Read MongoDB connection parameters from environment variables

.dockerignore

@@ -0,0 +1,8 @@
+# This file specifies directories/files we want Docker to omit from the Docker build context.
+# Docs: https://docs.docker.com/build/concepts/context/#dockerignore-files
+
+# Omit Python virtual environments.
+.venv/
+
+# Omit Python cache files.
+__pycache__/

.env.example

@@ -0,0 +1,7 @@
+# See `src/config.py` for documentation about these environment variables.
+
+MONGO_HOST=mongo
+MONGO_PORT=27017
+MONGO_USERNAME=admin
+MONGO_PASSWORD=root
+MONGO_DATABASE=bertron

.github/workflows/ci.yml

@@ -1,16 +1,25 @@
 name: CI
 
+# Configure GHA to run this workflow whenever either:
+# (a) a pull request is opened or updated, or
+# (b) a commit(s) is pushed into the `main` branch.
 on:
-  push:
-#    branches:
-#      - main
-  pull_request:
+  pull_request: { }
+  push: { branches: [ main ] }
 
 jobs:
   pr-check:
     name: PR checks
     runs-on: ubuntu-latest
 
+    # Define environment variables that will be available to all steps within this job.
+    env:
+      MONGO_HOST: mongo
+      MONGO_PORT: 27017
+      MONGO_USERNAME: admin
+      MONGO_PASSWORD: root
+      MONGO_DATABASE: bertron
+
     steps:
       - uses: actions/checkout@v4
 
@@ -32,16 +41,17 @@ jobs:
       - name: Linting
         run: uv run ruff check -- src/
 
-      - name: Run tests locally
-        run: uv run pytest tests
-
       # Note: This spins up containers running the default services.
+      #       This does not include the `ingest` or `test` services,
+      #       which aren't part of the default profile.
       - name: Spin up Docker Compose stack in background
         run: docker compose up --detach
 
-      # Note: This spins up the "test" container.
+      # Note: The `--exit-code-from test` option applies the exit code of the `test` container
+      #       to the `docker compose` process, so that the GHA step fails if tests fail.
+      #       Reference: https://docs.docker.com/reference/cli/docker/compose/up/
       - name: Spin up `test` container
-        run: docker compose up test
+        run: docker compose up --exit-code-from test test
 
       # Note: This spins everything down.
       - name: Spin down Docker Compose stack

.gitignore

@@ -8,12 +8,17 @@
 # Ignore unrelated files created by Finder on macOS.
 .DS_Store
 
-# Jupter Notebook checkpoints.
+# Jupyter Notebook checkpoints.
 **/.ipynb_checkpoints/
 
-# TODO: Document these.
-.venv
+# Ignore Python virtual environments.
+.venv/
+
+# TODO: Document this.
 **/*.egg-info/
 
 # Ignore Python bytecode files.
-__pycache__
+__pycache__
+
+# Top-level environment configuration file.
+/.env

CONTRIBUTING.md

@@ -10,49 +10,59 @@ Check https://docs.astral.sh/uv/#installation for alternative installation metho
 
 ## Preparing dev environment
 
+Create your `.env` file (if you haven't already) and edit its contents to reflect
+your environment.
+
+```sh
+cp .env.example .env
+
+# (Optional) Edit its contents.
+# vi .env
 ```
+> Note: Git will ignore your `.env` file.
+
+Create and activate a Python virtual environment.
+
+```sh
 uv venv
 source .venv/bin/activate
 ```
 
-You can also run commands without activating the dev environment with `uv run`,
-for example you can run `ruff check` with
-```
+You can also run commands without first activating the Python virtual environment, by prefixing them with `uv run`. For example, you can run `ruff check` with:
+```sh
 uv run ruff check
 ```
-but the dev environment is useful if you're using an IDE.
 
 ## Adding dependencies
 
 You can add dependencies with `uv add`. The following command:
-```
+```sh
 uv add polars duckdb
 ```
 adds `polars` and `duckdb` as project dependencies.
 
 For dev dependencies (only needed for development, but not for using the package later),
 use `--dev`:
-```
+```sh
 uv add --dev ruff ipykernel pytest pytest-cov mypy
 ```
 
 ## Updating dependencies
 
 You can run
-```
+```sh
 uv sync --upgrade
 ```
 to install the latest dependency version that matches the version range in `pyproject.toml`.
 This will also update `uv.lock` to make the installation reproducible.
 
