Find the checksum for data in Zarr

Provides a new `hexdigest` method for Zarr Arrays that creates a
checksum from the array based on its encoded data and its metadata. Uses
the sha1 checksum by default due to its speed and reliability.

Author: jakirkham

zarr/core.py

@@ -2,6 +2,7 @@
 from __future__ import absolute_import, print_function, division
 import operator
 import itertools
+import hashlib
 import re
 
 
@@ -87,6 +88,7 @@ class Array(object):
     set_mask_selection
     get_coordinate_selection
     set_coordinate_selection
+    hexdigest
     resize
     append
     view
@@ -1835,6 +1837,24 @@ def bytestr(n):
 
         return items
 
+    def hexdigest(self, hashname="sha1"):
+        """
+        Compute a checksum for the data. Default uses sha1 for speed.
+        """
+
+        h = hashlib.new(hashname)
+
+        for i in itertools.product(*[range(s) for s in self.cdata_shape]):
+            h.update(self.chunk_store.get(self._chunk_key(i), b""))
+
+        h.update(self.store.get(self._key_prefix + array_meta_key, b""))
+
+        h.update(self.store.get(self.attrs.key, b""))
+
+        checksum = h.hexdigest()
+
+        return checksum
+
     def __getstate__(self):
         return (self._store, self._path, self._read_only, self._chunk_store,
                 self._synchronizer, self._cache_metadata)