chore(mypy): fix mypy lint errors (#540)

* chore(mypy): fix mypy lint errors

* chore: add config for mypy

Author: badmonster0

pyproject.toml

@@ -9,7 +9,12 @@ description = "With CocoIndex, users declare the transformation, CocoIndex creat
 authors = [{ name = "CocoIndex", email = "[email protected]" }]
 readme = "README.md"
 requires-python = ">=3.11"
-dependencies = ["sentence-transformers>=3.3.1", "click>=8.1.8", "rich>=14.0.0", "python-dotenv>=1.1.0"]
+dependencies = [
+    "sentence-transformers>=3.3.1",
+    "click>=8.1.8",
+    "rich>=14.0.0",
+    "python-dotenv>=1.1.0",
+]
 license = "Apache-2.0"
 urls = { Homepage = "https://cocoindex.io/" }
 
@@ -23,4 +28,9 @@ module-name = "cocoindex._engine"
 features = ["pyo3/extension-module"]
 
 [project.optional-dependencies]
-test = ["pytest"]
+test = ["pytest"]
+
+[tool.mypy]
+python_version = "3.11"
+# Disable for not as we have missing type annotations. See https://github.com/cocoindex-io/cocoindex/issues/539
+#   strict = true

python/cocoindex/auth_registry.py

@@ -5,7 +5,7 @@
 from dataclasses import dataclass
 from typing import Generic, TypeVar
 
-from . import _engine
+from . import _engine  # type: ignore
 from .convert import dump_engine_object
 
 T = TypeVar("T")

python/cocoindex/cli.py

@@ -90,6 +90,8 @@ def _load_user_app(app_target: str) -> types.ModuleType:
                 raise ImportError(f"Could not create spec for file: {app_path}")
             module = importlib.util.module_from_spec(spec)
             sys.modules[spec.name] = module 
+            if spec.loader is None:
+                raise ImportError(f"Could not create loader for file: {app_path}")
             spec.loader.exec_module(module)
             return module
         except (ImportError, FileNotFoundError, PermissionError) as e:
@@ -145,20 +147,21 @@ def ls(app_target: str | None):
     If APP_TARGET is omitted, lists all flows that have a persisted
     setup in the backend.
     """
+    persisted_flow_names = flow_names_with_setup()
     if app_target:
         app_ref = _get_app_ref_from_specifier(app_target)
         _load_user_app(app_ref)
 
         current_flow_names = set(flow.flow_names())
-        persisted_flow_names = set(flow_names_with_setup())
 
         if not current_flow_names:
             click.echo(f"No flows are defined in '{app_ref}'.")
             return
 
         has_missing = False
+        persisted_flow_names_set = set(persisted_flow_names)
         for name in sorted(current_flow_names):
-            if name in persisted_flow_names:
+            if name in persisted_flow_names_set:
                 click.echo(name)
             else:
                 click.echo(f"{name} [+]")
@@ -170,13 +173,11 @@ def ls(app_target: str | None):
             click.echo('  [+]: Flows present in the current process, but missing setup.')
 
     else:
-        persisted_flow_names = sorted(flow_names_with_setup())
-
         if not persisted_flow_names:
             click.echo("No persisted flow setups found in the backend.")
             return
 
-        for name in persisted_flow_names:
+        for name in sorted(persisted_flow_names):
             click.echo(name)
 
 @cli.command()

python/cocoindex/convert.py

@@ -7,7 +7,7 @@
 import uuid
 
 from enum import Enum
-from typing import Any, Callable, get_origin
+from typing import Any, Callable, get_origin, Mapping
 from .typing import analyze_type_info, encode_enriched_type, is_namedtuple_type, TABLE_TYPES, KEY_FIELD_NAME
 
