Handling empty set in MySQL CASE statement - mysql

MySQL server version 5.0.45. Consider the following:
( SELECT CASE
WHEN t.group_id = 12 THEN 'yes'
ELSE 'no'
END
FROM sample_table t
WHERE t.user_id = 2
AND t.group_id = 12 ) as foo
This subquery of a larger statement works as I'd expect, yielding a 'yes' or 'no' string value most of the time. It's not ideal, but that's what you get for working on someone else's code!
However, sometimes the select can legitimately return an empty set, in which case foo is set to NULL. This does not actually break the application, but it's irritating. Is there a way I can guarantee that foo will always be either 'yes' or 'no', even if there are no matching rows in the table?


If it's a scalar subquery, (i. e. you use in a SELECT or WHERE clause, not in FROM clause), then use this:
IF(
EXISTS
(
SELECT NULL
FROM sample_table t
WHERE t.user_id = 2
AND t.group_id = 12
), 'yes', 'no'
)
or even this:
COALESCE(
(
SELECT 'yes'
FROM sample_table t
WHERE t.user_id = 2
AND t.group_id = 12
), 'no')


If you want the empty set to represent a "no" case, you could probably do this instead:
select ... where ... (foo = 'no' or foo is null)
This would handle both casees without having to drastically change your subquery.

Related

How return a value with mysql query depending on condition?

I have the next mysql table:
Table(cl_convenios_do_medico)
id auto_increment
med_id int
conv_id int
I´d like a sql to return '1':
If I found at least one "med_id" having a "conv_id".
If the table is empty
otherwise return '0'
I have tried something as:
select IFNULL( (select '1' from cl_convenios_do_medico CC where
CC.med_id=<my_med_id> and CC.conv_id=<my_conv_id> )
and exists(select '1' from cl_convenios_do_medico CC where CC.med_id=<my_med_id> limit 1),'0') as result
If seems to work when the table is not empty, but fail if table is empty.
Maybe someone has a simpler solution.

Do something like this
SELECT MAX(result)
FROM (
SELECT CASE WHEN COUNT(*) > 0 THEN 1 ELSE 0 END AS result
FROM cl_convenios_do_medico
WHERE med_id = 8
AND conv_id = 9
UNION
SELECT CASE WHEN COUNT(*) = 0 THEN 1 ELSE 0 END
FROM cl_convenios_do_medico
) c
DEMO

You can use CASE statement
CASE
WHEN condition1 THEN result1
WHEN condition2 THEN result2
WHEN conditionN THEN resultN
ELSE result
END;
for example
SELECT CASE WHEN (SELECT COUNT(*) FROM employee) > 1 THEN '1' ELSE '0' END

How to write the a new select query with backward compatibility after adding a new column to a MySQL database

I have a MySQL table running for 4 months and I have a select statement in that table, like below.
SELECT
CONCAT(
YEAR(FROM_UNIXTIME(creation_time)),
'-',
IF(
MONTH(FROM_UNIXTIME(creation_time)) < 10,
CONCAT('0', MONTH(FROM_UNIXTIME(creation_time))),
MONTH(FROM_UNIXTIME(creation_time))
)
) AS Period,
(
COUNT(CASE
WHEN system_name = 'System' THEN 1
ELSE NULL
END)
) AS "Some data",
FROM table_name
GROUP BY
Period
ORDER BY
Period DESC
Lately, I've added a new feature and a column, let's say is_rerun. This value is just added and not exist previously. Now, i would like to write a query with the current statement which checks the system_name and also the is_rerun field and if this field exists and value is 1 then return 1 and if the column not exist or it its value is zero, then return null.
I tried IF EXISTS re_run THEN 1 ELSE NULL, but no luck. I can also insert values for the previous runs but i don't want to do that. Is there any solution. Thanks.
SELECT
CONCAT(
YEAR(FROM_UNIXTIME(creation_time)),
'-',
IF(
MONTH(FROM_UNIXTIME(creation_time)) < 10,
CONCAT('0', MONTH(FROM_UNIXTIME(creation_time))),
MONTH(FROM_UNIXTIME(creation_time))
)
) AS Period,
(
COUNT(CASE
WHEN system_name = 'System' AND IF EXISTS is_rerun THEN 1
ELSE NULL
END)
) AS "Some data",
FROM table_name
GROUP BY
Period
ORDER BY
Period DESC

