开发自定义源组件

开发人员可以通过 SQL ServerIntegration Services 编写可连接到自定义数据源的源组件，并可以将这些自定义数据源中的数据提供给数据流任务中的其他组件。如果必须连接到无法使用现有 Integration Services 源之一进行访问的数据源，则创建自定义源是非常重要的。

源组件有一个或多个输出，没有输入。在设计时，源组件用于创建和配置连接、从外部数据源读取列元数据并基于外部数据源配置源的输出列。在执行过程中，源组件连接到外部数据源，并将行添加到输出缓冲区。然后，数据流任务会将此数据行缓冲区提供给下游组件。

有关源组件示例，请参阅 Codeplex 上的 Integration Services 示例。有关数据流组件开发的一般概述，请参阅开发自定义数据流组件。

设计时

实现源组件的设计时功能包括指定与外部数据源的连接、添加和配置反映数据源的输出列以及验证该组件是否可以执行。根据定义，源组件没有输入，有一个或多个异步输出。

创建组件

源组件使用包中定义的 ConnectionManager 对象连接到外部数据源。源组件将元素添加到 ComponentMetaData 属性的 RuntimeConnectionCollection 集合，以指明自己需要连接管理器。此集合有两个用途：一是保存对组件所用包中的连接管理器的引用；二是通知设计器此集合需要连接管理器。IDTSRuntimeConnection100 添加到该集合后，**“高级编辑器”将显示“连接属性”**选项卡，用户可以在其中选择或创建包中的连接。

下面的代码示例演示 ProvideComponentProperties 的实现，该实现添加一个输出，然后将 IDTSRuntimeConnection100 对象添加到 RuntimeConnectionCollection。

using System;
using System.Collections;
using System.Data;
using System.Data.SqlClient;
using System.Data.OleDb;
using Microsoft.SqlServer.Dts.Runtime;
using Microsoft.SqlServer.Dts.Runtime.Wrapper;
using Microsoft.SqlServer.Dts.Pipeline;
using Microsoft.SqlServer.Dts.Pipeline.Wrapper;

namespace Microsoft.Samples.SqlServer.Dts
{
    [DtsPipelineComponent(DisplayName = "MySourceComponent",ComponentType = ComponentType.SourceAdapter)]
    public class MyComponent : PipelineComponent
    {
        public override void ProvideComponentProperties()
        {
            // Reset the component.
            base.RemoveAllInputsOutputsAndCustomProperties();
            ComponentMetaData.RuntimeConnectionCollection.RemoveAll();

            IDTSOutput100 output = ComponentMetaData.OutputCollection.New();
            output.Name = "Output";

            IDTSRuntimeConnection100 connection = ComponentMetaData.RuntimeConnectionCollection.New();
            connection.Name = "ADO.NET";
        }

Imports System.Data
Imports System.Data.SqlClient
Imports Microsoft.SqlServer.Dts.Runtime
Imports Microsoft.SqlServer.Dts.Runtime.Wrapper
Imports Microsoft.SqlServer.Dts.Pipeline
Imports Microsoft.SqlServer.Dts.Pipeline.Wrapper

<DtsPipelineComponent(DisplayName:="MySourceComponent", ComponentType:=ComponentType.SourceAdapter)> _
Public Class MySourceComponent
    Inherits PipelineComponent

    Public Overrides Sub ProvideComponentProperties()

        ' Allow for resetting the component.
        RemoveAllInputsOutputsAndCustomProperties()
        ComponentMetaData.RuntimeConnectionCollection.RemoveAll()

        Dim output As IDTSOutput100 = ComponentMetaData.OutputCollection.New()
        output.Name = "Output"

        Dim connection As IDTSRuntimeConnection100 = ComponentMetaData.RuntimeConnectionCollection.New()
        connection.Name = "ADO.NET"

    End Sub
End Class

连接外部数据源

将连接添加到 RuntimeConnectionCollection 后，请重写 AcquireConnections 方法以建立与外部数据源的连接。此方法在设计和执行期间调用。组件应先建立与连接管理器（由运行时连接指定）的连接，再建立与外部数据源的连接。

连接建立后，组件应在内部缓存该连接，并在调用 ReleaseConnections 方法时释放该连接。ReleaseConnections 方法与 AcquireConnections 方法一样，在设计时和执行时调用。开发人员可重写此方法，释放在 AcquireConnections 过程中由组件建立的连接。

下面的代码示例演示在 AcquireConnections 方法中连接到 ADO.NET 连接，然后在 ReleaseConnections 方法中关闭该连接的组件。

private SqlConnection sqlConnection;

public override void AcquireConnections(object transaction)
{
    if (ComponentMetaData.RuntimeConnectionCollection[0].ConnectionManager != null)
    {
        ConnectionManager cm = Microsoft.SqlServer.Dts.Runtime.DtsConvert.GetWrapper(ComponentMetaData.RuntimeConnectionCollection[0].ConnectionManager);
        ConnectionManagerAdoNet cmado = cm.InnerObject as ConnectionManagerAdoNet;

        if (cmado == null)
            throw new Exception("The ConnectionManager " + cm.Name + " is not an ADO.NET connection.");

        sqlConnection = cmado.AcquireConnection(transaction) as SqlConnection;
        sqlConnection.Open();
    }
}

public override void ReleaseConnections()
{
    if (sqlConnection != null && sqlConnection.State != ConnectionState.Closed)
        sqlConnection.Close();
}

Private sqlConnection As SqlConnection

Public Overrides Sub AcquireConnections(ByVal transaction As Object)

    If Not IsNothing(ComponentMetaData.RuntimeConnectionCollection(0).ConnectionManager) Then

        Dim cm As ConnectionManager = Microsoft.SqlServer.Dts.Runtime.DtsConvert.GetWrapper(ComponentMetaData.RuntimeConnectionCollection(0).ConnectionManager)
        Dim cmado As ConnectionManagerAdoNet = CType(cm.InnerObject, ConnectionManagerAdoNet)

        If IsNothing(cmado) Then
            Throw New Exception("The ConnectionManager " + cm.Name + " is not an ADO.NET connection.")
        End If

        sqlConnection = CType(cmado.AcquireConnection(transaction), SqlConnection)
        sqlConnection.Open()

    End If
End Sub

Public Overrides Sub ReleaseConnections()

    If Not IsNothing(sqlConnection) And sqlConnection.State <> ConnectionState.Closed Then
        sqlConnection.Close()
    End If

End Sub

创建和配置输出列

源组件的输出列反映在执行该组件的过程中从外部数据源添加到数据流的列。在设计时，在组件配置为连接到外部数据源后添加输出列。组件用来将列添加到其输出集合的设计时方法会因所需组件而异，但不应在 Validate 或 AcquireConnections 执行期间添加列。例如，自定义属性中包含用于控制组件数据集的 SQL 语句的组件可在 SetComponentProperty 方法中添加其输出列。组件会检查自己是否具有缓存的连接，如果有，则连接到数据源并生成输出列。

创建输出列后，调用 SetDataTypeProperties 方法可设置该列的数据类型属性。此方法是必需的，因为 DataType、Length、Precision 和 CodePage 属性是只读的，并且每个属性都依赖于其他属性的设置。此方法强制使这些值的设置保持一致，并且数据流任务会验证是否正确设置了这些值。

列的 DataType 决定了为其他属性设置的值。下表说明了对每个 DataType 的依赖属性的要求。未列出的数据类型的依赖属性设置为零。

DataType	Length	Scale	Precision	CodePage
DT_DECIMAL	0	大于 0 且小于或等于 28。	0	0
DT_CY	0	0	0	0
DT_NUMERIC	0	大于 0 且小于或等于 28，并且小于 Precision。	大于或等于 1 且小于或等于 38。	0
DT_BYTES	大于 0。	0	0	0
DT_STR	大于 0 且小于 8000。	0	0	不为 0，并且是一个有效的代码页。
DT_WSTR	大于 0 且小于 4000。	0	0	0

