My current project involves working on an Android application that receives a large amount of dynamic JSON data. Initially, I wrote my custom methods to handle this using the standard org.json packages, but I found it to be messy and plagued with bugs.
After exploring different solutions, I decided to refactor most of my methods using Gson. The JSON data consists of nested objects, arrays, and even further nested objects.
I've structured a top-level container class like this
public class MovieContainer implements Serializable {
private List<Movie> movies;
public List<Movie> getMovies() {
return movies;
}
}
The movie class includes
public class Movie implements Serializable {
private static final long serialVersionUID = 1L;
@SerializedName("rating")
private Rating ratings;
@SerializedName("tmdb_id")
private int tmdbID;
@SerializedName("actor_roles")
private Actor actors;
// ...20 or so items and getters/setters removed for brevity
I've also created separate classes for handling Actors, Ratings, and other related objects.
To parse the JSON, I have implemented 12 different classes which can sometimes make my Java code look convoluted like
movieItem.getImages().getBackdrop().getURL()
This structure doesn't feel very intuitive from a readability perspective.
While I have used custom Deserializer methods to handle dynamic objects, I feel like I'm back to square one with messy code. The added classes and complexity don't seem to justify the initial problem-solving approach I took.
Am I utilizing GSON incorrectly? Have I missed the core concept altogether? Are there established patterns I should follow but am unable to find examples of?
I would greatly appreciate any guidance on the correct way to tackle this issue.
A snippet of the input JSON can be viewed here (it's lengthy, hence not included here).