Modify getsize to return total size, not just the top level

Author: benjeffery

docs/release.rst

@@ -24,6 +24,13 @@ Enhancements
   By :user:`Deepak Cherian <dcherian>`.
 
 
+Bug fixes
+~~~~~~~~~
+
+* ``getsize`` now returns the total size of all nested arrays. 
+  By :user:`Ben Jeffery <benjeffery>` :issue:`253`.
+
+
 Docs
 ~~~~
 

zarr/storage.py

@@ -30,7 +30,6 @@
 from collections import OrderedDict
 from collections.abc import MutableMapping
 from functools import lru_cache
-from os import scandir
 from pickle import PicklingError
 from threading import Lock, RLock
 from typing import Sequence, Mapping, Optional, Union, List, Tuple, Dict, Any
@@ -269,9 +268,15 @@ def _getsize(store: BaseStore, path: Path = None) -> int:
             # also include zarr.json?
             # members += ['zarr.json']
         else:
-            members = listdir(store, path)
-            prefix = _path_to_prefix(path)
-            members = [prefix + k for k in members]
+            to_visit = [path]
+            members = []
+            while to_visit:
+                print(to_visit)
+                current_path = to_visit.pop()
+                current_members = listdir(store, current_path)
+                prefix = _path_to_prefix(current_path)
+                members.extend([prefix + k for k in current_members])
+                to_visit.extend([prefix + k for k in current_members])
         for k in members:
             try:
                 v = store[k]
@@ -971,8 +976,12 @@ def getsize(self, path: Path = None):
         elif isinstance(value, self.cls):
             # total size for directory
             size = 0
-            for v in value.values():
-                if not isinstance(v, self.cls):
+            to_visit = list(value.values())
+            while to_visit:
+                v = to_visit.pop()
+                if isinstance(v, self.cls):
+                    to_visit.extend(v.values())
+                else:
                     size += buffer_size(v)
             return size
 
@@ -1269,9 +1278,10 @@ def getsize(self, path=None):
             return os.path.getsize(fs_path)
         elif os.path.isdir(fs_path):
             size = 0
-            for child in scandir(fs_path):
-                if child.is_file():
-                    size += child.stat().st_size
+            for root, dirs, files in os.walk(fs_path):
+                for file in files:
+                    file_path = os.path.join(root, file)
+                    size += os.path.getsize(file_path)
             return size
         else:
             return 0
@@ -1903,29 +1913,19 @@ def listdir(self, path=None):
     def getsize(self, path=None):
         path = normalize_storage_path(path)
         with self.mutex:
-            children = self.listdir(path)
-            if children:
-                size = 0
-                for child in children:
-                    if path:
-                        name = path + "/" + child
-                    else:
-                        name = child
-                    try:
-                        info = self.zf.getinfo(name)
-                    except KeyError:
-                        pass
-                    else:
-                        size += info.compress_size
-                return size
-            elif path:
+            to_visit = [path] if path else self.listdir(path)
+            total_size = 0
+            while to_visit:
+                current_path = to_visit.pop()
                 try:
-                    info = self.zf.getinfo(path)
-                    return info.compress_size
+                    info = self.zf.getinfo(current_path)
+                    total_size += info.compress_size
                 except KeyError:
-                    return 0
-            else:
-                return 0
+                    children = self.listdir(current_path)
+                    for child in children:
+                        full_path = current_path + "/" + child if current_path else child
+                        to_visit.append(full_path)
+            return total_size
 
     def clear(self):
         if self.mode == "r":
@@ -2488,6 +2488,8 @@ def listdir(self, path: Path = None):
                 return listing
 
     def getsize(self, path=None) -> int:
+        print("WYF")
+        print(self._store, path)
         return getsize(self._store, path=path)
 
     def _pop_value(self):
@@ -2745,10 +2747,9 @@ def getsize(self, path=None):
         size = self.cursor.execute(
             """
             SELECT COALESCE(SUM(LENGTH(v)), 0) FROM zarr
-            WHERE k LIKE (? || "%") AND
-                  0 == INSTR(LTRIM(SUBSTR(k, LENGTH(?) + 1), "/"), "/")
+            WHERE k LIKE (? || "%")
             """,
-            (path, path),
+            (path,),
         )
         for (s,) in size:
             return s

zarr/tests/test_storage.py

@@ -366,19 +366,10 @@ def test_hierarchy(self):
 
         # test getsize (optional)
         if hasattr(store, "getsize"):
-            # TODO: proper behavior of getsize?
-            #       v3 returns size of all nested arrays, not just the
-            #       size of the arrays in the current folder.
-            if self.version == 2:
-                assert 6 == store.getsize()
-            else:
-                assert 15 == store.getsize()
+            assert 15 == store.getsize()
             assert 3 == store.getsize("a")
             assert 3 == store.getsize("b")
-            if self.version == 2:
-                assert 3 == store.getsize("c")
-            else:
-                assert 9 == store.getsize("c")
+            assert 9 == store.getsize("c")
             assert 3 == store.getsize("c/d")
             assert 6 == store.getsize("c/e")
             assert 3 == store.getsize("c/e/f")
@@ -2256,7 +2247,7 @@ def test_getsize():
     store["foo"] = b"aaa"
     store["bar"] = b"bbbb"
     store["baz/quux"] = b"ccccc"
-    assert 7 == getsize(store)
+    assert 12 == getsize(store)
     assert 5 == getsize(store, "baz")
 
     store = KVStore(dict())