Steven Swafford: Overworked and Underpaid : XML

How to use a MemoryStream to Perform Downloads and Bypass the File System and the Problems Between Browsers and Servers

sswafford — Thu, 21 Sep 2006 12:29:00 GMT

I was recently was tasked to develop a solution to download data from an Oracle database and allow the client to save the data as an XML document without the use of the file system. I turned to the System.IO Namespace to accomplish this specifically using the MemoryStream Class. First I will demonstrate how I accomplished this task and keep in mind that I am using a DataAccess Layer (DAL) for communication to the database.

DAL ExtractData Class

/// <summary>
/// Exports the admin extract.
/// </summary>
/// <param name="tableName">Name of the table to extract.</param>
/// <param name="dataSetName">Name of the dataset to create.</param>
/// <returns></returns>
public MemoryStream exportAdminExtract(string tableName, string dataSetName,
   string storedProcedureName)
{
    Oracle.DataAccess.Client.OracleDataAdapter da = new OracleDataAdapter();
    MemoryStream returnStream = new MemoryStream();
    DataSet extractDataSet = new DataSet(dataSetName);

using (Oracle.DataAccess.Client.OracleConnection dbConn = new OracleConnection(_oraMainConnString))
{
Oracle.DataAccess.Client.OracleCommand cmd = new Oracle.DataAccess.Client.OracleCommand();
cmd.Connection = dbConn;
cmd.CommandText = storedProcedureName;
cmd.CommandType = CommandType.StoredProcedure;

Oracle.DataAccess.Client.OracleParameter paramExtractCursor
= new OracleParameter("extractsCursor", OracleDbType.RefCursor);
paramExtractCursor.Direction = ParameterDirection.Output;
cmd.Parameters.Add(paramExtractCursor);

try
{
     dbConn.Open();
     da.SelectCommand = cmd;
     da.Fill(extractDataSet, tableName.ToUpper());
     extractDataSet.WriteXml(returnStream, XmlWriteMode.WriteSchema);
}
catch (Exception ex)
{
     bool rethrow = ExceptionPolicy.HandleException(ex, "Global Exception Policy");
     if (rethrow)
     {
  throw;
     }
}
finally
{
     da.Dispose();
     returnStream.Close();
     returnStream.Dispose();
     cmd.Dispose();
     dbConn.Close();
     dbConn.Dispose();
}
    }
    return returnStream;
}

The class above is fairly straight forward. As you can see exportAdminExtract() takes three parameters which are tableName, dataSetName, and storedProcedureName, connects to an Oracle database executes the appropriate commands and returns the MemoryStream object.

Now to make use of this class let us assume you will be using an image to fire the event to allow the client to perform the download. In your web form you would have something similiar to the following:

<asp:ImageButton ID="btnExport" runat="server" ToolTip="Export" CommandName="btnExport_Click" ImageUrl="extractdata.gif"></asp:ImageButton>

Next in the code behind you must define the event, if you look at the ImageButton I defined above you will see the event name is "btnExport_Click".

/// <summary>
/// Handles the Click event of the btnExport control.
/// </summary>
/// <param name="sender">The source of the event.</param>
/// <param name="e">The <see cref="T:System.Web.UI.ImageClickEventArgs"/> instance containing the event data.</param>
private void btnExport_Click(object sender, System.Web.UI.ImageClickEventArgs e)
{
    string _tableName = "Offices";
    string _dataSetName = "MyAppOffices";
    string _storedProcedureName = "MYAPP_ADMIN.GETOFFICES";

ExportData extracts = new ExportData();
System.IO.MemoryStream mstream = extracts.exportAdminExtract(_tableName, _dataSetName,
_storedProcedureName);

//Convert the memorystream to an array of bytes.
byte[] byteArray = mstream.ToArray();

    //Clean up the memory stream
    mstream.Flush();
    mstream.Close();

// Clear all content output from the buffer stream
Response.Clear();

// Add a HTTP header to the output stream that specifies the default filename for the browser's download dialog
Response.AddHeader("Content-Disposition", "attachment; filename=" + _tableName + ".xml");

// Add a HTTP header to the output stream that contains the content length(File Size).
Response.AddHeader("Content-Length", byteArray.Length.ToString());

// Set the HTTP MIME type of the output stream
Response.ContentType = "text/xml";

    // Write the data out to the client.
    Response.BinaryWrite(byteArray);
    Response.End();
}

That is all there is to it! Now when the client clicks the image to perform the download they will be presented a file save-as dialog which in turn allows them to save the data extract where they desire.

While this is not an earth shattering example there were some issues that I ran into that I felt I should share to prevent others from dealing with the same issues. While some of the problems I encountered were resolved, there we others that I am still seeking an answer for.

