全部插入子查询

Question

i want to insert all data in datashop where shopdata with single dataid for shopid , i have two table

1st data ,

 id 
 1
 2
 3
 4
 5

2nd datashop

 dataid   | shopid
 -----------------
    1     |    1 
    2     |    1 
    3     |    1 
    4     |    1 
    5     |    1 
    1     |    2 
    2     |    2 
    3     |    2

now i want insert all data in datashop with shopid

i try using a query but is not work ...

 INSERT INTO datashop( dataid , shopid ) 
 (Select id
 from (Select id from data
 Union all
 Select dataid from datashop )Z
 group by id as dataid , 3 as shopid) 

this query is not run show multiple errors

5 errors were found during analysis.

Unexpected token. (near "(" at position 49)

An expression was expected. (near "(" at position 71)

Unexpected token. (near "(" at position 71)

This type of clause was previously parsed. (near "Select" at position 72)

Unrecognized statement type. (near "all" at position 111)

Answer 1

This is the syntax you want:

insert into datashop(dataid, shopid) 
    Select id, 3 as shopid
    from (Select id from data
          Union all
          Select dataid from datashop
         ) Z
    group by id;

You had several small issues:

• You were listing columns after the query. This was probably the cause of the syntax problems.
• You had an alias in the group by clause (related to previous point).
• Your select did not have the appropriate number of columns.
• The parentheses around the select are unnecessary.

Also note: the query could be simplified to:

insert into datashop(dataid, shopid) 
    Select id, 3
    from data
    Union
    Select dataid, 3
    from datashop;

union removes duplicates.

Answer 2

insert into datashop(dataid, shopid)  
Select id as dataid, 3 as shopid
  from from data
Union 
Select id as dataid, 4 as shopid
  from from data
Union 
Select id as dataid, 5 as shopid
  from from data

Answer 3

The specification isn't very clear.

It looks like you have a table, let's call it (for sake of example) "shop" And that table contains a single column "id" (integer type?) and there are five rows in the table, with values 1 thru 5. It looks like the "id" column may be the primary key of the table, or at least guaranteed to be unique and non-null.

Then it looks like you have a second table which you want to populate with some rows. That table contains two columns, one named id, the other named shopid.

We're not given any idea of what this table represents in the model, so we're really just guessing. It looks to me like you might be looking for a Cartesian product, every shop.id matched to every other shop.id, including rows where the two columns are equal.

For the lack of another name, we'll refer to this second table as "shop_to_shop".

Again, this is all just a guess... but maybe you are looking to populate the shop_to_shop table with the result from a query something like this:

  SELECT r.id   AS id
       , s.id   AS shopid
    FROM shop r
   CROSS
    JOIN shot s
   ORDER
      BY r.id
       , s.id

If "id" is unique in "shop" (again, just a guess) there won't be any duplicate tuples in the result. If there are already some rows in the "shop_to_shop" table and we don't want to create any duplicates of rows that are already in the table, we can use a anti-join

  SELECT r.id   AS id
       , s.id   AS shopid
    FROM shop r
   CROSS
    JOIN shot s
    LEFT
    JOIN shop_to_shop t
      ON t.id     = r.id
     AND t.shopid = s.id
   WHERE t.id IS NULL
   ORDER
      BY r.id
       , s.id

Once we get a SELECT statement that's returning the rows that we want to insert, whatever that is, convert that into an INSERT statement by preceding it with "insert tablename (colname,colname) "

  INSERT INTO shop_to_shop (id, shopid)
  SELECT ...