求一个SQL查询写法

表如下：
表1
brand  codekey  content1
a         1             ccc
a         2             bbb
a         3             ddd
b         5             aaa
...
表2
brand  codekey  content2
a        1              xxx
a        1              yyy
a        2              zzz
b        8              hhh
...

我想查询的是，当条件brand = 'a' 时，查询结果如下：
brand  codekey  content
a         1             cccxxxyyy
a         2             bbbzzz
a         3             ddd

解释一下，返回的结果是key中不应该有重复的，对应的content列是相应的codekey的content1中的所有与content2中所有的字符串连接。

问题已解决，贴出解决办法，以供参考：

select  brand, 
       codekey, 
        replace (path,  '   ' ,  '' )  as  content 
from  (  select  brand, 
              codekey, 
               max (SYS_CONNECT_BY_PATH(content,  '   ' ))  as  path 
        from  (  select  brand, 
                     codekey, 
                     content,
                     (row_number()  over ( order   by  brand, codekey, content)  +  dense_rank()  over ( order   by  brand, codekey)) rn, 
                      min (content)  over (partition  by  brand, codekey) content1 
               from  (  select  brand, 
                            codekey, 
                            content1  as  content 
                      from  table1 
                      union   all   select  brand, 
                                      codekey, 
                                      content2  as  content 
                                from  table2 
                                order   by  brand, codekey
                   )
            ) 
       start  with  content  =  content1 
       connect  by  prior rn  =  rn  -   1  
        group   by  brand, codekey 
        having  brand  =   ' a '
)