Merge branch 'main' into fix/3050

Author: d-v-b

.pre-commit-config.yaml

@@ -6,7 +6,7 @@ ci:
 default_stages: [pre-commit, pre-push]
 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.9.9
+    rev: v0.11.9
     hooks:
       - id: ruff
         args: ["--fix", "--show-fixes"]
@@ -40,7 +40,7 @@ repos:
           - hypothesis
           - s3fs
   - repo: https://github.com/scientific-python/cookie
-    rev: 2025.01.22
+    rev: 2025.05.02
     hooks:
       - id: sp-repo-review
   - repo: https://github.com/pre-commit/pygrep-hooks

changes/2913.feature.rst

@@ -0,0 +1 @@
+Added a `print_debug_info` function for bug reports.

changes/3049.misc.rst

@@ -0,0 +1 @@
+Added tests for ``AsyncArray``, ``Array`` and removed duplicate argument parsing.

changes/3062.bugfix.rst

@@ -0,0 +1 @@
+Using various functions to open data with ``mode='a'`` no longer deletes existing data in the store.

pyproject.toml

@@ -310,7 +310,7 @@ extend-select = [
     "RUF",
     "SIM",  # flake8-simplify
     "SLOT", # flake8-slots
-    "TCH",  # flake8-type-checking
+    "TC",  # flake8-type-checking
     "TRY",  # tryceratops
     "UP",   # pyupgrade
     "W",    # pycodestyle warnings
@@ -338,6 +338,7 @@ ignore = [
     "Q003",
     "COM812",
     "COM819",
+    "TC006",
 ]
 
 [tool.ruff.lint.extend-per-file-ignores]

src/zarr/__init__.py

@@ -37,6 +37,54 @@
 # in case setuptools scm screw up and find version to be 0.0.0
 assert not __version__.startswith("0.0.0")
 
+
+def print_debug_info() -> None:
+    """
+    Print version info for use in bug reports.
+    """
+    import platform
+    from importlib.metadata import version
+
+    def print_packages(packages: list[str]) -> None:
+        not_installed = []
+        for package in packages:
+            try:
+                print(f"{package}: {version(package)}")
+            except ModuleNotFoundError:
+                not_installed.append(package)
+        if not_installed:
+            print("\n**Not Installed:**")
+            for package in not_installed:
+                print(package)
+
+    required = [
+        "packaging",
+        "numpy",
+        "numcodecs",
+        "typing_extensions",
+        "donfig",
+    ]
+    optional = [
+        "botocore",
+        "cupy-cuda12x",
+        "fsspec",
+        "numcodecs",
+        "s3fs",
+        "gcsfs",
+        "universal-pathlib",
+        "rich",
+        "obstore",
+    ]
+
+    print(f"platform: {platform.platform()}")
+    print(f"python: {platform.python_version()}")
+    print(f"zarr: {__version__}\n")
+    print("**Required dependencies:**")
+    print_packages(required)
+    print("\n**Optional dependencies:**")
+    print_packages(optional)
+
+
 __all__ = [
     "Array",
     "AsyncArray",
@@ -67,6 +115,7 @@
     "open_consolidated",
     "open_group",
     "open_like",
+    "print_debug_info",
     "save",
     "save_array",
     "save_group",

src/zarr/api/asynchronous.py

@@ -88,7 +88,7 @@
 
 _READ_MODES: tuple[AccessModeLiteral, ...] = ("r", "r+", "a")
 _CREATE_MODES: tuple[AccessModeLiteral, ...] = ("a", "w", "w-")
-_OVERWRITE_MODES: tuple[AccessModeLiteral, ...] = ("a", "r+", "w")
+_OVERWRITE_MODES: tuple[AccessModeLiteral, ...] = ("w",)
 
 
 def _infer_overwrite(mode: AccessModeLiteral) -> bool:
@@ -817,7 +817,6 @@ async def open_group(
         warnings.warn("chunk_store is not yet implemented", RuntimeWarning, stacklevel=2)
 
     store_path = await make_store_path(store, mode=mode, storage_options=storage_options, path=path)
-
     if attributes is None:
         attributes = {}
 
@@ -1019,11 +1018,6 @@ async def create(
         warnings.warn("object_codec is not yet implemented", RuntimeWarning, stacklevel=2)
     if read_only is not None:
         warnings.warn("read_only is not yet implemented", RuntimeWarning, stacklevel=2)
-    if dimension_separator is not None and zarr_format == 3:
-        raise ValueError(
-            "dimension_separator is not supported for zarr format 3, use chunk_key_encoding instead"
-        )
-
     if order is not None:
         _warn_order_kwarg()
     if write_empty_chunks is not None:

src/zarr/api/synchronous.py

@@ -858,6 +858,7 @@ def create_array(
         Ignored otherwise.
     overwrite : bool, default False
         Whether to overwrite an array with the same name in the store, if one exists.
+        If `True`, all existing paths in the store will be deleted.
     config : ArrayConfigLike, optional
         Runtime configuration for the array.
     write_data : bool

src/zarr/core/array.py

@@ -140,7 +140,8 @@ def parse_array_metadata(data: Any) -> ArrayMetadata:
     if isinstance(data, ArrayMetadata):
         return data
     elif isinstance(data, dict):
-        if data["zarr_format"] == 3:
+        zarr_format = data.get("zarr_format")
+        if zarr_format == 3:
             meta_out = ArrayV3Metadata.from_dict(data)
             if len(meta_out.storage_transformers) > 0:
                 msg = (
@@ -149,9 +150,11 @@ def parse_array_metadata(data: Any) -> ArrayMetadata:
                 )
                 raise ValueError(msg)
             return meta_out
-        elif data["zarr_format"] == 2:
+        elif zarr_format == 2:
             return ArrayV2Metadata.from_dict(data)
-    raise TypeError
+        else:
+            raise ValueError(f"Invalid zarr_format: {zarr_format}. Expected 2 or 3")
+    raise TypeError  # pragma: no cover
 
 
 def create_codec_pipeline(metadata: ArrayMetadata) -> CodecPipeline:
@@ -160,8 +163,7 @@ def create_codec_pipeline(metadata: ArrayMetadata) -> CodecPipeline:
     elif isinstance(metadata, ArrayV2Metadata):
         v2_codec = V2Codec(filters=metadata.filters, compressor=metadata.compressor)
         return get_pipeline_class().from_codecs([v2_codec])
-    else:
-        raise TypeError
+    raise TypeError  # pragma: no cover
 
 
 async def get_array_metadata(
@@ -268,17 +270,6 @@ def __init__(
         store_path: StorePath,
         config: ArrayConfigLike | None = None,
     ) -> None:
-        if isinstance(metadata, dict):
-            zarr_format = metadata["zarr_format"]
-            # TODO: remove this when we extensively type the dict representation of metadata
-            _metadata = cast(dict[str, JSON], metadata)
-            if zarr_format == 2:
-                metadata = ArrayV2Metadata.from_dict(_metadata)
-            elif zarr_format == 3:
-                metadata = ArrayV3Metadata.from_dict(_metadata)
-            else:
-                raise ValueError(f"Invalid zarr_format: {zarr_format}. Expected 2 or 3")
-
         metadata_parsed = parse_array_metadata(metadata)
         config_parsed = parse_array_config(config)
 

src/zarr/testing/stateful.py

@@ -17,6 +17,7 @@
 import zarr
 from zarr import Array
 from zarr.abc.store import Store
+from zarr.codecs.bytes import BytesCodec
 from zarr.core.buffer import Buffer, BufferPrototype, cpu, default_buffer_prototype
 from zarr.core.sync import SyncMixin
 from zarr.storage import LocalStore, MemoryStore
@@ -108,7 +109,15 @@ def add_array(
         assume(self.can_add(path))
         note(f"Adding array:  path='{path}'  shape={array.shape}  chunks={chunks}")
         for store in [self.store, self.model]:
-            zarr.array(array, chunks=chunks, path=path, store=store, fill_value=fill_value)
+            zarr.array(
+                array,
+                chunks=chunks,
+                path=path,
+                store=store,
+                fill_value=fill_value,
+                # Chose bytes codec to avoid wasting time compressing the data being written
+                codecs=[BytesCodec()],
+            )
         self.all_arrays.add(path)
 
     # @precondition(lambda self: bool(self.all_groups))