in reply to RE: RE: reversing a hash -- printing it
in thread reversing a hash -- printing it

Simple - table 1 contains your actions: 
id   action
---------------
 1     ls
 2     cd
 3     file
 4     zonk

[download]
Table 2 contains your list of products: 
id   product
-----------------
 1     /usr/bin
 2     /home
 3     /usr/tmp

[download]
And finally Table 3 is a cross reference of the two, called product_actions: 
p_id   a_id
-------------
 2      3
 1      3
 2      4

[download]
so something like: 
SELECT actions.action, products.product
FROM   actions, products, product_actions
WHERE  actions.id = product_actions.a_id
AND    product.id = product_actions.p_id

[download]
Would produce: 
file /home
file /usr/bin
zonk /home  #don't try this one at home kids!

[download]
At least that's what I think that extremely had in mind. Your results will vary, and from what I hear you saying, it sounds to me like you are working with a very un- normalized database.

Hope this helps,
Jeff

Replies are listed 'Best First'.
RE: RE: RE: RE: reversing a hash -- printing it
by jptxs (Curate) on Oct 14, 2000 at 07:39 UTC

    I see what you're saying now.

    1. Yes, it's de-normalized, but I don't think speed was the reason, if you catch my drift : )

    2. I like this idea and will see if I can coerce the DBA to try it...I just need to figure out how to add the third layer here : ) This also needs to have "appointment ids" in table 3, and the select would need to grab everything for a particular "appointment id" and then sort id for output like: 

    Appointment:
product: action, action, action
product: action
product: action etc. etc.

    [download]

    3. The biggest trouble I could see with this would be I would need to do many many inserts for appointments with a lot of products involved in order to populate the records properly. Not too big a deal, but more than I had planned. Off to test : )

    -- I'm a solipsist, and so is everyone else. (think about it)

[reply]
[d/l]

In Section Seekers of Perl Wisdom