 
@@ -104,21 +104,20 @@ def _make_engine_struct_value_decoder(
 
     src_name_to_idx = {f['name']: i for i, f in enumerate(src_fields)}
 
-    is_dataclass = dataclasses.is_dataclass(dst_struct_type)
-    is_namedtuple = is_namedtuple_type(dst_struct_type)
-    
-    if is_dataclass:
+    parameters: Mapping[str, inspect.Parameter]
+    if dataclasses.is_dataclass(dst_struct_type):
         parameters = inspect.signature(dst_struct_type).parameters
-    elif is_namedtuple:
+    elif is_namedtuple_type(dst_struct_type):
         defaults = getattr(dst_struct_type, '_field_defaults', {})
+        fields = getattr(dst_struct_type, '_fields', ())
         parameters = {
             name: inspect.Parameter(
                 name=name,
                 kind=inspect.Parameter.POSITIONAL_OR_KEYWORD,
                 default=defaults.get(name, inspect.Parameter.empty),
                 annotation=dst_struct_type.__annotations__.get(name, inspect.Parameter.empty)
             )
-            for name in dst_struct_type._fields
+            for name in fields
         }
     else:
         raise ValueError(f"Unsupported struct type: {dst_struct_type}")

python/cocoindex/flow.py

@@ -17,7 +17,7 @@
 from rich.text import Text
 from rich.tree import Tree
 
-from . import _engine
+from . import _engine  # type: ignore
 from . import index
 from . import op
 from . import setting

python/cocoindex/lib.py

@@ -4,7 +4,8 @@
 import warnings
 from typing import Callable, Any
 
-from . import _engine, flow, query, setting
+from . import _engine  # type: ignore
+from . import flow, query, setting
 from .convert import dump_engine_object
 
 

python/cocoindex/op.py

@@ -10,7 +10,7 @@
 
 from .typing import encode_enriched_type, resolve_forward_ref
 from .convert import encode_engine_value, make_engine_value_decoder
-from . import _engine
+from . import _engine  # type: ignore
 
 class OpCategory(Enum):
     """The category of the operation."""
@@ -101,7 +101,7 @@ def behavior_version(self):
 
     class _WrappedClass(executor_cls, _Fallback):
         _args_decoders: list[Callable[[Any], Any]]
-        _kwargs_decoders: dict[str, Callable[[str, Any], Any]]
+        _kwargs_decoders: dict[str, Callable[[Any], Any]]
         _acall: Callable
 
         def __init__(self, spec):
@@ -247,13 +247,13 @@ def __call__(self, *args, **kwargs):
                 return fn(*args, **kwargs)
 
         class _Spec(FunctionSpec):
-            pass
+            def __call__(self, *args, **kwargs):
+                return fn(*args, **kwargs)
 
         _Spec.__name__ = op_name
         _Spec.__doc__ = fn.__doc__
         _Spec.__module__ = fn.__module__
         _Spec.__qualname__ = fn.__qualname__
-        _Spec.__wrapped__ = fn
 
         _register_op_factory(
             category=OpCategory.FUNCTION,

python/cocoindex/query.py

@@ -4,7 +4,7 @@
 
 from . import flow as fl
 from . import index
-from . import _engine
+from . import _engine  # type: ignore
 
 _handlers_lock = Lock()
 _handlers: dict[str, _engine.SimpleSemanticsQueryHandler] = {}

python/cocoindex/setup.py

@@ -1,6 +1,6 @@
 from . import flow
 from . import setting
-from . import _engine
+from . import _engine  # type: ignore
 
 def sync_setup() -> _engine.SetupStatus:
     flow.ensure_all_flows_built()

python/cocoindex/typing.py

@@ -5,7 +5,7 @@
 import types
 import inspect
 import uuid
-from typing import Annotated, NamedTuple, Any, TypeVar, TYPE_CHECKING, overload, Sequence, Protocol, Generic, Literal
+from typing import Annotated, NamedTuple, Any, TypeVar, TYPE_CHECKING, overload, Sequence, Generic, Literal, Protocol
 
 class VectorInfo(NamedTuple):
     dim: int | None
@@ -34,8 +34,11 @@ def __init__(self, key: str, value: Any):
     T_co = TypeVar('T_co', covariant=True)
     Dim_co = TypeVar('Dim_co', bound=int, covariant=True)
 
-    class Vector(Sequence[T_co], Generic[T_co, Dim_co], Protocol):
+    class Vector(Protocol, Generic[T_co, Dim_co]):
         """Vector[T, Dim] is a special typing alias for a list[T] with optional dimension info"""
+        def __getitem__(self, index: int) -> T_co: ...
+        def __len__(self) -> int: ...
+
 else:
     class Vector:  # type: ignore[unreachable]
         """ A special typing alias for a list[T] with optional dimension info """