Mei link ml

README.md

@@ -10,8 +10,8 @@ data access.
 This project uses a **PEP 420 namespace package** structure with three main
 components:
 
-- **`ca_biositing.datamodels`**: Shared SQLModel database models and database
-  configuration
+- **`ca_biositing.datamodels`**: Shared LinkML/SQLModel database models and
+  database configuration
 - **`ca_biositing.pipeline`**: ETL pipelines orchestrated with Prefect, deployed
   via Docker
 - **`ca_biositing.webservice`**: FastAPI REST API for data access
@@ -152,6 +152,7 @@ Key tasks:
 - **Development**: `test`, `test-cov`, `pre-commit`, `pre-commit-all`
 - **Applications**: `start-webservice`, `qgis`
 - **Database**: `access-db`, `check-db-health`
+- **Datamodels**: `update-schema`, `migrate`
 
 ## Architecture
 
@@ -181,6 +182,15 @@ Pipeline architecture:
 
 ### Database Models
 
+We use a **LinkML-first approach** for defining our data schema. The workflow
+is:
+
+1.  **LinkML Schema**: The schema is defined in YAML files (source of truth).
+2.  **SQLAlchemy Generation**: Python classes are automatically generated from
+    LinkML.
+3.  **Alembic Migrations**: Database migrations are generated from the Python
+    classes.
+
 SQLModel-based models provide:
 
 - Type-safe database operations

alembic/env.py

@@ -13,17 +13,18 @@
 load_dotenv(dotenv_path=PROJECT_ROOT / ".env")  # Looks for .env in the project root by default
 
 # --- Import your models so Alembic knows about them ---
-from ca_biositing.datamodels.biomass import *
-from ca_biositing.datamodels.data_and_references import *
-from ca_biositing.datamodels.experiments_analysis import *
-from ca_biositing.datamodels.external_datasets import *
-from ca_biositing.datamodels.geographic_locations import *
-from ca_biositing.datamodels.metadata_samples import *
-from ca_biositing.datamodels.organizations import *
-from ca_biositing.datamodels.people_contacts import *
-from ca_biositing.datamodels.sample_preprocessing import *
-from ca_biositing.datamodels.specific_aalysis_results import *
-from ca_biositing.datamodels.user import *
+# from ca_biositing.datamodels.biomass import *
+# from ca_biositing.datamodels.data_and_references import *
+# from ca_biositing.datamodels.experiments_analysis import *
+# from ca_biositing.datamodels.external_datasets import *
+# from ca_biositing.datamodels.geographic_locations import *
+# from ca_biositing.datamodels.metadata_samples import *
+# from ca_biositing.datamodels.organizations import *
+# from ca_biositing.datamodels.people_contacts import *
+# from ca_biositing.datamodels.sample_preprocessing import *
+# from ca_biositing.datamodels.specific_aalysis_results import *
+# from ca_biositing.datamodels.user import *
+from ca_biositing.datamodels.schemas.generated.ca_biositing import *
 from sqlmodel import SQLModel
 import importlib.util
 from pathlib import Path
@@ -51,7 +52,8 @@
 # See: https://alembic.sqlalchemy.org/en/latest/autogenerate.html#affecting-the-autogenerate-process
 
 # Start with the default SQLModel metadata
-target_metadata = SQLModel.metadata
+# target_metadata = SQLModel.metadata
+target_metadata = Base.metadata
 
 # Manually merge metadata from generated modules
 # for table in census_metadata.tables.values():

alembic/versions/301006b4795f_initial_migration.py → alembic/versions/184417c6d044_deleted_duplicate_yaml_files.py

@@ -1,8 +1,8 @@
-"""Initial migration
+"""Deleted duplicate .yaml files
 
-Revision ID: 301006b4795f
-Revises: e15bda744fef
-Create Date: 2025-11-12 20:54:01.573577
+Revision ID: 184417c6d044
+Revises: bfa07351a507
+Create Date: 2025-12-17 19:03:32.278635
 
 """
 from typing import Sequence, Union
@@ -12,8 +12,8 @@
 
 
 # revision identifiers, used by Alembic.
-revision: str = '301006b4795f'
-down_revision: Union[str, Sequence[str], None] = 'e15bda744fef'
+revision: str = '184417c6d044'
+down_revision: Union[str, Sequence[str], None] = 'bfa07351a507'
 branch_labels: Union[str, Sequence[str], None] = None
 depends_on: Union[str, Sequence[str], None] = None