chore: replace hard-coded next values with constants throughout all the source code (#2056)

tests/processor/test_batch_conversion.py

@@ -2,7 +2,7 @@ import torch
 
 from lerobot.processor import DataProcessorPipeline, TransitionKey
 from lerobot.processor.converters import batch_to_transition, transition_to_batch
-from lerobot.utils.constants import ACTION, OBS_IMAGE, OBS_PREFIX, OBS_STATE
+from lerobot.utils.constants import ACTION, DONE, OBS_IMAGE, OBS_PREFIX, OBS_STATE, REWARD, TRUNCATED
 
 
 def _dummy_batch():
@@ -12,9 +12,9 @@ def _dummy_batch():
         f"{OBS_IMAGE}.right": torch.randn(1, 3, 128, 128),
         OBS_STATE: torch.tensor([[0.1, 0.2, 0.3, 0.4]]),
         ACTION: torch.tensor([[0.5]]),
-        "next.reward": 1.0,
-        "next.done": False,
-        "next.truncated": False,
+        REWARD: 1.0,
+        DONE: False,
+        TRUNCATED: False,
         "info": {"key": "value"},
     }
 
@@ -38,9 +38,9 @@ def test_observation_grouping_roundtrip():
 
     # Check other fields
     assert torch.allclose(batch_out[ACTION], batch_in[ACTION])
-    assert batch_out["next.reward"] == batch_in["next.reward"]
-    assert batch_out["next.done"] == batch_in["next.done"]
-    assert batch_out["next.truncated"] == batch_in["next.truncated"]
+    assert batch_out[REWARD] == batch_in[REWARD]
+    assert batch_out[DONE] == batch_in[DONE]
+    assert batch_out[TRUNCATED] == batch_in[TRUNCATED]
     assert batch_out["info"] == batch_in["info"]
 
 
@@ -51,9 +51,9 @@ def test_batch_to_transition_observation_grouping():
         f"{OBS_IMAGE}.left": torch.randn(1, 3, 128, 128),
         OBS_STATE: [1, 2, 3, 4],
         ACTION: torch.tensor([0.1, 0.2, 0.3, 0.4]),
-        "next.reward": 1.5,
-        "next.done": True,
-        "next.truncated": False,
+        REWARD: 1.5,
+        DONE: True,
+        TRUNCATED: False,
         "info": {"episode": 42},
     }
 
@@ -115,9 +115,9 @@ def test_transition_to_batch_observation_flattening():
 
     # Check other fields are mapped to next.* format
     assert batch[ACTION] == "action_data"
-    assert batch["next.reward"] == 1.5
-    assert batch["next.done"]
-    assert not batch["next.truncated"]
+    assert batch[REWARD] == 1.5
+    assert batch[DONE]
+    assert not batch[TRUNCATED]
     assert batch["info"] == {"episode": 42}
 
 
@@ -125,9 +125,9 @@ def test_no_observation_keys():
     """Test behavior when there are no observation.* keys."""
     batch = {
         ACTION: torch.tensor([1.0, 2.0]),
-        "next.reward": 2.0,
-        "next.done": False,
-        "next.truncated": True,
+        REWARD: 2.0,
+        DONE: False,
+        TRUNCATED: True,
         "info": {"test": "no_obs"},
     }
 
@@ -146,9 +146,9 @@ def test_no_observation_keys():
     # Round trip should work
     reconstructed_batch = transition_to_batch(transition)
     assert torch.allclose(reconstructed_batch[ACTION], torch.tensor([1.0, 2.0]))
-    assert reconstructed_batch["next.reward"] == 2.0
-    assert not reconstructed_batch["next.done"]
-    assert reconstructed_batch["next.truncated"]
+    assert reconstructed_batch[REWARD] == 2.0
+    assert not reconstructed_batch[DONE]
+    assert reconstructed_batch[TRUNCATED]
     assert reconstructed_batch["info"] == {"test": "no_obs"}
 
 
@@ -173,9 +173,9 @@ def test_minimal_batch():
     reconstructed_batch = transition_to_batch(transition)
     assert reconstructed_batch[OBS_STATE] == "minimal_state"
     assert torch.allclose(reconstructed_batch[ACTION], torch.tensor([0.5]))
-    assert reconstructed_batch["next.reward"] == 0.0
-    assert not reconstructed_batch["next.done"]
-    assert not reconstructed_batch["next.truncated"]
+    assert reconstructed_batch[REWARD] == 0.0
+    assert not reconstructed_batch[DONE]
+    assert not reconstructed_batch[TRUNCATED]
     assert reconstructed_batch["info"] == {}
 
 
@@ -197,9 +197,9 @@ def test_empty_batch():
     # Round trip
     reconstructed_batch = transition_to_batch(transition)
     assert reconstructed_batch[ACTION] is None
-    assert reconstructed_batch["next.reward"] == 0.0
-    assert not reconstructed_batch["next.done"]
-    assert not reconstructed_batch["next.truncated"]
+    assert reconstructed_batch[REWARD] == 0.0
+    assert not reconstructed_batch[DONE]
+    assert not reconstructed_batch[TRUNCATED]
     assert reconstructed_batch["info"] == {}
 
 
@@ -210,9 +210,9 @@ def test_complex_nested_observation():
         f"{OBS_IMAGE}.left": {"image": torch.randn(1, 3, 128, 128), "timestamp": 1234567891},
         OBS_STATE: torch.randn(7),
         ACTION: torch.randn(8),
-        "next.reward": 3.14,
-        "next.done": False,
-        "next.truncated": True,
+        REWARD: 3.14,
+        DONE: False,
+        TRUNCATED: True,
         "info": {"episode_length": 200, "success": True},
     }
 
@@ -240,9 +240,9 @@ def test_complex_nested_observation():
     assert torch.allclose(batch[ACTION], reconstructed_batch[ACTION])
 
     # Check other fields
-    assert batch["next.reward"] == reconstructed_batch["next.reward"]
-    assert batch["next.done"] == reconstructed_batch["next.done"]
-    assert batch["next.truncated"] == reconstructed_batch["next.truncated"]
+    assert batch[REWARD] == reconstructed_batch[REWARD]
+    assert batch[DONE] == reconstructed_batch[DONE]
+    assert batch[TRUNCATED] == reconstructed_batch[TRUNCATED]
     assert batch["info"] == reconstructed_batch["info"]
 
 
@@ -267,13 +267,13 @@ def test_custom_converter():
     batch = {
         OBS_STATE: torch.randn(1, 4),
         ACTION: torch.randn(1, 2),
-        "next.reward": 1.0,
-        "next.done": False,
+        REWARD: 1.0,
+        DONE: False,
     }
 
     result = processor(batch)
 
     # Check the reward was doubled by our custom converter
-    assert result["next.reward"] == 2.0
+    assert result[REWARD] == 2.0
     assert torch.allclose(result[OBS_STATE], batch[OBS_STATE])
     assert torch.allclose(result[ACTION], batch[ACTION])