I read a lot of similar questions but didn't find a solution for me. Basically I would like to have a where clause (AND PEKP.VORGANGS_ART = 'BE') only if a special condition (PFSP.EINKAUFS_KZ = 2) is true.
I tried in many ways like that:
SELECT *
FROM PFSP
LEFT OUTER JOIN PFAK on PFSP.RUECKMELDE_NR = PFAK.RUECKMELDE_NR
LEFT OUTER JOIN PEKP ON (PFSP.BESTELL_NR=PEKP.VORGANGS_NR)
AND (PFSP.BESTELL_POS_NR=PEKP.VORGANGS_POS_NR)
LEFT OUTER JOIN PMLB ON PFSP.KOMPONENTEN_ARTIKEL_NR=PMLB.ARTIKEL_NR
WHERE PFAK.KD_VORGANGS_NR = '910-001213'
AND PFSP.RUECKMELDE_STATUS = '3'
AND PFSP.BESCHAFFUNGSKENNER = 'F'
AND CASE PFSP.EINKAUFS_KZ
WHEN 2 THEN PEKP.VORGANGS_ART = 'BE'
END
but I keep getting errors:
wrong syntax near '='"
Skip the CASE, use AND/OR instead:
AND (PFSP.EINKAUFS_KZ <> 2 OR PEKP.VORGANGS_ART = 'BE')
If PFSP.EINKAUFS_KZ = 2, then PEKP.VORGANGS_ART must be equal to 'BE'.
If PFSP.EINKAUFS_KZ <> 2, it doesn't matter what PEKP.VORGANGS_ART is.
With this CASE expression:
AND 1 = CASE
WHEN PFSP.EINKAUFS_KZ = 2 AND PEKP.VORGANGS_ART = 'BE' THEN 1
WHEN PFSP.EINKAUFS_KZ = 2 THEN 0
ELSE 1
END
The order of the conditions is important.
You probably want something like:
AND (
(PFSP.EINKAUFS_KZ = 2 AND PEKP.VORGANGS_ART='BE')
OR PFSP.EINKAUFS_KZ <> 2 --You may need to catch NULL as well
)
SELECT *
FROM PFSP
LEFT OUTER JOIN PFAK on PFSP.RUECKMELDE_NR = PFAK.RUECKMELDE_NR
LEFT OUTER JOIN PEKP ON (PFSP.BESTELL_NR=PEKP.VORGANGS_NR)
AND (PFSP.BESTELL_POS_NR=PEKP.VORGANGS_POS_NR)
LEFT OUTER JOIN PMLB ON PFSP.KOMPONENTEN_ARTIKEL_NR=PMLB.ARTIKEL_NR
WHERE PFAK.KD_VORGANGS_NR = '910-001213'
AND PFSP.RUECKMELDE_STATUS = '3'
AND PFSP.BESCHAFFUNGSKENNER = 'F'
AND ((PFSP.EINKAUFS_KZ = 2 AND PEKP.VORGANGS_ART = 'BE') OR (PFSP.EINKAUFS_KZ <> 2) )
Related
I'm trying to add a conditional SELECT column in a query and I'm having trouble writing it out:
SELECT v.GLTypeID,
v.something,
v.somethignElse,
CASE WHEN (SELECT TOP 1 Value FROM InterfaceParam WHERE InterfaceId = 0 AND Descr = 'gf') = 1 THEN a.CreditID ELSE NULL END AS CreditMemoID
FROM vGLDetails2 v
....
LEFT OUTER JOIN AssociationFund f
ON v.FundID = f.FundID
LEFT JOIN dbo.APLedger a ON v.RelID = a.APLedgerID AND v.RelTypeID IN (39, 40)
....
ORDER BY v.Code;
The query above works, however if the CASE statement is still returning an additional column regardless of the result of the subquery. How can I remove it if the subquery doesn't return a row?
How can I do this?
