preserve filters, order, fill_value on copy zarr-to-zarr

Author: alimanfoo

zarr/convenience.py

@@ -793,19 +793,25 @@ def _copy(log, source, dest, name, root, shallow, without_attrs, if_exists,
                         kws.setdefault('compression', source.compression)
                         kws.setdefault('compression_opts', source.compression_opts)
                         kws.setdefault('shuffle', source.shuffle)
+                        kws.setdefault('fletcher32', source.fletcher32)
+                        kws.setdefault('fillvalue', source.fillvalue)
                     else:
                         # h5py -> zarr; use zarr default compression options
-                        pass
+                        kws.setdefault('fill_value', source.fillvalue)
                 else:
                     if dest_h5py:
                         # zarr -> h5py; use some vaguely sensible defaults
                         kws.setdefault('chunks', True)
                         kws.setdefault('compression', 'gzip')
                         kws.setdefault('compression_opts', 1)
                         kws.setdefault('shuffle', False)
+                        kws.setdefault('fillvalue', source.fill_value)
                     else:
                         # zarr -> zarr; preserve compression options by default
                         kws.setdefault('compressor', source.compressor)
+                        kws.setdefault('filters', source.filters)
+                        kws.setdefault('order', source.order)
+                        kws.setdefault('fill_value', source.fill_value)
 
                 # create new dataset in destination
                 ds = dest.create_dataset(name, shape=source.shape,

zarr/tests/test_convenience.py

@@ -315,8 +315,14 @@ def setUp(self):
         foo.attrs['experiment'] = 'weird science'
         baz = foo.create_dataset('bar/baz', data=np.arange(100), chunks=(50,))
         baz.attrs['units'] = 'metres'
+        if self.source_h5py:
+            extra_kws = dict(compression='gzip', compression_opts=3, fillvalue=84,
+                             shuffle=True, fletcher32=True)
+        else:
+            extra_kws = dict(compressor=Zlib(3), order='F', fill_value=42,
+                             filters=[Adler32()])
         source.create_dataset('spam', data=np.arange(100, 200).reshape(20, 5),
-                              chunks=(10, 2))
+                              chunks=(10, 2), dtype='i2', **extra_kws)
         self.source = source
 
     def test_copy_array(self):
@@ -326,6 +332,8 @@ def test_copy_array(self):
         # copy array with default options
         copy(source['foo/bar/baz'], dest)
         check_copied_array(source['foo/bar/baz'], dest['baz'])
+        copy(source['spam'], dest)
+        check_copied_array(source['spam'], dest['spam'])
 
     def test_copy_array_name(self):
         source = self.source