Merge pull request #87 from labthings/pydantic-2-10

Work with pydantic 2.10

Author: rwb27

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "labthings-fastapi"
-version = "0.0.6"
+version = "0.0.7"
 authors = [
   { name="Richard Bowman", email="richard.bowman@cantab.net" },
 ]

src/labthings_fastapi/thing_description/__init__.py

@@ -15,10 +15,10 @@
 
 from __future__ import annotations
 from collections.abc import Mapping, Sequence
-from typing import Any, Optional, Union
+from typing import Any, Optional
 import json
 
-from pydantic import TypeAdapter, ValidationError, BaseModel
+from pydantic import TypeAdapter, ValidationError
 from .model import DataSchema
 
 
@@ -192,7 +192,7 @@ def jsonschema_to_dataschema(
     return output
 
 
-def type_to_dataschema(t: Union[type, BaseModel], **kwargs) -> DataSchema:
+def type_to_dataschema(t: type, **kwargs) -> DataSchema:
     """Convert a Python type to a Thing Description DataSchema
 
     This makes use of pydantic's `schema_of` function to create a
@@ -205,9 +205,14 @@ def type_to_dataschema(t: Union[type, BaseModel], **kwargs) -> DataSchema:
     is passed in. Typically you'll want to use this for the
     `title` field.
     """
-    if isinstance(t, BaseModel):
+    if hasattr(t, "model_json_schema"):
+        # The input should be a `BaseModel` subclass, in which case this works:
         json_schema = t.model_json_schema()
     else:
+        # In principle, the below should work for any type, though some
+        # deferred annotations can go wrong.
+        # Some attempt at looking up the environment of functions might help
+        # here.
         json_schema = TypeAdapter(t).json_schema()
     schema_dict = jsonschema_to_dataschema(json_schema)
     # Definitions of referenced ($ref) schemas are put in a

src/labthings_fastapi/types/numpy.py

@@ -30,14 +30,15 @@ def double(arr: NDArray) -> NDArray:
     WrapSerializer,
 )
 from typing import Annotated, Any, List, Union
+from typing_extensions import TypeAlias
 from collections.abc import Mapping, Sequence
 
 
 # Define a nested list of floats with 0-6 dimensions
 # This would be most elegantly defined as a recursive type
 # but the below gets the job done for now.
-Number = Union[int, float]
-NestedListOfNumbers = Union[
+Number: TypeAlias = Union[int, float]
+NestedListOfNumbers: TypeAlias = Union[
     Number,
     List[Number],
     List[List[Number]],
@@ -68,10 +69,12 @@ def listoflists_to_np(lol: Union[NestedListOfNumbers, np.ndarray]) -> np.ndarray
 
 
 # Define an annotated type so Pydantic can cope with numpy
-NDArray = Annotated[
+NDArray: TypeAlias = Annotated[
     np.ndarray,
     PlainValidator(listoflists_to_np),
-    PlainSerializer(np_to_listoflists, when_used="json-unless-none"),
+    PlainSerializer(
+        np_to_listoflists, when_used="json-unless-none", return_type=NestedListOfNumbers
+    ),
     WithJsonSchema(NestedListOfNumbersModel.model_json_schema(), mode="validation"),
 ]
 

tests/test_numpy_type.py

@@ -1,13 +1,17 @@
 from __future__ import annotations
 
-from pydantic import BaseModel
+from pydantic import BaseModel, RootModel
 import numpy as np
 
 from labthings_fastapi.types.numpy import NDArray, DenumpifyingDict
 from labthings_fastapi.thing import Thing
 from labthings_fastapi.decorators import thing_action
 
 
+class ArrayModel(RootModel):
+    root: NDArray
+
+
 def check_field_works_with_list(data):
     class Model(BaseModel):
         a: NDArray
@@ -86,3 +90,10 @@ def test_denumpifying_dict():
     assert dump["e"] is None
     assert dump["f"] == 1
     d.model_dump_json()
+
+
+def test_rootmodel():
+    for input in [[0, 1, 2], np.arange(3)]:
+        m = ArrayModel(root=input)
+        assert isinstance(m.root, np.ndarray)
+        assert (m.model_dump() == [0, 1, 2]).all()