I have this legacy code:
private Worksheet _xlSheet;
. . .
_xlSheet.PageSetup.PrintArea = "A1:" +
GetExcelTextColumnName(_grandTotalsColumn) + finalRow;
. . .
protected string GetExcelTextColumnName(int columnNum)
{
StringBuilder sb = new StringBuilder();
if (columnNum > 26)
{
int firstLetter = ((columnNum - 1) / 26) + 64;
int secondLetter = ((columnNum - 1) % 26) + 65;
sb.Append((char)firstLetter);
sb.Append((char)secondLetter);
}
else
{
sb.Append((char)(64 + (columnNum - 1)));
}
return sb.ToString();
}
I simplified the code to this extent:
_xlSheet.PageSetup.PrintArea = "A1:" + GetExcelTextColumnName(_xlSheet.UsedRange.Columns.Count) + _xlSheet.UsedRange.Rows.Count;
...but wonder if there is a way to simplify the "Column-number-to-column-Alpha" fancy-pants code. Is there, or is the existing GetExcelTextColumnName() actually elegant and as-concise-and-maintainable-as-possible code?
I would use a iterative algorithm, which will support columns beyond ZZ which you find in recent versions of Excel.
private const string Alphabet = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
public string GetColumnName(int columnIndex)
{
if (columnIndex < 0) throw new ArgumentOutOfRangeException("columnIndex", columnIndex, "Column index index may not be negative.");
string result = "";
for (; ; )
{
result = Alphabet[(columnIndex) % 26] + result;
if (columnIndex < 26) break;
columnIndex = columnIndex / 26 - 1;
}
return result;
}
I also think the use of a StringBuilder
is overkill, since you know you're likely to have only one or two concatenations.
You could also consider using a Range object to avoid the need to do the calculation - probably slightly slower, but arguably more readable. Something like:
_xlSheet.PageSetup.PrintArea =
_xlSheet.Range("A1")
.Resize(finalRow, _grandTotalsColumn).Address;