refactor: move MetadataProperty schemas to his own file and unify can…

…onical MetadataProperty schema adding dataset_id (#4489)

# Description

This PR includes the following changes:
* Move `MetadataProperty` related schemas for API v1 to his own file at
`schemas/v1/metadata_properties.py`.
* Use only one `MetadataProperty` canonical schema removing
duplications.
* `MetadataProperty ` schema will include always `dataset_id` as
attribute.

Refs #4407 

**Type of change**

(Please delete options that are not relevant. Remember to title the PR
according to the type of change)

- [ ] Bug fix (non-breaking change which fixes an issue)
- [x] New feature (non-breaking change which adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to not work as expected)
- [x] Refactor (change restructuring the codebase without changing
functionality)
- [ ] Improvement (change adding some improvement to an existing
functionality)
- [ ] Documentation update

**How Has This Been Tested**

- [ ] Modifying and running unit tests.

**Checklist**

- [ ] I added relevant documentation
- [x] follows the style guidelines of this project
- [x] I did a self-review of my code
- [ ] I made corresponding changes to the documentation
- [ ] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] I filled out [the contributor form](https://tally.so/r/n9XrxK)
(see text above)
- [x] I have added relevant notes to the CHANGELOG.md file (See
https://keepachangelog.com/)

CHANGELOG.md

@@ -25,6 +25,7 @@ These are the section headers that we use:
 - API v1 responses returning `Response` schema now always include `record_id` as attribute. ([#4482](https://github.com/argilla-io/argilla/pull/4482))
 - API v1 responses returning `Question` schema now always include `dataset_id` attribute. ([#4487](https://github.com/argilla-io/argilla/pull/4487))
 - API v1 responses returning `Field` schema now always include `dataset_id` attribute. ([#4488](https://github.com/argilla-io/argilla/pull/4488))
+- API v1 responses returning `MetadataProperty` schema now always include `dataset_id` attribute. ([#4489](https://github.com/argilla-io/argilla/pull/4489))
 
 ### Changed
 

src/argilla/server/apis/v1/handlers/datasets/datasets.py

@@ -30,14 +30,12 @@
     DatasetMetrics,
     Datasets,
     DatasetUpdate,
-    MetadataProperties,
-    MetadataProperty,
-    MetadataPropertyCreate,
     VectorSettings,
     VectorSettingsCreate,
     VectorsSettings,
 )
 from argilla.server.schemas.v1.fields import Field, FieldCreate, Fields
+from argilla.server.schemas.v1.metadata_properties import MetadataProperties, MetadataProperty, MetadataPropertyCreate
 from argilla.server.schemas.v1.questions import Question, QuestionCreate, Questions
 from argilla.server.search_engine import (
     SearchEngine,

src/argilla/server/contexts/datasets.py

@@ -53,13 +53,12 @@
 from argilla.server.models.suggestions import SuggestionCreateWithRecordId
 from argilla.server.schemas.v1.datasets import (
     DatasetCreate,
-    MetadataPropertyCreate,
 )
 from argilla.server.schemas.v1.datasets import (
     VectorSettings as VectorSettingsSchema,
 )
 from argilla.server.schemas.v1.fields import FieldCreate
-from argilla.server.schemas.v1.metadata_properties import MetadataPropertyUpdate
+from argilla.server.schemas.v1.metadata_properties import MetadataPropertyCreate, MetadataPropertyUpdate
 from argilla.server.schemas.v1.questions import QuestionCreate
 from argilla.server.schemas.v1.records import (
     RecordCreate,

src/argilla/server/schemas/v1/datasets.py

@@ -13,16 +13,15 @@
 #  limitations under the License.
 
 from datetime import datetime
-from typing import Any, Dict, Generic, List, Literal, Optional, TypeVar, Union
+from typing import List, Literal, Optional, Union
 from uuid import UUID
 
 from fastapi import Query
 
-from argilla.server.enums import DatasetStatus, MetadataPropertyType, SimilarityOrder, SortOrder
-from argilla.server.pydantic_v1 import BaseModel, PositiveInt, constr, root_validator
-from argilla.server.pydantic_v1 import Field as PydanticField
-from argilla.server.pydantic_v1.generics import GenericModel
+from argilla.server.enums import DatasetStatus, SimilarityOrder, SortOrder
+from argilla.server.pydantic_v1 import BaseModel, Field, PositiveInt, constr, root_validator
 from argilla.server.schemas.base import UpdateSchema
+from argilla.server.schemas.v1.metadata_properties import MetadataPropertyName
 from argilla.server.schemas.v1.questions import QuestionName
 from argilla.server.schemas.v1.records import Record, RecordFilterScope
 from argilla.server.schemas.v1.responses import ResponseFilterScope
@@ -39,21 +38,12 @@
 DATASET_GUIDELINES_MIN_LENGTH = 1
 DATASET_GUIDELINES_MAX_LENGTH = 10000
 
