Pydantic v2 Meilisearch

dbs/meilisearch/.env.example

@@ -1,13 +1,13 @@
 # Master key must be at least 16 bytes, composed of valid UTF-8 characters
 MEILI_MASTER_KEY = ""
-MEILI_VERSION = "v1.1.1"
+MEILI_VERSION = "v1.2.0"
 MEILI_PORT = 7700
 MEILI_URL = "localhost"
 MEILI_SERVICE = "meilisearch"
 API_PORT = 8003
 
 # Container image tag
-TAG = "0.1.0"
+TAG = "0.2.0"
 
 # Docker project namespace (defaults to the current folder name if not set)
 COMPOSE_PROJECT_NAME = meili_wine

dbs/meilisearch/api/config.py

@@ -1,12 +1,14 @@
-from pydantic import BaseSettings
+from pydantic_settings import BaseSettings, SettingsConfigDict
 
 
 class Settings(BaseSettings):
+    model_config = SettingsConfigDict(
+        env_file=".env",
+        extra="allow",
+    )
+
     meili_service: str
     meili_master_key: str
     meili_port: int
     meili_url: str
     tag: str
-
-    class Config:
-        env_file = ".env"

dbs/meilisearch/api/routers/rest.py

@@ -1,5 +1,6 @@
-from meilisearch_python_async import Client
 from fastapi import APIRouter, HTTPException, Query, Request
+from meilisearch_python_async import Client
+
 from schemas.retriever import (
     FullTextSearch,
     TopWinesByCountry,

dbs/meilisearch/requirements.txt

@@ -1,6 +1,8 @@
-meilisearch-python-async==1.2.0
-pydantic[dotenv]>=1.10.7, <2.0.0
-fastapi>=0.95.0, <1.0.0
+meilisearch-python-async~=1.4.0
+pydantic~=2.0.0
+pydantic-settings~=2.0.0
+python-dotenv>=1.0.0
+fastapi~=0.100.0
 httpx>=0.24.0
 aiohttp>=3.8.4
 uvicorn>=0.21.0, <1.0.0

dbs/meilisearch/schemas/retriever.py

@@ -1,18 +1,9 @@
-from pydantic import BaseModel
+from pydantic import BaseModel, ConfigDict
 
 
 class FullTextSearch(BaseModel):
-    id: int
-    country: str
-    title: str
-    description: str | None
-    points: int
-    price: float | str | None
-    variety: str | None
-    winery: str | None
-
-    class Config:
-        schema_extra = {
+    model_config = ConfigDict(
+        json_schema_extra={
             "example": {
                 "id": 3845,
                 "country": "Italy",
@@ -24,9 +15,23 @@ class Config:
                 "winery": "Castellinuzza e Piuca",
             }
         }
+    )
+
+    id: int
+    country: str
+    title: str
+    description: str | None
+    points: int
+    price: float | str | None
+    variety: str | None
+    winery: str | None
 
 
 class TopWinesByCountry(BaseModel):
+    model_config = ConfigDict(
+        validate_assignment=True,
+    )
+
     id: int
     country: str
     title: str
@@ -36,11 +41,12 @@ class TopWinesByCountry(BaseModel):
     variety: str | None
     winery: str | None
 
-    class Config:
-        validate_assignment = True
-
 
 class TopWinesByProvince(BaseModel):
+    model_config = ConfigDict(
+        validate_assignment=True,
+    )
+
     id: int
     country: str
     province: str
@@ -50,6 +56,3 @@ class TopWinesByProvince(BaseModel):
     price: float | str | None = "Not available"
     variety: str | None
     winery: str | None
-
-    class Config:
-        validate_assignment = True

dbs/meilisearch/schemas/wine.py

@@ -1,26 +1,13 @@
-from pydantic import BaseModel, root_validator
+from pydantic import BaseModel, ConfigDict, Field, model_validator
 
