Little optimization

The file source is now being handled by native java io. The file stream is not converted to view (extra overhead). Overall 217seconds to citylots.json. Which is a significant improvement? from 227seconds. It's to micro optimize the entire code. Before that documentation is important. But THIS IS THE FASTEST VARIANT OF THE PARSER. LAZY VERSION COMING SOON.

Easy API

It's easier to write type-safe code now. You still have to know the data type though. So make sure you do it right. Errors and exception added in the implicit for the same.

Lazy evaluation

Due to some bad implementation, the library was not working lazily. Tokenizer was passing a stream which was eagerly evaluated by the Lexical Analyzer due to usage of "val" binding. Fixed the same. Benchmarked with citylots.json. Parsed and constructed immutable map in 227s for 413122 number of lines of nested Json.

API Changes

Now added the ability to use apply() method on all of the concrete classes of JsonArray and JsonObject. Basic values however does not supports them and will raise IllegalAccessException on calling apply. This allows for a smoother syntax that is more close to how we would deal with Json in practice

First release

Supports basic parsing of Json files