As a starter: you have a group by query, so you need to put is_rerun in an aggregate function.
Based on your description, I think that something like case(case when is_rerun = 1 then 1 end) should do the work: it returns 1 if any is_rerun in the group is 1, else null.
Or if you can live with 0 instead of null, then you can use a simpler expression: max(is_rerun = 1).
Note that your query could be largely simplified as for the date formating logic and the conditional count. I would phrase it as:
select
date_format(from_unixtime(creation_time),'%Y-%m') period,
sum(system_name = 'System') some_data,
max(is_rerun = 1) is_rerun
from mytable
group by period
order by period desc

MySQL execute SELECT based on condition

How to use CASE appropriately for executing SELECT query based on user input ?
On executing following query :
SELECT
CASE WHEN user_input_variable = 'y' THEN
(SELECT * FROM table_foo WHERE bar = '6f322766-0ec0-4d24-840f-c857a82a6efe')
ELSE
(SELECT 0)
END
If the user has selected 'y' then it should return records from the table, else it should return an empty result set.
I am getting error:
Operand should contain 1 column(s)

The sub-query must return not more than 1 column. Do a UNION ALL instead.
Note that the number of columns must be the same, that's why you need to select null's in the second select.
SELECT * FROM table_foo WHERE bar = '6f322766-0ec0-4d24-840f-c857a82a6efe')
and user_input_variable = 'y'
union all
SELECT 0, null, null... where user_input_variable <> 'y'

You could use the IF condition also.
IF user_input_variable = 'y'
SELECT * FROM table_foo WHERE bar = '6f322766-0ec0-4d24-840f-c857a82a6efe')

how to use user defined variables in where clause

I have this query, and i m trying to use the user defined variable #noVar in my where clause to show only the records with value 'Yes' on that variable.
but when I use Having #noVar = 'Yes' as in the query below, it returns 0 result.
SELECT svcreqdetail.id, svcreqcheckin.stime as checkin, #etime:= time(timestampadd(minute, svcreqdetail.hours*60 , concat(caredate,' ', caretime))) as endtime, svcreqcheckout.stime as checkout, time_to_sec( if(svcreqcheckout.stime > svcreqcheckin.stime,
timediff(svcreqcheckout.stime, svcreqcheckin.stime),
addtime(timediff(svcreqcheckout.stime, svcreqcheckin.stime), '24:00:00.000000')))/3600 AS wrkHrs, svcreqdetail.hours,
svcreqstatus.status, #checkoutvar:= time_to_sec(timediff(svcreqcheckout.stime, #etime))/60 as checkoutvar,#noVar:= if (#checkoutvar <= 15,'Yes', 'No') as noVar, qualif
FROM svcreqdetail
LEFT JOIN svcreqcheckin ON svcreqcheckin.reqid = svcreqdetail.id
LEFT JOIN svcreqcheckout ON svcreqcheckout.reqid = svcreqdetail.id
JOIN svcreqstatus ON svcreqstatus.reqdid = svcreqdetail.id
WHERE (yearweek( caredate ) = yearweek( date_sub( CURRENT_DATE, INTERVAL 1 week ) )
AND svcreqstatus.status != 'Incompleted'
AND svcreqstatus.status != 'Deleted')
having #noVar = 'Yes'
is there anyway i can test against that variable in my where clause. and thank you

I don't think you can use user-defined variables in the HAVING clause like that.
One option would be to put your existing query (without the HAVING clause) into a sub-query and filter the results using a WHERE clause outside the sub-query, like so:
SELECT *
FROM
(
<your existing query goes here>
) AS sub_query
WHERE noVar = 'Yes'

