Question

I have a table with, let’s say, the following columns:

Name , Parent Name, ID

Let’s also say that there are three entries where Parent Name is Null (meaning they are the top-most parent) – F_one, G_one, and H_one.

If I want to delete all the descendants of one of those parents (G_one, why not?) meaning all the children of G_one, all the children of those children, and the children of those, and so on all the way until the terminal level where, that row’s Name does not exist as a Parent Name for any other entry.

Is that possible to be done easily, maybe with a single query?

Bonus, is there a way to select all of the G_one lineage so I can manipulate it to my whim and will?

Can assume:
-No Children are shared among parents

Cannot assume:
-A discrete or even consistent number of sub-levels.

Answer 1

As @Marc B’s suggestion, a FORIEGN KEY with ON DELETE CASCADE would achieve this.

If you haven’t one, you can add it now:

1. If there is a UNIQUE constraint on Name (I assume the PRIMARY key os ID), skip thi sstep. If there ism’t one, create it:

   ALTER TABLE tableX
     ADD CONSTRAINT unique_Name
       UNIQUE (Name) ;
   

2. If the previous step succeeded, add the FOREIGN KEY:

   ALTER TABLE tableX
     ADD CONSTRAINT fk_Name_ParentName
       FOREIGN KEY (ParentName)
          REFERENCES tableX (Name)
          ON UPDATE CASCADE
          ON DELETE CASCADE ;
   

3. If the previous step succeeded, you can now delete your rows with one statement:

   DELETE
   FROM tableX
   WHERE ParentName = 'G_one' ;
   

   This should result in: Y rows affected.