chore: use pyright as static type checker (#227)

* use pyright as static type checker

* make pyright happy

Author: zen-xu

.github/workflows/lint.yaml

@@ -10,12 +10,28 @@ on:
       - synchronize
 
 jobs:
-  taplo-lint:
-    name: taplo lint
+  taplo:
+    name: taplo
     runs-on: ubuntu-22.04
     steps:
       - uses: actions/checkout@v4
       - uses: uncenter/setup-taplo@v1
         with:
           version: "0.9.3"
       - run: taplo fmt --check
+
+  pyright:
+    name: pyright
+    runs-on: ubuntu-22.04
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v4
+        with:
+          python-version: "3.11"
+          cache: pyright
+      - run: |
+          python -m venv .venv
+          source .venv/bin/activate
+          pip install pandas numpy scipy sparse
+      - run: echo "$PWD/.venv/bin" >> $GITHUB_PATH
+      - uses: jakebailey/pyright-action@v2

.pre-commit-config.yaml

@@ -1,8 +1,10 @@
 ci:
   autofix_prs: false
+  skip: [pyright]
 
 default_language_version:
   python: python3.11
+  node: 23.9.0
 
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
@@ -25,7 +27,7 @@ repos:
         args: [--fix]
       - id: ruff-format
 
-  - repo: https://github.com/pre-commit/mirrors-mypy
-    rev: v1.15.0
+  - repo: https://github.com/RobertCraigie/pyright-python
+    rev: v1.1.400
     hooks:
-      - id: mypy
+      - id: pyright

pixi.lock

@@ -1,5 +1,3 @@
-# mypy: disable-error-code="overload-overlap,misc,type-arg"
-
 import datetime as dt
 import sys
 
@@ -63,7 +61,6 @@ from .types import (
     _IndexT,
     _RunEndType,
     _Size,
-    _ValueT,
 )
 
 _T = TypeVar("_T")
@@ -1627,25 +1624,25 @@ class RunEndEncodedArray(Array[scalar.RunEndEncodedScalar[_RunEndType, _BasicVal
     def from_arrays(
         run_ends: Int16Array,
         values: Array,
-        type: _ValueT | None = None,
+        type: DataType | None = None,
     ) -> RunEndEncodedArray[types.Int16Type, _BasicValueT]: ...
     @overload
     @staticmethod
     def from_arrays(
         run_ends: Int32Array,
         values: Array,
-        type: _ValueT | None = None,
+        type: DataType | None = None,
     ) -> RunEndEncodedArray[types.Int32Type, _BasicValueT]: ...
     @overload
     @staticmethod
     def from_arrays(
         run_ends: Int64Array,
         values: Array,
-        type: _ValueT | None = None,
+        type: DataType | None = None,
     ) -> RunEndEncodedArray[types.Int64Type, _BasicValueT]: ...
     @staticmethod
-    def from_buffers(  # type: ignore[override]
-        type: _ValueT,
+    def from_buffers(
+        type: DataType,
         length: int,
         buffers: list[Buffer],
         null_count: int = -1,

pyarrow-stubs/__lib_pxi/array.pyi

@@ -1,4 +1,3 @@
-# mypy: disable-error-code="name-defined"
 from types import ModuleType
 from typing import Any, Iterable, TypeGuard
 

pyarrow-stubs/__lib_pxi/pandas_shim.pyi

@@ -1,4 +1,3 @@
-# mypy: disable-error-code="overload-overlap,misc,type-arg"
 import collections.abc
 import datetime as dt
 import sys

pyarrow-stubs/__lib_pxi/scalar.pyi

@@ -1,5 +1,3 @@
-# mypy: disable-error-code="overload-overlap,type-arg,misc"
-
 import datetime as dt
 import sys
 

pyarrow-stubs/__lib_pxi/table.pyi

@@ -1,5 +1,3 @@
-# mypy: disable-error-code="import-untyped"
-
 import sys
 
 if sys.version_info >= (3, 11):
@@ -10,8 +8,8 @@ else:
 import numpy as np
 
 from pyarrow.lib import _Weakrefable
-from pydata.sparse import COO  # type: ignore[import-not-found]
 from scipy.sparse import coo_matrix, csr_matrix
+from sparse import COO
 
 class Tensor(_Weakrefable):
     @classmethod

pyarrow-stubs/__lib_pxi/tensor.pyi

@@ -29,7 +29,7 @@ Compression: TypeAlias = Literal[
 NullEncoding: TypeAlias = Literal["mask", "encode"]
 NullSelectionBehavior: TypeAlias = Literal["drop", "emit_null"]
 Mask: TypeAlias = Sequence[bool | None] | NDArray[np.bool_] | BooleanArray
-Indices: TypeAlias = Sequence[int] | NDArray[np.integer] | IntegerArray
+Indices: TypeAlias = Sequence[int] | NDArray[np.integer[Any]] | IntegerArray
 PyScalar: TypeAlias = (
     bool | int | float | Decimal | str | bytes | dt.date | dt.datetime | dt.time | dt.timedelta
 )

pyarrow-stubs/_stubs_typing.pyi

@@ -1,4 +1,3 @@
-# mypy: disable-error-code="misc,type-var,var-annotated"
 # ruff: noqa: I001
 from typing import Literal, TypeAlias, TypeVar, overload, Any, Iterable, ParamSpec, Sequence
 from collections.abc import Callable

pyarrow-stubs/compute.pyi

@@ -48,12 +48,12 @@ pyarrow-stubs = { path = ".", editable = true }
 ipython       = "*"
 scipy         = "*"
 pre-commit    = "*"
-mypy          = ">=1.11"
 ruff          = ">=0.5"
 types-cffi    = "*"
 pandas-stubs  = "*"
 hatchling     = "*"
 fsspec        = "*"
+sparse        = "*"
 pyright       = { version = ">=1.1.385,<2", extras = ["nodejs"] }
 
 [tool.pixi.tasks]
@@ -85,9 +85,6 @@ lines-between-types = 1
 [tool.ruff.format]
 docstring-code-format = true
 
-[tool.mypy]
-explicit_package_bases = true
-files                  = "pyarrow-stubs"
-namespace_packages     = true
-show_error_codes       = true
-disable_error_code     = ["overload-overlap", "import-not-found"]
+[tool.pyright]
+typeCheckingMode     = "basic"
+reportMissingImports = false