Merge branch 'zarr-developers:main' into main

gorloffslava · web-flow · commit ed3e79e5b772 · 2024-11-21T12:28:42.000+05:00
diff --git a/.github/workflows/ci.yaml b/.github/workflows/ci.yaml
@@ -78,7 +78,7 @@ jobs:
         run: |
           conda activate env
           # TODO: remove --pre option when zarr v3 is out
-          python -m pip install --pre zarr
+          python -m pip install --pre zarr>=3.0.0b2
 
       # This is used to test with zfpy, which does not yet support numpy 2.0
       - name: Install older numpy and zfpy
@@ -100,7 +100,7 @@ jobs:
           conda activate env
           pytest -v
 
-      - uses: codecov/codecov-action@v4
+      - uses: codecov/codecov-action@v5
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           verbose: true
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -15,7 +15,7 @@ repos:
     - id: debug-statements
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.7.2
+    rev: v0.7.4
     hooks:
     - id: ruff
       args: ["--fix", "--show-fixes"]
@@ -25,3 +25,10 @@ repos:
     rev: 2024.08.19
     hooks:
     - id: sp-repo-review
+
+  -   repo: https://github.com/pre-commit/mirrors-mypy
+      rev: 'v1.13.0'
+      hooks:
+      -   id: mypy
+          args: [--config-file, pyproject.toml]
+          additional_dependencies: [numpy, pytest, crc32c, zfpy, 'zarr>=3.0.0b2']
diff --git a/.readthedocs.yaml b/.readthedocs.yaml
@@ -9,7 +9,7 @@ build:
     python: "3.12"
   jobs:
     post_install:
-      - python -m pip install --pre 'zarr'
+      - python -m pip install --pre 'zarr>=3.0.0b2'
 
 sphinx:
   configuration: docs/conf.py
diff --git a/docs/conf.py b/docs/conf.py
@@ -12,7 +12,6 @@
 # All configuration values have a default; values that are commented out
 # serve to show the default.
 
-
 import os
 import sys
 from unittest.mock import Mock as MagicMock
@@ -232,7 +231,7 @@ def __getattr__(cls, name):
 
 # -- Options for LaTeX output ---------------------------------------------
 
-latex_elements = {
+latex_elements: dict[str, str] = {
     # The paper size ('letterpaper' or 'a4paper').
     #'papersize': 'letterpaper',
     # The font size ('10pt', '11pt' or '12pt').
diff --git a/docs/contributing.rst b/docs/contributing.rst
@@ -1,5 +1,5 @@
-Contributing to NumCodecs
-=========================
+Contributing
+============
 
 NumCodecs is a community maintained project. We welcome contributions in the form of bug
 reports, bug fixes, documentation, enhancement proposals and more. This page provides
diff --git a/docs/release.rst b/docs/release.rst
@@ -11,21 +11,38 @@ Release notes
 Unreleased
 ----------
 
-Fix
-~~~
-* Fix in-place mutation of input array in `BitRound`.
-  By :user:`Sam Levang <slevang>`, :issue:`608`
-* Fix an issue where importing numcodecs would lock the state of `multiprocessing`
-  and prevent user code to call `multiprocessing.set_start_method("spawn")`
-  subsequently.
-  By :user:`Clément Robert <neutrinoceros>` :issue:`522`
+Fixes
+~~~~~
+* Cleanup ``crc32c`` soft dependency.
+  By :user:`John Kirkham <jakirkham>`, :issue:`637`
+
+Improvements
+~~~~~~~~~~~~
+* Add `noexcept` to `_utils` C-equiv functions
+  By :user:`John Kirkham <jakirkham>`, :issue:`641`.
+
+.. _release_0.14.0:
+
+0.14.0
+------
 
 Enhancements
 ~~~~~~~~~~~~
 * Add Crc32c checksum codec.
   By :user:`Norman Rzepka <normanrz>`, :issue:`613`.
 * Add codec wrappers for Zarr 3.
   By :user:`Norman Rzepka <normanrz>`, :issue:`524`
+* Added mypy type checking to continuous integration.
+  By :user:`David Stansby <dstansby>`, :issue:`460`.
+
+Fixes
+~~~~~
+* Fix in-place mutation of input array in `BitRound`.
+  By :user:`Sam Levang <slevang>`, :issue:`608`
+* Fix an issue where importing numcodecs would lock the state of `multiprocessing`
+  and prevent user code to call `multiprocessing.set_start_method("spawn")`
+  subsequently.
+  By :user:`Clément Robert <neutrinoceros>` :issue:`522`
 
 Maintenance
 ~~~~~~~~~~~
diff --git a/numcodecs/__init__.py b/numcodecs/__init__.py
@@ -117,13 +117,17 @@
 
     register_codec(MsgPack)
 
