Recognize .b2b extension as BatchStore in DictStore

Author: FrancescAlted

src/blosc2/dict_store.py

@@ -36,6 +36,8 @@ class DictStore:
       are stored as .b2nd files.
     - blosc2.SChunk: super-chunks. When persisted externally they are stored
       as .b2f files.
+    - blosc2.BatchStore: batched variable-length containers. When persisted
+      externally they are stored as .b2b files.
     - blosc2.C2Array: columnar containers. These are always kept inside the
       embedded store (never externalized).
     - numpy.ndarray: converted to blosc2.NDArray on assignment.
@@ -91,7 +93,7 @@ class DictStore:
     Notes
     -----
     - External persistence uses the following file extensions:
-      .b2nd for NDArray and .b2f for SChunk.
+      .b2nd for NDArray, .b2f for SChunk, and .b2b for BatchStore.
     """
 
     def __init__(
@@ -181,8 +183,11 @@ def _init_read_mode(self, dparams: blosc2.DParams | None = None):
                 dparams=dparams,
             )
             for filepath in self.offsets:
-                if filepath.endswith((".b2nd", ".b2f")):
-                    key = "/" + filepath[: -5 if filepath.endswith(".b2nd") else -4]
+                if filepath.endswith((".b2nd", ".b2f", ".b2b")):
+                    if filepath.endswith(".b2nd"):
+                        key = "/" + filepath[:-5]
+                    else:
+                        key = "/" + filepath[:-4]
                     self.map_tree[key] = filepath
         else:  # .b2d
             if not os.path.isdir(self.localpath):
@@ -228,14 +233,14 @@ def _update_map_tree(self):
         for root, _, files in os.walk(self.working_dir):
             for file in files:
                 filepath = os.path.join(root, file)
-                if filepath.endswith((".b2nd", ".b2f")):
+                if filepath.endswith((".b2nd", ".b2f", ".b2b")):
                     # Convert filename to key: remove extension and ensure starts with /
                     rel_path = os.path.relpath(filepath, self.working_dir)
                     # Normalize path separators to forward slashes for cross-platform consistency
                     rel_path = rel_path.replace(os.sep, "/")
                     if rel_path.endswith(".b2nd"):
                         key = rel_path[:-5]
-                    elif rel_path.endswith(".b2f"):
+                    elif rel_path.endswith(".b2b") or rel_path.endswith(".b2f"):
                         key = rel_path[:-4]
                     else:
                         continue
@@ -264,6 +269,8 @@ def _is_external_value(value: blosc2.Array | SChunk | blosc2.VLArray | blosc2.Ba
     def _external_ext(value: blosc2.Array | SChunk | blosc2.VLArray | blosc2.BatchStore) -> str:
         if isinstance(value, blosc2.NDArray):
             return ".b2nd"
+        if isinstance(value, blosc2.BatchStore):
+            return ".b2b"
         return ".b2f"
 
     def __setitem__(

src/blosc2/tree_store.py

@@ -668,8 +668,15 @@ def _persist_vlmeta(self) -> None:
         """
         if hasattr(self, "_vlmeta_key"):
             vlmeta_key = self._vlmeta_key
-            # Only embedded case is expected; handle it safely.
-            if hasattr(self, "_estore") and vlmeta_key in self._estore:
+            if vlmeta_key in self.map_tree:
+                filepath = self.map_tree[vlmeta_key]
+                dest_path = os.path.join(self.working_dir, filepath)
+                parent_dir = os.path.dirname(dest_path)
+                if parent_dir and not os.path.exists(parent_dir):
+                    os.makedirs(parent_dir, exist_ok=True)
+                with open(dest_path, "wb") as f:
+                    f.write(self._vlmeta.to_cframe())
+            elif hasattr(self, "_estore") and vlmeta_key in self._estore:
                 # Replace the stored snapshot
                 with contextlib.suppress(KeyError):
                     del self._estore[vlmeta_key]

tests/test_tree_store.py

@@ -654,6 +654,34 @@ def test_external_vlarray_support():
     os.remove("test_vlarray_external.b2z")
 
 
+def test_external_batchstore_support(tmp_path):
+    store_path = tmp_path / "test_batchstore_external.b2d"
+
+    with TreeStore(str(store_path), mode="w", threshold=0) as tstore:
+        bstore = blosc2.BatchStore(blocksize_max=2)
+        bstore.extend([[{"id": 1}, {"id": 2}], [{"id": 3}]])
+        tstore["/data/batchstore"] = bstore
+
+        batchstore_path = store_path / "data" / "batchstore.b2b"
+        assert batchstore_path.exists()
+
+    with TreeStore(str(store_path), mode="r") as tstore:
+        retrieved = tstore["/data/batchstore"]
+        assert isinstance(retrieved, blosc2.BatchStore)
+        assert [batch[:] for batch in retrieved] == [[{"id": 1}, {"id": 2}], [{"id": 3}]]
+
+
+def test_treestore_vlmeta_externalized_b2d(tmp_path):
+    store_path = tmp_path / "test_vlmeta_externalized.b2d"
+
+    with TreeStore(str(store_path), mode="w", threshold=0) as tstore:
+        tstore["/data"] = np.array([1, 2, 3])
+        tstore.vlmeta["schema_manifest"] = {"version": 1, "fields": {"a": {"kind": "fixed"}}}
+
+    with TreeStore(str(store_path), mode="r") as tstore:
+        assert tstore.vlmeta["schema_manifest"] == {"version": 1, "fields": {"a": {"kind": "fixed"}}}
+
+
 def test_walk_topdown_argument_ordering():
     """Ensure walk supports topdown argument mimicking os.walk order semantics."""
     with TreeStore("test_walk_topdown.b2z", mode="w") as tstore: