Why does Dimension.End.Column return an index larger than the last column with any values?

Question

Reading spreadsheets with EPPlus, I was surprised to see Dimension.End.Column returning a higher (1-based) index than the column I know to be the last.

For example, in a spreadsheet like this:

   A       B       C       D...
1  value

I would expect Dimension.End.Column to be 1, corresponding to the first column.

However, it's often larger. Why is that?

Answer 1

To answer this question it is needed to drill deep down into the source code of EpPlus.

Anyway, here is an extension method that I use to get an actual (valued) dimension and it works just fine even for edge cases:

public static ExcelAddressBase GetValuedDimension(this ExcelWorksheet worksheet)
{
    var dimension = worksheet.Dimension;
    if (dimension == null) return null;
    var cells = worksheet.Cells[dimension.Address];
    Int32 minRow = 0, minCol = 0, maxRow = 0, maxCol = 0;
    var hasValue = false;
    foreach (var cell in cells.Where(cell => cell.Value != null))
    {
        if (!hasValue)
        {
            minRow = cell.Start.Row;
            minCol = cell.Start.Column;
            maxRow = cell.End.Row;
            maxCol = cell.End.Column;
            hasValue = true;
        }
        else
        {
            if (cell.Start.Column < minCol)
            {
                minCol = cell.Start.Column;
            }
            if (cell.End.Row > maxRow)
            {
                maxRow = cell.End.Row;
            }
            if (cell.End.Column > maxCol)
            {
                maxCol = cell.End.Column;
            }
        }
    }
    return hasValue ? new ExcelAddressBase(minRow, minCol, maxRow, maxCol) : null;
}