Question

For example, you have a simple table with just one column.
ie.

CREATE TABLE movies   (title VARCHAR2(255 BYTE))

set up with the following data:

INSERT INTO movies   (title) VALUES ('Scream');
INSERT INTO movies   (title) VALUES ('Blair Witch');
INSERT INTO movies   (title) VALUES ('Friday the 13th');
INSERT INTO movies   (title) VALUES ('Scary Movie');
INSERT INTO movies   (title) VALUES ('Hide and Seek');
INSERT INTO movies   (title) VALUES ('Alien vs Predator');

Is there a single query or PL/SQL that will do the following dynamically (ie without having to manually do a “UNION select ‘scream’ from dual…” for every value)?

Obviously this query is wrong but you get the idea:

Select * from movies
where title in (
'Scream',
'Scary Movie',
'Exorcist',
'Dracula',
'Saw',
'Hide and Seek'
)

Desired result being a record for every value in the “WHERE TITLE IN” clause where the record is not present in the table.
ie.

'Exorcist'
'Dracula'
'Saw'

Answer 1

If you’re using 10g or higher, you can build a function which converts a CSV string into a dynamic table. Check out the code for a string tokenizer in this other response.

You would use it like this:

select * from movies
where title NOT in (
         select * 
          from table (string_tokenizer
                      (
                          'Scream, Scary Movie,Exorcist,Dracula,Saw,Hide and Seek'
                        )

                  )
     )
/

---

Here is a slightly simpler implementation which doesn’t require any additional infrastructure:

SQL> select * from table(sys.dbms_debug_vc2coll('Scream',
'Scary Movie',
'Exorcist',
'Dracula',
'Saw',
'Hide and Seek'
 ))
/
  2    3    4    5    6    7    8  
COLUMN_VALUE
--------------------------------------------------------------------------------
Scream
Scary Movie
Exorcist
Dracula
Saw
Hide and Seek

6 rows selected.

SQL> 

This is similar to the Table Value Constructor, but it does only work for single column “tables”.