The current converter classes produce and consume DOM trees. This is memory-inefficient, and memory-prohibitive for situations such as dumping a large database table (where each row could be a separate object).
Adding this feature will break JDK 1.5 compatibility. I think the solution is to create a parallel development branch (dev_2_0), which supports JDK 1.6, and have the two branches evolve in lock-step. At the same time, I'd want to add a toStax() method to the builder.
This will require a lot of merging, which might be the push I need to convert the repository to git.
In the interim, I think some refactoriing is in order: creating event-like source and result classes that could pull their data from either the DOM or STAX. Also, a short-term change to improve the memory efficiency of Bean->XML conversions is to use XmlBuilder rather than DOM.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Adding this feature will break JDK 1.5 compatibility. I think the solution is to create a parallel development branch (dev_2_0), which supports JDK 1.6, and have the two branches evolve in lock-step. At the same time, I'd want to add a toStax() method to the builder.
This will require a lot of merging, which might be the push I need to convert the repository to git.
In the interim, I think some refactoriing is in order: creating event-like source and result classes that could pull their data from either the DOM or STAX. Also, a short-term change to improve the memory efficiency of Bean->XML conversions is to use XmlBuilder rather than DOM.