由于对数据类型属性的限制是基于输出列的数据类型的，因此使用托管类型时必须选择正确的 SSIS 数据类型。基类提供了三个帮助器方法：ConvertBufferDataTypeToFitManaged、BufferTypeToDataRecordType 和 DataRecordTypeToBufferType，它们可协助托管组件开发人员在给定托管类型的情况下选择 SSIS 数据类型。这些方法用于将托管数据类型转换为 SSIS 数据类型，反之亦然。

下面的代码示例演示如何基于表架构来填充组件的输出列集合。基类的帮助器方法用于设置列的数据类型，并基于该数据类型设置依赖属性。

SqlCommand sqlCommand;

private void CreateColumnsFromDataTable()
{
    // Get the output.
    IDTSOutput100 output = ComponentMetaData.OutputCollection[0];

    // Start clean, and remove the columns from both collections.
    output.OutputColumnCollection.RemoveAll();
    output.ExternalMetadataColumnCollection.RemoveAll();

    this.sqlCommand = sqlConnection.CreateCommand();
    this.sqlCommand.CommandType = CommandType.Text;
    this.sqlCommand.CommandText = (string)ComponentMetaData.CustomPropertyCollection["SqlStatement"].Value;
    SqlDataReader schemaReader = this.sqlCommand.ExecuteReader(CommandBehavior.SchemaOnly);
    DataTable dataTable = schemaReader.GetSchemaTable();

    // Walk the columns in the schema, 
    // and for each data column create an output column and an external metadata column.
    foreach (DataRow row in dataTable.Rows)
    {
        IDTSOutputColumn100 outColumn = output.OutputColumnCollection.New();
        IDTSExternalMetadataColumn100 exColumn = output.ExternalMetadataColumnCollection.New();

        // Set column data type properties.
        bool isLong = false;
        DataType dt = DataRecordTypeToBufferType((Type)row["DataType"]);
        dt = ConvertBufferDataTypeToFitManaged(dt, ref isLong);
        int length = 0;
        int precision = (short)row["NumericPrecision"];
        int scale = (short)row["NumericScale"];
        int codepage = dataTable.Locale.TextInfo.ANSICodePage;

        switch (dt)
        {
            // The length cannot be zero, and the code page property must contain a valid code page.
            case DataType.DT_STR:
            case DataType.DT_TEXT:
                length = precision;
                precision = 0;
                scale = 0;
                break;

            case DataType.DT_WSTR:
                length = precision;
                codepage = 0;
                scale = 0;
                precision = 0;
                break;

            case DataType.DT_BYTES:
                precision = 0;
                scale = 0;
                codepage = 0;
                break;

            case DataType.DT_NUMERIC:
                length = 0;
                codepage = 0;

                if (precision > 38)
                    precision = 38;

                if (scale > 6)
                    scale = 6;
                break;

            case DataType.DT_DECIMAL:
                length = 0;
                precision = 0;
                codepage = 0;
                break;

            default:
                length = 0;
                precision = 0;
                codepage = 0;
                scale = 0;
                break;

        }

        // Set the properties of the output column.
        outColumn.Name = (string)row["ColumnName"];
        outColumn.SetDataTypeProperties(dt, length, precision, scale, codepage);
    }
}

Private sqlCommand As SqlCommand

Private Sub CreateColumnsFromDataTable()

    ' Get the output.
    Dim output As IDTSOutput100 = ComponentMetaData.OutputCollection(0)

    ' Start clean, and remove the columns from both collections.
    output.OutputColumnCollection.RemoveAll()
    output.ExternalMetadataColumnCollection.RemoveAll()

    Me.sqlCommand = sqlConnection.CreateCommand()
    Me.sqlCommand.CommandType = CommandType.Text
    Me.sqlCommand.CommandText = CStr(ComponentMetaData.CustomPropertyCollection("SqlStatement").Value)

    Dim schemaReader As SqlDataReader = Me.sqlCommand.ExecuteReader(CommandBehavior.SchemaOnly)
    Dim dataTable As DataTable = schemaReader.GetSchemaTable()

    ' Walk the columns in the schema, 
    ' and for each data column create an output column and an external metadata column.
    For Each row As DataRow In dataTable.Rows

        Dim outColumn As IDTSOutputColumn100 = output.OutputColumnCollection.New()
        Dim exColumn As IDTSExternalMetadataColumn100 = output.ExternalMetadataColumnCollection.New()

