BUG: fix + test open_mfdataset fails on variable attributes with list…

doc/whats-new.rst

@@ -61,8 +61,12 @@ Bug fixes
   By `Tom Nicholas <http://github.com/TomNicholas>`_.
 - Fixed crash when applying ``distributed.Client.compute()`` to a DataArray
   (:issue:`3171`). By `Guido Imperiale <https://github.com/crusaderky>`_.
-
-
+- Better error message when using groupby on an empty DataArray (:issue:`3037`).
+  By `Hasan Ahmad <https://github.com/HasanAhmadQ7>`_.
+- Fix error that arises when using open_mfdataset on a series of netcdf files
+  having differing values for a variable attribute of type list. (:issue:`3034`)
+  By `Hasan Ahmad <https://github.com/HasanAhmadQ7>`_.
+
 .. _whats-new.0.12.3:
 
 v0.12.3 (10 July 2019)
@@ -103,8 +107,6 @@ Bug fixes
 - Fix HDF5 error that could arise when reading multiple groups from a file at
   once (:issue:`2954`).
   By `Stephan Hoyer <https://github.com/shoyer>`_.
-- Better error message when using groupby on an empty DataArray (:issue:`3037`).
-  By `Hasan Ahmad <https://github.com/HasanAhmadQ7>`_.
 
 .. _whats-new.0.12.2:
 

xarray/core/utils.py

@@ -129,18 +129,31 @@ def maybe_wrap_array(original, new_array):
 
 def equivalent(first: T, second: T) -> bool:
     """Compare two objects for equivalence (identity or equality), using
-    array_equiv if either object is an ndarray
+    array_equiv if either object is an ndarray. If both objects are lists,
+    equivalent is sequentially called on all the elements.
     """
     # TODO: refactor to avoid circular import
     from . import duck_array_ops
     if isinstance(first, np.ndarray) or isinstance(second, np.ndarray):
         return duck_array_ops.array_equiv(first, second)
+    elif isinstance(first, list) or isinstance(second, list):
+        return list_equiv(first, second)
     else:
         return ((first is second) or
                 (first == second) or
                 (pd.isnull(first) and pd.isnull(second)))
 
 
+def list_equiv(first, second):
+    equiv = True
+    if len(first) != len(second):
+        return False
+    else:
+        for f, s in zip(first, second):
+            equiv = equiv and equivalent(f, s)
+    return equiv
+
+
 def peek_at(iterable: Iterable[T]) -> Tuple[T, Iterator[T]]:
     """Returns the first value from iterable, as well as a new iterator with
     the same content as the original iterable

xarray/tests/test_backends.py

@@ -2361,6 +2361,30 @@ def test_open_mfdataset_manyfiles(readengine, nfiles, parallel, chunks,
             assert_identical(original, actual)
 
 
+@requires_netCDF4
+def test_open_mfdataset_list_attr():
+    """
+    Case when an attribute of type list differs across the multiple files
+    """
+    from netCDF4 import Dataset
+    with create_tmp_files(2) as nfiles:
+        for i in range(2):
+            f = Dataset(nfiles[i], "w")
+            f.createDimension("x", 3)
+            vlvar = f.createVariable("test_var", np.int32, ("x"))
+            # here create an attribute as a list
+            vlvar.test_attr = ["string a {}".format(i),
+                               "string b {}".format(i)]
+            vlvar[:] = np.arange(3)
+            f.close()
+        ds1 = open_dataset(nfiles[0])
+        ds2 = open_dataset(nfiles[1])
+        original = xr.concat([ds1, ds2], dim='x')
+        with xr.open_mfdataset([nfiles[0], nfiles[1]], combine='nested',
+                               concat_dim='x') as actual:
+            assert_identical(actual, original)
+
+
 @requires_scipy_or_netCDF4
 @requires_dask
 class TestOpenMFDatasetWithDataVarsAndCoordsKw: