Excel Work Book - Read from C# substantially slow ?

Question

was experimenting with reading from an excel workbook and noticed it takes a long time to read a sheet with 3560 rows and 7 columns, about 1 minute and 17 seconds. All I did was loop through the whole sheet and store the values in a list.

Is this normal, or am I doing something wrong ?

 static void Main(string[] args)
    {
        List<string> testList = new List<string>();
        Excel.Application excelApp = new Excel.Application();
        Excel.Workbook workbook = excelApp.Workbooks.Open(@"C:\Users\rnewell\Desktop\FxData.xlsx");
        Excel.Worksheet worksheet = workbook.Sheets[1];
        Excel.Range range = worksheet.UsedRange;

        int rowCount = range.Rows.Count;
        int colCount = range.Columns.Count;



        int rowCounter = 1;
        int colCounter = 1;

        while (rowCounter < rowCount)
        {
            colCounter = 1;
            while (colCounter <= colCount)
            {
                //Console.Write(range.Cells[rowCounter, colCounter].Value2.ToString() + " ");
                testList.Add(range.Cells[rowCounter, colCounter].Value2.ToString());
                colCounter++;
            }
            Console.WriteLine();
            rowCounter++;


        }



        Console.ReadKey();
        excelApp.Workbooks.Close();


    }

Answer 1

@TimWilliams' comment is the correct answer. Reading a single cell takes as long as reading a range of any size. This is the overhead of talking to the COM layer, and you are incurring it thousands of times. You should write the range to an object[,] , and then access that array cell by cell.

    int rowCount = range.Rows.Count;
    int colCount = range.Columns.Count;

    object[,] values= range.Value2;

    int rowCounter = 1;
    int colCounter = 1;

    while (rowCounter < rowCount)
    {
        colCounter = 1;
        while (colCounter <= colCount)
        {
            // check for null?
            testList.Add(values[rowCounter, colCounter].ToString());
        }
    }

Note that the array will be one-based instead of zero-based like normal C# arrays. The data will go from 1 to rowCount and from 1 to colCount, but Rows and Columns properties will return rowCount and colCount, not 1 + rowCount and 1 + colCount. If you want to write data back, you can use a zero-based array of the right size (in fact you have to AFAIK since you can't create a one-based array) and it will work fine.

Answer 2

Since you are loading data from the Open XML (*.xlsx) file format, I would suggest you use Open XML SDK . It doesn't start Excel in the background which is always a good thing, in particular if you need to run your code non-interactively .

I've also written a blog post on different methods of accessing data in Excel which you might find useful.

Answer 3

In general, it should be a matter of seconds.

But as you are creating an instance of Excel itself including its addons it may take a long time to initialize everything in your instance.

For your purpose you can use any public domain excel sheet reading library which doesn't launch Excel.