First and most important is you may not use the following with my example (at least in some cases):

   1:  Response.Cache.SetCacheability(HttpCacheability.NoCache);

I must thank Walter Wang for pointing out the following Article titled "Internet Explorer Cannot Download: Error Message When You Use an HTTPS URL to Open an Office Document or PDF File". Also take note of the workaround as your code modifcation may or may not be enough to correct this problem. In my case it was and my downloads work perfectly. Also a thanks goes out to WenJun Zhang for the suggestion of using Wfetch to see exactly what was occuring, see the article titled "HOW TO: Use Wfetch.exe to Troubleshoot HTTP Connections".

I mentioned earlier that there were some problems that I am still seeking answers for. These are:

Downloads worked perfectly with FireFox 1.5 even with the Response.Cache.SetCacheability set on IIS5 and IIS6 both under SSL and no SSL.
Downloads worked perfectly with IE6 even with the Response.Cache.SetCacheability set on IIS5 on XP PRO SP2 with no SSL.
Downloads worked perfectly with IE6 even with the Response.Cache.SetCacheability set on IIS6 running as IIS5 Isolation both under SSL and no SSL.

How to edit an XML file that contains a namespace

sswafford — Mon, 27 Mar 2006 03:16:00 GMT

I must admit using xpath to query and find a specific node in an XML file is not my strongest point and when a namespace was part of the XML file it threw me off even more. A huge thank you goes out to Adam Sills for pointing out my errors.

Take the following XML file that was generated using the June 2005 Patterns and Practices library:

<?xml version="1.0" encoding="utf-8"?>
<dataConfiguration>
<xmlSerializerSection type="Microsoft.Practices.EnterpriseLibrary.Data.Configuration.DatabaseSettings, Microsoft.Practices.EnterpriseLibrary.Data , Version=1.1.0.0, Culture=neutral, PublicKeyToken=null">
    <enterpriseLibrary.databaseSettings xmlns:xsd="http://www.w3.org/2001/XMLSchema " xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" defaultInstance="SQL Server Instance" xmlns=" http://www.microsoft.com/practices/enterpriselibrary/08-31-2004/data">
      <databaseTypes>
        <databaseType name="Oracle" type="Microsoft.Practices.EnterpriseLibrary.Data.Oracle.OracleDatabase , Microsoft.Practices.EnterpriseLibrary.Data, Version=1.1.0.0, Culture=neutral, PublicKeyToken=null" />
        <databaseType name="SQL Server" type="Microsoft.Practices.EnterpriseLibrary.Data.Sql.SqlDatabase , Microsoft.Practices.EnterpriseLibrary.Data, Version=1.1.0.0, Culture=neutral, PublicKeyToken=null" />
      </databaseTypes>
      <instances>
        <instance name="Oracle Instance" type="Oracle" connectionString="Oracle Connection String" />
        <instance name="SQL Server Instance" type="SQL Server" connectionString="SQL Server Connection String" />
      </instances>
      <connectionStrings>
        <connectionString xsi:type="OracleConnectionStringData" name="Oracle Connection String">
          <parameters>
            <parameter name="Data Source" value="null" isSensitive="false" />
            <parameter name="Password" value="null" isSensitive="false" />
            <parameter name="User Id" value="null" isSensitive="false" />
            <parameter name="Host" value="null" isSensitive="false" />
          </parameters>
          <packages />
        </connectionString>
        <connectionString name="SQL Server Connection String">
          <parameters>
            <parameter name="database" value="null" isSensitive="false" />
            <parameter name="Integrated Security" value="True" isSensitive="false" />
            <parameter name="server" value="(local)" isSensitive="false" />
          </parameters>
        </connectionString>
      </connectionStrings>
    </enterpriseLibrary.databaseSettings>
</xmlSerializerSection>
</dataConfiguration>

Now what if I wanted to edit the Data Source value of the connectionString node that has a value of "Oracle Connection String"? To be brief let us assume I have a command button on a web form. The following code is placed in the Click event procedure:

// Instantiate a XmlDocument Object
System.Xml.XmlDocument xmlDoc = new System.Xml.XmlDocument();
xmlDoc.Load(file); //file that is passed in

// Instantiate a XmlNameSpaceManager Object
XmlNamespaceManager xmlNsMgr = new XmlNamespaceManager(xmlDoc.NameTable);

// Add the namespace
xmlNsMgr.AddNamespace("x", "http://www.microsoft.com/practices/enterpriselibrary/08-31-2004/data ");

// Data Source XPath String
string _xpathDS = "//x:connectionStrings/x:connectionString[@name='Oracle Connection String']/x:parameters/x:parameter[@name='Data Source']";

// Instantiate a XmlNode Object
System.Xml.XmlNode node1 = xmlDoc.SelectSingleNode(_xpathDS, xmlNsMgr);
if (node1 != null)
{
node1.Attributes["value"].Value= "oracledb";
}

xmlDoc.Save(file);

That is all there is to it. As long as node1 is not null then the value of the Data Source is set to oracledb.

Free XML e-Book

sswafford — Sat, 01 Jan 2005 20:39:00 GMT

I just ran accross a free e-book titled Essential XML Quick Reference: A Programmer's Reference to XML, XPath, XSLT, XML Schema, SOAP, and More at TheServerSide.NET.

Be sure to grab your download as the price is right!