Question

EDIT To simplify this further.

Let’s take a look at this structure:

Table1: idProduct (int) | Name (varchar)

Table2: idDay (int) | idProduct (int) | date_start(datetime) | date_end (datetime)

idProduct is a common column in both tables. Table two contains data similar to this:

0, 2, 11/22/2012, 11/22/2012
1, 2, 11/23/2012, 11/23/2012
2, 2, 11/24/2012, 11/24/2012
3, 2, 00/00/0000, 11/25/2012
4, 2, 11/25/2012, 00/00/0000

Note that date_start, date_end or both may be all zeroes, in which case this upper or lower limit is considered unbound/unlimited.

In the above case I would like to select idProduct with ID 2 ONLY if today’s date is between the period – earliest date in date_start and latest date in date_end.
If date_start is 00/00/0000 always return the product BUT if date_end is in the future. The reverse applies for date_end being all zeroes and date_start being set. If both columns are zeroes return the product.

Answer 1

I don’t see the “LEFT JOIN” requirement in your question : just in your title. So

SELECT t1.idProduct, t1.Name
FROM table1 t1
INNER JOIN table2 t2 ON t1.idProduct = t2.idProduct
WHERE DATE(t2.date_start) < DATE(NOW())--this will also take values with DATETIME = 0000
AND (DATE(t2.date_end) > DATE(NOW()) OR YEAR(t2.date_end) = 0)
GROUP BY t1.idProduct, t1.Name

or if you want to check if product has at least one start_date smaller than today (or 0) and at least one end_date bigger than today(or 0), even if start_date and end_date don’t share the same idDay :

SELECT t1.idProduct, t1.Name
FROM table1 t1
WHERE t1.idProduct IN
    (SELECT t2.IdProduct
    FROM table t2
    GROUP BY t2.IdProduct
    HAVING 
    MIN(t2.date_start) <= DATE(NOW()) AND 
    (MAX(t2.date_end) >= DATE(NOW()) OR YEAR(MAX(t2.date_end)) = 0))

It will take :

min start_date = 0 and max end_date = 0

min start_date = 0 and max end_date >= today

min start_date <= today and max end_date =0

min start_date <= today and max end_date >= today