Privacy policies in JSON

[alexmbraga]

• In previous version, policies were represented as files or hardcoded.
• Now, privacy policies should be represented in JSON format
• @tbasso or @brenofranca83 can give details

[alexmbraga]

*This comment comes from @brenofranca83
*It was move from issue #35 in other prepository: TMA-framework-a
Hi all (@mrmassis @tbasso ),
Some additional info for this issue:

The new interface to be provided for the anonymizer should address a ResultSet as a dataset.
So, I understand the new public method should be something like
public ResultSet run(ResultSet dataset, String policy, int k, float riskThreshold);

In this sense, we'll need to construct the Data object for arx as the method create in
https://github.com/arx-deidentifier/arx/blob/master/src/main/org/deidentifier/arx/Data.java

There are several arx examples and one of them should do something similar in
https://github.com/arx-deidentifier/arx/tree/master/src/example/org/deidentifier/arx/examples

Note the riskThreshold (rmax) is now out of the policy file as a matter of semantics/cohesion.
Besides, the policy file should be formated as a JSON file. So, please use the native Java JSON API:
http://www.oracle.com/technetwork/articles/java/json-1973242.html

The JSON format should follow the idea of the policy format, identifying non-sensitive fields,
identifying/PKs/SKs fields, and anonymization mechanisms (SR, SL, DT, AG, CT).