        ' Set column data type properties.
        Dim isLong As Boolean = False
        Dim dt As DataType = DataRecordTypeToBufferType(CType(row("DataType"), Type))
        dt = ConvertBufferDataTypeToFitManaged(dt, isLong)
        Dim length As Integer = 0
        Dim precision As Integer = CType(row("NumericPrecision"), Short)
        Dim scale As Integer = CType(row("NumericScale"), Short)
        Dim codepage As Integer = dataTable.Locale.TextInfo.ANSICodePage

        Select Case dt

            ' The length cannot be zero, and the code page property must contain a valid code page.
            Case DataType.DT_STR
            Case DataType.DT_TEXT
                length = precision
                precision = 0
                scale = 0

            Case DataType.DT_WSTR
                length = precision
                codepage = 0
                scale = 0
                precision = 0

            Case DataType.DT_BYTES
                precision = 0
                scale = 0
                codepage = 0

            Case DataType.DT_NUMERIC
                length = 0
                codepage = 0

                If precision > 38 Then
                    precision = 38
                End If

                If scale > 6 Then
                    scale = 6
                End If

            Case DataType.DT_DECIMAL
                length = 0
                precision = 0
                codepage = 0

            Case Else
                length = 0
                precision = 0
                codepage = 0
                scale = 0
        End Select

        ' Set the properties of the output column.
        outColumn.Name = CStr(row("ColumnName"))
        outColumn.SetDataTypeProperties(dt, length, precision, scale, codepage)
    Next
End Sub

验证组件

您应验证源组件，并验证源组件的输出列集合中定义的列是否与外部数据源中的列相匹配。有时，根据外部数据源验证输出列是不可能，例如处于断开连接状态时，或者希望避免与服务器之间费时的往返通信时。在这些情况下，仍然可以使用输出对象的 ExternalMetadataColumnCollection 来验证输出中的列。有关详细信息，请参阅验证数据流组件。

此集合同时存在于输入对象和输出对象中，可以使用来自外部数据源的列填充此集合。此集合可用于在 SSIS 设计器处于离线状态、组件断开连接或 ValidateExternalMetadata 属性为 false 时验证输出列。应在创建输出列的同时对该集合进行首次填充。将外部元数据列添加到集合相对比较容易，因为外部元数据列最初就应与输出列相匹配。列的数据类型属性应已正确设置，这些属性可直接复制到 IDTSExternalMetadataColumn100 对象。

下面的示例代码添加一个外部元数据列，该列基于新创建的输出列。假定已经创建了该输出列。

private void CreateExternalMetaDataColumn(IDTSOutput100 output, IDTSOutputColumn100 outputColumn)
{
    
    // Set the properties of the external metadata column.
    IDTSExternalMetadataColumn100 externalColumn = output.ExternalMetadataColumnCollection.New();
    externalColumn.Name = outputColumn.Name;
    externalColumn.Precision = outputColumn.Precision;
    externalColumn.Length = outputColumn.Length;
    externalColumn.DataType = outputColumn.DataType;
    externalColumn.Scale = outputColumn.Scale;

    // Map the external column to the output column.
    outputColumn.ExternalMetadataColumnID = externalColumn.ID;

}

Private Sub CreateExternalMetaDataColumn(ByVal output As IDTSOutput100, ByVal outputColumn As IDTSOutputColumn100)


        ' Set the properties of the external metadata column.
        Dim externalColumn As IDTSExternalMetadataColumn100 = output.ExternalMetadataColumnCollection.New()
        externalColumn.Name = outputColumn.Name
        externalColumn.Precision = outputColumn.Precision
        externalColumn.Length = outputColumn.Length
        externalColumn.DataType = outputColumn.DataType
        externalColumn.Scale = outputColumn.Scale

        ' Map the external column to the output column.
        outputColumn.ExternalMetadataColumnID = externalColumn.ID

    End Sub

运行时

执行期间，组件会将行添加到输出缓冲区中，输出缓冲区由数据流任务创建并在 PrimeOutput 中提供给组件。只为源组件调用一次该方法，该方法会为连接到下游组件的组件的每个 IDTSOutput100 接收一个输出缓冲区。

在缓冲区中查找列

组件的输出缓冲区中包含组件定义的列以及向下游组件的输出添加的所有列。例如，如果源组件在其输出中提供三列，下一个组件添加了第四个输出列，则供源组件使用的输出缓冲区将包含这四列。

缓冲区行中的列顺序不是由输出列集合中输出列的索引确定的。只有使用 BufferManager 的 FindColumnByLineageID 方法才能在缓冲区行中准确地定位输出列。此方法在指定缓冲区中查找具有指定沿袭 ID 的列，并返回其在行中的位置。输出列的索引通常在 PreExecute 方法中查找，并存储以供在 PrimeOutput 中使用。

下面的代码示例在调用 PreExecute 期间在输出缓冲区中查找输出列的位置，并将其存储在内部结构中。列的名称也存储在该结构中，在本主题下一部分的 PrimeOutput 方法的代码示例中将使用该名称。

ArrayList columnInformation;

private struct ColumnInfo
{
    public int BufferColumnIndex;
    public string ColumnName;
}

public override void PreExecute()
{
    this.columnInformation = new ArrayList();
    IDTSOutput100 output = ComponentMetaData.OutputCollection[0];

    foreach (IDTSOutputColumn100 col in output.OutputColumnCollection)
    {
        ColumnInfo ci = new ColumnInfo();
        ci.BufferColumnIndex = BufferManager.FindColumnByLineageID(output.Buffer, col.LineageID);
        ci.ColumnName = col.Name;
        columnInformation.Add(ci);
    }
}

Public Overrides Sub PreExecute()

    Me.columnInformation = New ArrayList()
    Dim output As IDTSOutput100 = ComponentMetaData.OutputCollection(0)

    For Each col As IDTSOutputColumn100 In output.OutputColumnCollection

        Dim ci As ColumnInfo = New ColumnInfo()
        ci.BufferColumnIndex = BufferManager.FindColumnByLineageID(output.Buffer, col.LineageID)
        ci.ColumnName = col.Name
        columnInformation.Add(ci)
    Next
End Sub

处理行

调用 AddRow 方法可将行添加到输出缓冲区，该方法会创建一个其各列为空值的新缓冲区行。然后，组件将为各列赋值。提供给组件的输出缓冲区由数据流任务创建并监视。输出缓冲区满后，缓冲区中的行会移到下一个组件中。无法确定一批行何时发送到下一个组件，因为数据流任务执行的行移动对组件开发人员是透明的，并且输出缓冲区的 RowCount 属性始终为零。源组件完成将行添加到其输出缓冲区后，它将调用 PipelineBuffer 的 SetEndOfRowset 方法通知数据流任务，缓冲区中的其余行会传递给下一个组件。

源组件从外部数据源读取行时，可以调用 IncrementPipelinePerfCounter 方法来更新“Rows read”或“BLOB bytes read”性能计数器。有关详细信息，请参阅监视数据流引擎的性能。

下面的代码示例演示在 PrimeOutput 中将行添加到输出缓冲区的组件。缓冲区中输出列的索引已在前面的代码示例中使用 PreExecute 找到。

public override void PrimeOutput(int outputs, int[] outputIDs, PipelineBuffer[] buffers)
{
    IDTSOutput100 output = ComponentMetaData.OutputCollection[0];
    PipelineBuffer buffer = buffers[0];

    SqlDataReader dataReader = sqlCommand.ExecuteReader();

    // Loop over the rows in the DataReader, 
    // and add them to the output buffer.
    while (dataReader.Read())
    {
        // Add a row to the output buffer.
        buffer.AddRow();

        for (int x = 0; x < columnInformation.Count; x++)
        {
            ColumnInfo ci = (ColumnInfo)columnInformation[x];
            int ordinal = dataReader.GetOrdinal(ci.ColumnName);

            if (dataReader.IsDBNull(ordinal))
                buffer.SetNull(ci.BufferColumnIndex);
            else
            {
                buffer[ci.BufferColumnIndex] = dataReader[ci.ColumnName];
            }
        }
    }
    buffer.SetEndOfRowset();
}

