>>>Hi Antonio,
>>>
>>>Here is the explanation. Some customers have cost centers as numbers (e.g. 800.0045) but my program has a table where these cost centers are characters (e.g. "800.0045") (because different customers have a different pattern of values).
>>>I was converting from an excel spreadsheet that customer sent me to my table. I was trying to create a generic code that would work with any set of values.
>>>And thank you for the sample code.
>>
>>Ok, Dmitry, it's becoming clearer to me. If I got it completely right, you must respect whatever string of characters the Excel file has in the cell / column that holds the cost center id and insert it as a character in your table.
>>
>>Can you export the Excel data to CSV and just then append the rows to your tables, in that way enforcing the target type as character instead of number?
>
>Yes, I suppose I can export the excel to CSV. But I found that using the library created by Greg Greene makes it easier. Since I can do it all in once step, code that directly converts the excel to a DBF table.
>Thank you.

I had suspicion that this "database" was source of the problem. Forcing some columns into string early on helps a lot for other tight places like leading zeroes deleted from strings meant as phone numbers / area or cell prefix code...

And yes, with some clients you do not have any pull and sometimes clients trick cells out of string interpretation