HSSFRow.RemoveAllCells() currently assumes that all cells are used in a row.
Given that I have a row where the first cells is unused and then I have couple of cells that are used the current code blows up here:
```
int initialLen = cells.Count;
for (int i = 0; i < initialLen; i++) // assumes that all cells up until count is used
{
RemoveCell(cells[i], true); // throws KeyNotFoundException since with key 0 does not exist.
}
```
It should iterate over the keys (or values) instead to access all cells that are actually used on a row.
Something like this:
```
for (var key in cells.Keys.ToList())
{
RemoveCell(cells[i], true);
}
```
Comments: ** Comment from web user: tonyqus **
Given that I have a row where the first cells is unused and then I have couple of cells that are used the current code blows up here:
```
int initialLen = cells.Count;
for (int i = 0; i < initialLen; i++) // assumes that all cells up until count is used
{
RemoveCell(cells[i], true); // throws KeyNotFoundException since with key 0 does not exist.
}
```
It should iterate over the keys (or values) instead to access all cells that are actually used on a row.
Something like this:
```
for (var key in cells.Keys.ToList())
{
RemoveCell(cells[i], true);
}
```
Comments: ** Comment from web user: tonyqus **
which version of NPOI are you using? 1.2.5?