Apply suggestions from code review

Co-authored-by: Joe Hamman <[email protected]>

Author: rabernat

src/zarr/codecs/__init__.py

@@ -34,7 +34,7 @@
 ]
 
 
-def get_default_array_bytes_codec(
+def _get_default_array_bytes_codec(
     np_dtype: np.dtype[Any],
 ) -> BytesCodec | VLenUTF8Codec | VLenBytesCodec:
     dtype = DataType.from_numpy(np_dtype)

src/zarr/codecs/vlen_utf8.py

@@ -19,8 +19,8 @@
 
 
 # can use a global because there are no parameters
-vlen_utf8_codec = VLenUTF8()
-vlen_bytes_codec = VLenBytes()
+_vlen_utf8_codec = VLenUTF8()
+_vlen_bytes_codec = VLenBytes()
 
 
 @dataclass(frozen=True)

src/zarr/core/buffer/core.py

@@ -313,7 +313,6 @@ class NDBuffer:
     """
 
     def __init__(self, array: NDArrayLike) -> None:
-        # assert array.dtype != object
         self._data = array
 
     @classmethod
@@ -466,7 +465,6 @@ def all_equal(self, other: Any, equal_nan: bool = True) -> bool:
             # Handle None fill_value for Zarr V2
             return False
         # use array_equal to obtain equal_nan=True functionality
-        # Note from Ryan: doesn't this lead to a huge amount of unnecessary memory allocation on every single chunk?
         # Since fill-value is a scalar, isn't there a faster path than allocating a new array for fill value
         # every single time we have to write data?
         _data, other = np.broadcast_arrays(self._data, other)