Question

Possible Duplicate:
How to get the range of occupied cells in excel sheet

Im trying to find the last used row in an Excel worksheet. For doing this, I’m using this code:

int lastUsedRow = currentWS.Cells.SpecialCells(Excel.XlCellType.xlCellTypeLastCell,Type.Missing).Row;

Most of the time it works fine, but sometimes Excel thinks that there’s more rows in the sheet than theres suppose to be.

Fx: If I copy data from sheet1, containing 12 rows, to an empty sheet2, then deleting all data in sheet2, by right clicking and press “Delete…” and copy data from sheet3, containing 5 rows, to sheet2, then lastUsedRow will give me the value of 12 which should have been 5.

The image example above is suppose to give my the value of 22 as row count, but instead i’ll get 634. Notice the scroll bar to the right.

It seems like Excel thinks that some cells are filled even though I just deleted all cells, before copying new data with fewer rows into the sheet.

Are there any way to “resize” the view of the data in the sheet, so that i’ll get the right number of used cells or maybe another way to find the number of the last used row?

Thanks.

Answer 1

Here is the code I use:

public static string GetMinimalUsedRangeAddress(Excel.Worksheet sheet)
{
    string address = String.Empty;
    try
    {
        int rowMax = 0;
        int colMax = 0;

        Excel.Range usedRange = sheet.UsedRange;
        Excel.Range lastCell = usedRange.SpecialCells(Excel.XlCellType.xlCellTypeLastCell, Type.Missing);
        int lastRow = lastCell.Row;
        int lastCol = lastCell.Column;
        int rowMin = lastRow + 1;
        int colMin = lastCol + 1;

        int rr = usedRange.Rows.Count;
        int cc = usedRange.Columns.Count;
        for (int r = 1; r <= rr; r++)
        {
            for (int c = 1; c <= cc; c++)
            {
                Excel.Range cell = usedRange.Cells[r, c] as Excel.Range;
                if (cell != null && cell.Value != null && !String.IsNullOrEmpty(cell.Value.ToString()))
                {
                    if (cell.Row > rowMax)
                        rowMax = cell.Row;
                    if (cell.Column > colMax)
                        colMax = cell.Column;
                    if (cell.Row < rowMin)
                        rowMin = cell.Row;
                    if (cell.Column < colMin)
                        colMin = cell.Column;
                }
                MRCO(cell);
            }
        }

        if (!(rowMax == 0 || colMax == 0 || rowMin == lastRow + 1 || colMin == lastCol + 1))
            address = Cells2Address(rowMin, colMin, rowMax, colMax);

        MRCO(lastCell);
        MRCO(usedRange);
    }
    catch (Exception ex)
    {
        // log as needed
    }
    return address; // caller should test return for String.Empty
}


public static string Cells2Address(int row1, int col1, int row2, int col2)
{
    return ColNum2Letter(col1) + row1.ToString() + ":" + ColNum2Letter(col2) + row2.ToString();
}


public static string ColNum2Letter(int colNum)
{
    if (colNum <= 26) 
        return ((char)(colNum + 64)).ToString();

    colNum--; //decrement to put value on zero based index
    return ColNum2Letter(colNum / 26) + ColNum2Letter((colNum % 26) + 1);
}


public static void MRCO(object obj)
{
    if (obj == null) { return; }
    try
    {
        System.Runtime.InteropServices.Marshal.ReleaseComObject(obj);
    }
    catch
    {
        // ignore, cf: http://support.microsoft.com/default.aspx/kb/317109
    }
    finally
    {
        obj = null;
    }
}

Note: you might be tempted to replace all the individual cell value checks with CountA but that will fail in certain cases. For example, if a cell contains the formula =IF(A1=55,"Y",""), a resulting empty string will count as a non-blank cell using CountA.