-METADATA_PROPERTY_CREATE_NAME_REGEX = r"^(?=.*[a-z0-9])[a-z0-9_-]+$"
-METADATA_PROPERTY_CREATE_NAME_MIN_LENGTH = 1
-METADATA_PROPERTY_CREATE_NAME_MAX_LENGTH = 200
-METADATA_PROPERTY_CREATE_TITLE_MIN_LENGTH = 1
-METADATA_PROPERTY_CREATE_TITLE_MAX_LENGTH = 500
-
 VECTOR_SETTINGS_CREATE_NAME_REGEX = r"^(?=.*[a-z0-9])[a-z0-9_-]+$"
 VECTOR_SETTINGS_CREATE_NAME_MIN_LENGTH = 1
 VECTOR_SETTINGS_CREATE_NAME_MAX_LENGTH = 200
 VECTOR_SETTINGS_CREATE_TITLE_MIN_LENGTH = 1
 VECTOR_SETTINGS_CREATE_TITLE_MAX_LENGTH = 500
 
-TERMS_METADATA_PROPERTY_VALUES_MIN_ITEMS = 1
-TERMS_METADATA_PROPERTY_VALUES_MAX_ITEMS = 250
-
 TERMS_FILTER_VALUES_MIN_ITEMS = 1
 TERMS_FILTER_VALUES_MAX_ITEMS = 250
 
@@ -85,12 +75,12 @@ class Datasets(BaseModel):
 
 DatasetName = Annotated[
     constr(regex=DATASET_NAME_REGEX, min_length=DATASET_NAME_MIN_LENGTH, max_length=DATASET_NAME_MAX_LENGTH),
-    PydanticField(..., description="Dataset name"),
+    Field(..., description="Dataset name"),
 ]
 
 DatasetGuidelines = Annotated[
     constr(min_length=DATASET_GUIDELINES_MIN_LENGTH, max_length=DATASET_GUIDELINES_MAX_LENGTH),
-    PydanticField(..., description="Dataset guidelines"),
+    Field(..., description="Dataset guidelines"),
 ]
 
 
@@ -151,12 +141,12 @@ class VectorsSettings(BaseModel):
         min_length=VECTOR_SETTINGS_CREATE_TITLE_MIN_LENGTH,
         max_length=VECTOR_SETTINGS_CREATE_TITLE_MAX_LENGTH,
     ),
-    PydanticField(..., description="The title of the vector settings"),
+    Field(..., description="The title of the vector settings"),
 ]
 
 
 class VectorSettingsCreate(BaseModel):
-    name: str = PydanticField(
+    name: str = Field(
         ...,
         regex=VECTOR_SETTINGS_CREATE_NAME_REGEX,
         min_length=VECTOR_SETTINGS_CREATE_NAME_MIN_LENGTH,
@@ -167,107 +157,6 @@ class VectorSettingsCreate(BaseModel):
     dimensions: PositiveInt
 