-from numcodecs.checksum32 import CRC32, CRC32C, Adler32, JenkinsLookup3
+from numcodecs.checksum32 import CRC32, Adler32, JenkinsLookup3
 
 register_codec(CRC32)
-register_codec(CRC32C)
 register_codec(Adler32)
 register_codec(JenkinsLookup3)
 
+with suppress(ImportError):
+    from numcodecs.checksum32 import CRC32C
+
+    register_codec(CRC32C)
+
 from numcodecs.json import JSON
 
 register_codec(JSON)
diff --git a/numcodecs/_utils.pxd b/numcodecs/_utils.pxd
@@ -8,14 +8,14 @@
 from libc.stdint cimport uint8_t, uint32_t
 
 
-cdef inline void store_le32(uint8_t c[4], uint32_t i) nogil:
+cdef inline void store_le32(uint8_t c[4], uint32_t i) noexcept nogil:
     c[0] = i & 0xFF
     c[1] = (i >> 8) & 0xFF
     c[2] = (i >> 16) & 0xFF
     c[3] = (i >> 24) & 0xFF
 
 
-cdef inline uint32_t load_le32(const uint8_t c[4]) nogil:
+cdef inline uint32_t load_le32(const uint8_t c[4]) noexcept nogil:
     return (
         c[0] |
         (c[1] << 8) |
diff --git a/numcodecs/abc.py b/numcodecs/abc.py
@@ -29,13 +29,14 @@
 """
 
 from abc import ABC, abstractmethod
+from typing import Optional
 
 
 class Codec(ABC):
     """Codec abstract base class."""
 
     # override in sub-class
-    codec_id = None
+    codec_id: Optional[str] = None
     """Codec identifier."""
 
     @abstractmethod
diff --git a/numcodecs/checksum32.py b/numcodecs/checksum32.py
@@ -1,19 +1,29 @@
 import struct
 import zlib
-from typing import Literal
+from collections.abc import Callable
+from contextlib import suppress
+from types import ModuleType
+from typing import TYPE_CHECKING, Literal, Optional
 
 import numpy as np
 
 from .abc import Codec
 from .compat import ensure_contiguous_ndarray, ndarray_copy
 from .jenkins import jenkins_lookup3
 
+_crc32c: Optional[ModuleType] = None
+with suppress(ImportError):
+    import crc32c as _crc32c  # type: ignore[no-redef]
+
+if TYPE_CHECKING:  # pragma: no cover
+    from typing_extensions import Buffer
+
 CHECKSUM_LOCATION = Literal['start', 'end']
 
 
 class Checksum32(Codec):
     # override in sub-class
-    checksum = None
+    checksum: Callable[["Buffer", int], int] | None = None
     location: CHECKSUM_LOCATION = 'start'
 
     def __init__(self, location: CHECKSUM_LOCATION | None = None):
@@ -72,28 +82,6 @@ class CRC32(Checksum32):
     location = 'start'
 
 
-class CRC32C(Checksum32):
-    """Codec add a crc32c checksum to the buffer.
-
-    Parameters
-    ----------
-    location : 'start' or 'end'
-        Where to place the checksum in the buffer.
-    """
-
-    codec_id = 'crc32c'
-
-    def checksum(self, buf):
-        try:
-            from crc32c import crc32c as crc32c_
-
-            return crc32c_(buf)
-        except ImportError:  # pragma: no cover
-            raise ImportError("crc32c must be installed to use the CRC32C checksum codec.")
-
-    location = 'end'
-
-
 class Adler32(Checksum32):
     """Codec add a adler32 checksum to the buffer.
 
@@ -164,3 +152,19 @@ def decode(self, buf, out=None):
             out.view("uint8")[:] = b[:-4]
             return out
         return memoryview(b[:-4])
+
+
+if _crc32c:
+
+    class CRC32C(Checksum32):
+        """Codec add a crc32c checksum to the buffer.
+
+        Parameters
+        ----------
+        location : 'start' or 'end'
+            Where to place the checksum in the buffer.
+        """
+
+        codec_id = 'crc32c'
+        checksum = _crc32c.crc32c  # type: ignore[union-attr]
+        location = 'end'
diff --git a/numcodecs/compat.py b/numcodecs/compat.py
@@ -100,7 +100,7 @@ def ensure_contiguous_ndarray_like(buf, max_buffer_size=None, flatten=True) -> N
 
