How to separate variables and values, then insert in a table? - sql-server

Question

Problem
A stored procedure is receiving list of variables and values, and the delimiter. This stored procedure needs to insert those in a table.
--Example table
create table #tempo
(
Variable1 int,
Variable2 int,
Variable3 int
)
These are the parameters to the stored procedure:
declare #variableList varchar(100)
declare #valueList varchar(100)
declare #separator char(1)
set #variableList = 'Variable1#Variable2#Variable3'
set #valueList = '1111#2222#3333'
set #separator = '#'
Result
What I want to achieve is this:
select * from #tempo
+---------+---------+---------+
|Variable1|Variable2|Variable3|
+---------+---------+---------+
|1111 |2222 |3333 |
+---------+---------+---------+
One way to do it
I can use a loop and build dynamic SQL but I want to avoid it. Other than the obvious reasons for not using dynamic SQL, the loop structure is hard to maintain, explain and testing can become an issue too.
Ideal way
I am thinking about a more elegant way to do this, for example with string_split or coalesce etc. But cannot figure out a way without using dynamic SQL or loops.

Answer 1

If you always have same set of column names then it is very easy to do with pivoting, but if columns are changing then you can use the same script with dynamically adjusted list of variables, provided as a parameter or from direct reading from temp table:
INSERT INTO #tempo SELECT *
FROM (
SELECT [value], rv = 'Variable' + CAST(Row_Number() OVER ( ORDER BY (SELECT 1)) as VARCHAR)
FROM STRING_SPLIT(#valueList,#separator)
) AS src
PIVOT (MAX([value]) FOR rv IN (Variable1,Variable2,Variable3)) AS pvt;

Answer 2

You can always try pivoting out the data. This is just the select, but could easily have an insert wrapped into it.
We use a split string with a row ID to allow matching of two split data sets. Function is :
CREATE FUNCTION [dbo].[Split] (#RowData NVARCHAR(MAX), #SplitOn NVARCHAR(5))
RETURNS #RtnValue TABLE (Id INT IDENTITY(1, 1), Data NVARCHAR(100))
AS
BEGIN
DECLARE #Cnt INT;
SET #Cnt = 1;
WHILE (CHARINDEX(#SplitOn, #RowData) > 0)
BEGIN
INSERT INTO #RtnValue (Data)
SELECT Data = LTRIM(RTRIM(SUBSTRING(#RowData, 1, CHARINDEX(#SplitOn, #RowData) - 1)));
SET #RowData = SUBSTRING(#RowData, CHARINDEX(#SplitOn, #RowData) + 1, LEN(#RowData));
SET #Cnt = #Cnt + 1;
END;
INSERT INTO #RtnValue (Data)
SELECT Data = LTRIM(RTRIM(#RowData));
RETURN;
END;
You can then join the two sets together to give some key value pairs, and from there pivot out the data to give the format you requested. If you replace the last select with a select from any of the previous cte's then you can see how the logic unfolds.
DECLARE #variableList VARCHAR(100);
DECLARE #valueList VARCHAR(100);
DECLARE #separator CHAR(1);
SET #variableList = 'Variable1,Variable2,Variable3';
SET #valueList = '1111, 2222, 3333';
SET #separator = ',';
WITH cteVar AS (SELECT Id, Data FROM dbo.Split(#variableList, #separator) )
, cteVal AS (SELECT Id, Data FROM dbo.Split(#valueList, #separator) )
, cteData AS
(SELECT cteVar.Data VariableData
, cteVal.Data ValueData
FROM cteVar
JOIN cteVal ON cteVal.Id = cteVar.Id)
, ctePivot AS
(SELECT *
FROM cteData
PIVOT ( MAX(ValueData)
FOR VariableData IN ([Variable1], [Variable2], [Variable3])) AS PivotTable)
SELECT *
FROM ctePivot;
This is quite a long approach to it but hopefully it well help you understand the steps involved. Its worth looking at the Pivot function in general anyway, its well documented.