Public Overrides Sub PrimeOutput(ByVal outputs As Integer, ByVal outputIDs As Integer(), ByVal buffers As PipelineBuffer())

    Dim output As IDTSOutput100 = ComponentMetaData.OutputCollection(0)
    Dim buffer As PipelineBuffer = buffers(0)

    Dim dataReader As SqlDataReader = sqlCommand.ExecuteReader()

    ' Loop over the rows in the DataReader, 
    ' and add them to the output buffer.
    While (dataReader.Read())

        ' Add a row to the output buffer.
        buffer.AddRow()

        For x As Integer = 0 To columnInformation.Count

            Dim ci As ColumnInfo = CType(columnInformation(x), ColumnInfo)

            Dim ordinal As Integer = dataReader.GetOrdinal(ci.ColumnName)

            If (dataReader.IsDBNull(ordinal)) Then
                buffer.SetNull(ci.BufferColumnIndex)
            Else
                buffer(ci.BufferColumnIndex) = dataReader(ci.ColumnName)

            End If
        Next

    End While

    buffer.SetEndOfRowset()
End Sub

示例

下面的示例演示一个简单的源组件，该组件使用文件连接管理器将文件的二进制内容加载到数据流中。此示例未演示本主题中讨论的全部方法和功能。它演示了每个自定义源组件都必须重写的重要方法，但不包含用于设计时验证的代码。有关更完整的源组件示例，请参阅 Codeplex 上的 Integration Services 示例。

using System;
using System.IO;
using Microsoft.SqlServer.Dts.Pipeline;
using Microsoft.SqlServer.Dts.Pipeline.Wrapper;
using Microsoft.SqlServer.Dts.Runtime.Wrapper;

namespace BlobSrc
{
  [DtsPipelineComponent(DisplayName = "BLOB Inserter Source", Description = "Inserts files into the data flow as BLOBs")]
  public class BlobSrc : PipelineComponent
  {
    IDTSConnectionManager100 m_ConnMgr;
    int m_FileNameColumnIndex = -1;
    int m_FileBlobColumnIndex = -1;

    public override void ProvideComponentProperties()
    {
      IDTSOutput100 output = ComponentMetaData.OutputCollection.New();
      output.Name = "BLOB File Inserter Output";

      IDTSOutputColumn100 column = output.OutputColumnCollection.New();
      column.Name = "FileName";
      column.SetDataTypeProperties(DataType.DT_WSTR, 256, 0, 0, 0);

      column = output.OutputColumnCollection.New();
      column.Name = "FileBLOB";
      column.SetDataTypeProperties(DataType.DT_IMAGE, 0, 0, 0, 0);

      IDTSRuntimeConnection100 conn = ComponentMetaData.RuntimeConnectionCollection.New();
      conn.Name = "FileConnection";
    }

    public override void AcquireConnections(object transaction)
    {
      IDTSRuntimeConnection100 conn = ComponentMetaData.RuntimeConnectionCollection[0];
      m_ConnMgr = conn.ConnectionManager;
    }

    public override void ReleaseConnections()
    {
      m_ConnMgr = null;
    }

    public override void PreExecute()
    {
      IDTSOutput100 output = ComponentMetaData.OutputCollection[0];

      m_FileNameColumnIndex = (int)BufferManager.FindColumnByLineageID(output.Buffer, output.OutputColumnCollection[0].LineageID);
      m_FileBlobColumnIndex = (int)BufferManager.FindColumnByLineageID(output.Buffer, output.OutputColumnCollection[1].LineageID);
    }

    public override void PrimeOutput(int outputs, int[] outputIDs, PipelineBuffer[] buffers)
    {
      string strFileName = (string)m_ConnMgr.AcquireConnection(null);

      while (strFileName != null)
      {
        buffers[0].AddRow();

        buffers[0].SetString(m_FileNameColumnIndex, strFileName);

