Can anyone tell me which one better to use for xml parsing - DOM or Castor (with criterion being lesser memory usage)?
Both will create an object graph in memory containing all the unmarshalled objects. I wouldn't expect a big advantage from one over the other in terms of memory use.