Allow _FillValue and missing_value to differ (Fixes pydata#1749)

The CF standard permits both values, and them to have different values,
so we should not be treating this as an error--just mask out all of
them.

Author: dopplershift

xarray/coding/variables.py

@@ -152,26 +152,22 @@ def encode(self, variable, name=None):
     def decode(self, variable, name=None):
         dims, data, attrs, encoding = unpack_for_decoding(variable)
 
+        raw_fill_values = []
         if 'missing_value' in attrs:
-            # missing_value is deprecated, but we still want to support it as
-            # an alias for _FillValue.
-            if ('_FillValue' in attrs and
-                not utils.equivalent(attrs['_FillValue'],
-                                     attrs['missing_value'])):
-                raise ValueError("Conflicting _FillValue and missing_value "
-                                 "attrs on a variable {!r}: {} vs. {}\n\n"
-                                 "Consider opening the offending dataset "
-                                 "using decode_cf=False, correcting the "
-                                 "attrs and decoding explicitly using "
-                                 "xarray.decode_cf()."
-                                 .format(name, attrs['_FillValue'],
-                                         attrs['missing_value']))
-            attrs['_FillValue'] = attrs.pop('missing_value')
+            missing = pop_to(attrs, encoding, 'missing_value', name=name)
+            # Need to handle single and multiple values in 'missing_value'
+            try:
+                raw_fill_values.extend(missing)
+            except TypeError:
+                raw_fill_values.append(missing)
 
         if '_FillValue' in attrs:
-            raw_fill_value = pop_to(attrs, encoding, '_FillValue', name=name)
+            raw_fill_values.append(pop_to(attrs, encoding, '_FillValue',
+                                          name=name))
+
+        if raw_fill_values:
             encoded_fill_values = [
-                fv for fv in np.ravel(raw_fill_value) if not pd.isnull(fv)]
+                fv for fv in np.ravel(raw_fill_values) if not pd.isnull(fv)]
 
             if len(encoded_fill_values) > 1:
                 warnings.warn("variable {!r} has multiple fill values {}, "

xarray/tests/test_conventions.py

@@ -156,12 +156,16 @@ def test(self):
 
 
 def test_decode_cf_with_conflicting_fill_missing_value():
-    var = Variable(['t'], np.arange(10),
+    expected = Variable(['t'], [np.nan, np.nan, 2], {'units': 'foobar'})
+    expected[[0, 1]] = np.nan
+    var = Variable(['t'], np.arange(3),
                    {'units': 'foobar',
                     'missing_value': 0,
                     '_FillValue': 1})
-    with raises_regex(ValueError, "_FillValue and missing_value"):
-        conventions.decode_cf_variable('t', var)
+    with warnings.catch_warnings(record=True) as w:
+        actual = conventions.decode_cf_variable('t', var)
+        assert_identical(actual, expected)
+        assert 'has multiple fill' in str(w[0].message)
 
     expected = Variable(['t'], np.arange(10), {'units': 'foobar'})