fix(event-handler): preserve metadata constraints in parameter validation (#7609)

* Fixing regression with annotated types

* Fixing regression with annotated types

* Fixing regression with annotated types

Author: leandrodamascena

aws_lambda_powertools/event_handler/openapi/compat.py

@@ -14,6 +14,7 @@
 # We use this for forward reference, as it allows us to handle forward references in type annotations.
 from pydantic._internal._typing_extra import eval_type_lenient
 from pydantic._internal._utils import lenient_issubclass
+from pydantic.fields import FieldInfo as PydanticFieldInfo
 from pydantic_core import PydanticUndefined, PydanticUndefinedType
 from typing_extensions import Annotated, Literal, get_args, get_origin
 
@@ -186,8 +187,36 @@ def model_rebuild(model: type[BaseModel]) -> None:
 def copy_field_info(*, field_info: FieldInfo, annotation: Any) -> FieldInfo:
     # Create a shallow copy of the field_info to preserve its type and all attributes
     new_field = copy(field_info)
-    # Update only the annotation to the new one
-    new_field.annotation = annotation
+
+    # Recursively extract all metadata from nested Annotated types
+    def extract_metadata(ann: Any) -> tuple[Any, list[Any]]:
+        """Extract base type and all non-FieldInfo metadata from potentially nested Annotated types."""
+        if get_origin(ann) is not Annotated:
+            return ann, []
+
+        args = get_args(ann)
+        base_type = args[0]
+        metadata = list(args[1:])
+
+        # If base type is also Annotated, recursively extract its metadata
+        if get_origin(base_type) is Annotated:
+            inner_base, inner_metadata = extract_metadata(base_type)
+            all_metadata = [m for m in inner_metadata + metadata if not isinstance(m, PydanticFieldInfo)]
+            return inner_base, all_metadata
+        else:
+            constraint_metadata = [m for m in metadata if not isinstance(m, PydanticFieldInfo)]
+            return base_type, constraint_metadata
+
+    # Extract base type and constraints
+    base_type, constraints = extract_metadata(annotation)
+
+    # Set the annotation with base type and all constraint metadata
+    # Use tuple unpacking for Python 3.9+ compatibility
+    if constraints:
+        new_field.annotation = Annotated[(base_type, *constraints)]
+    else:
+        new_field.annotation = base_type
+
     return new_field
 
 

aws_lambda_powertools/event_handler/openapi/params.py

@@ -1110,6 +1110,10 @@ def get_field_info_annotated_type(annotation, value, is_path_param: bool) -> tup
     type_annotation = annotated_args[0]
     powertools_annotations = [arg for arg in annotated_args[1:] if isinstance(arg, FieldInfo)]
 
+    # Preserve non-FieldInfo metadata (like annotated_types constraints)
+    # This is important for constraints like Interval, Gt, Lt, etc.
+    other_metadata = [arg for arg in annotated_args[1:] if not isinstance(arg, FieldInfo)]
+
     # Determine which annotation to use
     powertools_annotation: FieldInfo | None = None
     has_discriminator_with_param = False
@@ -1124,6 +1128,11 @@ def get_field_info_annotated_type(annotation, value, is_path_param: bool) -> tup
     else:
         powertools_annotation = next(iter(powertools_annotations), None)
 
+    # Reconstruct type_annotation with non-FieldInfo metadata if present
+    # This ensures constraints like Interval are preserved
+    if other_metadata and not has_discriminator_with_param:
+        type_annotation = Annotated[(type_annotation, *other_metadata)]
+
     # Process the annotation if it exists
     field_info: FieldInfo | None = None
     if isinstance(powertools_annotation, FieldInfo):  # pragma: no cover

tests/functional/event_handler/_pydantic/test_openapi_params.py

@@ -1,7 +1,9 @@
+import json
 from dataclasses import dataclass
 from datetime import datetime
 from typing import List, Optional, Tuple
 
+import pytest
 from pydantic import BaseModel, Field
 from typing_extensions import Annotated
 
@@ -1044,3 +1046,185 @@ def complex_handler(params: Annotated[QueryParams, Query()]):
     assert type_mapping["int_field"] == "integer"
     assert type_mapping["float_field"] == "number"
     assert type_mapping["bool_field"] == "boolean"
+
+
+@pytest.mark.parametrize(
+    "body_value,expected_value",
+    [
+        ("50", 50),  # Valid: within range
+        ("0", 0),  # Valid: at lower bound
+        ("100", 100),  # Valid: at upper bound
+    ],
+)
+def test_annotated_types_interval_constraints_in_body_params(body_value, expected_value):
+    """
+    Test for issue #7600: Validate that annotated_types.Interval constraints
+    are properly enforced in Body parameters with valid values.
+    """
+    from annotated_types import Interval
+
+    # GIVEN an APIGatewayRestResolver with validation enabled
+    app = APIGatewayRestResolver(enable_validation=True)
+
+    # AND a constrained type using annotated_types.Interval
+    ConstrainedInt = Annotated[int, Interval(ge=0, le=100)]
+
+    @app.post("/items")
+    def create_item(value: Annotated[ConstrainedInt, Body()]):
+        return {"value": value}
+
+    # WHEN sending a request with a valid value
+    event = {
+        "resource": "/items",
+        "path": "/items",
+        "httpMethod": "POST",
+        "body": body_value,
+        "isBase64Encoded": False,
+    }
+
+    # THEN the request should succeed
+    result = app(event, {})
+    assert result["statusCode"] == 200
+    body = json.loads(result["body"])
+    assert body["value"] == expected_value
+
+
+@pytest.mark.parametrize(
+    "body_value",
+    [
+        "-1",  # Invalid: below range
+        "101",  # Invalid: above range
+    ],
+)
+def test_annotated_types_interval_constraints_in_body_params_invalid(body_value):
+    """
+    Test for issue #7600: Validate that annotated_types.Interval constraints
+    reject invalid values in Body parameters.
+    """
+    from annotated_types import Interval
+
+    # GIVEN an APIGatewayRestResolver with validation enabled
+    app = APIGatewayRestResolver(enable_validation=True)
+
+    # AND a constrained type using annotated_types.Interval
+    constrained_int = Annotated[int, Interval(ge=0, le=100)]
+
+    @app.post("/items")
+    def create_item(value: Annotated[constrained_int, Body()]):
+        return {"value": value}
+
+    # WHEN sending a request with an invalid value
+    event = {
+        "resource": "/items",
+        "path": "/items",
+        "httpMethod": "POST",
+        "body": body_value,
+        "isBase64Encoded": False,
+    }
+
+    # THEN validation should fail
+    result = app(event, {})
+    assert result["statusCode"] == 422
+
+
+@pytest.mark.parametrize(
+    "query_value,expected_value",
+    [
+        ("50", 50),  # Valid: within range
+        ("0", 0),  # Valid: at lower bound
+        ("100", 100),  # Valid: at upper bound
+    ],
+)
+def test_annotated_types_interval_constraints_in_query_params(query_value, expected_value):
+    """
+    Test for issue #7600: Validate that annotated_types.Interval constraints
+    are properly enforced in Query parameters with valid values.
+    """
+    from annotated_types import Interval
+
+    # GIVEN an APIGatewayRestResolver with validation enabled
+    app = APIGatewayRestResolver(enable_validation=True)
+
+    # AND a constrained type using annotated_types.Interval
+    constrained_int = Annotated[int, Interval(ge=0, le=100)]
+
+    @app.get("/items")
+    def list_items(limit: Annotated[constrained_int, Query()]):
+        return {"limit": limit}
+
+    # WHEN sending a request with a valid value
+    event = {
+        "resource": "/items",
+        "path": "/items",
+        "httpMethod": "GET",
+        "queryStringParameters": {"limit": query_value},
+        "isBase64Encoded": False,
+    }
+
+    # THEN the request should succeed
+    result = app(event, {})
+    assert result["statusCode"] == 200
+    body = json.loads(result["body"])
+    assert body["limit"] == expected_value
+
+
+@pytest.mark.parametrize(
+    "query_value",
+    [
+        "-1",  # Invalid: below range
+        "101",  # Invalid: above range
+    ],
+)
+def test_annotated_types_interval_constraints_in_query_params_invalid(query_value):
+    """
+    Test for issue #7600: Validate that annotated_types.Interval constraints
+    reject invalid values in Query parameters.
+    """
+    from annotated_types import Interval
+
+    # GIVEN an APIGatewayRestResolver with validation enabled
+    app = APIGatewayRestResolver(enable_validation=True)
+
+    # AND a constrained type using annotated_types.Interval
+    constrained_int = Annotated[int, Interval(ge=0, le=100)]
+
+    @app.get("/items")
+    def list_items(limit: Annotated[constrained_int, Query()]):
+        return {"limit": limit}
+
+    # WHEN sending a request with an invalid value
+    event = {
+        "resource": "/items",
+        "path": "/items",
+        "httpMethod": "GET",
+        "queryStringParameters": {"limit": query_value},
+        "isBase64Encoded": False,
+    }
+
+    # THEN validation should fail
+    result = app(event, {})
+    assert result["statusCode"] == 422
+
+
+def test_annotated_types_interval_in_openapi_schema():
+    """
+    Test that annotated_types.Interval constraints are reflected in the OpenAPI schema.
+    """
+    from annotated_types import Interval
+
+    app = APIGatewayRestResolver()
+    constrained_int = Annotated[int, Interval(ge=0, le=100)]
+
+    @app.get("/items")
+    def list_items(limit: Annotated[constrained_int, Query()] = 10):
+        return {"limit": limit}
+
+    schema = app.get_openapi_schema()
+
+    # Verify the Query parameter schema includes constraints
+    get_operation = schema.paths["/items"].get
+    limit_param = next(p for p in get_operation.parameters if p.name == "limit")
+
+    assert limit_param.schema_.type == "integer"
+    assert limit_param.schema_.default == 10
+    assert limit_param.required is False