     # check for datetime or timedelta ndarray, the buffer interface doesn't support those
     if arr.dtype.kind in "Mm":
-        arr = arr.view(np.int64)
+        arr = arr.view(np.int64)  # type: ignore[arg-type]
 
     # check memory is contiguous, if so flatten
     if arr.flags.c_contiguous or arr.flags.f_contiguous:
@@ -117,7 +117,7 @@ def ensure_contiguous_ndarray_like(buf, max_buffer_size=None, flatten=True) -> N
     return arr
 
 
-def ensure_contiguous_ndarray(buf, max_buffer_size=None, flatten=True) -> np.array:
+def ensure_contiguous_ndarray(buf, max_buffer_size=None, flatten=True) -> np.ndarray:
     """Convenience function to coerce `buf` to a numpy array, if it is not already a
     numpy array. Also ensures that the returned value exports fully contiguous memory,
     and supports the new-style buffer interface. If the optional max_buffer_size is
diff --git a/numcodecs/compat_ext.pyx b/numcodecs/compat_ext.pyx
@@ -12,7 +12,7 @@ from .compat import ensure_contiguous_ndarray
 cdef class Buffer:
     """Convenience class for buffer interface."""
 
-    def __cinit__(self, obj, flags):
+    def __cinit__(self, obj, int flags):
         PyObject_GetBuffer(obj, &(self.buffer), flags)
         self.acquired = True
         self.ptr = <char *> self.buffer.buf
diff --git a/numcodecs/lzma.py b/numcodecs/lzma.py
@@ -1,11 +1,14 @@
-import contextlib
+from types import ModuleType
+from typing import Optional
 
-_lzma = None
+_lzma: Optional[ModuleType] = None
 try:
     import lzma as _lzma
 except ImportError:  # pragma: no cover
-    with contextlib.suppress(ImportError):
-        from backports import lzma as _lzma
+    try:
+        from backports import lzma as _lzma  # type: ignore[no-redef]
+    except ImportError:
+        pass
 
 
 if _lzma:
diff --git a/numcodecs/ndarray_like.py b/numcodecs/ndarray_like.py
@@ -1,7 +1,7 @@
 from typing import Any, ClassVar, Optional, Protocol, runtime_checkable
 
 
-class _CachedProtocolMeta(Protocol.__class__):
+class _CachedProtocolMeta(Protocol.__class__):  # type: ignore[name-defined]
     """Custom implementation of @runtime_checkable
 
     The native implementation of @runtime_checkable is slow,
diff --git a/numcodecs/pcodec.py b/numcodecs/pcodec.py
@@ -1,7 +1,6 @@
 from typing import Literal, Optional
 
-import numcodecs
-import numcodecs.abc
+from numcodecs.abc import Codec
 from numcodecs.compat import ensure_contiguous_ndarray
 
 try:
@@ -13,7 +12,7 @@
 DEFAULT_MAX_PAGE_N = 262144
 
 
-class PCodec(numcodecs.abc.Codec):
+class PCodec(Codec):
     """
     PCodec (or pco, pronounced "pico") losslessly compresses and decompresses
     numerical sequences with high compression ratio and fast speed.
diff --git a/numcodecs/registry.py b/numcodecs/registry.py
@@ -2,11 +2,13 @@
 applications to dynamically register and look-up codec classes."""
 
 import logging
-from importlib.metadata import entry_points
+from importlib.metadata import EntryPoints, entry_points
+
+from numcodecs.abc import Codec
 
 logger = logging.getLogger("numcodecs")
-codec_registry = {}
-entries = {}
+codec_registry: dict[str, Codec] = {}
+entries: dict[str, "EntryPoints"] = {}
 
 
 def run_entrypoints():
diff --git a/numcodecs/tests/test_checksum32.py b/numcodecs/tests/test_checksum32.py
diff --git a/numcodecs/tests/test_entrypoints_backport.py b/numcodecs/tests/test_entrypoints_backport.py
diff --git a/numcodecs/tests/test_lzma.py b/numcodecs/tests/test_lzma.py
diff --git a/numcodecs/tests/test_zarr3.py b/numcodecs/tests/test_zarr3.py
diff --git a/numcodecs/tests/test_zfpy.py b/numcodecs/tests/test_zfpy.py
diff --git a/numcodecs/zarr3.py b/numcodecs/zarr3.py
diff --git a/numcodecs/zfpy.py b/numcodecs/zfpy.py
diff --git a/pyproject.toml b/pyproject.toml
