How to use clipboard to copy data from Excel Sheet to DataTable?

Question

I have a Winform project, created on Microsoft Framework 3.5. The users may have installed Windows 7 or Windows XP, and Office 2007 or above.

I'm working on in a procedure to get the clipboard data and put in on a C# DataTable. I already created a method to get the raw data from the clipboard and upload it in a DataTable.

But in some cases, the Excel data shows a value, but internally have another:

I'm investigating a method to get the raw data from Excel:

string XmlFmt = "XML Spreadsheet";
var clipboard = Clipboard.GetDataObject();

if (clipboard.GetDataPresent(XmlFmt))
{
    var clipData = clipboard.GetData(XmlFmt);
    StreamReader streamReader = new StreamReader((MemoryStream)clipData);
    streamReader.BaseStream.SetLength(streamReader.BaseStream.Length - 1);

    string xmlText = streamReader.ReadToEnd();
    var stream = new StringReader(xmlText);

    XmlDocument xmlDocument = new XmlDocument();
    xmlDocument.LoadXml(xmlText);

    DataSet dsExcelData = new DataSet();
    dsExcelData.ReadXml(new XmlNodeReader(xmlDocument));
}

But, this method retrieves me a DataSet with multiples tables with the configuration of each part of the Excel Data:

Basically, I want to convert these structures to a simple DataTable with only the raw data. Someone could help me with a hint how achieve this? ...I don't want to use a third party library in this implementation.

Answer 1

I found a clean and bullet-proof solution. Here the code:

First, a extension to convert a XmlDocument to XElement:

/// <summary> Convert XML Document to XDocument </summary>
/// <param name="xmlDocument">Attached XML Document</param>
public static XDocument fwToXDocument(this XmlDocument xmlDocument)
{
    using (XmlNodeReader xmlNodeReader = new XmlNodeReader(xmlDocument))
    {
        xmlNodeReader.MoveToContent();
        return XDocument.Load(xmlNodeReader);
    }
}

The complete function:

private DataTable clipboardExcelToDataTable(bool blnFirstRowHasHeader = false)
{
    string strTime = "S " + DateTime.Now.ToString("mm:ss:fff");
    var clipboard = Clipboard.GetDataObject();
    if (!clipboard.GetDataPresent("XML Spreadsheet")) return null;

    strTime += "\r\nRead " + DateTime.Now.ToString("mm:ss:fff");
    StreamReader streamReader = new StreamReader((MemoryStream)clipboard.GetData("XML Spreadsheet"));
    strTime += "\r\nFinish read " + DateTime.Now.ToString("mm:ss:fff");
    streamReader.BaseStream.SetLength(streamReader.BaseStream.Length - 1);

    XmlDocument xmlDocument = new XmlDocument();
    xmlDocument.LoadXml(streamReader.ReadToEnd());
    strTime += "\r\nRead XML Document " + DateTime.Now.ToString("mm:ss:fff");

    XNamespace ssNs = "urn:schemas-microsoft-com:office:spreadsheet";
    DataTable dtData = new DataTable();

    var linqRows = xmlDocument.fwToXDocument().Descendants(ssNs + "Row").ToList<XElement>();

    for (int x = 0; x < linqRows.Max(a => a.Descendants(ssNs + "Cell").Count()); x++)
        dtData.Columns.Add("Column " + (x + 1).ToString());

    int intCol = 0;

    DataRow drCurrent;

    linqRows.ForEach(rowElement =>
        {
            intCol = 0;
            drCurrent = dtData.Rows.Add();
            rowElement.Descendants(ssNs + "Cell")
                .ToList<XElement>()
                .ForEach(cell => drCurrent[intCol++] = cell.Value);
        });

    if (blnFirstRowHasHeader)
    {
        int x = 0;
        foreach (DataColumn dcCurrent in dtData.Columns)
            dcCurrent.ColumnName = dtData.Rows[0][x++].ToString();

        dtData.Rows.RemoveAt(0);
    }

    strTime += "\r\nF " + DateTime.Now.ToString("mm:ss:fff");

    return dtData;
}

The process takes ~15 seconds to read ~25,000 rows.

Works perfectly for any kind of data. Basically, the method creates a grid with the same structure of the Excel WorkSheet. Merge of rows or columns will fill up the first cell able. All columns will be string DataType by default.

Answer 2

If they are flat data you could do it like this.

private class Field
{
        public string Valor { get; set; }
}

private class Row
{
        public List<Field> Fields { get; set; }

        public Row(string value)
        {
            Fields = new List<Field>();
            var fieldsString = value.Split(new char[] {'\t'});
            foreach (string f in fieldsString)
            {
                Fields.Add(new Field {Valor = f});
            }
    }
}

public Parse()
{
    var data = Clipboard.GetDataObject();
    var datos = (string)data.GetData(DataFormats.Text);
    var stringRows = datos.Split(new Char[] {'\r', '\n'}, StringSplitOptions.RemoveEmptyEntries);
    var table = new List<Row>(stringRows.Length) ;
    foreach (string stringRow in stringRows)
    {
        table.Add( new Row(stringRow) );
    }
}

Answer 3

in some cases, the Excel data shows a value, but internally have another.

Using the XML method is the reason for that internal structure of multiple datatables. Try this method instead:

private void PasteFromExcel()
{
    DataTable tbl = new DataTable();
    tbl.TableName = "ImportedTable";
    List<string> data = new List<string>(ClipboardData.Split('\n'));
    bool firstRow = true;

    if (data.Count > 0 && string.IsNullOrWhiteSpace(data[data.Count - 1]))
    {
        data.RemoveAt(data.Count - 1);
    }

    foreach (string iterationRow in data)
    {
        string row = iterationRow;
        if (row.EndsWith("\r"))
        {
            row = row.Substring(0, row.Length - "\r".Length);
        }

        string[] rowData = row.Split(new char[] { '\r', '\x09' });
        DataRow newRow = tbl.NewRow();
        if (firstRow)
        {
            int colNumber = 0;
            foreach (string value in rowData)
            {
                if (string.IsNullOrWhiteSpace(value))
                {
                    tbl.Columns.Add(string.Format("[BLANK{0}]", colNumber));
                }
                else if (!tbl.Columns.Contains(value))
                {
                    tbl.Columns.Add(value);
                }
                else
                {
                    tbl.Columns.Add(string.Format("Column {0}", colNumber));
                }
                colNumber++;
            }
            firstRow = false;
        }
        else
        {
            for (int i = 0; i < rowData.Length; i++)
            {
                if (i >= tbl.Columns.Count) break;
                newRow[i] = rowData[i];
            }
            tbl.Rows.Add(newRow);
        }
    }

    DataGridView1.DataSource = tbl;
}

Ref: http://www.seesharpdot.net/?p=221

Edit:

I've done some tests and even using the "XML Spreadsheet" Clipboard format the data can get stored in exponential notation:

You could detect and convert these numbers: Parse a Number from Exponential Notation