Merge pull request #213 from jakirkham/support_digest

Add a `digest` method for Zarr Arrays

Author: alimanfoo

docs/api/core.rst

@@ -14,6 +14,7 @@ The Array class (``zarr.core``)
     .. automethod:: set_coordinate_selection
     .. automethod:: get_orthogonal_selection
     .. automethod:: set_orthogonal_selection
+    .. automethod:: digest
     .. automethod:: hexdigest
     .. automethod:: resize
     .. automethod:: append

zarr/core.py

@@ -1,5 +1,6 @@
 # -*- coding: utf-8 -*-
 from __future__ import absolute_import, print_function, division
+import binascii
 import operator
 import itertools
 import hashlib
@@ -88,6 +89,7 @@ class Array(object):
     set_mask_selection
     get_coordinate_selection
     set_coordinate_selection
+    digest
     hexdigest
     resize
     append
@@ -1855,22 +1857,23 @@ def bytestr(n):
 
         return items
 
-    def hexdigest(self, hashname="sha1"):
+    def digest(self, hashname="sha1"):
         """
         Compute a checksum for the data. Default uses sha1 for speed.
 
         Examples
         --------
+        >>> import binascii
         >>> import zarr
         >>> z = zarr.empty(shape=(10000, 10000), chunks=(1000, 1000))
-        >>> z.hexdigest()
-        '041f90bc7a571452af4f850a8ca2c6cddfa8a1ac'
+        >>> binascii.hexlify(z.digest())
+        b'041f90bc7a571452af4f850a8ca2c6cddfa8a1ac'
         >>> z = zarr.zeros(shape=(10000, 10000), chunks=(1000, 1000))
-        >>> z.hexdigest()
-        '7162d416d26a68063b66ed1f30e0a866e4abed60'
+        >>> binascii.hexlify(z.digest())
+        b'7162d416d26a68063b66ed1f30e0a866e4abed60'
         >>> z = zarr.zeros(shape=(10000, 10000), dtype="u1", chunks=(1000, 1000))
-        >>> z.hexdigest()
-        'cb387af37410ae5a3222e893cf3373e4e4f22816'
+        >>> binascii.hexlify(z.digest())
+        b'cb387af37410ae5a3222e893cf3373e4e4f22816'
         """
 
         h = hashlib.new(hashname)
@@ -1882,7 +1885,33 @@ def hexdigest(self, hashname="sha1"):
 
         h.update(self.store.get(self.attrs.key, b""))
 
-        checksum = h.hexdigest()
+        checksum = h.digest()
+
+        return checksum
+
+    def hexdigest(self, hashname="sha1"):
+        """
+        Compute a checksum for the data. Default uses sha1 for speed.
+
+        Examples
+        --------
+        >>> import zarr
+        >>> z = zarr.empty(shape=(10000, 10000), chunks=(1000, 1000))
+        >>> z.hexdigest()
+        '041f90bc7a571452af4f850a8ca2c6cddfa8a1ac'
+        >>> z = zarr.zeros(shape=(10000, 10000), chunks=(1000, 1000))
+        >>> z.hexdigest()
+        '7162d416d26a68063b66ed1f30e0a866e4abed60'
+        >>> z = zarr.zeros(shape=(10000, 10000), dtype="u1", chunks=(1000, 1000))
+        >>> z.hexdigest()
+        'cb387af37410ae5a3222e893cf3373e4e4f22816'
+        """
+
+        checksum = binascii.hexlify(self.digest(hashname=hashname))
+
+        # This is a bytes object on Python 3 and we want a str.
+        if type(checksum) is not str:
+            checksum = checksum.decode('utf8')
 
         return checksum