How To Import Text Files With The Same Name And Schema But Different Directories Into Database?

August 07, 2024 Post a Comment

I require to import multiple txt files with the same name and same schemas into the same table in SQL Server 2008 database. The problem that I have is that they are all in differen

Solution 1:

Yes. You will want to use a Foreach File Container and then check the Traverse Subfolder option.

Edit

Apparently my answer wasn't cromulent enough, so please accept this working code which illustrates what my brief original answer stated.

Source data

I created 3 folders as described above to contain files sample1.txt and sample2.txt

C:\>MKDIR SSISDATA\SO\TEST\201304C:\>MKDIR SSISDATA\SO\TEST\201305C:\>MKDIR SSISDATA\SO\TEST\201306

The contents of the file are below. Each version of the file in each folder has the ID value incremented along with the text values altered to prove it has picked up the new file.

ID,value
1,ABC

Package generation

This part assumes you have BIDS Helper installed. It is not required for the solution but simply provides a common framework future readers could use to reproduce this solution

I created a BIML file with the following content. Even though I have the table create step in there, I needed to have that run on the target server prior to generating the package.

<Bimlxmlns="http://schemas.varigence.com/biml.xsd"><!-- Create a basic flat file source definition --><FileFormats><FlatFileFormatName="FFFSrc"CodePage="1252"RowDelimiter="CRLF"IsUnicode="false"FlatFileType="Delimited"ColumnNamesInFirstDataRow="true"
        ><Columns><ColumnName="ID"DataType="Int32"Delimiter=","ColumnType="Delimited"
                /><ColumnName="value"DataType="AnsiString"Delimiter="CRLF"InputLength="20"MaximumWidth="20"Length="20"CodePage="1252"ColumnType="Delimited"
                    /></Columns></FlatFileFormat></FileFormats><!-- Create a connection that uses the flat file format defined above--><Connections><FlatFileConnectionName="FFSrc"FileFormat="FFFSrc"FilePath="C:\ssisdata\so\TEST\201306\sample1.txt"DelayValidation="true"
        /><OleDbConnectionName="tempdb"ConnectionString="Data Source=localhost\dev2012;Initial Catalog=tempdb;Provider=SQLNCLI11.1;Integrated Security=SSPI;Auto Translate=False;"
        /></Connections><!-- Create a package to illustrate how to apply an expression on the Connection Manager --><Packages><PackageName="so_19957451"ConstraintMode="Linear"
        ><Connections><ConnectionConnectionName="tempdb"/><ConnectionConnectionName="FFSrc"><Expressions><!-- Assign a variable to the ConnectionString property. 
                        The syntax for this is ConnectionManagerName.Property --><ExpressionPropertyName="FFSrc.ConnectionString">@[User::CurrentFileName]</Expression></Expressions></Connection></Connections><!-- Create a single variable that points to the current file --><Variables><VariableName="CurrentFileName"DataType="String">C:\ssisdata\so\TEST\201306\sample1.txt</Variable><VariableName="FileMask"DataType="String">*.txt</Variable><VariableName="SourceFolder"DataType="String">C:\ssisdata\so\TEST</Variable><VariableName="RowCountInput"DataType="Int32">0</Variable><VariableName="TargetTable"DataType="String">[dbo].[so_19957451]</Variable></Variables><!-- Add a foreach file enumerator. Use the above --><Tasks><ExecuteSQLName="SQL Create Table"ConnectionName="tempdb"><DirectInput>
                        IF NOT EXISTS (SELECT * FROM sys.tables T WHERE T.name = 'so_19957451' and T.schema_id = schema_id('dbo'))
                        BEGIN
                            CREATE TABLE dbo.so_19957451(ID int NOT NULL, value varchar(20) NOT NULL);
                        END
                    </DirectInput></ExecuteSQL><ForEachFileLoopName="FELC Consume files"FileSpecification="*.csv"ProcessSubfolders="true"RetrieveFileNameFormat="FullyQualified"Folder="C:\"ConstraintMode="Linear"
                ><!-- Define the expressions to make the input folder and the file mask 
                    driven by variable values --><Expressions><ExpressionPropertyName="Directory">@[User::SourceFolder]</Expression><ExpressionPropertyName="FileSpec">@[User::FileMask]</Expression></Expressions><VariableMappings><!-- Notice that we use the convention of User.Variable name here --><VariableMappingName="0"VariableName="User.CurrentFileName"
                        /></VariableMappings><Tasks><DataflowName="DFT Import file"DelayValidation="true"><Transformations><FlatFileSourceName="FFS Sample"ConnectionName="FFSrc"/><RowCountName="RC Source"VariableName="User.RowCountInput"/><OleDbDestinationName="OLE_DST"ConnectionName="tempdb"><TableFromVariableOutputVariableName="User.TargetTable"/></OleDbDestination></Transformations></Dataflow></Tasks></ForEachFileLoop></Tasks></Package></Packages></Biml>

Right click on the biml file and select Generate SSIS Package. At this point, you should have a package named so_19957451 added to your current SSIS project.

Package configuration

There's no need for any configuration because it's already been done via BIML but moar screenshots make for better answers.

This is the basic package

Control flow

Here are my variables

Variables

Configuration of the Foreach Loop, as called out in the MSDN article as well as my note of select the Traverse subfolder

Traverse Subfolder

Assign the value generated per loop to the variable Current

assign variables

The flat file source has an expression applied to the ConnectionString property to ensure it uses the Variable @User::CurrentFileName. This changes the source per execution of the loop.

enter image description here

Execution results

Results from the database