表中 null 而不是 null 列的案例表达式

Question

My table:

CREATE TABLE StudentScore (
    Student_ID INT,
    Student_Name NVARCHAR (50),
    Student_Score INT) 
GO

INSERT INTO StudentScore VALUES (1,'Ali', NULL)
INSERT INTO StudentScore VALUES (2,'Zaid', 770)
INSERT INTO StudentScore VALUES (3,'Mohd', 1140)
INSERT INTO StudentScore VALUES (4,NULL, 770)
INSERT INTO StudentScore VALUES (5,'John', 1240)
INSERT INTO StudentScore VALUES (6,'Mike', 1140)
INSERT INTO StudentScore VALUES (7,'Goerge', NULL)

Query tried

select * from StudentScore
Select TYPE1 =
CASE WHEN ANY(SELECT COLUMN IS NULL) THEN 'AT least 1 NULL'
    ELSE 'NON-NULL'
END

Basically I want that if there is any single null bvalue for any column in table StudentScore, then the type of that column should be null else it should be not null (note that this is part of interview question and I cannot use information_schema etc. I need to do this using case. Can anyone help

For eg here ID will be NON-NULL, rest two will be type 'At least 1 null'

EDIT after seeing answers to clarify:

I want that my code should check all rows of columns and return 'Non-null' if all the rows in a column are not null. All the columns should be checked individually.

For eg this code gives output as below:

select case when Student_Score is null then 'non-null'
else 'non-null'
end TYPE1 from StudentScore

TYPE1
non-null
non-null
non-null
non-null
non-null
non-null
non-null

The above is not my desired output .

My desired output is

Not null columns : ID, At least 1 null value (in all the rows corresponding to one column): Student_Score, Student_name.

So the code should return 'at least 1 null value' if for a particular column, there is at least one null value present in all the rows .

For eg it should check all 8 rows corresponding to each column and if there is no null value in all rows corresponding to one column, then only that column is going to 'Not null'

Also I removed primary key to make question more generic.

Answer 1

You can't use the all columns operator (*) in this case
you could try using a case when with OR condition for each explicit named column

    select case when col1 is null 
            OR col2 is null 
            OR  col3 is NULL then 'AT least 1 NULL'
            ELSE 'NON-NULL' END type
    from StudentScore

in your case

    select Student_ID, Student_Name, Student_Score
    , case when Student_Name is null 
                OR Student_Score is null 
                then 'AT least 1 NULL'
                ELSE 'NON-NULL' END type
        from StudentScore

Answer 2

I guess you want 1 row and 1 column as result, so you can use EXISTS:

select 
  case 
    when exists (select 1 from StudentScore where Student_Name is null) then 'AT least 1 NULL'
    when exists (select 1 from StudentScore where Student_Score is null) then 'AT least 1 NULL'
    else 'NON-NULL'
  end TYPE1 

or:

select 
  case 
    when exists (select 1 from StudentScore where Student_Name is null)
    or exists (select 1 from StudentScore where Student_Score is null) then 'AT least 1 NULL'
    else 'NON-NULL'
  end TYPE1

See the demo .
Result:

> | TYPE1           |
> | :-------------- |
> | AT least 1 NULL |

If you want 1 result for each of the columns:

select 
  case when exists (select 1 from StudentScore where Student_ID is null) then 'AT least 1 NULL' else 'NON-NULL' end ID,
  case when exists (select 1 from StudentScore where Student_Name is null) then 'AT least 1 NULL' else 'NON-NULL' end Student_Name,
  case  when exists (select 1 from StudentScore where Student_Score is null) then 'AT least 1 NULL' else 'NON-NULL' end Student_Score

See the demo .
Results:

> ID       | Student_Name    | Student_Score  
> :------- | :-------------- | :--------------
> NON-NULL | AT least 1 NULL | AT least 1 NULL

Or if you want 2 rows, 1 for each type and the column names as a comma separated list:

select type, string_agg(colname, ',') columns
from (
  select 'id' colname, case when exists (select 1 from StudentScore where Student_ID is null) then 'AT least 1 NULL' else 'NON-NULL' end type
  union all
  select 'Student_Name', case when exists (select 1 from StudentScore where Student_Name is null) then 'AT least 1 NULL' else 'NON-NULL' end
  union all
  select 'Student_Score', case when exists (select 1 from StudentScore where Student_Score is null) then 'AT least 1 NULL' else 'NON-NULL' end
) t
group by type

This code works for SQL Server 2017+.
See the demo .
Results:

> type            | columns                   
> :-------------- | :-------------------------
> AT least 1 NULL | Student_Name,Student_Score
> NON-NULL        | id   

Answer 3

The simplest method is to put this into separate columns:

select (case when count(*) = count(student_id) then 'No Null Values' else 'Null Values' end) as student_id,
       (case when count(*) = count(student_name) then 'No Null Values' else 'Null Values' end) as student_name,
       (case when count(*) = count(student_score) then 'No Null Values' else 'Null Values' end) as student_score
from studentscores;

This should be the simplest and most performance way to do what you want.

If you want this in separate rows, I would just unpivot these results:

select v.*
from (select (case when count(*) = count(student_id) then 'No Null Values' else 'Null Values' end) as student_id,
             (case when count(*) = count(student_name) then 'No Null Values' else 'Null Values' end) as student_name,
             (case when count(*) = count(student_score) then 'No Null Values' else 'Null Values' end) as student_score
      from studentscores
     ) ss cross apply
     (values ('student_id', student_id), ('student_name', student_name), ('student_score', student_score)
     ) v(col, str)