Change the location of AS. For example:
SELECT v.GLTypeID,
v.something,
v.somethignElse,
CASE WHEN (
SELECT TOP 1 Value
FROM InterfaceParam
WHERE InterfaceId = 0 AND Descr = 'creditMemo') = 1
THEN a.CreditID -- AS is not valid here
END AS CreditMemoID -- AS is valid here
FROM vGLDetails2 v
....
LEFT OUTER JOIN AssociationFund f
ON v.FundID = f.FundID
LEFT JOIN dbo.APLedger a ON v.RelID = a.APLedgerID AND v.RelTypeID IN (39, 40)
....
ORDER BY v.Code;
Note: I removed ELSE NULL since this is the default behavior of CASE.
I can't quite get my conditions correct to alter an existing query that works.
SELECT dd.GasYear,D.DealCash_IsAgency,D.DealCash_IsNetBack, dd.GasYearName,dd.MonthName,dd.Season,dd.FirstOfMonthDate,Deal_OrigDate,DATENAME(WEEKDAY, Deal_OrigDate) AS Orig_DayofWeek
,Deal_StartDate,DATENAME(WEEKDAY, Deal_StartDate) AS Start_DayofWeek,Deal_StopDate,DATENAME(WEEKDAY, Deal_StopDate) AS Stop_DayofWeek
,CASE
WHEN CAST(DATEDIFF(d, Deal_StartDate, Deal_StopDate) AS CHAR(10)) > 6 THEN 'All Other'
END AS DealType
,Deal_Owner,d.Deal_Id,CP_Cmp_ShortName
,CASE
WHEN Qty_MeasurementType_Short = 'GJ' THEN DealCashTran_Qty / 2.055098
ELSE DealCashTran_Qty
END DealCashTran_Dth_Qty
FROM uvDealCash d
LEFT OUTER JOIN uvDealCashTran AS dct
ON d.Deal_Id = dct.Deal_Id
INNER JOIN Cmp c
ON d.CP_Cmp_ShortName = c.Cmp_ShortName
INNER JOIN [uvCmp_CmpType_CmpCPType] ct
ON c.Cmp_Id = ct.Cmp_Id
INNER JOIN BISandbox.dbo.dimDate dd
ON d.Deal_StartDate = dd.FullDate
WHERE ((Deal_Owner IN ('TYK', 'JTT', 'MML', 'CGG', 'LIO', 'MAT')
AND ct.CmpCPType_Descript = 'Producer')
--US Entities Kristen's wants added
OR (CP_Cmp_ShortName IN ('SEV', 'CNAT', 'TNMKT', 'CENT')))
AND Deal_LengthInMonths >= 0
AND DATEDIFF(d, Deal_StartDate, Deal_StopDate) > 6
AND Deal_StartDate > '4/1/2018' --=<Parameters.StartDate>
AND DealPurchaseSellType_Short = 'P'
AND d.DealCashType_Short IN ('DFM', 'FM')
In addition to the above I want to exclude results if DealCash_IsAgency = 1 or D.DealCash_IsNetBack = 1
I tried several variations by adding this to the end of the existing WHERE statement.
--exclude AMA and NB
AND (D.DealCash_IsAgency = 1 OR D.DealCash_IsNetBack = 1)
It still returns rows with 1's in both of the columns.
Greatly appreciate any suggestions.
Thanks
Brent
if you want to exclude then you have to add a condition like this by adding not:
AND NOT (D.DealCash_IsAgency = 1 OR D.DealCash_IsNetBack = 1)
So I have just upgraded to MySQL 8 to test if it would be a viable option, but I have stumbled upon quiet the riddle.
It throws an error at 'table_product_features AS pf' but I can't seem to figure out why as in the documentation it is done exactly the same way: https://dev.mysql.com/doc/refman/8.0/en/join.html
Error:
You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'groups ON pf.parent_id = groups.feature_id LEFT JOIN table_product_features_des' at line 1 (1064)
I also checked the bug tracker but that doesnt seem to have anything related to my issue.
