adam2809 / jsontypeinference

Json Type Inference

Project description

This project provides you with a way to automatically create data models of JSON files and functions to read into them.

What is and is not supported

Arbitrary nesting of objects is supported but only arrays that contain boolean, string or number values will work properly. If they don't a List<Any> type will be assumed. As many APIs send back an array of objects the usage I recommend is to only give the program the first element of the array and when using a deserialization library tell it that the JSON file has structure List<[Generated data class name]>. Another quirk is that if the JSON file provided is not an object then the data class generated will have a topLevel property with an appropriate type.

Usage

To use this project add an entry to the JSON array in src/main/resources/config.json with this structure:

{
    "source": "Path to a file with a sample JSON based on which data classes will be created",
    "dest": "Path to a file where the generated data classes will be placed",
    "className": "Name of the top level data class"
}

Run the program to update all files provided in config.json