[DRAFT] [REFACTOR] Upgrade Pydantic to v2

argilla-server/pyproject.toml

@@ -21,7 +21,8 @@ maintainers = [{ name = "argilla", email = "[email protected]" }]
 dependencies = [
     # Basic dependencies
     "fastapi >= 0.103.1,< 1.0.0",
-    "pydantic >= 1.10.7,< 2.0",
+    "pydantic >= 2.9.1",
+    "pydantic-settings >= 2.5.2",
     "uvicorn[standard] >= 0.15.0,< 0.25.0",
     "opensearch-py ~= 2.0.0",
     "elasticsearch8[async] ~= 8.7.0",

argilla-server/src/argilla_server/__init__.py

@@ -12,11 +12,5 @@
 #  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
-import warnings
 
-from argilla_server.pydantic_v1 import PYDANTIC_MAJOR_VERSION
-
-if PYDANTIC_MAJOR_VERSION >= 2:
-    warnings.warn("The argilla_server package is not compatible with Pydantic 2. " "Please use Pydantic 1.x instead.")
-else:
-    from argilla_server._app import app  # noqa
+from argilla_server._app import app  # noqa

argilla-server/src/argilla_server/api/handlers/v1/oauth2.py

@@ -25,7 +25,9 @@
 from argilla_server.enums import UserRole
 from argilla_server.errors.future import NotFoundError
 from argilla_server.models import User
-from argilla_server.pydantic_v1 import Field
+
+# from argilla_server.pydantic_v1 import Field
+from pydantic import Field
 from argilla_server.security.authentication.oauth2 import OAuth2ClientProvider
 from argilla_server.security.authentication.userinfo import UserInfo
 from argilla_server.security.settings import settings

argilla-server/src/argilla_server/api/schemas/v1/commons.py

@@ -14,7 +14,8 @@
 
 from typing import Any, Dict, Set, Union
 
-from argilla_server.pydantic_v1 import BaseModel, root_validator
+# from argilla_server.pydantic_v1 import BaseModel, root_validator
+from pydantic import model_validator, BaseModel
 
 
 class UpdateSchema(BaseModel):
@@ -25,7 +26,8 @@ class UpdateSchema(BaseModel):
 
     __non_nullable_fields__: Union[Set[str], None] = None
 
-    @root_validator(pre=True)
+    @model_validator(mode="before")
+    @classmethod
     def validate_non_nullable_fields(cls, values: Dict[str, Any]) -> Dict[str, Any]:
         if cls.__non_nullable_fields__ is None:
             return values

argilla-server/src/argilla_server/api/schemas/v1/datasets.py

@@ -12,32 +12,44 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 
+from uuid import UUID
 from datetime import datetime
 from typing import List, Literal, Optional, Union
-from uuid import UUID
+from typing_extensions import Annotated
 
 from argilla_server.api.schemas.v1.commons import UpdateSchema
 from argilla_server.enums import DatasetDistributionStrategy, DatasetStatus
-from argilla_server.pydantic_v1 import BaseModel, Field, constr
+
+# from argilla_server.pydantic_v1 import BaseModel, Field, constr
+from pydantic import StringConstraints, ConfigDict, BaseModel, Field
 
 try:
     from typing import Annotated
 except ImportError:
     from typing_extensions import Annotated
 
-DATASET_NAME_REGEX = r"^(?!-|_)[a-zA-Z0-9-_ ]+$"
+# Pydantic v2 error: look-around, including look-ahead and look-behind, is not supported so rewriting it:
+# DATASET_NAME_REGEX = r"^(?!-|_)[a-zA-Z0-9-_ ]+$"
+DATASET_NAME_REGEX = r"^[a-zA-Z0-9 ][a-zA-Z0-9-_ ]*$"
 DATASET_NAME_MIN_LENGTH = 1
 DATASET_NAME_MAX_LENGTH = 200
 DATASET_GUIDELINES_MIN_LENGTH = 1
 DATASET_GUIDELINES_MAX_LENGTH = 10000
 
 DatasetName = Annotated[
-    constr(regex=DATASET_NAME_REGEX, min_length=DATASET_NAME_MIN_LENGTH, max_length=DATASET_NAME_MAX_LENGTH),
+    Annotated[
+        str,
+        StringConstraints(
+            pattern=DATASET_NAME_REGEX, min_length=DATASET_NAME_MIN_LENGTH, max_length=DATASET_NAME_MAX_LENGTH
+        ),
+    ],
     Field(..., description="Dataset name"),
 ]
 
 DatasetGuidelines = Annotated[
-    constr(min_length=DATASET_GUIDELINES_MIN_LENGTH, max_length=DATASET_GUIDELINES_MAX_LENGTH),
+    Annotated[
+        str, StringConstraints(min_length=DATASET_GUIDELINES_MIN_LENGTH, max_length=DATASET_GUIDELINES_MAX_LENGTH)
+    ],
     Field(..., description="Dataset guidelines"),
 ]
 
@@ -105,7 +117,7 @@ class UsersProgress(BaseModel):
 class Dataset(BaseModel):
     id: UUID
     name: str
-    guidelines: Optional[str]
+    guidelines: Optional[str] = None
     allow_extra_metadata: bool
     status: DatasetStatus
     distribution: DatasetDistribution
@@ -114,8 +126,7 @@ class Dataset(BaseModel):
     inserted_at: datetime
     updated_at: datetime
 
-    class Config:
-        orm_mode = True
+    model_config = ConfigDict(from_attributes=True)
 
 
 class Datasets(BaseModel):
@@ -124,7 +135,7 @@ class Datasets(BaseModel):
 
 class DatasetCreate(BaseModel):
     name: DatasetName
-    guidelines: Optional[DatasetGuidelines]
+    guidelines: Optional[DatasetGuidelines] = None
     allow_extra_metadata: bool = True
     distribution: DatasetDistributionCreate = DatasetOverlapDistributionCreate(
         strategy=DatasetDistributionStrategy.overlap,
@@ -134,9 +145,9 @@ class DatasetCreate(BaseModel):
 
 
 class DatasetUpdate(UpdateSchema):
-    name: Optional[DatasetName]
-    guidelines: Optional[DatasetGuidelines]
-    allow_extra_metadata: Optional[bool]
-    distribution: Optional[DatasetDistributionUpdate]
+    name: Optional[DatasetName] = None
+    guidelines: Optional[DatasetGuidelines] = None
+    allow_extra_metadata: Optional[bool] = None
+    distribution: Optional[DatasetDistributionUpdate] = None
 
     __non_nullable_fields__ = {"name", "allow_extra_metadata", "distribution"}