Combine Multiple Select Count Queries on same table - php

Question

I have a table with a bunch of order rows in it. I need to check the table to see if there's 1 or more rows matching 2 WHERE clauses. If there is, I set a variable to 'yes' that I use elsewhere on the page.
Currently I'm checking this with individual php mysql queries but wondering if it's possible to combine them into a single select query so i'm hitting the database one time rather than like 10 or more times.
Here's an example of the current code:
$query9 = "SELECT COUNT(*) as num FROM mojave_requests where status = 'Pending' AND local = 'L-BKED'";
$requestEDI = mysql_fetch_array(mysql_query($query9));
$requestEDI = $requestEDI[num];
if ($requestEDI> 0) {
$requestsEDI = 'yes';
}
$query11 = "SELECT COUNT(*) as num FROM mojave_requests where status = 'Pending' AND local = 'LOQ53'";
$requestLOQ53 = mysql_fetch_array(mysql_query($query11));
$requestLOQ53 = $requestLOQ53[num];
if ($requestLOQ53 > 0) {
$requestsLOQ53 = 'yes';
}
$query13 = "SELECT COUNT(*) as num FROM mojave_requests where status = 'Pending' AND local = 'L-YRYR'";
$requestLYRYR = mysql_fetch_array(mysql_query($query13));
$requestLYRYR = $requestLYRYR[num];
if ($requestLYRYR > 0) {
$requestsLYRYR = 'yes';
}
$query15 = "SELECT COUNT(*) as num FROM mojave_requests where status = 'Pending' AND local = 'L-CAL0611'";
$requestLCAL0611 = mysql_fetch_array(mysql_query($query15));
$requestLCAL0611 = $requestLCAL0611[num];
if ($requestLCAL0611 > 0) {
$requestsLCAL0611 = 'yes';
}
The only two things different between each query is the "local" column string I'm checking and the variable name I set to "Yes" if there's at least 1 match.
Thanks for looking.

Answer 1

You can use conditional sum as
SELECT
sum
(
case when status = 'Pending' AND local = 'L-BKED' then 1 else 0 end
) as `L-BKED_COUNT`,
sum
(
case when status = 'Pending' AND local = 'LOQ53' then 1 else 0 end
) as `LOQ53_COUNT`,
sum
(
case when status = 'Pending' AND local = 'L-YRYR' then 1 else 0 end
) as `L-YRYR_COUNT`,
sum
(
case when status = 'Pending' AND local = 'L-CAL0611' then 1 else 0 end
) as `L-CAL0611_COUNT`
FROM mojave_requests

Answer 2

This GROUP BY SELECT will give you the count for local matching the values that you need:
SELECT COUNT(*) as num,
local
FROM mojave_requests
WHERE status = 'Pending'
GROUP BY local
HAVING local IN ('L-BKED','LOQ53','L-YRYR','L-CAL0611')
SQL Fiddle
You should also change your PHP code and do not use those $request___ variables, but only a while cycle to check:
$result = mysql_query($sql);
while ($row = mysql_fetch_array($result)) {
switch ($row[1]) {
case 'L-BKED':
code to be executed if count for L-BKED > 0
break;
case 'LOQ53':
code to be executed if count for LOQ53 > 0
break;
case 'L-YRYR':
code to be executed if count for L-YRYR> 0
break;
...
}
}

Answer 3

You are looking for conditional aggregation. The simplest way to do this in MySQL is to take advantage of the fact that booleans are treated as numbers in a numeric context:
SELECT SUM(status = 'Pending' AND local = 'L-BKED') as num_bked,
SUM(status = 'Pending' AND local = 'LOQ53') as num_loc53,
SUM(status = 'Pending' AND local = 'L-YRYR') as num_yryr,
SUM(status = 'Pending' AND local = 'L-CAL0611') as num_cal0611
FROM mojave_requests;
This returns the values in one row. You can fetch the values by referencing the columns in php.
You can simplify this query by factoring out the first condition:
SELECT SUM(local = 'L-BKED') as num_bked,
SUM(local = 'LOQ53') as num_loc53,
SUM(local = 'L-YRYR') as num_yryr,
SUM(local = 'L-CAL0611') as num_cal0611
FROM mojave_requests
WHERE status = 'Pending';
If few of the status values are actually PENDING, then an index on mojave_requests(status, local) should noticeably improve performance.