MySQL
  Home arrow MySQL arrow Page 9 - Understanding SQL Joins
Dev Shed Forums 
Administration  
AJAX  
Apache  
BrainDump  
DHTML  
Flash  
Java  
JavaScript  
Multimedia  
MySQL  
Oracle  
Perl  
PHP  
Practices  
Python  
Reviews  
Security  
Style-Sheets  
Web Services  
XML  
Zend  
Zope  
Forums Sitemap 
IBM® developerWorks 
Sun Developer Network 
Dedicated Servers 
E-Commerce Hosting 
Linux Web Hosting 
Managed Hosting 
Small Business Hosting 
Moblin 
JMSL Numerical Library 
VPS Hosting 
Weekly Newsletter

 
Developer Updates  
Free Website Content 
 RSS  Articles
 RSS  Forums
 RSS  All Feeds
Write For Us Get Paid 
Request Media Kit
Contact Us 
Site Map 
Privacy Policy 
Support 
 USERNAME
 
 PASSWORD
 
 
  >>> SIGN UP!  
  Lost Password? 
MYSQL

Understanding SQL Joins
By: The Disenchanted Developer, (c) Melonfire
  • Search For More Articles!
  • Disclaimer
  • Author Terms
  • Rating: 4 stars4 stars4 stars4 stars4 stars / 214
    2002-08-20

    Table of Contents:
  • Understanding SQL Joins
  • Meeting The Family
  • Keeping It Simple
  • Crossed Wires
  • Finding Common Ground
  • One Step Left...
  • ...Two Steps Right
  • The Bookworm Turns
  • Up A Tree
  • A Long Goodbye

  • Rate this Article: Poor Best 
      ADD THIS ARTICLE TO:
      Del.ici.ous Digg
      Blink Simpy
      Google Spurl
      Y! MyWeb Furl
    Email Me Similar Content When Posted
    Add Developer Shed Article Feed To Your Site
    Email Article To Friend
    Print Version Of Article
    PDF Version Of Article
     
     
    ADVERTISEMENT


    Understanding SQL Joins - Up A Tree


    (Page 9 of 10 )

    In addition to inner and outer joins, SQL also allows a third type of join, known as a "self join". Typically, this type of join is used to extract data from a table whose records contain internal links to each other. Consider the following table, which illustrates what I mean:

    +----+------------------------+--------+ | id | label | parent | +----+------------------------+--------+ | 1 | Services | 0 | | 2 | Company | 0 | | 3 | Media Center | 0 | | 4 | Your Account | 0 | | 5 | Community | 0 | | 6 | For Content Publishers | 1 | | 7 | For Small Businesses | 1 | | 8 | Background | 2 | | 9 | Clients | 2 | | 10 | Addresses | 2 | | 11 | Jobs | 2 | | 12 | News | 2 | | 13 | Press Releases | 3 | | 14 | Media Kit | 3 | | 15 | Log In | 4 | | 16 | Columns | 5 | | 17 | Colophon | 16 | | 18 | Cut | 16 | | 19 | Boombox | 16 | +----+------------------------+--------+
    This is a simple menu structure, with each record identifying a unique node in the menu tree. Each record has a unique record ID, and also contains a parent ID - these two IDs are used to define the parent-child relationships between the branches of the menu tree. So, if I were to represent the data above hierarchically, it would look like this:

    <root> |-Services |-For Content Publishers |-For Small Businesses |-Company |-Background |-Clients |-Addresses |-Jobs |-News |-Media Center |-Press Releases |-Media Kit |-Your Account |-Log In |-Community |-Columns |-Colophon |-Cut |-Boombox
    Now, let's suppose I need to display a list of all the nodes in the tree, together with the names of their parents. What I'm looking for is a resultset resembling this:

    +--------------+------------------------+ | parent_label | child_label | +--------------+------------------------+ | Services | For Content Publishers | | Company | Background | | Your Account | Log In | +--------------+------------------------+
    If you think about it, you'll see that there's no easy way to obtain this resultset - since all the data is in a single table, a simple SELECT won't work, and neither will one of those complicated outer joins. What I really need here is something called a self join, which allows me to create a second, virtual copy of the first table, and then use a regular inner join to map the two together and get the output I need.

    Here's the query I would use:

    SELECT a.label AS parent_label, b.label AS child_label FROM menu AS a, menu AS b WHERE a.id = b.parent;
    Here's the output:

    +--------------+------------------------+ | parent_label | child_label | +--------------+------------------------+ | Services | For Content Publishers | | Services | For Small Businesses | | Company | Background | | Company | Clients | | Company | Addresses | | Company | Jobs | | Company | News | | Media Center | Press Releases | | Media Center | Media Kit | | Your Account | Log In | | Community | Columns | | Columns | Colophon | | Columns | Cut | | Columns | Boombox | +--------------+------------------------+
    Exactly what I need!

    Most of the magic here lies in the table aliasing - I've created two copies of the "menu" table, and aliased them as "a" and "b" respectively. This will result in the following two "virtual" tables.

    +----+------------------------+--------+ | TABLE a | +----+------------------------+--------+ | id | label | parent | +----+------------------------+--------+ | 1 | Services | 0 | | 2 | Company | 0 | | 3 | Media Center | 0 | | 4 | Your Account | 0 | | 5 | Community | 0 | | 6 | For Content Publishers | 1 | | 7 | For Small Businesses | 1 | | 8 | Background | 2 | | 9 | Clients | 2 | | 10 | Addresses | 2 | | 11 | Jobs | 2 | | 12 | News | 2 | | 13 | Press Releases | 3 | | 14 | Media Kit | 3 | | 15 | Log In | 4 | | 16 | Columns | 5 | | 17 | Colophon | 16 | | 18 | Cut | 16 | | 19 | Boombox | 16 | +----+------------------------+--------+ +----+------------------------+--------+ | TABLE b | +----+------------------------+--------+ | id | label | parent | +----+------------------------+--------+ | 1 | Services | 0 | | 2 | Company | 0 | | 3 | Media Center | 0 | | 4 | Your Account | 0 | | 5 | Community | 0 | | 6 | For Content Publishers | 1 | | 7 | For Small Businesses | 1 | | 8 | Background | 2 | | 9 | Clients | 2 | | 10 | Addresses | 2 | | 11 | Jobs | 2 | | 12 | News | 2 | | 13 | Press Releases | 3 | | 14 | Media Kit | 3 | | 15 | Log In | 4 | | 16 | Columns | 5 | | 17 | Colophon | 16 | | 18 | Cut | 16 | | 19 | Boombox | 16 | +----+------------------------+--------+
    Once these two tables have been created, it's a simple matter to join them together, using the node IDs as the common column, and to obtain a list of child and parent labels in the desired format.

    More MySQL Articles
    More By The Disenchanted Developer, (c) Melonfire


     

       

    MYSQL ARTICLES

    - Take Some Load off MySQL with MemCached
    - MySQL Table Prefix Changer Tool in PHP
    - Using the SIGNAL Statement for Error Handling
    - Error Handling Examples
    - Error Handling
    - Completing a Search Engine with MySQL and PH...
    - Paginating Result Sets for a Search Engine B...
    - Building a Search Engine with MySQL and PHP 5
    - Using Boolean Operators for Full Text and Bo...
    - PHP, MySQL and the PEAR Database
    - Working with PHP and MySQL
    - Getting PHP to Talk to MySQL
    - Creating an RSS Reader: the Reader
    - MySQL Security Overview
    - Creating the Admin Script for a PHP/MySQL Bl...





    © 2003-2008 by Developer Shed. All rights reserved. DS Cluster 6 hosted by Hostway