 
-NT = TypeVar("NT", int, float)
-
-
-class NumericMetadataProperty(GenericModel, Generic[NT]):
-    min: Optional[NT] = None
-    max: Optional[NT] = None
-
-    @root_validator(skip_on_failure=True)
-    def check_bounds(cls, values: Dict[str, Any]) -> Dict[str, Any]:
-        min = values.get("min")
-        max = values.get("max")
-
-        if min is not None and max is not None and min >= max:
-            raise ValueError(f"'min' ({min}) must be lower than 'max' ({max})")
-
-        return values
-
-
-class TermsMetadataPropertyCreate(BaseModel):
-    type: Literal[MetadataPropertyType.terms]
-    values: Optional[List[str]] = PydanticField(
-        None, min_items=TERMS_METADATA_PROPERTY_VALUES_MIN_ITEMS, max_items=TERMS_METADATA_PROPERTY_VALUES_MAX_ITEMS
-    )
-
-
-class IntegerMetadataPropertyCreate(NumericMetadataProperty[int]):
-    type: Literal[MetadataPropertyType.integer]
-
-
-class FloatMetadataPropertyCreate(NumericMetadataProperty[float]):
-    type: Literal[MetadataPropertyType.float]
-
-
-MetadataPropertyName = Annotated[
-    str,
-    PydanticField(
-        ...,
-        regex=METADATA_PROPERTY_CREATE_NAME_REGEX,
-        min_length=METADATA_PROPERTY_CREATE_NAME_MIN_LENGTH,
-        max_length=METADATA_PROPERTY_CREATE_NAME_MAX_LENGTH,
-    ),
-]
-
-MetadataPropertyTitle = Annotated[
-    constr(min_length=METADATA_PROPERTY_CREATE_TITLE_MIN_LENGTH, max_length=METADATA_PROPERTY_CREATE_TITLE_MAX_LENGTH),
-    PydanticField(..., description="The title of the metadata property"),
-]
-
-MetadataPropertySettingsCreate = Annotated[
-    Union[TermsMetadataPropertyCreate, IntegerMetadataPropertyCreate, FloatMetadataPropertyCreate],
-    PydanticField(..., discriminator="type"),
-]
-
-
-class MetadataPropertyCreate(BaseModel):
-    name: MetadataPropertyName
-    title: MetadataPropertyTitle
-    settings: MetadataPropertySettingsCreate
-    visible_for_annotators: bool = True
-
-
-class TermsMetadataProperty(BaseModel):
-    type: Literal[MetadataPropertyType.terms]
-    values: Optional[List[str]] = None
-
-
-class IntegerMetadataProperty(BaseModel):
-    type: Literal[MetadataPropertyType.integer]
-    min: Optional[int] = None
-    max: Optional[int] = None
-
-
-class FloatMetadataProperty(BaseModel):
-    type: Literal[MetadataPropertyType.float]
-    min: Optional[float] = None
-    max: Optional[float] = None
-
-
-MetadataPropertySettings = Annotated[
-    Union[TermsMetadataProperty, IntegerMetadataProperty, FloatMetadataProperty],
-    PydanticField(..., discriminator="type"),
-]
-
-
-class MetadataProperty(BaseModel):
-    id: UUID
-    name: str
-    title: str
-    settings: MetadataPropertySettings
-    visible_for_annotators: bool
-    inserted_at: datetime
-    updated_at: datetime
-
-    class Config:
-        orm_mode = True
-
-
-class MetadataProperties(BaseModel):
-    items: List[MetadataProperty]
-
-
 class MetadataParsedQueryParam:
     def __init__(self, string: str):
         k, *v = string.split(":", maxsplit=1)
@@ -277,7 +166,7 @@ def __init__(self, string: str):
 
 
 class MetadataQueryParams(BaseModel):
-    metadata: List[str] = PydanticField(Query([], pattern=r"^(?=.*[a-z0-9])[a-z0-9_-]+:(.+(,(.+))*)$"))
+    metadata: List[str] = Field(Query([], pattern=r"^(?=.*[a-z0-9])[a-z0-9_-]+:(.+(,(.+))*)$"))
 
     @property
     def metadata_parsed(self) -> List[MetadataParsedQueryParam]:
@@ -321,16 +210,14 @@ class MetadataFilterScope(BaseModel):
 
 FilterScope = Annotated[
     Union[RecordFilterScope, ResponseFilterScope, SuggestionFilterScope, MetadataFilterScope],
-    PydanticField(..., discriminator="entity"),
+    Field(..., discriminator="entity"),
 ]
 
 
 class TermsFilter(BaseModel):
     type: Literal["terms"]
     scope: FilterScope
-    values: List[str] = PydanticField(
-        ..., min_items=TERMS_FILTER_VALUES_MIN_ITEMS, max_items=TERMS_FILTER_VALUES_MAX_ITEMS
-    )
+    values: List[str] = Field(..., min_items=TERMS_FILTER_VALUES_MIN_ITEMS, max_items=TERMS_FILTER_VALUES_MAX_ITEMS)
 
 
 class RangeFilter(BaseModel):
@@ -352,13 +239,11 @@ def check_ge_and_le(cls, values: dict) -> dict:
         return values
 
 
-Filter = Annotated[Union[TermsFilter, RangeFilter], PydanticField(..., discriminator="type")]
+Filter = Annotated[Union[TermsFilter, RangeFilter], Field(..., discriminator="type")]
 
 
 class Filters(BaseModel):
-    and_: List[Filter] = PydanticField(
-        None, alias="and", min_items=FILTERS_AND_MIN_ITEMS, max_items=FILTERS_AND_MAX_ITEMS
-    )
+    and_: List[Filter] = Field(None, alias="and", min_items=FILTERS_AND_MIN_ITEMS, max_items=FILTERS_AND_MAX_ITEMS)
 
 
 class Order(BaseModel):
@@ -369,7 +254,7 @@ class Order(BaseModel):
 class SearchRecordsQuery(BaseModel):
     query: Optional[Query]
     filters: Optional[Filters]
-    sort: Optional[List[Order]] = PydanticField(
+    sort: Optional[List[Order]] = Field(
         None, min_items=SEARCH_RECORDS_QUERY_SORT_MIN_ITEMS, max_items=SEARCH_RECORDS_QUERY_SORT_MAX_ITEMS
     )