Add name kwarg to from_zarr (#1)

continuous_integration/appveyor/setup_conda_environment.cmd

@@ -26,7 +26,7 @@ copy NUL %CONDA_PREFIX%\conda-meta\pinned
 @rem Install optional dependencies for tests
 %CONDA_INSTALL% numpy pandas cloudpickle distributed
 %CONDA_INSTALL% bcolz bokeh h5py ipython lz4 psutil pytables s3fs scipy
-%CONDA_INSTALL% -c conda-forge fastparquet snappy
+%CONDA_INSTALL% -c conda-forge fastparquet snappy zarr
 
 %PIP_INSTALL% --no-deps --upgrade locket git+https://github.com/dask/partd
 %PIP_INSTALL% --no-deps --upgrade heapdict git+https://github.com/dask/cachey

dask/array/core.py

@@ -2149,7 +2149,7 @@ def from_array(x, chunks, name=None, lock=False, asarray=True, fancy=True,
     return Array(dsk, name, chunks, dtype=x.dtype)
 
 
-def from_zarr(url, component=None, storage_options=None, chunks=None, **kwargs):
+def from_zarr(url, component=None, storage_options=None, chunks=None,name=None, **kwargs):
     """Load array from the zarr storage format
 
     See https://zarr.readthedocs.io for details about the format.
@@ -2170,6 +2170,8 @@ def from_zarr(url, component=None, storage_options=None, chunks=None, **kwargs):
         Passed to ``da.from_array``, allows setting the chunks on
         initialisation, if the chunking scheme in the on-disc dataset is not
         optimal for the calculations to follow.
+    name : str, optional
+         An optional keyname for the dataframe.  Defaults to hashing the input
     kwargs: passed to ``zarr.Array``.
     """
     import zarr
@@ -2186,7 +2188,9 @@ def from_zarr(url, component=None, storage_options=None, chunks=None, **kwargs):
         mapper = url
         z = zarr.Array(mapper, read_only=True, path=component, **kwargs)
     chunks = chunks if chunks is not None else z.chunks
-    return from_array(z, chunks, name='zarr-%s' % url)
+    if name is None:
+        name = 'from-zarr-' + tokenize(z, component, storage_options, chunks, **kwargs)
+    return from_array(z, chunks, name=name )
 
 
 def to_zarr(arr, url, component=None, storage_options=None,

dask/array/tests/test_array_core.py

@@ -3437,6 +3437,20 @@ def test_normalize_chunks_nan():
     assert "auto" in str(info.value)
 
 
+def test_from_zarr_unique_name():
+    zarr = pytest.importorskip('zarr')
+    a = zarr.array([1, 2, 3])
+    b = zarr.array([4, 5, 6])
+
+    assert da.from_zarr(a).name != da.from_zarr(b).name
+
+
+def test_from_zarr_name():
+    zarr = pytest.importorskip('zarr')
+    a = zarr.array([1, 2, 3])
+    assert da.from_zarr(a, name='foo').name == 'foo'
+
+
 def test_zarr_roundtrip():
     pytest.importorskip('zarr')
     with tmpdir() as d: