Add PostgreSQL

We're going to need to use a database to store the state (and results)
of pdf conversions.

.env.example

@@ -4,3 +4,14 @@ S3_SECRET_ACCESS_KEY=${S3_SECRET_ACCESS_KEY}
 S3_ENDPOINT=${S3_ENDPOINT:-"https://nyc3.digitaloceanspaces.com"}
 S3_REGION=${S3_REGION:-"us-east-1"}
 S3_BUCKET=${S3_BUCKET}
+
+# Database setup
+PGHOST=${PGHOST:-"localhost"}
+PGUSER=${PGUSER:-"user"}
+PGPASSWORD=${PGPASSWORD:-"password"}
+PGDATABASE=${PGDATABASE:-"database"}
+PGPORT=${PGPORT:-:"5432"}
+
+## The schema being used for the application database.
+## Migrations will be run against this schema.
+APP_SCHEMA=${APP_SCHEMA:-"public"}

.gitlab-ci.yml

@@ -15,6 +15,8 @@ lint:
     - make lint
 
 test:
+  services:
+    - postgres
   stage: test
   script:
     - make test
@@ -24,3 +26,8 @@ test:
     S3_ENDPOINT: "https://example.com"
     S3_REGION: "us-east-1"
     S3_BUCKET: "example"
+    DATABASE_URL: "postgresql://test_user:password@localhost/test_database"
+    POSTGRES_DB: "test_database"
+    POSTGRES_USER: "test_user"
+    POSTGRES_PASSWORD: "password"
+    POSTGRES_HOST_AUTH_METHOD: trust

README.md

@@ -6,6 +6,7 @@ This project requires:
 
 - Python 3.10.x
 - make
+- PostgreSQL
 
 ## Dev setup
 

requirements-dev.txt

@@ -3,3 +3,4 @@ pyright==1.1.368
 pytest==8.2.2
 pytest-mock==3.14.0
 starlette==0.37.2
+sqlalchemy-stubs==0.4

requirements.txt

 boto3==1.34.131
 fastapi==0.111.0
+psycopg2-binary==2.9.9
 pydantic==2.7.4
 pydantic-settings==2.3.3
+SQLAlchemy==2.0.31
 uvicorn==0.30.1

service/core/database.py

+from sqlalchemy import Column, DateTime, create_engine, func, text
+from sqlalchemy.engine.url import URL
+from sqlalchemy.ext.declarative import declared_attr
+
+# We need to ignore type checks on this line due to https://github.com/dropbox/sqlalchemy-stubs/issues/250
+from sqlalchemy.orm import as_declarative, sessionmaker  # type: ignore
+
+from service.core.settings import settings
+
+database_url = URL(
+    "postgresql",
+    host=settings.PGHOST,
+    username=settings.PGUSER,
+    password=settings.PGPASSWORD,
+    database=settings.PGDATABASE,
+    port=settings.PGPORT,
+    query={},
+)
+
+engine = create_engine(database_url)
+SessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=engine)
+
+
+@as_declarative()
+class Base:
+    @declared_attr
+    def created_at(cls):
+        return Column(
+            DateTime(timezone=True), nullable=False, server_default=func.now()
+        )
+
+    @declared_attr
+    def updated_at(cls):
+        return Column(
+            DateTime(timezone=True), nullable=False, server_default=func.now()
+        )
+
+
+def get_db():
+    db = SessionLocal()
+    db.execute(text(f"set search_path to {settings.APP_SCHEMA}"))
+    try:
+        yield db
+    finally:
+        db.close()

service/core/settings.py

@@ -7,6 +7,12 @@ class Settings(BaseSettings):
         extra="allow",
     )
     APP_NAME: str = "PDF to Markdown Service"
+    PGHOST: str
+    PGUSER: str
+    PGPASSWORD: str
+    PGDATABASE: str
+    PGPORT: int = 5432
+    APP_SCHEMA: str = "public"
     S3_BUCKET: str
     S3_ACCESS_KEY_ID: str
     S3_SECRET_ACCESS_KEY: str