First, I have read about similar posts and have read the comments that this isn't an ideal solution and I get it but the boss (ie client) wants it this way. The parameters are as follows (for various reasons too bizarre to go into but trust me):
1. SQL Server Mgmt Studio 2016
2. NO parameters or pass throughs or temp tables. All has to be within contained code.
So here we go:
I need to create column headings that reflect dates:
1. Current date
2. Most recent quarter end prior to current date
3. Most recent quarter end prior to #2
4. Most recent quarter end prior to #3
5. Most recent quarter end prior to #4
6. Most recent quarter end prior to #5
So if using today's date, my column names would be as follows
12/18/2016 9/30/2016 6/30/2016 3/31/2016 12/31/2016 9/30/2015
I can easily do it in SAS but can't in SQL given the requirements stated above.
Help please with same code.
Thank you
Paula
Seems like a long way to go for something which really belongs in the presentation layer. That said, consider the following:
Let's assume you maintain a naming convention for your calculated fields, for example [CurrentDay], [QtrMinus1], [QtrMinus2], [QtrMinus3], [QtrMinus4],[QtrMinus5]. Then we can wrap your complicated query in some dynamic SQL.
Just as an illustration, let's assume your current query results looks like this
After the "wrap", the results will then look like so:
The code - Since you did NOT exclude Dynamic SQL.
Declare #S varchar(max)='
Select [CustName]
,['+convert(varchar(10),GetDate(),101)+'] = [CurrentDay]
,['+Convert(varchar(10),EOMonth(DateFromParts(Year(DateAdd(QQ,-1,GetDate())),DatePart(QQ,DateAdd(QQ,-1,GetDate()))*3,1)),101)+'] = [QtrMinus1]
,['+Convert(varchar(10),EOMonth(DateFromParts(Year(DateAdd(QQ,-2,GetDate())),DatePart(QQ,DateAdd(QQ,-2,GetDate()))*3,1)),101)+'] = [QtrMinus2]
,['+Convert(varchar(10),EOMonth(DateFromParts(Year(DateAdd(QQ,-3,GetDate())),DatePart(QQ,DateAdd(QQ,-3,GetDate()))*3,1)),101)+'] = [QtrMinus3]
,['+Convert(varchar(10),EOMonth(DateFromParts(Year(DateAdd(QQ,-4,GetDate())),DatePart(QQ,DateAdd(QQ,-4,GetDate()))*3,1)),101)+'] = [QtrMinus4]
,['+Convert(varchar(10),EOMonth(DateFromParts(Year(DateAdd(QQ,-5,GetDate())),DatePart(QQ,DateAdd(QQ,-5,GetDate()))*3,1)),101)+'] = [QtrMinus5]
From (
-- Your Complicated Query --
Select * from YourTable
) A
'
Exec(#S)
If it helps the visualization, the generated SQL is as follows:
Select [CustName]
,[12/18/2016] = [CurrentDay]
,[09/30/2016] = [QtrMinus1]
,[06/30/2016] = [QtrMinus2]
,[03/31/2016] = [QtrMinus3]
,[12/31/2015] = [QtrMinus4]
,[09/30/2015] = [QtrMinus5]
From (
-- Your Complicated Query --
Select * from YourTable
) A
Here is one way using dynamic query
DECLARE #prior_quarters INT = 4,
#int INT =1,
#col_list VARCHAR(max)=Quotename(CONVERT(VARCHAR(20), Getdate(), 101))
WHILE #int <= #prior_quarters
BEGIN
SELECT #col_list += Concat(',', Quotename(CONVERT(VARCHAR(20), Eomonth(Getdate(), ( ( ( ( Month(Getdate()) - 1 ) % 3 ) + 1 ) * -1 ) * #int), 101)))
SET #int+=1
END
--SELECT #col_list -- for debugging
EXEC ('select '+#col_list+' from yourtable')
Related
I have a Visual studio based stored procedure that generates a report for a monthly audit process. In the database being queried, all data for each month lives in its own individual table (Contacts_month_1, Contacts_month_2, etc.)
The SQL used in this report generation has some minor logic included, to allow it to work dynamically, rather than use hard coded dates. The problem arose at the start of January 2017, when I started receiving not just the results for the prior month, but additionally the prior year as well. To be specific, the audit report for December 2016 included data for both 12/2016 and 12/2015. Initially I thought it was a fluke of some kind based on the turn of the year, and we have not had this automated process during the turn as of yet. Unfortunately when I came in to the office today, inside the output file for January 2017, I also received the results for January 2016.
