Case
I have a list of teacher and student combinations with one combination per record and I want to created a comma delimited list of students per teacher.
Solution
If your source is a database then the easiest solution is a TSQL query like this:
If your source is for example a flat file or a database that doesn't support a query like this, then there are also options within SSIS. For this solution I use a asynchronous Script Component.
1) Sorted source
We need a sorted source because we are comparing records with each other. In this case make sure the source is sorted on teacher first (and optional secondly on student).
2) Script Component
Add a Script Component (type transformation) and select the Teacher and Student columns as ReadOnly input columns.
3) Asynchronous
We need to make the Script Component asynchronous because it throws out a different number of rows than there are incomming. Go to the Inputs and Outputs pane, click on Output 0 and change the SynchronousInputID to None.
4) Output
We now need to create an output for the Script Component. Expand the Output 0 and add two columns:
Teacher (same data type and size as the input column teacher)
Students (same data type as the input column student, but larger to fit multiple student names)
5) The Script
Copy the three variables and the two methods to your Script Component (and remove any other existing methods).
or in VB.Net
Note: You can change the delimiter in C# line 70 and VB.Net line 64.
6) Testing
For testing purposes I added a derived column and a couple of data viewer.
Conclusion: both methods have the same result. For a SQL database source, the T-SQL method is probably a little faster. If you can do the sort in the source the performance differences will diminish.
I have a list of teacher and student combinations with one combination per record and I want to created a comma delimited list of students per teacher.
Solution
If your source is a database then the easiest solution is a TSQL query like this:
--TSQL Query WITH UniqueTeachers AS ( SELECT DISTINCT Teacher FROM TeacherStudentTable ) SELECT Teacher , Students = STUFF(( SELECT ',' + Student FROM TeacherStudentTable WHERE Teacher = UniqueTeachers.Teacher ORDER BY Student FOR XML PATH( '' ), TYPE).value( '.' , 'varchar(100)' ), 1, 1, '' ) FROM UniqueTeachers ORDER BY Teacher |
The query in SSIS as source |
If your source is for example a flat file or a database that doesn't support a query like this, then there are also options within SSIS. For this solution I use a asynchronous Script Component.
1) Sorted source
We need a sorted source because we are comparing records with each other. In this case make sure the source is sorted on teacher first (and optional secondly on student).
Add Sort transformation if source isn't sorted |
2) Script Component
Add a Script Component (type transformation) and select the Teacher and Student columns as ReadOnly input columns.
Input columns: Teacher and Student |
3) Asynchronous
We need to make the Script Component asynchronous because it throws out a different number of rows than there are incomming. Go to the Inputs and Outputs pane, click on Output 0 and change the SynchronousInputID to None.
Asynchonous |
4) Output
We now need to create an output for the Script Component. Expand the Output 0 and add two columns:
Teacher (same data type and size as the input column teacher)
Students (same data type as the input column student, but larger to fit multiple student names)
Output columns |
5) The Script
Copy the three variables and the two methods to your Script Component (and remove any other existing methods).
// C# Code using System; using System.Data; using Microsoft.SqlServer.Dts.Pipeline.Wrapper; using Microsoft.SqlServer.Dts.Runtime.Wrapper; [Microsoft.SqlServer.Dts.Pipeline.SSISScriptComponentEntryPointAttribute] public class ScriptMain : UserComponent { bool initialRow = true ; // Indicater for the first row string teacher = "" ; // Name of the teacher to track teacherchanges between rows string students = "" ; // The comma delimited list of students public override void Input0_ProcessInput(Input0Buffer Buffer) { // Loop through buffer while (Buffer.NextRow()) { // Process an input row Input0_ProcessInputRow(Buffer); // Change the indicator after the first row has been processed initialRow = false ; } // Check if this is the last row if (Buffer.EndOfRowset()) { // Fill the columns of the existing output row with values // from the variable before closing this Script Component Output0Buffer.Teacher = teacher; Output0Buffer.Students = students; } } public override void Input0_ProcessInputRow(Input0Buffer Row) { if (initialRow) { // This is for the first input row only // Create a new output row Output0Buffer.AddRow(); // Now fill the variables with the values from the input row teacher = Row.Teacher; students = Row.Student; } else if ((!initialRow) & (teacher != Row.Teacher)) { // This isn't the first row, but the teacher did change // Fill the columns of the existing output row with values // from the variable before creating a new output row Output0Buffer.Teacher = teacher; Output0Buffer.Students = students; // Create a new output row Output0Buffer.AddRow(); // Now fill the variables with the values from the input row teacher = Row.Teacher; students = Row.Student; } else if ((!initialRow) & (teacher == Row.Teacher)) { // This isn't the first row, and the teacher did not change // Concatenate the studentsname to the variable students += "," + Row.Student; } } // Little explanation: // Rows are created in memory with .AddRow() // and will be submitted to the output when a // new / subsequent row is created or when // the last buffer has been finished. } |
or in VB.Net
'VB.Net code Imports System Imports System.Data Imports System.Math Imports Microsoft.SqlServer.Dts.Pipeline.Wrapper Imports Microsoft.SqlServer.Dts.Runtime.Wrapper <Microsoft.SqlServer.Dts.Pipeline.SSISScriptComponentEntryPointAttribute()> _ <CLSCompliant( False )> _ Public Class ScriptMain Inherits UserComponent Private initialRow As Boolean = True ' Indicater for the first row Private teacher As String = "" ' Name of the teacher to track teacherchanges between rows Private students As String = "" ' The comma delimited list of students Public Overrides Sub Input0_ProcessInput(Buffer As Input0Buffer) ' Loop through buffer While Buffer.NextRow() ' Process an input row Input0_ProcessInputRow(Buffer) ' Change the indicator after the first row has been processed initialRow = False End While ' Check if this is the last row If Buffer.EndOfRowset() Then ' Fill the columns of the existing output row with values ' from the variable before closing this Script Component Output0Buffer.Teacher = teacher Output0Buffer.Students = students End If End Sub Public Overrides Sub Input0_ProcessInputRow(Row As Input0Buffer) If initialRow Then ' This is for the first input row only ' Create a new output row Output0Buffer.AddRow() ' Now fill the variables with the values from the input row teacher = Row.Teacher students = Row.Student ElseIf ( Not initialRow) And (teacher <> Row.Teacher) Then ' This isn't the first row, but the teacher did change ' Fill the columns of the existing output row with values ' from the variable before creating a new output row Output0Buffer.Teacher = teacher Output0Buffer.Students = students ' Create a new output row Output0Buffer.AddRow() ' Now fill the variables with the values from the input row teacher = Row.Teacher students = Row.Student ElseIf ( Not initialRow) And (teacher = Row.Teacher) Then ' This isn't the first row, and the teacher did not change ' Concatenate the studentsname to the variable students += "," & Convert.ToString(Row.Student) End If End Sub ' Little explanation: ' Rows are created in memory with .AddRow() ' and will be submitted to the output when a ' new / subsequent row is created or when ' the last buffer has been finished. End Class |
Note: You can change the delimiter in C# line 70 and VB.Net line 64.
6) Testing
For testing purposes I added a derived column and a couple of data viewer.
The result |
Conclusion: both methods have the same result. For a SQL database source, the T-SQL method is probably a little faster. If you can do the sort in the source the performance differences will diminish.
No comments:
Post a Comment