StaxMate 1.3: a sizable leap for High Performance XML Processing on Java Platform
So here it is: a significant revision of StaxMate library, version 1.3. Given that it is "only" a minor version bump, why is it important?
Because this time around, I focused on improving those small things that are irritating on day-to-day usage, based on my own experiences as well as on my co-workers' experiences; and resolving of which can actually improve both enjoyability and productivity of development. Things like:
- Lack of decent Javadocs: mostly resolved now
- Missing Typed Access methods (for booleans, ints, longs), as well as mismatch between element/attribute access, and reader/writer-side support.
- Verbosity of constructing root-level instances (root cursor, output document/fragment): addressed by allowing factories to be constructed with matching Stax factory, and adding more convenience methods that construct not only StaxMate objects, but also underlying stream reader/writer
- Unexpected exceptions for some methods: for example, added SMInputCursor.getStreamLocation() that works even if cursor itself is not value
- Provide SMInputCursor.getPathDesc() (which works best if "element tracking" is enabled; if so, can get XPath-like description)
On the other hand, I did not yet address all items in the backlog (see StaxMate Jira for list of remaining ideas). This is mostly because I really wanted to get a new "official" release out so that this first batch of improvements can be properly tried out by users. And if all goes well, maybe I will get to write an article or two about how to use them. In the meantime, feel free to check out updated version of the Sample Uuid Generator web app included with StaxMate 1.3 distribution package (under dir ./sample-webapp).
With that said, please go and check out StaxMate 1.3, since: "using Stax without StaxMate is like drinking tonic water without gin!"