[Enhancement] Add Warehouse Optimization Environment for RL

.github/workflows/docker-build.yml

@@ -85,6 +85,8 @@ jobs:
             dockerfile: src/envs/textarena_env/server/Dockerfile
           - name: browsergym-env
             dockerfile: src/envs/browsergym_env/server/Dockerfile
+          - name: warehouse-env
+            dockerfile: src/envs/warehouse_env/server/Dockerfile
 
     steps:
       - name: Checkout code

examples/warehouse_simple.py

@@ -0,0 +1,216 @@
+"""
+Simple example script demonstrating the Warehouse Environment.
+
+This script shows basic usage of the warehouse environment with both
+random and greedy agents.
+"""
+
+import os
+import random
+import sys
+
+# Add src directory to path for imports
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..", "src"))
+
+from envs.warehouse_env import WarehouseAction, WarehouseEnv
+
+
+def random_agent_example():
+    """Run a simple random agent in the warehouse."""
+    print("=" * 60)
+    print("RANDOM AGENT EXAMPLE")
+    print("=" * 60)
+
+    # Connect to server (assumes server is running on localhost:8000)
+    # Or use from_docker_image() to automatically start a container
+    try:
+        env = WarehouseEnv(base_url="http://localhost:8000")
+    except Exception as e:
+        print(f"Error connecting to server: {e}")
+        print("\nPlease start the server first:")
+        print("  docker run -p 8000:8000 warehouse-env:latest")
+        print("Or run the server locally:")
+        print(
+            "  python -m uvicorn envs.warehouse_env.server.app:app --host 0.0.0.0 --port 8000"
+        )
+        return
+
+    # Run one episode
+    result = env.reset()
+    print(f"\nStarting episode...")
+
+    # Debug: print what we got
+    print(f"Grid data: {result.observation.grid}")
+    print(f"Robot position: {result.observation.robot_position}")
+
+    if result.observation.grid and len(result.observation.grid) > 0:
+        print(
+            f"Grid size: {len(result.observation.grid)}x{len(result.observation.grid[0])}"
+        )
+    else:
+        print("Warning: Grid is empty!")
+
+    print(f"Packages to deliver: {result.observation.total_packages}")
+    print(f"Max steps: {result.observation.time_remaining}")
+
+    step = 0
+    done = False
+
+    while not done and step < 50:  # Limit to 50 steps for demo
+        # Random action
+        action = WarehouseAction(action_id=random.randint(0, 5))
+        result = env.step(action)
+
+        step += 1
+        if result.reward != -0.1:  # Only print interesting events
+            print(
+                f"Step {step}: {action.action_name} -> {result.observation.message} (reward: {result.reward:.1f})"
+            )
+
+        done = result.done
+
+    # Print final results
+    print(f"\nEpisode finished!")
+    print(f"Steps taken: {step}")
+    print(
+        f"Packages delivered: {result.observation.packages_delivered}/{result.observation.total_packages}"
+    )
+    print(f"Total reward: {env.state().cum_reward:.2f}")
+
+    env.close()
+
+
+def greedy_agent_example():
+    """Run a simple greedy agent that moves toward targets."""
+    print("\n" + "=" * 60)
+    print("GREEDY AGENT EXAMPLE")
+    print("=" * 60)
+
+    def get_greedy_action(obs):
+        """Simple greedy policy: move toward nearest target."""
+        robot_x, robot_y = obs.robot_position
+
+        # Determine target location
+        if obs.robot_carrying is None:
+            # Not carrying: move toward nearest waiting package
+            for pkg in obs.packages:
+                if pkg["status"] == "waiting":
+                    target_x, target_y = pkg["pickup_location"]
+                    break
+            else:
+                # No packages waiting, try to pick up
+                return 4
+        else:
+            # Carrying: move toward dropoff zone
+            pkg = next((p for p in obs.packages if p["id"] == obs.robot_carrying), None)
+            if pkg:
+                target_x, target_y = pkg["dropoff_location"]
+            else:
+                return 4  # Try action
+
+        # Simple pathfinding: move closer on one axis at a time
+        if robot_x < target_x:
+            return 3  # RIGHT
+        elif robot_x > target_x:
+            return 2  # LEFT
+        elif robot_y < target_y:
+            return 1  # DOWN
+        elif robot_y > target_y:
+            return 0  # UP
+        else:
+            # At target location
+            return 4 if obs.robot_carrying is None else 5
+
+    try:
+        env = WarehouseEnv(base_url="http://localhost:8000")
+    except Exception as e:
+        print(f"Error connecting to server: {e}")
+        return
+
+    # Run 3 episodes
+    for episode in range(3):
+        result = env.reset()
+        print(f"\nEpisode {episode + 1}")
+        print(f"Packages: {result.observation.total_packages}")
+
+        done = False
+        steps = 0
+
+        while not done and steps < 200:
+            action_id = get_greedy_action(result.observation)
+            action = WarehouseAction(action_id=action_id)
+            result = env.step(action)
+            steps += 1
+
+            # Print delivery events
+            if "delivered" in result.observation.message.lower():
+                print(f"  Step {steps}: {result.observation.message}")
+
+            done = result.done
+
+        state = env.state()
+        print(
+            f"  Result: {state.packages_delivered}/{state.total_packages} delivered, "
+            f"reward: {state.cum_reward:.2f}, steps: {steps}"
+        )
+
+    env.close()
+
+
+def visualization_example():
+    """Demonstrate ASCII visualization."""
+    print("\n" + "=" * 60)
+    print("VISUALIZATION EXAMPLE")
+    print("=" * 60)
+
+    try:
+        env = WarehouseEnv(base_url="http://localhost:8000")
+    except Exception as e:
+        print(f"Error connecting to server: {e}")
+        return
+
+    # Reset and show initial state
+    result = env.reset()
+    print("\nInitial warehouse state:")
+    print(env.render_ascii())
+
+    # Take a few actions and show updates
+    actions = [3, 3, 1, 1, 4]  # RIGHT, RIGHT, DOWN, DOWN, PICKUP
+    for i, action_id in enumerate(actions):
+        action = WarehouseAction(action_id=action_id)
+        result = env.step(action)
+        print(f"\nAfter action {i+1} ({action.action_name}):")
+        print(env.render_ascii())
+
+        if result.done:
+            break
+
+    env.close()
+
+
+if __name__ == "__main__":
+    print("Warehouse Environment - Example Script")
+    print("=" * 60)
+    print("\nThis script demonstrates the warehouse environment.")
+    print("Make sure the server is running on http://localhost:8000")
+    print("\nTo start the server:")
+    print("  docker run -p 8000:8000 warehouse-env:latest")
+    print("\n" + "=" * 60)
+
+    # Run examples
+    try:
+        random_agent_example()
+        greedy_agent_example()
+        visualization_example()
+
+        print("\n" + "=" * 60)
+        print("All examples completed successfully!")
+        print("=" * 60)
+
+    except KeyboardInterrupt:
+        print("\n\nInterrupted by user")
+    except Exception as e:
+        print(f"\n\nError: {e}")
+        import traceback
+
+        traceback.print_exc()