Data Schema with characters incompatible with ascii

# Overview

The Data Schema JSON will throw an error if it includes characters not compatible with ascii.
This happens due to the use of Swedish characters like "Ö" and "Ä".


This is the [line of code](https://github.com/frictionlessdata/ckanext-validation/blob/master/ckanext/validation/validators.py#L24) where it happens. I made a quick workaround by replacing it with:

`descriptor = json.loads(value.encode('utf-8'))`

Just wanted to get some feedback on how to approach this, if there should be somewhere an option to declare the encoding of the file? Or you want a PR with this work around?

---

Please preserve this line to notify @amercader (maintainer of this repository)