: toplists
Current user: root@localhost
SSL: Not in use
Current pager: stdout
Using outfile: ''
Using delimiter: ;
Server version: 5.0.95 Source distribution
Protocol version: 10
Connection: Localhost via UNIX socket
Server characterset: utf8
Db characterset: utf8
Client characterset: utf8
Conn. characterset: utf8
UNIX socket: /var/lib/mysql/mysql.sock
Uptime: 39 sec

Threads: 1 Questions: 12 Slow queries: 0 Opens: 15 Flush tables: 1 Open tables: 9 Queries per second avg: 0.308
--------------

mysql> show variables like "char%";
+--------------------------+----------------------------+
| Variable_name | Value |
+--------------------------+----------------------------+
| character_set_client | utf8 |
| character_set_connection | utf8 |
| character_set_database | utf8 |
| character_set_filesystem | binary |
| character_set_results | utf8 |
| character_set_server | utf8 |
| character_set_system | utf8 |
| character_sets_dir | /usr/share/mysql/charsets/ |
+--------------------------+----------------------------+
8 rows in set (0.00 sec)

mysql> show variables like "colla%";
+----------------------+-----------------+
| Variable_name | Value |
+----------------------+-----------------+
| collation_connection | utf8_general_ci |
| collation_database | utf8_general_ci |
| collation_server | utf8_general_ci |
+----------------------+-----------------+
3 rows in set (0.00 sec)

将mysql编码改成utf8，并在命令行中指定编码为utf8，执行导出报错：

[zhouhh@Hadoop46 ~]$ sqoop export --connect "jdbc:mysql://Hadoop48/toplists?useUnicode=true&characterEncoding=utf-8" -m 1 --table award --export-dir /user/hive/warehouse/hive_myaward/000000_0 --input-null-string "\\\\N" --input-null-non-string "\\\\N" --input-fields-terminated-by "\\01" --input-lines-terminated-by "\\n"

12/07/20 13:17:22 INFO mapred.JobClient: Task Id : attempt_201207191159_0233_m_000000_0, Status : FAILED
java.io.IOException: java.sql.SQLException: Incorrect string value: '\xE6\x9D\x80\xE7\xA0\xB4...' for column 'nick' at row 1

原来我创建的表缺省编码不是utf8的，也需改变一下。

mysql> ALTER TABLE award CONVERT TO CHARACTER SET utf8 COLLATE utf8_general_ci;
Query OK, 0 rows affected (0.00 sec)
Records: 0 Duplicates: 0 Warnings: 0

再导出

[zhouhh@Hadoop46 ~]$ sqoop export --connect "jdbc:mysql://Hadoop48/toplists?useUnicode=true&characterEncoding=utf-8" -m 1 --table award --export-dir /user/hive/warehouse/hive_myaward/000000_0 --input-null-string "\\\\N" --input-null-non-string "\\\\N" --input-fields-terminated-by "\\01" --input-lines-terminated-by "\\n"
...
12/07/20 14:03:10 INFO mapred.JobClient: map 0% reduce 0%
12/07/20 14:03:24 INFO mapred.JobClient: map 100% reduce 0%
...
12/07/20 14:03:29 INFO mapreduce.ExportJobBase: Exported 2 record

检查表
mysql> select * from award;
+-------------------------------+-----------+-----------+------+-----------+-----------+--------+------+-------------+-----------------+---------+
| rowkey | productid | matchid | rank | tourneyid | userid | gameid | gold | loginid | nick | plat |
+-------------------------------+-----------+-----------+------+-----------+-----------+--------+------+-------------+-----------------+---------+
| 2012-04-27 06:55:00:402713629 | 5947 | 433203828 | 2 | 4027102 | 402713629 | 1001 | NULL | 715878221 | 杀破天A | ios |
| 2012-04-27 06:55:00:406788559 | 778 | 433203930 | 19 | 4017780 | 406788559 | 1001 | 1 | 13835155880 | 亲牛牛旦旦 | android |
+-------------------------------+-----------+-----------+------+-----------+-----------+--------+------+-------------+-----------------+---------+
2 rows in set (0.00 sec)