Data for avro testing is generated given the following inputs:
- a .avsc avro schema file (describing the record structure).
- a .json file with the actual data values.

These are the inputs for generating the actual avro file to be tested. This is
done by manually invoking the avro compiler, giving the 2 as inputs.

For example, from this current directory, invoke as follows:

  java -jar <path to>/avro-tools-1.7.4.jar fromjson simple.json --schema-file simple.avsc > simple.avro
  
Creating an avro file with compression can be done by adding --codec <snappy|deflate|bzip2> to the command line:

  java -jar <path to>/avro-tools-1.7.4.jar fromjson --codec snappy simple.json --schema-file simple.avsc > simple.snappy.avro

