Implement full-dimension chunking support

Author: BrianMichell

src/mdio/builder/templates/base.py

@@ -148,14 +148,29 @@ def coordinate_names(self) -> tuple[str, ...]:
     @property
     def full_chunk_shape(self) -> tuple[int, ...]:
         """Returns the chunk shape for the variables."""
-        return copy.deepcopy(self._var_chunk_shape)
+        # If dimension sizes are not set yet, return the stored shape as-is
+        if len(self._dim_sizes) != len(self._dim_names):
+            return self._var_chunk_shape
+
+        # Expand -1 values to full dimension sizes
+        return tuple(
+            dim_size if chunk_size == -1 else chunk_size
+            for chunk_size, dim_size in zip(self._var_chunk_shape, self._dim_sizes, strict=False)
+        )
 
     @full_chunk_shape.setter
     def full_chunk_shape(self, shape: tuple[int, ...]) -> None:
         """Sets the chunk shape for the variables."""
-        if len(shape) != len(self._dim_sizes):
-            msg = f"Chunk shape {shape} does not match dimension sizes {self._dim_sizes}"
+        if len(shape) != len(self._dim_names):
+            msg = f"Chunk shape {shape} has {len(shape)} dimensions, expected {len(self._dim_names)}"
             raise ValueError(msg)
+
+        # Validate that all values are positive integers or -1
+        for chunk_size in shape:
+            if chunk_size != -1 and chunk_size <= 0:
+                msg = f"Chunk size must be positive integer or -1, got {chunk_size}"
+                raise ValueError(msg)
+
         self._var_chunk_shape = shape
 
     @property

tests/unit/v1/templates/test_seismic_templates.py

@@ -23,9 +23,49 @@ def test_chunk_shape_assignment_exception(self) -> None:
         template = Seismic2DPostStackTemplate("time")
         template.build_dataset("test", (50, 50))
 
-        with pytest.raises(ValueError, match="Chunk shape.*does not match dimension sizes"):
+        with pytest.raises(ValueError, match="Chunk shape.*has.*dimensions, expected"):
             template.full_chunk_shape = (32, 32, 32)
 
+    def test_chunk_shape_with_minus_one_before_build(self) -> None:
+        """Test that chunk shape can be set with -1 before build_dataset."""
+        template = Seismic2DPostStackTemplate("time")
+
+        # Should be able to set chunk shape with -1 before build_dataset
+        template.full_chunk_shape = (32, -1)
+
+        # Before build_dataset, getter should return unexpanded values
+        assert template.full_chunk_shape == (32, -1)
+        assert template._var_chunk_shape == (32, -1)
+
+    def test_chunk_shape_with_minus_one_after_build(self) -> None:
+        """Test that -1 values are expanded after build_dataset."""
+        template = Seismic2DPostStackTemplate("time")
+        template.full_chunk_shape = (32, -1)
+
+        # Build dataset with specific sizes
+        template.build_dataset("test", (100, 200))
+
+        # After build_dataset, getter should expand -1 to dimension size
+        assert template.full_chunk_shape == (32, 200)
+        assert template._var_chunk_shape == (32, -1)  # Internal storage unchanged
+
+    def test_chunk_shape_validation_invalid_values(self) -> None:
+        """Test that chunk shape setter rejects invalid values."""
+        template = Seismic2DPostStackTemplate("time")
+        template.build_dataset("test", (50, 50))
+
+        # Test rejection of 0
+        with pytest.raises(ValueError, match="Chunk size must be positive integer or -1"):
+            template.full_chunk_shape = (32, 0)
+
+        # Test rejection of negative values other than -1
+        with pytest.raises(ValueError, match="Chunk size must be positive integer or -1"):
+            template.full_chunk_shape = (32, -2)
+
+        # Test that positive values and -1 are accepted
+        template.full_chunk_shape = (32, -1)  # Should not raise
+        template.full_chunk_shape = (32, 16)  # Should not raise
+
     def test_all_templates_inherit_from_abstract(self) -> None:
         """Test that all concrete templates inherit from AbstractDatasetTemplate."""
         registry = TemplateRegistry()