SQL Select syntax error

thanks for your help yesterday.
I am now trying to incorporate the query from yesterday into an existing query so I can show the highest itemcode's reporting group in the existing query..but I have a syntax error somewhere at my Select statement.
ERROR: Keyword SELECT not expected.
I have tried putting brackets at every possible place but still no go..can you please help? (ps-this whole query has been giving me nightmares!)
WITH CALC1 AS (SELECT OTQUOT, OTIT01 AS ITEMS, ROUND(OQCQ01 * OVRC01,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT01 <> ''
UNION ALL
SELECT OTQUOT, OTIT02 AS ITEMS, ROUND(OQCQ02 * OVRC02,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT02 <> ''
UNION ALL
SELECT OTQUOT, OTIT03 AS ITEMS, ROUND(OQCQ03 * OVRC03,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT03 <> ''
UNION ALL
SELECT OTQUOT, OTIT04 AS ITEMS, ROUND(OQCQ04 * OVRC04,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT04 <> ''
UNION ALL
SELECT OTQUOT, OTIT05 AS ITEMS, ROUND(OQCQ05 * OVRC05,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT05 <> ''
UNION ALL
SELECT OTQUOT, OTIT06 AS ITEMS, ROUND(OQCQ06 * OVRC06,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT06 <> ''
UNION ALL
SELECT OTQUOT, OTIT07 AS ITEMS, ROUND(OQCQ07 * OVRC07,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT07 <> ''
UNION ALL
SELECT OTQUOT, OTIT08 AS ITEMS, ROUND(OQCQ08 * OVRC08,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT08 <> ''
UNION ALL
SELECT OTQUOT, OTIT09 AS ITEMS, ROUND(OQCQ09 * OVRC09,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT09 <> ''
UNION ALL
SELECT OTQUOT, OTIT10 AS ITEMS, ROUND(OQCQ10 * OVRC10,2) AS COST
FROM #LIB#.RTQOTA
WHERE OTIT10 <> ''
)
(SELECT OTQUOT, DESC
FROM (
SELECT OTQUOT, ITEMS, B.IXRPGP AS GROUP, C.OTRDSC AS DESC, COST, ROW_NUMBER() OVER
(ORDER BY COST DESC) AS RN
FROM CALC1 AS A INNER JOIN #LIB#.ITMCON AS B ON (A.ITEMS = B.IKITMC) INNER JOIN
DATAGRP.GDSGRP AS C ON (B.IXRPGP = C.OKRPGP)
) T
WHERE T.RN >= 1)
SELECT
A.OKPBRN,
A.OCAREA,
A.OTCCDE,
A.OTCNAM,
A.OTSMAN,
A.OKPBRN||A.OAPNUM AS OTQUOT,
A.OTONUM,
A.OTCAD1,
A.OTCAD2,
A.OTCAD3,
A.OTPCDE,
A.OTDEL1,
A.OTDEL2,
A.OTDEL3,
CHAR(DATE(CASE WHEN SUBSTR(A.ODOQDT,5,4) = '0000' THEN '0001' ELSE SUBSTR(A.ODOQDT,5,4)
END ||'-'||
CASE WHEN SUBSTR(A.ODOQDT,4,2) = '00' THEN '01' ELSE SUBSTR(A.ODOQDT,3,2) END ||'-'||
CASE WHEN SUBSTR(A.ODOQDT,1,2) = '00' THEN '01' ELSE SUBSTR(A.ODOQDT,1,2) END), ISO) AS
ODOQDT_CCYYMMDD,
CHAR(DATE(CASE WHEN SUBSTR(A.ODDELD,7,2) = '' THEN '0001' ELSE '20'||SUBSTR(A.ODDELD,7,2)
END ||'-'||
CASE WHEN SUBSTR(A.ODDELD,4,2) = '' THEN '01' ELSE SUBSTR(A.ODDELD,4,2) END ||'-'||
CASE WHEN SUBSTR(A.ODDELD,1,2) = '' THEN '01' ELSE SUBSTR(A.ODDELD,1,2) END), ISO) AS
ODDELD_CCYYMMDD,
B.DESC,
A.OVQTVL
FROM
#LIB#.RTQCTL AS A INNER JOIN CALC1 AS B ON (A.OKPBRN||A.OAPNUM = B.OTQUOT)
WHERE
A.OKPBRN = '#OKPBRN#'
AND A.OTCCDE NOT LIKE '*DEP%'
AND CHAR(DATE(CASE WHEN SUBSTR(A.ODOQDT,5,4) = '0000' THEN '0001' ELSE SUBSTR
(A.ODOQDT,5,4) END ||'-'||
CASE WHEN SUBSTR(A.ODOQDT,4,2) = '00' THEN '01' ELSE SUBSTR(A.ODOQDT,3,2) END ||'-'||
CASE WHEN SUBSTR(A.ODOQDT,1,2) = '00' THEN '01' ELSE SUBSTR(A.ODOQDT,1,2) END), ISO) >=
CHAR(CURDATE() - 3 MONTH, ISO)
AND A.OCQF01 = '0'
AND A.OCQF02 = '0'
AND A.OCQF04 = '0'
AND A.OCQF05 = '0'
AND A.OCQF06 = '0'
AND A.OCQF07 = '0'
AND A.OCQF08 = '0'
AND A.OCQF09 = '0'
AND A.OCQF10 = '1'
AND A.OTCGRP LIKE 'S/%'
ORDER BY
A.OTSMAN ASC,
A.OVQTVL DESC,
CHAR(DATE(CASE WHEN SUBSTR(A.ODDELD,7,2) = '' THEN '0001' ELSE '20'||SUBSTR(A.ODDELD,7,2) END ||'-'||
CASE WHEN SUBSTR(A.ODDELD,4,2) = '' THEN '01' ELSE SUBSTR(A.ODDELD,4,2) END ||'-'||
CASE WHEN SUBSTR(A.ODDELD,1,2) = '' THEN '01' ELSE SUBSTR(A.ODDELD,1,2) END),ISO) ASC

