While browsing through the source code for ResultSetIterator in version 1.6.1 I found
the following code snippet:
public boolean hasNext() {
try {
advance();
return (!eof);
} catch (SQLException e) {
throw new RuntimeException("hasNext(): SQLException: " + e);
}
}
public Object next() {
try {
advance();
if (eof) {
throw new NoSuchElementException();
}
current = false;
return (this);
} catch (SQLException e) {
throw new RuntimeException("next(): SQLException: " + e);
}
}
If someone were to use it in the normal format in which the hasNext() and next()
methods are used, the person would skip over one row in the resultset for every
iteration of the iterator since both methods call unconditionally the advance()
method, which advances the resultset by one.
I propose that a flag variable be introduced into the iterator class so that the
next() method would conditionally advance the resultset by one only if no previous
hasNext() method has been called. The following is my proposed code change:
private boolean hasNextCalled = false;
/**
* <p>Return <code>true</code> if the iteration has more elements.</p>
*/
public boolean hasNext() {
try {
advance();
hasNextCalled = true;
return (!eof);
} catch (SQLException e) {
throw new RuntimeException("hasNext(): SQLException: " + e);
}
}
/**
* <p>Return the next element in the iteration.</p>
*/
public Object next() {
try {
if(!hasNextCalled)
{
advance();
}
else
{
hasNextCalled = false;
}
if (eof) {
throw new NoSuchElementException();
}
current = false;
return (this);
} catch (SQLException e) {
throw new RuntimeException("next(): SQLException: " + e);
}
}
What do you guys think?
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
