Station API Datagen

A library mod for Station API which allows modders to programmatically generate JSON files.

Note

Currently, the API does not make hashes of files to check for changes. As of now, all files are regenerated every time you run data. Along with this, stale files are not removed. I recommend completely clearing your data generated folder and re-running data before making a release just to be certain that there are no stale files.

Usage

The buildscript codeblocks are Groovy, however it should be very similar (if not then exactly the same) in Kotlin DSL.

1. Include the datagen library. You can do this via Jitpack:

   dependencies {
       modImplementation("com.github.emmathemartian:stapi-datagen:version")
   }

2. Configure source sets to include a new generated source set. Make a directory at src/generated/, then add this to your buildscript:

   sourceSets {
       main {
           java {
               srcDir("src/main/java")
           }
           resources {
               srcDir("src/test/resources")
               srcDir("src/generated/resources")
           }
       }
   }

3. Create a new run configuration to run the data generator:

   loom {
       ...
       
       runs {
           register("data") {
               property("datagen.run", "MODID") // replace with your mod's id
               property("datagen.path", project.projectDir.toPath().resolve("src/generated/resources/").toAbsolutePath().toString())
               client()
           }
       }
   }

4. Create an entrypoint for the data generator to find:

   public class ModData implements DataEntrypoint {
      @Entrypoint.Namespace
      private static final Namespace NAMESPACE = Null.get();
      
      @Override
      public void run() {
         DataGenContext context = new DataGenContext(NAMESPACE);
         
         context.run(new CraftingRecipeProvider(context) {
            @Override
            public void run(DataGenContext context) {
               shapeless()
                     .ingredient(Ingredient.of(Block.DIRT.asItem()))
                     .ingredient(Ingredient.of(Item.WATER_BUCKET))
                     .result(new ItemStack(Block.CLAY.asItem()))
                     .save("dirt_to_clay", this, context);
            }
         });
      }
   }

   // fabric.mod.json
   {
      "entrypoints": {
         "data": [
            "yourpackage.ModData"
         ]
      }
   }

5. Reload Gradle and test using gradlew runData. You should see the file at src/generated/resources/assets/example_mod/stationapi/recipes/crafting/dirt_to_clay.json.

See src/test/java/emmathemartian/datagen/test/ for a more comprehensive example.

Troubleshooting

The window does not appear or it closes after loading

This is intentional when running the data generator. Due to when data is made and when the mod's JAR is made, you can't use any freshly generated files until the next run.

Old recipes/models/whatever still exist even after removing the code to generate them

Read note.

"Entry ... is a duplicate but no duplicate handling strategy has been set."

Either you have a literal duplicate file across two source sets, or you are experiencing this bug. If you do not have any duplicate files, but you have duplicate folders, it will be the latter.

Example: src/main/resouces/assets/ and src/generated/resources/assets/. The assets folder is flagged as a duplicate for some reason. Thanks Gradle.

To fix the latter, just put this line in processResources (or whatever other task is failing):

duplicatesStrategy = DuplicatesStrategy.EXCLUDE

If you prefer not to hunt down individual tasks to define duplicatesStrategy in, then you can add this code to your buildscript, which should configure the duplicates strategy for most tasks where it is important:

tasks.withType(Jar).configureEach {
	duplicatesStrategy = DuplicatesStrategy.EXCLUDE
}

tasks.withType(Copy).configureEach {
	duplicatesStrategy = DuplicatesStrategy.EXCLUDE
}