I attempted to include a year check to the process, however I am still getting the same result output. Any ideas would be appreciated.
Declare #GetMonth TinyInt
,#SessionTable varchar(50)
,#ContactTable varchar(50)
,#TableVersion varchar(2)
Declare #GetYear SmallInt
,#SessionTable_year varchar(50)
,#ContactTable_year varchar(50)
,#TableVersion_year varchar(4)
Set #GetMonth=MONTH(cast(GetDate() as Datetime))-1
Set #GetYear=YEAR(cast(GetDate() as Datetime))
If (#getmonth=0) Set #GetMonth=12 + (#GetYear-1)
Set #TableVersion=CAST(#getMonth as varchar(2))
Set #SessionTable='[CentralDWH].[dbo].[Sessions_month_' +#tableversion +']'
Set #ContactTable ='[CentralDWH].[dbo].[Contacts_month_' +#tableversion +']'
-- Select #GetMonth,#GetYear (DEBUGGING STATEMENT)
-- Select #SessionTable,#ContactTable (DEBUGGING STATEMENT)
Exec('SELECT [PBX_id] as AgentID
,[p22_value] as Skill
,''Athens'' as Location
,Convert(varchar(20),c.[local_start_time],120) as local_start_time
,convert(varchar(20),c.[local_end_time],120) as local_end_time
,U.[USER_NAME]
,call_id
FROM '+#SessionTable +' S
Inner join dbo.Users U on S.user_key=U.user_key
inner Join '+ #ContactTable+' C on S.contact_key=C.contact_key
Where is_screen > 0
And Unit_Num between 398003 and 398005
and P22_value is not null
and c.[local_start_time] > ' + #GetYear
+ ' order by local_start_time')
As I understand, the #GetMonth variable is used for returning the previous month
Set #GetMonth = MONTH(CAST(GetDate() AS Datetime)) - 1
After a quick look after you procedure my first issue was this line of code:
IF (#getmonth = 0)
SET #GetMonth = 12 + (#GetYear - 1)
I don't understand why are you setting the #GetMonth variable to 12 + current year -1 and I assume this is the cause to the problem.
Did you want to get the 12th month of the previous year when the current month is 1 (January)? If yes then you can easily change the If block to
If #GetMonth = 0
Begin
Set #GetMonth = 12
Set #GetYear = #GetYear - 1
End
Other issues:
It's recommended to keep the consistency of the names of the variables #GetMonth, #getmonth, this will cause an error if the database collation is case sensitive.
#GetMonth is declared as TinyInt and this will cause an arithmetic overflow if you try to store the year
I recommend testing the dynamic SQL statement that you are composing here with some hard coded values to check the results returned, you can use January and 2016 to check if the actual issue in your procedure or it's in your query.
Hope it helps
Thanks for your help, I figured out the root of the problem, and it was because i was not casting the GetYear as a varchar when trying to run the T-SQL statement. This in turn caused the variable to be completely ignored. I also cleaned up the query a little bit after realizing i was goofing up pretty hard.
Below is the cleaned up functional query, so that it may help someone in the future:
Declare #GetMonth SmallInt,
#SessionTable varchar(50),
#ContactTable varchar(50),
#TableVersion varchar(2),
#GetYear SmallInt,
#YearCheck varchar(4)
Set #GetMonth=MONTH(cast(GetDate() as Datetime))-1
Set #GetYear=YEAR(cast(GetDate() as Datetime))-1
If (#GetMonth=0)
Begin
Set #GetMonth =12
Set #GetYear =#GetYear - 1
End
Set #TableVersion=CAST(#GetMonth as varchar(2))
Set #SessionTable='[CentralDWH].[dbo].[Sessions_month_' +#tableversion +']'
Set #ContactTable ='[CentralDWH].[dbo].[Contacts_month_' +#tableversion +']'
Set #YearCheck=CAST(#GetYear as varchar(4))
--Select #GetMonth,#GetYear,#YearCheck (DEBUGGING STATEMENT)
-- Select #SessionTable,#ContactTable (DEBUGGING STATEMENT)
Exec('SELECT
[PBX_id] as AgentID,
[p22_value] as Skill,
''Athens'' as Location,
Convert(varchar(20),c.[local_start_time],120) as local_start_time,
convert(varchar(20),c.[local_end_time],120) as local_end_time,
U.[USER_NAME],
call_id
FROM '+#SessionTable +' S
Inner join dbo.Users U on S.user_key=U.user_key
inner Join '+ #ContactTable+' C on S.contact_key=C.contact_key
Where is_screen>0
And Unit_Num between 398003 and 398005
And P22_value is not null
And year(c.[local_start_time]) > '+#YearCheck+'
order by local_start_time')
Once I cleaned all this up and remembered to cast the year properly, everything fell into place.
