Hacker News new | ask | show | jobs
by cess11 472 days ago
The low hanging fruit in this area is to do partial unmarshalling or using a SAX parser on a stream. It's likely you'll have to do this to retrieve data and put it in whatever succinct or otherwise efficient data structure.

In Java, which I consider to have the best tooling for advanced XML applications, you'd look into JAXB on streams, StAX or SAX. On complicated and heavily nested XML it might take some effort and profiling to figure out the optimal state machines for exhaustive traversal, if that's what you do.

I'd also like to mention that XSLT is an often underappreciated approach.