基本语法：

<span style="font-size:18px;"><em>select ... from <TableName>where <Conditional-1>start with <Conditional-2>connect by <Conditional-3>;<Conditional-1>：过滤条件，用于对返回的所有记录进行过滤。<Conditional-2>：查询结果重起始根结点的限定条件。<Conditional-3>：连接条件</em></span>

数据表结构如下：

create table table_test(priorId number,id number,name varchar(5),description varchar(10));

测试数据：

insert into table_test(priorId,id,name,description) values(0,1,'f','f父类');insert into table_test(priorId,id,name,description) values(1,2,'f1','f子类1');insert into table_test(priorId,id,name,description) values(1,3,'f2','f子类2');insert into table_test(priorId,id,name,description) values(0,4,'F','F父类');insert into table_test(priorId,id,name,description) values(4,5,'F1','F子类1');insert into table_test(priorId,id,name,description) values(4,6,'F2','F子类2');

获取完整树：

select * from table_test start with priorId = 0 connect by prior id = priorId;

获取特定子树：

select * from table_test start with id = 1 connect by prior id = priorId;select * from table_test start with id = 4 connect by prior id = priorId;

如果connect by prior中的prior被省略，则查询将不进行深层递归：

select * from table_test start with priorId = 0 connect by id = priorId;select * from table_test start with id = 1 connect by id = priorId;

Oracle start with...connect by prior 实现递归查询