Re: How to navigate tree without CONNECT BY?

From: Andrei Ivanov <andrei(dot)ivanov(at)ines(dot)ro>
To: "D(dot) Dante Lorenso" <dante(at)lorenso(dot)com>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: How to navigate tree without CONNECT BY?
Date: 2003-12-18 21:28:48
Message-ID: Pine.LNX.4.58L0.0312182327330.2128@webdev.ines.ro
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general


See http://gppl.terminal.ru/readme.html

On Thu, 18 Dec 2003, D. Dante Lorenso wrote:

> I have a simple table that I'd like to query to pull
> out a heirarchy from a tree relationship. What is the
> best way to do this without a 'CONNECT BY' clause like
> Oracle has?
>
> Example
>
> mytable
> +----------+-----------+
> | child_id | parent_id |
> +----------+-----------+
> | 1 | NULL |
> | 2 | NULL |
> | 3 | 1 |
> | 4 | 1 |
> | 5 | 2 |
> | 6 | 4 |
> | 7 | 4 |
> | 8 | 7 |
> | 9 | 3 |
> | 10 | 9 |
> +----------+-----------+
>
> I want to be able to select the child_id, parent_id, and the up-stream
> heirarchy level when starting at a given child...
>
> In Oracle you'd use a statement like
>
> SELECT *
> FROM account
> START WITH child_id = 10
> CONNECT BY PRIOR parent_id = child_id;
> (* note: may not be exactly correct *)
>
> I was thinking that PL/PGSQL could return a set using a function like
> 'get_tree_relation(child_id INTEGER)'
>
> Example 1:
>
> SELECT *
> FROM get_tree_relation(10)
> ORDER BY level ASC;
>
> +----------+-----------+-------+
> | child_id | parent_id | level |
> +----------+-----------+-------+
> | 10 | 9 | 1 |
> | 9 | 3 | 2 |
> | 3 | 1 | 3 |
> | 1 | NULL | 4 |
> +----------+-----------+-------+
>
> Example 2:
>
> SELECT *
> FROM get_tree_relation(2)
> ORDER BY level ASC;
>
> +----------+-----------+-------+
> | child_id | parent_id | level |
> +----------+-----------+-------+
> | 2 | NULL | 1 |
> +----------+-----------+-------+
>
> Example 2:
>
> SELECT *
> FROM get_tree_relation(11)
> ORDER BY level ASC;
>
> +----------+-----------+-------+
> | child_id | parent_id | level |
> +----------+-----------+-------+
> +----------+-----------+-------+
>
> I have a PL/PGSQL function that does this for me with some nested
> selects inside a loop, but my NEW problem is that I need to be able
> to detect circular loops. For example, if child_id refers to itself
> or if a parent_id refers to a child_id that is already in the
> heirarchy we don't want to get into an infinite loop. So I modified
> my function to use a TEMP table to store the records I had already
> seen, but then I had problems with the temp table:
>
> http://archives.postgresql.org/pgsql-bugs/2003-05/msg00084.php
>
> Without having to recompile any database code, can this process be
> build using out-of-the-box PostgreSQL features?
>
> There's gotta be an easy way to do this. It's a fairly common
> problem, isn't it?
>
> --Dante
>
>
>
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: the planner will ignore your desire to choose an index scan if your
> joining column's datatypes do not match
>

In response to

Browse pgsql-general by date

  From Date Subject
Next Message John Sidney-Woollett 2003-12-18 21:35:52 Re: Replication
Previous Message Bruno Wolff III 2003-12-18 21:25:18 Re: Select groupping by one column