reset order of encoding/decoding but special case times in CFMaskCoder

pydata · May 8, 2023 · 23c565c · 23c565c
1 parent f5db449
commit 23c565c
Show file tree

Hide file tree

Showing 2 changed files with 29 additions and 32 deletions.
diff --git a/xarray/coding/variables.py b/xarray/coding/variables.py
@@ -236,20 +236,19 @@ def encode(self, variable: Variable, name: T_Name = None):
                 f"Variable {name!r} has conflicting _FillValue ({fv}) and missing_value ({mv}). Cannot encode data."
             )
 
+        # cast to correct dtype in case of times
+        # see GH 7817
+        units = attrs.get("units", None)
+        if isinstance(units, str) and "since" in units:
+            encoded_dtype = encoding.pop("dtype")
+            if encoded_dtype is not None and encoded_dtype != data.dtype:
+                data = np.asarray(data, dtype=encoded_dtype)
+
         if fv_exists:
             # Ensure _FillValue is cast to same dtype as data's
             encoding["_FillValue"] = dtype.type(fv)
             fill_value = pop_to(encoding, attrs, "_FillValue", name=name)
-            # retrieve _FillValue in case of np.datetime64
-            # see GH 7817
-            if np.issubdtype(data.dtype, np.datetime64):
-                units = encoding.get("units", None)
-                if isinstance(units, str) and "since" in units:
-                    delta, _ = times._unpack_netcdf_time_units(units)
-                    delta = times._netcdf_to_numpy_timeunit(delta)
-                    fill_value = np.datetime64(fill_value.item(), delta).astype(
-                        "datetime64[ns]"
-                    )
+
             if not pd.isnull(fill_value):
                 data = duck_array_ops.fillna(data, fill_value)
 
@@ -284,19 +283,19 @@ def decode(self, variable: Variable, name: T_Name = None):
                     SerializationWarning,
                     stacklevel=3,
                 )
-
-            dtype, decoded_fill_value = dtypes.maybe_promote(data.dtype)
-            # retrieve _FillValue in case of np.datetime64
-            # see GH 7817
-            if np.issubdtype(data.dtype, np.datetime64) and decoded_fill_value.astype(
-                np.int64
-            ) == np.datetime64("NaT").astype(np.int64):
-                delta, _ = times._unpack_netcdf_time_units(encoding["units"])
-                delta = times._netcdf_to_numpy_timeunit(delta)
-                encoded_fill_values = {
-                    np.datetime64(encfill.item(), delta).astype("datetime64[ns]")
-                    for encfill in encoded_fill_values
-                }
+            units = attrs.get("units", None)
+            # try to cast to correct dtypes for data and fill_value
+            # GH 7817
+            if ((
+                isinstance(units, str)
+                and "since" in units
+                and np.issubdtype(data.dtype, np.integer)) or np.issubdtype(data.dtype, np.datetime64)
+            ):
+                dtype, decoded_fill_value = data.dtype, np.datetime64("NaT").astype(
+                    data.dtype
+                )
+            else:
+                dtype, decoded_fill_value = dtypes.maybe_promote(data.dtype)
             if encoded_fill_values:
                 transform = partial(
                     _apply_mask,

diff --git a/xarray/conventions.py b/xarray/conventions.py
@@ -174,10 +174,10 @@ def encode_cf_variable(
     ensure_not_multiindex(var, name=name)
 
     for coder in [
-        variables.CFScaleOffsetCoder(),
-        variables.CFMaskCoder(),
         times.CFDatetimeCoder(),
         times.CFTimedeltaCoder(),
+        variables.CFScaleOffsetCoder(),
+        variables.CFMaskCoder(),
         variables.UnsignedIntegerCoder(),
         variables.NonStringCoder(),
         variables.DefaultFillvalueCoder(),
@@ -263,13 +263,6 @@ def decode_cf_variable(
             var = strings.CharacterArrayCoder().decode(var, name=name)
         var = strings.EncodedStringCoder().decode(var)
 
-    # time decoding before masking
-    # GH 7817
-    if decode_timedelta:
-        var = times.CFTimedeltaCoder().decode(var, name=name)
-    if decode_times:
-        var = times.CFDatetimeCoder(use_cftime=use_cftime).decode(var, name=name)
-
     if mask_and_scale:
         for coder in [
             variables.UnsignedIntegerCoder(),
@@ -278,6 +271,11 @@ def decode_cf_variable(
         ]:
             var = coder.decode(var, name=name)
 
+    if decode_timedelta:
+        var = times.CFTimedeltaCoder().decode(var, name=name)
+    if decode_times:
+        var = times.CFDatetimeCoder(use_cftime=use_cftime).decode(var, name=name)
+
     if decode_endianness and not var.dtype.isnative:
         var = variables.EndianCoder().decode(var)
         original_dtype = var.dtype