Count all child nodes of hierarchical data in a table

Question

I want to count number of all child nodes under any level of tree structure maintained in a table using adjacency model (parent-child key). Table structure and data looks like this:

id -  item-   parentid    
1  -  A   -   
2  -  B   -   1   
3  -  C   -   1   
4  -  D   -   2   
5  -  E   -   2   
6  -  F   -   3   
7  -  G   -   3   
8  -  H   -   5   
9  -  I   -   5   
10 -  J   -   9   
11 -  K   -   4   

For example B has following child and grand child structure:

• B
  • E
    • H
    • I
      • J
  • F
    • K

Now if you want to count "All child nodes of B" my answer should be 6.

Any pure SQL Query based solution would be of great help. Or mysql/php will also work.

Thanks!

Answer 1

The way you store your data will not allow for a simple query to get the total child count. But have a look at:

http://en.wikipedia.org/wiki/Nested_set_model

Where a query like this would be possible.

Answer 2

Can be done fairly simply with a non-recursive stored procedure as follows:

Example calls

mysql> call category_hier(1);
+--------------+
| num_children |
+--------------+
|            3 |
+--------------+
1 row in set (0.00 sec)

Query OK, 0 rows affected (0.00 sec)

mysql> call category_hier(2);
+--------------+
| num_children |
+--------------+
|            2 |
+--------------+
1 row in set (0.00 sec)

Query OK, 0 rows affected (0.00 sec)

Full script

drop table if exists categories;
create table categories
(
cat_id smallint unsigned not null auto_increment primary key,
name varchar(255) not null,
parent_cat_id smallint unsigned null,
key (parent_cat_id)
)
engine = innodb;

insert into categories (name, parent_cat_id) values
('Location',null), 
('Color',null), 
   ('USA',1), 
      ('Illinois',3), 
      ('Chicago',3), 
   ('Black',2), 
   ('Red',2);


drop procedure if exists category_hier;
delimiter #

create procedure category_hier
(
in p_cat_id smallint unsigned
)
begin

declare v_done tinyint unsigned default 0;
declare v_depth smallint unsigned default 0;

create temporary table hier(
 parent_cat_id smallint unsigned, 
 cat_id smallint unsigned, 
 depth smallint unsigned default 0
)engine = memory;

insert into hier select parent_cat_id, cat_id, v_depth from categories where cat_id = p_cat_id;
create temporary table tmp engine=memory select * from hier;

/* http://dev.mysql.com/doc/refman/5.0/en/temporary-table-problems.html */

while not v_done do

    if exists( select 1 from categories c
        inner join tmp on c.parent_cat_id = tmp.cat_id and tmp.depth = v_depth) then

        insert into hier select c.parent_cat_id, c.cat_id, v_depth + 1 from categories c
            inner join tmp on c.parent_cat_id = tmp.cat_id and tmp.depth = v_depth;

        set v_depth = v_depth + 1;          

        truncate table tmp;
        insert into tmp select * from hier where depth = v_depth;

    else
        set v_done = 1;
    end if;

end while;

/*
select 
 c.cat_id,
 c.name as category_name,
 p.cat_id as parent_cat_id,
 p.name as parent_category_name,
 hier.depth
from 
 hier
inner join categories c on hier.cat_id = c.cat_id
left outer join categories p on hier.parent_cat_id = p.cat_id
order by
 hier.depth;
*/

select count(*) as num_children from hier where parent_cat_id is not null;

drop temporary table if exists hier;
drop temporary table if exists tmp;

end #

delimiter ;

call category_hier(1);

call category_hier(2);

You can easily adapt this example to suit your requirements.

Hope it helps :)

Answer 3

Below is a PHP based solution:

function countChildren($startId) {
    $directDescendents = *_query("SELECT id FROM Table WHERE parentid = ?", array( $startId ));
    $count = *_num_rows($directDescendents);
    while($row = *_fetch_array($directDescendents))
        $count += countChildren($row['id']);
    return $count;
}

$numChildren = countChildren(2); // Number of Children for 'B'

Replace *_num_rows and *_fetch_array with whatever functions for the SQL extension you are using. This won't be as efficient as a pure SQL solution, but it will work. The way I'm querying in the function is assuming bound parameters, but execute the query as you like.