I've been reading about COLLADA lately, and I was wondering if anyone knew of any good libraries out there for parsing COLLADA files. I've looked for COLLADA DOM, and fCOLLADA, but couldn't find them. What I'd really like (but doubt exists) is if some one decided to write some sort of serialization library that'll load the parts of the COLLADA file you need, then output it to a binary file format organized to my liking.
Any ideas on how to work with COLLADA files? I've thought about parsing through it using tinyXML to load what I need into memory for output, but I really dunno.