Implementing XmlTextWriter in new XmlRecordsetWriter for Streams

问题

For background see my question here.

So the problem now isn't that I can't send a DataSet to classic ASP but that it can't do anything with it. So I found some code to create a recordset xml structure from a DataSet.

I have tweaked it a little from it's original source. The problem is that I can't seem to extract the base stream and use it instead of having to write to a file. What am I missing?

Here is how I am trying to test the class:

[Test]
    public void TestWriteToStream()
    {            
        MemoryStream theStream = new MemoryStream();
        XmlRecordsetWriter theWriter = new XmlRecordsetWriter(theStream);
        theWriter.WriteRecordset( SomeFunctionThatReturnsADataSet() );
        theStream = (MemoryStream)theWriter.BaseStream;
        string xmlizedString = UTF8ByteArrayToString(theStream.ToArray());
        xmlizedString = xmlizedString.Substring(1);

        //Assert.AreEqual(m_XMLNotNull, xmlizedString);
        Console.WriteLine(xmlizedString);
    }

Here is my class:

using System;
using System.Collections.Generic;
using System.Data;
using System.IO;
using System.Text;
using System.Xml;

namespace Core{

public class XmlRecordsetWriter : XmlTextWriter
{
    #region Constructors
    // Constructor(s)
    public XmlRecordsetWriter(string filename) : base(filename, null) { SetupWriter(); }

    public XmlRecordsetWriter(Stream s) : base(s, null) { SetupWriter(); }

    public XmlRecordsetWriter(TextWriter tw) : base(tw) { SetupWriter(); }

    protected void SetupWriter()
    {
        base.Formatting = Formatting.Indented;
        base.Indentation = 3;
    }
    #endregion

    #region Methods

    // WriteRecordset
    public void WriteRecordset(DataSet ds) { WriteRecordset(ds.Tables[0]); }

    public void WriteRecordset(DataSet ds, string tableName) { WriteRecordset(ds.Tables[tableName]); }

    public void WriteRecordset(DataView dv) { WriteRecordset(dv.Table); }

    public void WriteRecordset(DataTable dt)
    {
        WriteStartDocument();
        WriteSchema(dt);
        WriteContent(dt);
        WriteEndDocument();
    }

    // WriteStartDocument
    public void WriteStartDocument()
    {
        base.WriteStartDocument();
        base.WriteComment("Created by XmlRecordsetWriter");

        base.WriteStartElement("xml");
        base.WriteAttributeString("xmlns", "s", null, "uuid:BDC6E3F0-6DA3-11d1-A2A3-00AA00C14882");
        base.WriteAttributeString("xmlns", "dt", null, "uuid:C2F41010-65B3-11d1-A29F-00AA00C14882");
        base.WriteAttributeString("xmlns", "rs", null, "urn:schemas-microsoft-com:rowset");
        base.WriteAttributeString("xmlns", "z", null, "#RowsetSchema");
    }

    // WriteSchema
    public void WriteSchema(DataSet ds) { WriteSchema(ds.Tables[0]); }

    public void WriteSchema(DataSet ds, string tableName) { WriteSchema(ds.Tables[tableName]); }

    public void WriteSchema(DataView dv){ WriteSchema(dv.Table); }

    public void WriteSchema(DataTable dt)
    {
        // Open the schema tag (XDR)
        base.WriteStartElement("s", "Schema", null);
        base.WriteAttributeString("id", "RowsetSchema");
        base.WriteStartElement("s", "ElementType", null);
        base.WriteAttributeString("name", "row");
        base.WriteAttributeString("content", "eltOnly");

        // Write the column info 
        int index=0;
        foreach(DataColumn dc in dt.Columns)
        {
            index ++;
            base.WriteStartElement("s", "AttributeType", null);
            base.WriteAttributeString("name", dc.ColumnName);
            base.WriteAttributeString("rs", "number", null, index.ToString());
            base.WriteEndElement();
        }

        // Close the schema tag 
        base.WriteStartElement("s", "extends", null); 
        base.WriteAttributeString("type", "rs:rowbase");
        base.WriteEndElement();
        base.WriteEndElement();
        base.WriteEndElement();
    }

    // WriteContent
    public void WriteContent(DataSet ds) { WriteContent(ds.Tables[0]); }

    public void WriteContent(DataSet ds, string tableName) { WriteContent(ds.Tables[tableName]); }

    public void WriteContent(DataView dv) { WriteContent(dv.Table); }

    public void WriteContent(DataTable dt)
    {
        // Write data
        base.WriteStartElement("rs", "data", null);
        foreach(DataRow row in dt.Rows)
        {
            base.WriteStartElement("z", "row", null);
            foreach(DataColumn dc in dt.Columns)
                base.WriteAttributeString(dc.ColumnName, row[dc.ColumnName].ToString());
            base.WriteEndElement();
        }
        base.WriteEndElement();
    }

    // WriteEndDocument
    public void WriteEndDocument()
    {
        base.WriteEndDocument();
        base.Flush();
        base.Close();
    }
    #endregion
}

}

回答1:

I think you want to work with Data-based objects and XML based data. If it's true, I suggest using ADODB Classes (It is in COM References: Microsoft ActiveX Data Objects 6.0 Library -Or in other versions like 2.8-).

You can convert your DataTable to a ADODB.Recordset by this code: Simplest code to convert an ADO.NET DataTable to an ADODB.Recordset. So you have a ConvertToRecordset() method to use in next code.

Now you need only save() method to have your XML file:

using ADODB;
using System;
using System.Data;
using System.IO;

namespace ConsoleApplicationTests
{
    class Program
    {
        static void Main(string[] args)
        {
            Recordset rs = new Recordset();
            DataTable dt = sampleDataTable();   //-i. -> SomeFunctionThatReturnsADataTable() 

            //-i. Convert DataTable to Recordset 
            rs = ConvertToRecordset(dt);

            //-i. Sample Output File
            String filename = @"C:\yourXMLfile.xml";
            FileStream fstream = new FileStream(filename, FileMode.Create);

            rs.Save(fstream, PersistFormatEnum.adPersistXML);
        }
    }
}

The power of ADODB.Recordset is here that you can open that saved XML file very easy:

rs.Open(fstream);

I hope it works!, Actually I wrote this answer to complete it later and if I'm in a right direction.

回答2:

First, the line:

theStream = (MemoryStream)theWriter.BaseStream;

is redundant as the MemoryStream should already be the base stream for the writer.

What it sounds like you want is just:

theWriter.Close();
theStream.Position = 0; // So you can start reading from the begining

string xml = null;
using (StringReader read = new StringReader(theStream))
{
   xml = read.ReadToEnd();
}

Then xml will be your xml string that you can load into an XPathDocument or XmlDocument and play with however you want.

来源：https://stackoverflow.com/questions/301045/implementing-xmltextwriter-in-new-xmlrecordsetwriter-for-streams

标签

xml

dataset

recordset