What exactly am I doing wrong here?
P.S. I put it in a code snippet so that it would remain formatted.
SELECT pf.feature_id,
pf.company_id,
pf.feature_type,
pf.parent_id,
pf.display_on_product,
pf.display_on_catalog,
pf.display_on_header,
table_product_features_descriptions.description,
table_product_features_descriptions.lang_code,
table_product_features_descriptions.prefix,
table_product_features_descriptions.suffix,
pf.categories_path,
table_product_features_descriptions.full_description,
pf.status,
pf.comparison,
pf.position,
groups.position AS group_position,
table_product_features_values.value,
table_product_features_values.variant_id,
table_product_features_values.value_int
FROM table_product_features AS pf
LEFT JOIN table_product_features AS groups
ON pf.parent_id = groups.feature_id
LEFT JOIN table_product_features_descriptions
ON table_product_features_descriptions.feature_id = pf.feature_id
AND table_product_features_descriptions.lang_code = 'en'
INNER JOIN table_product_features_values
ON table_product_features_values.feature_id = pf.feature_id
AND table_product_features_values.product_id = 230
AND table_product_features_values.lang_code = 'en'
INNER JOIN table_ult_objects_sharing
ON ( table_ult_objects_sharing.share_object_id = pf.feature_id
AND table_ult_objects_sharing.share_company_id = 1
AND table_ult_objects_sharing.share_object_type =
'product_features' )
WHERE 1
AND pf.feature_type != 'G'
AND pf.status IN ( 'A' )
AND pf.display_on_product = 'Y'
AND ( pf.categories_path = ''
OR Isnull(pf.categories_path)
OR Find_in_set(203, pf.categories_path)
OR Find_in_set(215, pf.categories_path)
OR Find_in_set(216, pf.categories_path) )
GROUP BY pf.feature_id
ORDER BY group_position,
pf.position,
table_product_features_descriptions.description
If your not using aggregate functions I don't see why you need the GROUP BY.
Also groups is a reserved word, change it to tgroups or something else.
SELECT pf.feature_id,
pf.company_id,
pf.feature_type,
pf.parent_id,
pf.display_on_product,
pf.display_on_catalog,
pf.display_on_header,
tdesc.description,
tdesc.lang_code,
tdesc.prefix,
tdesc.suffix,
pf.categories_path,
tdesc.full_description,
pf.status,
pf.comparison,
pf.position,
tgroups.position group_position,
tvals.value,
tvals.variant_id,
tvals.value_int
FROM table_product_features pf
LEFT JOIN table_product_features tgroups ON pf.parent_id = tgroups.feature_id
LEFT JOIN table_product_features_descriptions tdesc ON tdesc.feature_id = pf.feature_id AND tdesc.lang_code = 'en'
INNER JOIN table_product_features_values tvals ON tvals.feature_id = pf.feature_id AND tvals.product_id = 230 AND tvals.lang_code = 'en'
INNER JOIN table_ult_objects_sharing tshar ON (tshar.share_object_id = pf.feature_id AND tshar.share_company_id = 1 AND tshar.share_object_type = 'product_features')
WHERE 1
AND pf.feature_type != 'G'
AND pf.status IN ('A')
AND pf.display_on_product = 'Y'
AND (pf.categories_path = '' OR Isnull(pf.categories_path) OR Find_in_set(203, pf.categories_path) OR Find_in_set(215, pf.categories_path) OR Find_in_set(216, pf.categories_path))
ORDER BY group_position, pf.position, tdesc.description
Can anyone help me try to solve this equation? The "Lost_Weight" is a calculation field and I'm trying to get the result of HWeight - renWeight = '0' if the renWeight is 0. The where clause has a controversial statement that put a Challenge to me to make that happen. See below.
