List:General Discussion« Previous MessageNext Message »
From:Peter Brawley Date:March 31 2009 1:27am
Subject:Re: Data structure for matching for company data
View as plain text  
Choi,

 >I found that I cannot use FK for Myisam tables. Then,
 >this kind of data structuring is no use at all for
 >Mysiam tables. Is this correct?

You can use this structure with MyISAM tables. It will work fine except 
you won't have the advantage of database-level enforcement of foreign 
key constraints--do it with code.

PB

----


ChoiSaehoon wrote:
> Thanks, PB.
>
>  
>
>
>
>  
>
>  
>   
>> Date: Fri, 27 Mar 2009 09:54:47 -0500
>> From: peter.brawley@stripped
>> To: saehoony@stripped
>> CC: mysql@stripped
>> Subject: Re: Data structure for matching for company data
>>
>> Choi
>>
>>     
>>> 1. company (3 cols) - company id(pk), company name
>>> 2. industry (3 cols) - industry id(pk), industry, sub-industry
>>> 3. matching table (3 cols?) - match id(pk), company id(fk), industry
> id(fk)...? 
>>>       
>> Yes, you've got it. In the matching (usually called "bridging") table, 
>> any company or industry can appear multiple times, but no combo can 
>> appear more than once.
>>
>> PB
>> http://www.artfulsoftware.com
>>
>> -----
>>
>> ChoiSaehoon wrote:
>>     
>>> I'm a complete newbie, so please feel free to help me here!
>>>
>>>
>>>
>>> I simplified my problem as below:
>>> 1. Thousands of companies (i.e. company names)
>>> 2. 200 industries. Each company has tree-type(?) industry info. (Industry,
> Sub-industry. Say, each industry has 10 sub-industry information. (10X20))
>>>
>>>
>>>
>>> The problem is one company can have 1,2,3... or 15 industries and/or
> sub-industries... (as in a real-life case)
>>>
>>>
>>>
>>> What would be a smart way create tables for this situation? (both company and
> industry must be searchable)
>>>
>>>
>>>
>>> (One rough thought (after having studying MySQL for 2 weeks now) is create 3
> tables of 
>>> 1. company (3 cols) - company id(pk), company name
>>> 2. industry (3 cols) - industry id(pk), industry, sub-industry
>>> 3. matching table (3 cols?) - match id(pk), company id(fk), industry
> id(fk)...? But, one company can have more than 1 industry info! Ah... Then, if that's the
> case, different match id & same company id & different industry id??)
>>>
>>> I'm stuck here! :( 
>>>
>>>
>>>
>>> Any general advice would be appreciated... :)
>>>
>>> _________________________________________________________________
>>> ???? ???, ???? ?? ???? ???? ??? Windows Live Hotmail! ?? ???? ???!
>>> http://www.hotmail.com
>>>
>>> ------------------------------------------------------------------------
>>>
>>>
>>> No virus found in this incoming message.
>>> Checked by AVG - www.avg.com 
>>> Version: 8.0.238 / Virus Database: 270.11.30/2026 - Release Date: 03/27/09
> 07:13:00
>>>
>>>
>>>       
>
> _________________________________________________________________
> MSN ???? ??? ??, Windows Live Messenger!
> http://windowslive.msn.co.kr/wlm/messenger/
>   
> ------------------------------------------------------------------------
>
>
> No virus found in this incoming message.
> Checked by AVG - www.avg.com 
> Version: 8.0.238 / Virus Database: 270.11.32/2030 - Release Date: 03/30/09 08:40:00
>
>   

Thread
Data structure for matching for company dataChoiSaehoon27 Mar
  • Re: Data structure for matching for company dataPeter Brawley27 Mar
    • Re: Data structure for matching for company dataArthur Fuller28 Mar
      • RE: Data structure for matching for company dataChoiSaehoon31 Mar
    • RE: Data structure for matching for company dataChoiSaehoon31 Mar
      • Re: Data structure for matching for company dataPeter Brawley31 Mar
        • Re: Data structure for matching for company dataAndy Shellam31 Mar