On Wed, May 20, 2009 at 2:22 PM, Julian Reschke <[email protected]> wrote:
> [email protected] wrote:
>>
>> + /**
>> + * Returns the nodes in the given query result as an {...@link Iterable}
>> + * for use in a Java 5 for-each loop. The return value encapsulates
>> + * the {...@link QueryResult#getNodes()} method call. Potential
>> + * {...@link RepositoryException}s are converted to {...@link
>> RuntimeException}s.
>> + *
>> + * @param result query result
>> + * @return nodes in the query result
>> + */
>> + public static Iterable<Node> getNodes(final QueryResult result) {
>> + return new Iterable<Node>() {
>> + �...@suppresswarnings("unchecked")
>> + public Iterator<Node> iterator() {
>> + try {
>> + return result.getNodes();
>> + } catch (RepositoryException e) {
>> + throw new RuntimeException(
>> + "Unable to access nodes in " + result, e);
>> + }
>> + }
>> + };
>> + }
>
> Jukka, is it really a good idea to catch the RepositoryException here,
> mapping it to an unchecked exception?
i don't like this either. how about:
public static Iterable<Node> getNodes(final QueryResult result) throws
RepositoryException {
final Iterator iter = result.getNodes();
return new Iterable<Node>() {
@SuppressWarnings("unchecked")
public Iterator<Node> iterator() {
return iter;
}
};
i don't now about the exact semantics of the Iterable interface - i.e.
if a new iterator needs to be fetched for each call. but for the sake
of the "foreach" loop, this should work.
regards, toby
