引用和读取表中的列

Question

I have a table which is something like this -

Column1     Column2       Column3
V1          V1             10
V2          V2             20
V3          V1+V2          30
V4          V2+V3          50
V5          V5             10
V6          V1+V4          60
V7          V1+V5+V3       50
...         ...            ...

I'm trying to create another table which identifies the dependencies dynamically. So my result should look something like

Column1      Dependent1      Dependent2     Dependent3
V3            V1              V2             null
V4            V3              V2             null
V6            V1              V4             null
V7            V1              V5              V3
...           ...             ...             ...

Is there a way to reference the same table and read the dependencies in sql.

Using Postgres and sqlServer

Answer 1

You can do this using XML -

create table  SourceTable   ( Column1 varchar(10) , Column2 varchar(30) ,      Column3 int)
go

insert into SourceTable
select 'V1'    ,      'V1'           ,  10
union all select 'V2'    ,      'V2'           ,  20
union all select 'V3'    ,      'V1+V2'        ,  30
union all select 'V4'    ,      'V2+V3'        ,  50
union all select 'V5'    ,      'V5'           ,  10
union all select 'V6'    ,      'V1+V4'        ,  60
union all select 'V7'    ,      'V1+V5+V3'     ,  50


 create table temp (
     Column1  varchar(10)       , Column2 xml  
)

insert into temp
select Column1 ,   '<Col>' + REPLACE( Column2,'+' , '</Col><Col>' ) +     '</Col>'  from SourceTable 

 select * , t1.Column2.value('(/Col)[1]','varchar(10)')Dependent1,   
 t1.Column2.value('(/Col)[2]','varchar(10)') Dependent2  ,
 t1.Column2.value('(/Col)[3]','varchar(10)') Dependent3
 from temp as t1

Answer 2

For Postgres:

select column1, 
       split_part(column2, '+', 1) as dependent1,
       split_part(column2, '+', 2) as dependent2,
       split_part(column2, '+', 3) as dependent3
from some_table;