Question

Looking to run a query using multiple left joins. The following query works.

SELECT directory.location, sales.covers, labor.hours
FROM t_directory directory
LEFT JOIN t_sales sales
ON sales.id = directory.id
AND sales.business_date = '2011-11-14'
LEFT JOIN t_labor labor
ON labor.id = directory.id
AND labor.business_date = '2011-11-14'
ORDER BY directory.id ASC

The trouble comes if I try to query between a date range. Adding a Group By clause has been duplicating rows. Right now I’m actually using two separate queries, which is probably better practice, and a better solution. This is just out of pure curiosity at this point.

Here is what I have so far.

 SELECT directory.location, sales.covers, labor.hours
 FROM t_directory directory
 LEFT JOIN t_sales sales
 ON sales.id = directory.id
 AND sales.business_date BETWEEN '2011-11-13' AND '2011-11-14'
 LEFT JOIN t_labor labor
 ON labor.id = directory.id
 AND labor.business_date BETWEEN '2011-11-13' AND '2011-11-14'
 GROUP BY directory.id, sales.business_date, labor.business_date
 ORDER BY directory.id ASC

Clearly something wrong with my GROUP BY clause, and perhaps other errors as well.

I am looking for a result like this:

| location | covers | labor |
=============================
  loc1     |  300   | 99.40
  loc1     |  325   | 100.50
  loc2     |  250   | 89.00
  loc2     |  275   | 90.20
  loc3     |  400   | 100.00
  loc3     |  500   | 122.90

And of course, what I am actually getting is:

| location | covers | labor |
=============================
  loc1     |  300   | 99.40
  loc1     |  300   | 100.50  
  loc1     |  325   | 99.40
  loc1     |  325   | 100.50   
  loc2     |  250   | 89.00
  loc2     |  250   | 90.20
  loc2     |  275   | 89.00
  loc2     |  275   | 90.20
  loc3     |  400   | 100.00
  loc3     |  400   | 122.90
  loc3     |  500   | 100.00
  loc3     |  500   | 122.90

Any help with getting the expected results would be greatly appreciated.

EDIT:
Here are the tables with sample data that I want to join. They all share a common id.

t_directory
| *id | location | 
|100  |   loc1   |
|101  |   loc2   |
|102  |   loc3   |

t_sales
| business_date | id | sales    | covers |
|  2011-11-13   | 103| 4000.00  | 300    |
|  2011-11-14   | 103| 4050.00  | 325    |

t_labor
|business_date | id | hours | 
| 2011-11-13   | 103| 99.40 |
| 2011-11-14   | 103| 100.50|

Answer 1

Your expected result doesn’t really make sense to me. There’s no relationship between sales and labor rows, so why should a covers value of 300 go next to a labor value of 99.40? What’s tying those two values together? That’s why you’re finding it hard to write the query to produce the results: you’re trying to output two columns where the data in each are not related.

I probably haven’t explained that very well. I’m sure someone else can!