I'm trying to extract a portion of a value out of a database column using SQL server.
The example below works in a simple context with a varchar field. The result is: &kickstart& which is what I want.
I now want to do the same when retrieving a column from the database.
But SQL does not like what I am doing. I'm thinking it is something easy that I am not seeing.
Declare #FileName varchar(20) = '&kickstart&.cfg'
Declare #StartPos integer = 0
Declare #FileNameNoExt varchar(20)
SELECT #FileNameNoExt = Left(#FileName,( (charindex('.', #FileName, 0)) - 1))
SELECT #FileNameNoExt
Here is the SQL statement that I can't seem to get to work for me:
Declare #FileNameNoExt as varchar(20)
SELECT
i.InstallFileType AS InstallFileType,
o.OSlabel AS OSLabel,
SELECT #FileNameNoExt = (LEFT(oi.FIleName,( (charindex('.', oi.FIleName, 0) ) - 1) )) AS FileNameNoExt,
oi.FIleName AS FIleName
FROM
dbo.OperatingSystemInstallFiles oi
JOIN dbo.InstallFileTypes i ON oi.InstallFileTypeId = i.InstallFileTypeId
JOIN dbo.OperatingSystems o ON oi.OperatingSystemId = o.OperatingSystemId
Why do you need the variable at all? What's wrong with:
SELECT
i.InstallFileType AS InstallFileType,
o.OSlabel AS OSLabel,
LEFT(oi.FIleName,( (charindex('.', oi.FIleName, 0) ) - 1) ) AS FileNameNoExt,
oi.FIleName AS FIleName
FROM
dbo.OperatingSystemInstallFiles oi
JOIN dbo.InstallFileTypes i ON oi.InstallFileTypeId = i.InstallFileTypeId
JOIN dbo.OperatingSystems o ON oi.OperatingSystemId = o.OperatingSystemId
You've put a SELECT inside another SELECT list without nesting, which is a syntax error in SQL Server.
You are also attempting to assign a variable while performing a data-retrieval operation. You can select all data to be shown, or all data into variables but not both at the same time.
When the two issues above are resolved, I think you may still run into issues when committing filenames into a variable which only allows 20 characters - but then I don't know anything about your dataset.
I have an MS-SQL table, with a column titled 'ImportCount'.
Data in this column follows the below format:
ImportCount
[Schedules] 1376 schedule items imported from location H:\FOLDERA\AA\XX...
[Schedules] 10201 schedule items imported from location H:\FOLDERZZ\PERS\YY...
[Schedules] 999 schedule items imported from location R:\PERS\FOLDERA\AA\XX...
[Schedules] 21 schedule items imported from location H:\FOLDERA\MM\2014ZZ...
What I would like to do is extract that numerical portion of the data (which varies in length), but am struggling to get the right result. Would appreciate any help on this!
Thanks.
Try
select left(ImportCount, patindex('%[^0-9]%', ImportCount+'.') - 1)
select SUBSTRING(ImportCount,13,patindex('% schedule items%',ImportCount)-13) from table name
Try this..You can declare it as a SQL function also.
DECLARE #intText INT
DECLARE #textAplhaNumeric varchar(100)
set #textAplhaNumeric = '1376 schedule items imported from location'
SET #intText = PATINDEX('%[^0-9]%', #textAplhaNumeric)
BEGIN
WHILE #intText > 0
BEGIN
SET #textAplhaNumeric = STUFF(#textAplhaNumeric, #intText, 1, '' )
SET #intText = PATINDEX('%[^0-9]%', #textAplhaNumeric)
END
END
Select #textAplhaNumeric //output is 1376
It will work in case of NULL or empty values.
Please try:
SELECT LEFT(Val,PATINDEX('%[^0-9]%', Val+'a')-1) from(
SELECT
STUFF(ImportCount, 1, PATINDEX('%[0-9]%', ImportCount)-1, '') Val
FROM YourTable
)x
Right now I have this code to find next and previous rows using SQL Server 2005. intID is the gallery id number using bigint data type:
SQL = "SELECT TOP 1 max(p.galleryID) as previousrec, min(n.galleryID) AS nextrec FROM gallery AS p CROSS JOIN gallery AS n where p.galleryid < '"&intID&"' and n.galleryid > '"&intID&"'"
Set rsRec = Server.CreateObject("ADODB.Recordset")
rsRec.Open sql, Conn
strNext = rsRec("nextrec")
strPrevious = rsRec("previousrec")
rsRec.close
set rsRec = nothing
Problem Number 1:
The newest row will return nulls on the 'next record' because there is none. The oldest row will return nulls because there isn't a 'previous record'. So if either the 'next record' or 'previous record' doesn't exist then it returns nulls for both.
