Reverse engineering

Reverse Engineering

2011-07-11

One thing I have recently realized about the design of YeS is that it may be hard to "reverse" engineer. One of the things I was hoping to do is write a Ruby Class generator that used a schema as an input and spit out Ruby script. But this may proved difficult to do for complex YPath selector. For example how does one generate a Ruby class to handle something as simple as:

---
//foo:
  type: str

Clearly it means there is a foo attribute somewhere, but where exactly? I suspect it can still be done, but it will prove harder than say simple schemas like Kwalify and Igny's nacent design.