| id | flight_manager(执管单位) | sub_feet（子机型） | all_match（是否匹配） | fleet（主机型） |
| 1 | TAO | 319-320 | 1 | 320 |
| 2 | KHN | 319-320-321-323 | 0 | 320 |
| 3 | 321-323-320-325 | 1 | 320 |
如上表,需求：现在有个机型 320 ，执管单位KHN 要去上表中匹配数据
规则是：allmatch ==1 是要匹配 subfeet 字段里的值 查询出subfeet的值
allmatch ==0 是要匹配 feet 字段里的值 查询出feet的值，我现在是分2次查询，如下：
sql1：
select sub_fleet from t_flight_manager_fleet t where 1 = 1
and t.flight_manager = 'KHN'
and t.all_match = '1' and t.sub_fleet like '%320%'；
sql2：
select fleet from t_flight_manager_fleet t1 where 1 = 1
and t1.flight_manager = 'KHN'
and t1.all_match = '0' and t1.fleet='320';
现在的解决方案就是 去库里分2个sql查询，将查询的结果合并到一个list中再做处理，但考虑这种方式要查询2次数据库，如果数据量较大的话开销比较到，运行也慢，请问有没有写一条sql满足我的需求？（注意：不能使用union,union all）请大家指点一下。
直接用case是不是你想要的结果呢？或者Oracle的话直接用decode：
[code="sql"]
select case t.all_match when '1' then t.sub_fleet when '0' then t.fleet
from t_flight_manager_fleet t
where 1 = 1
and t.flight_manager = 'KHN'
and t.sub_fleet like '%320%'
and t.fleet = '320';
[/code]
or
[code="sql"]
select decode(t.all_match, '1', t.sub_fleet, '0', t.fleet) fleet
from t_flight_manager_fleet t
where 1 = 1
and t.flight_manager = 'KHN'
and t.sub_fleet like '%320%'
and t.fleet = '320';
[/code]
select decode(t.all_match,'1',t.sub_fleet,'0',t.fleet,'无匹配') as fleet
from t_flight_manager_fleet t
where 1 = 1
and t.flight_manager = 'KHN'
and (t.sub_fleet like '%320%' or t.fleet='320')