Problem Number 2:
I want to create a stored procedure to call from the DB so intid can just be passed to it
TIA
This will yield NULL for previous on the first row, and NULL for next on the last row. Though your ordering seems backwards to me; why is "next" lower than "previous"?
CREATE PROCEDURE dbo.GetGalleryBookends
#GalleryID INT
AS
BEGIN
SET NOCOUNT ON;
;WITH n AS
(
SELECT galleryID, rn = ROW_NUMBER()
OVER (ORDER BY galleryID)
FROM dbo.gallery
)
SELECT
previousrec = MAX(nA.galleryID),
nextrec = MIN(nB.galleryID)
FROM n
LEFT OUTER JOIN n AS nA
ON nA.rn = n.rn - 1
LEFT OUTER JOIN n AS nB
ON nB.rn = n.rn + 1
WHERE n.galleryID = #galleryID;
END
GO
Also, it doesn't make sense to want an empty string instead of NULL. Your ASP code can deal with NULL values just fine, otherwise you'd have to convert the resulting integers to strings every time. If you really want this you can say:
previousrec = COALESCE(CONVERT(VARCHAR(12), MIN(nA.galleryID)), ''),
nextrec = COALESCE(CONVERT(VARCHAR(12), MAX(nB.galleryID)), '')
But this will no longer work well when you move from ASP to ASP.NET because types are much more explicit. Much better to just have the application code be able to deal with, instead of being afraid of, NULL values.
This seems like a lot of work to get the previous and next ID, without retrieving any information about the current ID. Are you implementing paging? If so I highly recommend reviewing this article and this follow-up conversation.
Try this (nb not tested)
SELECT TOP 1 max(p.galleryID) as previousrec, min(n.galleryID) AS nextrec
FROM gallery AS p
CROSS JOIN gallery AS n
where (p.galleryid < #intID or p.galleryid is null)
and (n.galleryid > #intID or n.galleryid is null)
I'm assuming you validate that intID is an integer before using this code.
As for a stored procedure -- are you asking how to write a stored procedure? If so there are many tutorials which are quite good on the web.
Since Hogan contributed with the SQL statement, let me contribute with the stored proc part:
CREATE PROCEDURE spGetNextAndPreviousRecords
-- Add the parameters for the stored procedure here
#intID int
AS
BEGIN
-- SET NOCOUNT ON added to prevent extra result sets from
-- interfering with SELECT statements.
SET NOCOUNT ON;
SELECT TOP 1 max(p.galleryID) as previousrec, min(n.galleryID) AS nextrec
FROM gallery AS p
CROSS JOIN gallery AS n
where (p.galleryid < #intID or p.galleryid is null)
and (n.galleryid > #intID or n.galleryid is null)
END
And you call this from code as follows (assuming VB.NET):
Using c As New SqlConnection(ConfigurationManager.ConnectionStrings("ConnectionString").ConnectionString)
c.Open()
Dim command = New SqlCommand("spGetNextAndPreviousRecords")
command.Parameters.AddWithValue("#intID", yourID)
Dim reader as SqlDataReader = command.ExecuteReader()
While(reader.Read())
' read the result here
End While
End Using
I have a stored Procedure that works fine joining 2 tables together. I needed to add a new field from a new table that was not included in the original SP. What I am trying to do is sum a field from the new table for each record that is a child record of the Parent table which is in the original SP.
I tested the Sum based on th parent table in a test query and it works fine:
select totaldollars from TTS_EmpTime where emptimedaytotal_id='32878'
so then the next step would be to integrate into the SP. I did so and have set the new portions of the SP to be bold so you can see what was added. IF the bold portions are removed the SP works fine if not I get this error:
*Msg 8120, Level 16, State 1, Procedure TTS_RptTest2, Line 11
Column 'TTS_EmpTimeDayTotal.EmployeeID' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause*.