SELECT TOP (100) PERCENT dbo.Name.ID,
dbo.Name.CO_ID,
dbo.Name.FULL_NAME,
dbo.Name.STATE_PROVINCE, dbo.Tops_Profile.START_WGHT,
dbo.Tops_Profile.RENEWAL_WEIGHT,
dbo.Tops_Profile.H_WEIGHT - dbo.Tops_Profile.RENEWAL_WEIGHT AS Lost_Weight,
dbo.Tops_Profile.H_WEIGHT,
dbo.Name.JOIN_DATE,
dbo.Name.BIRTH_DATE,
RIGHT(dbo.TOPS_AWARDS.AWARD_TYPE, 15) AS AWARDS_TYPE,
dbo.TOPS_AWARDS.AWARD_CATEGORY,
dbo.TOPS_AWARDS.AWARD_DATE,
dbo.vw_RegDirs.TARGET_ID AS
RD_ID,
dbo.vw_Coords.TARGET_ID AS Coord_ID,
dbo.vw_Coords.FULL_NAME AS
Coord_Name,
dbo.vw_AreaCapts.TARGET_ID AS AC_ID, dbo.Name.STATUS
FROM dbo.Name INNER JOIN dbo.Tops_Profile
ON dbo.Name.ID = dbo.Tops_Profile.ID
INNER JOIN dbo.vw_RegDirs
ON dbo.Name.CO_ID = dbo.vw_RegDirs.CHAPTER
INNER JOIN dbo.vw_Coords
ON dbo.Name.CO_ID = dbo.vw_Coords.CHAPTER
INNER JOIN dbo.vw_AreaCapts
ON dbo.Name.CO_ID = dbo.vw_AreaCapts.CHAPTER
LEFT OUTER JOIN dbo.TOPS_AWARDS
ON dbo.Name.ID = dbo.TOPS_AWARDS.ID
WHERE (dbo.Tops_Profile.RENEWAL_WEIGHT <> '0')
AND (dbo.Name.STATUS = 'a')
AND (dbo.Tops_Profile.H_WEIGHT - dbo.Tops_Profile.RENEWAL_WEIGHT >= 100')
OR (dbo.Name.STATUS = 'a') AND (dbo.TOPS_AWARDS.AWARD_TYPE LIKE '%Century%')
right now if the renewal_weight is 0 it automatically give me the number from the H_WEIGHT in the Lost_Weight field and I need the result to be 0 if the RENEWAL_WEIGHT is 0.
I'm guessing 0 is a special value in this situation. This can be fixed by using CASE
Replace
dbo.Tops_Profile.H_WEIGHT - dbo.Tops_Profile.RENEWAL_WEIGHT AS Lost_Weight
With
CASE WHEN RENEWAL_WEIGHT = 0 THEN 0
ELSE dbo.Tops_Profile.H_WEIGHT - dbo.Tops_Profile.RENEWAL_WEIGHT
END AS Lost_Weight
Here is my statement:
SELECT
CASE
WHEN #UserRole = 1 THEN 1
ELSE 0
END AS [CanEdit],
F.FundingStreamName
FROM FundingStream AS F
LEFT JOIN Projects AS P ON P.FundingStream = F.FundingStreamID
WHERE ProjectNumber = #ProjectNumber
I noticed if FundingStreamID is null, the case statement will return nothing as well, how can I get the case statement to execute regardless if there is a funding stream or not? Thanks.
I think the problem is that your where clause is "undoing" your left outer join. Try moving the condition to the on clause:
SELECT (CASE WHEN #UserRole = 1 THEN 1
ELSE 0
END) AS [CanEdit],
F.FundingStreamName
FROM FundingStream F LEFT JOIN
Projects P
ON P.FundingStream = F.FundingStreamID AND
P.ProjectNumber = #ProjectNumber ;
Using #MicSim answer:
SELECT
CASE
WHEN #UserRole = 1 THEN 1
ELSE 0
END AS [CanEdit],
F.FundingStreamName
FROM FundingStream AS F
RIGHT JOIN Projects AS P ON P.FundingStream = F.FundingStreamID
WHERE ProjectNumber = #ProjectNumber
Thanks again for the help!