oracle：更改表以使用计算列对行进行分类？

Question

I have this table Shoe :

Shoe size
---------
32
32
36
38
34
40
32

And I would like it to display Small for sizes 32-34, medium for 36 and big for 38 and above.

How should I change it using SQL?

I've done

ALTER TABLE Shoe 
  ADD Size varchar(50)

but I'm not sure how to do a if shoe size >= 32 and <= 34, size = small kind of command.

Is the only way to do it is to add them manually?

Answer 1

If you just want to view this classification of your data, you could just create a query or view using a CASE expression:

SELECT
    size,
    CASE WHEN size >= 32 AND size < 34 THEN 'small'
         WHEN size >= 34 AND size < 40 THEN 'medium'
         ELSE 'large' END AS label
FROM Shoe;

Answer 2

For Oracle:

Query to add computed column:

 ALTER TABLE shoe ADD  sizes varchar(10) AS (CASE WHEN shoesize in(32,33,34) THEN 'small'
           WHEN shoesize in(35,36,37) THEN 'medium'
           WHEN shoesize>=38 THEN 'big' END ) ;

Select statements:

 select * from shoe;

Output:

SHOESIZE	SIZES
32	small
32	small
36	medium
34	small
40	big
32	small
38	big

db<>fiddle here

If you are using sql server you can add an computed column as below:

Schema and insert statements:

 create table shoe(shoesize int)
 
 INSERT INTO shoe(shoesize) VALUES(32),(32),(36),(38),(34),(40),(32);

Query to add computed column:

  ALTER TABLE shoe ADD size AS (CASE WHEN shoesize in(32,33,34) THEN 'small'
          WHEN shoesize in(35,36,37) THEN 'medium'
          when shoesize>=38 THEN'big' END );

Select from Shoe:

 select * from shoe

Output:

shoesize	size
32	small
32	small
36	medium
38	big
34	small
40	big
32	small

db<>fiddle here

If you are using MySql then:

Schema and insert statements:

 create table shoe(shoesize int);
 
 INSERT INTO shoe VALUES (32);
 INSERT INTO shoe VALUES (32);
 INSERT INTO shoe VALUES (36);
 INSERT INTO shoe VALUES (38);
 INSERT INTO shoe VALUES (34);
 INSERT INTO shoe VALUES (40);
 INSERT INTO shoe VALUES (32);

Query to add computed column:

 ALTER TABLE shoe ADD COLUMN size VARCHAR(10) GENERATED ALWAYS AS (CASE WHEN shoesize in(32,33,34) THEN 'small'
          WHEN shoesize in(35,36,37) THEN 'medium'
          when shoesize>=38 THEN'big' END );

Select from shoe:

 select * from shoe

Output:

shoesize	size
32	small
32	small
36	medium
38	big
34	small
40	big
32	small

db<>fiddle here

Answer 3

As you mentioned ALTER , that's a DDL which modifies table's description . In that case, you'd use a virtual column.

For sample data you posted:

SQL> select * from shoe;

 SHOE_SIZE
----------
        32
        36
        38
        34
        40
        32

6 rows selected.

you'd then

SQL> alter table shoe add
  2    description varchar2(10) generated always as
  3      (case when shoe_size <= 34 then 'small'
  4            when shoe_size  > 34 and shoe_size <= 40 then 'medium'
  5            else 'large'
  6       end)
  7    virtual;

Table altered.

The result is:

SQL> select * from shoe;

 SHOE_SIZE DESCRIPTION
---------- -----------
        32 small
        36 medium
        38 medium
        34 small
        40 medium
        32 small

6 rows selected.

SQL>

---

However, that's probably not the best option you have. Why? Because you might change your mind and decide to modify descriptions. A simpler/better option would be to create a view:

SQL> create or replace view v_shoe as
  2    select shoe_size,
  3      case when shoe_size <= 34 then 'small'
  4           when shoe_size  > 34 and shoe_size <= 40 then 'medium'
  5           else 'large'
  6      end as description
  7  from shoe;

View created.

SQL> select * from v_shoe;

 SHOE_SIZE DESCRIPTIO
---------- ----------
        32 small
        36 medium
        38 medium
        34 small
        40 medium
        32 small

6 rows selected.

SQL>

---

Or, yet another option: a detail table (can't reference the shoe table via referential integrity constraint as shoe_size can't be primary nor unique key because it allows duplicates):

SQL> select * from shoe_description;

 SHOE_SIZE DESCRIPTION
---------- -----------
        28 extra small
        30 small
        32 small
        34 small
        36 medium
        38 medium
        40 medium
        42 large
        44 large
        46 large

10 rows selected.

SQL>
SQL> select s.shoe_size, d.description
  2  from shoe s join shoe_Description d on d.shoe_size = s.shoe_size;

 SHOE_SIZE DESCRIPTION
---------- -----------
        32 small
        32 small
        34 small
        36 medium
        38 medium
        40 medium

6 rows selected.

SQL>

---

Or, using a function which does the job; though, there's something potentially wrong with such an option - context switching (from SQL to PL/SQL) so performance might suffer if there are a LOT (really a lot) of shoes you're dealing with:

SQL> create or replace function f_shoe_desc(par_shoe_size in number)
  2    return varchar2
  3  is
  4    retval varchar2(20);
  5  begin
  6    retval := case when par_shoe_size <= 34 then 'small'
  7                   when par_shoe_size  > 34 and par_shoe_size <= 40 then 'medium'
  8                   else 'large'
  9              end;
 10    return retval;
 11  end;
 12  /

Function created.

SQL> select shoe_size, f_shoe_desc(shoe_size) description
  2  from shoe;

 SHOE_SIZE DESCRIPTION
---------- -----------
        32 small
        36 medium
        38 medium
        34 small
        40 medium
        32 small

6 rows selected.

SQL>

---

I think I prefer the view option.