Here is my Stored Proc:
USE [TTSTimeClock]
GO
/****** Object: StoredProcedure [dbo].[TTS_RptTest2] Script Date: 03/04/2011 12:29:59 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
ALTER procedure [dbo].[TTS_RptTest2]
#BureauID nvarchar(36),
#CompanyID nvarchar(36),
#DivisionID nvarchar(10) ,
#punchDate smalldatetime,
#PeriodDays integer,
#EmployeeID nvarchar(20) = null
As
--with DayTotals as(
select
DayTotal.DivisionID,
DayTotal.EmployeeID,
EmpData.EmployeeFirstName AS First,
EmpData.EmployeeLastName AS Last,
EmpData.employeetypeid AS EmpId,
DayTotal.ID as DayTotalID,
-- Format the Date as MM/DD DOW or 2Digit Month & 2Digit Day and the 3Char Day of the week Uppercase
convert(varchar(5),DayTotal.PunchDate,101) + ' ' + upper(left(datename(dw,DayTotal.Punchdate),3))as PunchDate,
-- Format the in and out time as non military time with AM or PM No Dates
substring(convert(varchar(20), DayTotal.FirstDayPunch, 9), 13, 5) + ' ' + substring(convert(varchar(30), DayTotal.FirstDayPunch, 9), 25, 2)as TimeIn,
substring(convert(varchar(20), DayTotal.LastDayPunch, 9), 13, 5) + ' ' + substring(convert(varchar(30), DayTotal.LastDayPunch, 9), 25, 2) as TimeOut,
DayTotal.RegularHours,
DayTotal.NonOvertimeHours,
DayTotal.OvertimeHours,
DayTotal.TotalDayHRS,
DayTotal.PeriodRegular,
DayTotal.PeriodOtherTime,
DayTotal.PeriodOvertime,
DayTotal.PeriodTotal,
**sum(cast(EmpTime.TotalDollars as float)) as TotalDayDollars**
from TTS_EmpTimeDayTotal as DayTotal
INNER JOIN TTS_PayrollEmployees AS EmpData
ON DayTotal.EmployeeID = EmpData.EmployeeID
**inner JOIN TTS_Emptime as EmpTime
ON DayTotal.id = emptime.emptimedaytotal_id**
where
DayTotal.BureauID = #BureauID
AND DayTotal.CompanyID = #CompanyID
AND (DayTotal.DivisionID = #DivisionID)
AND daytotal.periodstart =
-- Period start date
(SELECT DISTINCT PeriodStart
FROM TTS_EmpTimeDayTotal
WHERE(BureauID = #BureauID) AND (CompanyID = #CompanyID) AND ( (DivisionID = #DivisionID))
AND (PunchDate = #punchDate)and periodend = dateadd(d,(#PeriodDays - 1),(periodstart)))
AND daytotal.periodend =
-- Period End Date
(SELECT DISTINCT PeriodEnd
FROM TTS_EmpTimeDayTotal
WHERE(BureauID = #BureauID) AND (CompanyID = #CompanyID) AND ( (DivisionID = #DivisionID))
AND (PunchDate = #punchDate)and periodend = dateadd(d,(#PeriodDays-1),(periodstart)))
-- Optional all employees or just one
AND (( #EmployeeID is Null) or (DayTotal.EmployeeID = #EmployeeID))
order by Empdata.employeetypeid,DayTotal.punchdate
I am not grouping at all so this must be caused by something else?
Any Help will be appreciated
Is this SQL Server? Looks like it. You're using SUM, an aggregate function, which I don't believe you can use without a GROUP BY clause. Did you always have the SUM in there, or did you add it alongside the new table?
If the latter, that may well be your problem.
Update
Based on OP's comment:
Wow that could be a pain would I do
somehing like groupby field1,field2,
and so on? as in a coma delimited
list. Is there another way to include
this one field that would be better?
Yes, in SQL Server you must be explicit with groupings when using an aggregate function. One alternative in your case would be to do the grouping as a subquery, and join on that, i.e.:
FROM TTS_EmpTimeDayTotal AS DayTotal
INNER JOIN TTS_PayrollEmployees AS EmpData ON DayTotal.EmployeeID = EmpData.EmployeeID
INNER JOIN (SELECT EmpTimeDayTotal_id, SUM(CAST(TotalDollars AS FLOAT)) AS TotalDayDollars
FROM TTS_Emptime
GROUP BY EmpTimeDayTotal_id) AS EmpTime ON DayTotal.id = EmpTime.EmpTimeDayTotal_id
And then simply reference EmpTime.TotalDayDollars in the SELECT list, instead of performing the SUM there.