 
 class Wine(BaseModel):
-    id: int
-    points: int
-    title: str
-    description: str | None
-    price: float | None
-    variety: str | None
-    winery: str | None
-    vineyard: str | None
-    country: str | None
-    province: str | None
-    region_1: str | None
-    region_2: str | None
-    taster_name: str | None
-    taster_twitter_handle: str | None
-
-    class Config:
-        allow_population_by_field_name = True
-        validate_assignment = True
-        schema_extra = {
+    model_config = ConfigDict(
+        populate_by_name=True,
+        validate_assignment=True,
+        extra="allow",
+        str_strip_whitespace=True,
+        json_schema_extra={
             "example": {
                 "id": 45100,
                 "points": 85,
@@ -37,20 +24,51 @@ class Config:
                 "taster_name": "Michael Schachner",
                 "taster_twitter_handle": "@wineschach",
             }
-        }
+        },
+    )
 
-    @root_validator(pre=True)
-    def _get_vineyard(cls, values):
-        "Rename designation to vineyard"
-        vineyard = values.pop("designation", None)
-        if vineyard:
-            values["vineyard"] = vineyard.strip()
-        return values
+    id: int
+    points: int
+    title: str
+    description: str | None
+    price: float | None
+    variety: str | None
+    winery: str | None
+    vineyard: str | None = Field(..., alias="designation")
+    country: str | None
+    province: str | None
+    region_1: str | None
+    region_2: str | None
+    taster_name: str | None
+    taster_twitter_handle: str | None
 
-    @root_validator
+    @model_validator(mode="before")
     def _fill_country_unknowns(cls, values):
         "Fill in missing country values with 'Unknown', as we always want this field to be queryable"
         country = values.get("country")
-        if not country:
+        if country is None or country == "null":
             values["country"] = "Unknown"
         return values
+
+
+if __name__ == "__main__":
+    data = {
+        "id": 45100,
+        "points": 85,
+        "title": "Balduzzi 2012 Reserva Merlot (Maule Valley)",
+        "description": "Ripe in color and aromas, this chunky wine delivers heavy baked-berry and raisin aromas in front of a jammy, extracted palate. Raisin and cooked berry flavors finish plump, with earthy notes.",
+        "price": 10,  # Test if field is cast to float
+        "variety": "Merlot",
+        "winery": "Balduzzi",
+        "designation": "Reserva",  # Test if field is renamed
+        "country": "null",  # Test unknown country
+        "province": " Maule Valley ",  # Test if field is stripped
+        "region_1": "null",
+        "region_2": "null",
+        "taster_name": "Michael Schachner",
+        "taster_twitter_handle": "@wineschach",
+    }
+    from pprint import pprint
+
+    wine = Wine(**data)
+    pprint(wine.model_dump(), sort_dicts=False)

dbs/meilisearch/scripts/bulk_index.py

@@ -14,7 +14,6 @@
 from meilisearch_python_async import Client
 from meilisearch_python_async.index import Index
 from meilisearch_python_async.models.settings import MeilisearchSettings
-from pydantic.main import ModelMetaclass
 
 sys.path.insert(1, os.path.realpath(Path(__file__).resolve().parents[1]))
 from api.config import Settings
@@ -62,15 +61,14 @@ def get_json_data(data_dir: Path, filename: str) -> list[JsonBlob]:
 
 def validate(
     data: list[JsonBlob],
-    model: ModelMetaclass,
     exclude_none: bool = False,
 ) -> list[JsonBlob]:
-    validated_data = [model(**item).dict(exclude_none=exclude_none) for item in data]
+    validated_data = [Wine(**item).model_dump(exclude_none=exclude_none) for item in data]
     return validated_data
 
 
 def process_chunks(data: list[JsonBlob]) -> tuple[list[JsonBlob], str]:
-    validated_data = validate(data, Wine, exclude_none=True)
+    validated_data = validate(data, exclude_none=True)
     return validated_data