        FileInfo fileInfo = new FileInfo(strFileName);
        byte[] fileData = new byte[fileInfo.Length];
        FileStream fs = new FileStream(strFileName, FileMode.Open, FileAccess.Read, FileShare.Read);
        fs.Read(fileData, 0, fileData.Length);

        buffers[0].AddBlobData(m_FileBlobColumnIndex, fileData);

        strFileName = (string)m_ConnMgr.AcquireConnection(null);
      }

      buffers[0].SetEndOfRowset();
    }
  }
}

Imports System 
Imports System.IO 
Imports Microsoft.SqlServer.Dts.Pipeline 
Imports Microsoft.SqlServer.Dts.Pipeline.Wrapper 
Imports Microsoft.SqlServer.Dts.Runtime.Wrapper 
Namespace BlobSrc 

 <DtsPipelineComponent(DisplayName="BLOB Inserter Source", Description="Inserts files into the data flow as BLOBs")> _ 
 Public Class BlobSrc 
 Inherits PipelineComponent 
   Private m_ConnMgr As IDTSConnectionManager100 
   Private m_FileNameColumnIndex As Integer = -1 
   Private m_FileBlobColumnIndex As Integer = -1 

   Public  Overrides Sub ProvideComponentProperties() 
     Dim output As IDTSOutput100 = ComponentMetaData.OutputCollection.New 
     output.Name = "BLOB File Inserter Output" 
     Dim column As IDTSOutputColumn100 = output.OutputColumnCollection.New 
     column.Name = "FileName" 
     column.SetDataTypeProperties(DataType.DT_WSTR, 256, 0, 0, 0) 
     column = output.OutputColumnCollection.New 
     column.Name = "FileBLOB" 
     column.SetDataTypeProperties(DataType.DT_IMAGE, 0, 0, 0, 0) 
     Dim conn As IDTSRuntimeConnection90 = ComponentMetaData.RuntimeConnectionCollection.New 
     conn.Name = "FileConnection" 
   End Sub 

   Public  Overrides Sub AcquireConnections(ByVal transaction As Object) 
     Dim conn As IDTSRuntimeConnection100 = ComponentMetaData.RuntimeConnectionCollection(0) 
     m_ConnMgr = conn.ConnectionManager 
   End Sub 

   Public  Overrides Sub ReleaseConnections() 
     m_ConnMgr = Nothing 
   End Sub 

   Public  Overrides Sub PreExecute() 
     Dim output As IDTSOutput100 = ComponentMetaData.OutputCollection(0) 
     m_FileNameColumnIndex = CType(BufferManager.FindColumnByLineageID(output.Buffer, output.OutputColumnCollection(0).LineageID), Integer) 
     m_FileBlobColumnIndex = CType(BufferManager.FindColumnByLineageID(output.Buffer, output.OutputColumnCollection(1).LineageID), Integer) 
   End Sub 

   Public  Overrides Sub PrimeOutput(ByVal outputs As Integer, ByVal outputIDs As Integer(), ByVal buffers As PipelineBuffer()) 
     Dim strFileName As String = CType(m_ConnMgr.AcquireConnection(Nothing), String) 
     While Not (strFileName Is Nothing) 
       buffers(0).AddRow 
       buffers(0).SetString(m_FileNameColumnIndex, strFileName) 
       Dim fileInfo As FileInfo = New FileInfo(strFileName) 
       Dim fileData(fileInfo.Length) As Byte 
       Dim fs As FileStream = New FileStream(strFileName, FileMode.Open, FileAccess.Read, FileShare.Read) 
       fs.Read(fileData, 0, fileData.Length) 
       buffers(0).AddBlobData(m_FileBlobColumnIndex, fileData) 
       strFileName = CType(m_ConnMgr.AcquireConnection(Nothing), String) 
     End While 
     buffers(0).SetEndOfRowset 
   End Sub 
 End Class 
End Namespace

使 Integration Services 保持最新 若要从 Microsoft 获得最新的下载内容、文章、示例和视频，以及从社区获得所选解决方案，请访问 MSDN 或 TechNet 上的 Integration Services 页： 访问 MSDN 上的 Integration Services 页 访问 TechNet 上的 Integration Services 页 若要获得有关这些更新的自动通知，请订阅该页上提供的 RSS 源。

请参阅

概念

开发自定义目标组件

使用脚本组件创建源