Maybe it's because I'm not fully awake yet that the problem seems less complex than it 

What is wrong with using a seperate table for each year?


On Thu, 26 Jun 2003 08:43:17 +0200, anders thoresson wrote:

> I'm working on a web based article planning system for a news paper, based 
>on PHP and MySQL. One vital table in the setup will contain release date 
>and number for each issue. It's a simple table with just three columns:
> id, which is the primary key and AUTO_INCREMENT
> date, which is the release date for each issue
> issue_number
> I need help with a solution to automatically create the issue_number. The 
>conditions is as follow:
> 1. The paper is published every Monday to Friday, except for national 
>holidays, so issue_number isn't the same as the day number
> 2. The first issue each year should be number 1, and then the rest should 
>follow in sequence
> The first condition is taken care of manually, where one of the editors 
>will input a start date and an end date between which the newspaper will be 
>released Mondays to Fridays. That way I'll handle the national holidays, 
>for each year the editors have to enter a couple of date sequences to 
>"fence out" the holidays.
> But keeping track of each and every issue_number manually isn't practical. 
>I've looked at the MySQL manual, but all I find is LAST_INSERT_ID(). But 
>from what I've understood, that function is only useful when it's the same 
>value that is being updated, and this isn't the case in my problem.
> So I guess I'll have to write my own function. The solution I think of 
>right now is adding new issues in two steps: First adding the date for the 
>new issue to the table, then checking the issue_number for the date before, 
>and then adding the issue_number next in turn to the new issue, or, if the 
>year differs start over from 1 again.
> Is there a smarter way of doing this?
>anders thoresson
>PHP Database Mailing List (
>To unsubscribe, visit:

PHP Database Mailing List (
To unsubscribe, visit:

Reply via email to