> Instead of clearing and reusing the the AttributeList object,
> wouldn't it be better to create a new attribute list object? If
> the old Attribute list isn't being referenced, it will be garbage
> collectible. If the old Attribute list is being reference, it
> won't be changed out from under the client. This way of doing it
> seems to offer the best of both worlds.
Depending on the virtual machine, this could be a killer. Remember
that a medium-sized XML document (such as a book) might have 10,000
elements: that would mean an extra 10,000 attribute lists allocated
and then garbage collected in what should be only a few seconds of
parsing.
In fact, since AttributeList is an interface, drivers often implement
it themselves rather than allocating an object for it, as in
import org.xml.sax.Parser;
import org.xml.sax.AttributeList;
public class MySAXDriver implements Parser, AttributeList
{
...
}
SAX was designed to be flexible so that people could write
highly-optimised implementations like this if they wanted. It's also
designed to work with languages that don't support GC out-of-the-box,
like C++.
However, we also recognised that people might want to keep around
attribute lists sometimes, so we made it very easy by adding the
org.xml.sax.helpers.AttributeListImpl class with a copy constructor:
AttributeList persistentAtts
public void startElement (String name, AttributeList atts)
{
persistentAtts = new AttributeListImpl(atts);
}
This approach does give you the best of both worlds -- it is trivially
easy to make a persistent copy when you need to, but you're not stuck
with the allocation/gc overhead when you don't need it.
All the best,
David
--
David Megginson david@megginson.com
http://www.megginson.com/