When using UNION, wrap your SELECT statements in parentheses.
e.g.
(SELECT * FROM foo)
UNION ALL
(SELECT * FROM bar)
Edit: Upon closer inspection of that monstrosity of a query, there are likely numerous other errors within it. I suggest ripping parts of it out and provide us with the smallest possible example that still causes the syntax error.

It looks like you have two totally separate queries, with no semicolon between them.
This is where I think the problem is:
. . .
WHERE T.RN >= 1)
SELECT A.OKPBRN, A.OCAREA, A.OTCCDE, A.OTCNAM, A.OTSMAN,
. . .
It would really help when you ask questions to identify the brand and version of the database you're using. Different databases support different SQL language features, so the right solution depends strongly on the technology you're using. Tag your question with the brand you use.
There are many things wrong with your query:
MySQL does not support WITH common table expressions. This is what made me think you were using Oracle or Microsoft SQL Server when you asked another question related to this query yesterday.
MySQL does not support the ROW_NUMBER() windowing function. My apologies for suggesting you use this function, I had assumed you were using a database that supports WITH so I assumed ROW_NUMBER() would be supported too. You should learn to use the LIMIT clause if you use MySQL.
Trying to put a subquery that returns multiple columns and multiple rows into a single column of your select-list.
If you're trying to make the subquery part of the select-list, then SELECT belongs in front of the subquery. You should give it a column alias too.
Complex date-conversion expressions. I would recommend storing dates in a canonical ISO form.
The original design of OVRC01, OVRC02, OVRC03, etc. This is a repeating group, which violates First Normal Form. The same is true for OCQF01, OCQF02, OCQF03, etc.
I'm not sure you're using LIKE wildcards correctly. * is not a standard wildcard for LIKE.
Indecipherable table naming and column naming conventions. It looks like an Oracle database from the 1980's. Without knowing anything about your database or the problem you're trying to solve, it's impossible to recommend another solution.

It looks you have two CTE tables but only one is named. The second starts (SELECT OTQUOT, DESC(. You could name it (i.e., change it to , CALC2 As (SELECT OTQUOT, DESC(... but I haven't the foggiest idea where you are actually wanting to use this table in the rest of your query.