-## Syncing dev environment
+## Syncing the Python virtual environment
 
-After adding or updating dependencies,
-run
-```
+After adding or updating dependencies, run
+```sh
 uv sync --all-extras --dev
 ```
-to make sure the dev environment has the updated dependencies.
+to make sure the Python virtual environment has the updated dependencies.
 
 ---
 

docker-compose.yml

@@ -3,6 +3,17 @@ services:
     # Use the container image built in the "development" stage of the Dockerfile.
     build: { context: ".", dockerfile: Dockerfile, target: development }
     restart: unless-stopped
+    # Propagate environment variables from the host into the container.
+    # Note: The `:?` suffix makes it so Docker Compose displays an error if the
+    #       environment variable is either (a) undefined, or (b) empty, in the
+    #       host environment.
+    # Docs: https://docs.docker.com/compose/how-tos/environment-variables/variable-interpolation/#additional-information
+    environment:
+      MONGO_HOST: ${MONGO_HOST:?}
+      MONGO_PORT: ${MONGO_PORT:?}
+      MONGO_USERNAME: ${MONGO_USERNAME:?}
+      MONGO_PASSWORD: ${MONGO_PASSWORD:?}
+      MONGO_DATABASE: ${MONGO_DATABASE:?}
     ports:
       # Map a host port (by default, 8000, but it can be overridden via an
       # environment variable) to port 8000 of the container; the latter being
@@ -30,8 +41,8 @@ services:
       - "${MONGO_PORT:-27017}:27017"
     restart: unless-stopped
     environment:
-      MONGO_INITDB_ROOT_USERNAME: admin
-      MONGO_INITDB_ROOT_PASSWORD: root
+      MONGO_INITDB_ROOT_USERNAME: "${MONGO_USERNAME:?}"
+      MONGO_INITDB_ROOT_PASSWORD: "${MONGO_PASSWORD:?}"
     volumes:
       - mongo_data:/data/db
 
@@ -40,9 +51,6 @@ services:
     build: { context: ".", dockerfile: Dockerfile, target: development }
     # This service should not start automatically - only run on demand
     profiles: ["tools"]
-    environment:
-      # Set the MongoDB connection string to connect to the mongo service
-      MONGO_URI: "mongodb://admin:root@mongo:27017"
     volumes:
       # Mount the root directory to access the ingest script and data files
       - ".:/app"
@@ -56,6 +64,12 @@ services:
     build: { context: ".", dockerfile: Dockerfile, target: development }
     # This service should not start automatically - only run on demand
     profiles: ["tools"]
+    environment:
+      MONGO_HOST: ${MONGO_HOST:?}
+      MONGO_PORT: ${MONGO_PORT:?}
+      MONGO_USERNAME: ${MONGO_USERNAME:?}
+      MONGO_PASSWORD: ${MONGO_PASSWORD:?}
+      MONGO_DATABASE: ${MONGO_DATABASE:?}  # the test suite will disregard this
     depends_on:
       - app
       - mongo

mongodb/README.md

@@ -22,8 +22,8 @@ python ingest_data.py --input your_data_file.json
 
 ### Command-line arguments
 
-- `--mongo-uri`: MongoDB connection URI (default: mongodb://localhost:27017)
-- `--db-name`: MongoDB database name (default: bertron)
+- `--mongo-uri`: MongoDB connection URI (default: `mongodb://localhost:27017`)
+- `--db-name`: MongoDB database name (default: `bertron`)
 - `--schema-path`: Path or URL to the schema JSON file (default: `bertron_schema.json` in the current directory)
 - `--input`: Path to input JSON file or directory containing JSON files (required)
 

pyproject.toml

@@ -26,6 +26,7 @@ dependencies = [
   "fastapi[standard]>=0.115.12",
   "jsonschema>=4.0.0",
   "nmdc-api-utilities>=0.3.9",
+  "pydantic-settings>=2.10.1",
   "pymongo>=4.13.1",
   "pytest>=8.4.0",
   "uvicorn>=0.34.3",

src/config.py

@@ -0,0 +1,34 @@
+from typing import Optional
+
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class Settings(BaseSettings):
+    r"""
+    Configuration settings for the application.
+
+    Order of precedence (highest to lowest) for settings values:
+    1. Process-level environment variable definitions.
+    2. Environment variable definitions loaded from a `.env` file.
+    3. Default values defined in this class.
+
+    References:
+    - https://docs.pydantic.dev/latest/concepts/pydantic_settings/#dotenv-env-support
+    """
+
+    # Load environment variable definitions from a `.env` file, if one exists,
+    # in the current working directory. We allow the `.env` file to contain
+    # extra environment variables (beyond those modeled below), since this
+    # repository currently contains a variety of independent applications.
+    model_config = SettingsConfigDict(env_file=".env", extra="ignore")
+
+    # MongoDB connection settings.
+    mongo_host: str = "localhost"
+    mongo_port: int = 27017  # note: env. var. value will be coerced into int
+    mongo_username: Optional[str] = None
+    mongo_password: Optional[str] = None
+    mongo_database: str = "bertron"
+
+
+# Instantiate a settings object that can be imported into other modules.
+settings = Settings()

src/server.py

@@ -10,13 +10,17 @@
 
 from lib.helpers import get_package_version
 from models import HealthResponse, VersionResponse
+from config import settings as cfg
 
 # Set up logging
 logger = logging.getLogger(__name__)
 
-# Connect to MongoDB.
-# TODO: Get these values from environment variables instead of hard-coding them.
-mongo_client = MongoClient("mongo:27017", username="admin", password="root")
+# Connect to the MongoDB server.
+mongo_client = MongoClient(
+    f"{cfg.mongo_host}:{cfg.mongo_port}",
+    username=cfg.mongo_username,
+    password=cfg.mongo_password,
+)
 
 app = FastAPI(
     title="BERtron API",
@@ -58,7 +62,7 @@ def get_version() -> VersionResponse:
 @app.get("/bertron")
 def get_all_entities():
     r"""Get all documents from the entities collection."""
-    db = mongo_client["bertron"]
+    db = mongo_client[cfg.mongo_database]
 
     # Check if the collection exists
     if "entities" not in db.list_collection_names():
@@ -104,7 +108,7 @@ def find_entities(query: MongoDBQuery):
         "sort": {"field1": 1, "field2": -1}
     }
     """
-    db = mongo_client["bertron"]
+    db = mongo_client[cfg.mongo_database]
 
     # Check if the collection exists
     if "entities" not in db.list_collection_names():
@@ -153,7 +157,7 @@ def find_nearby_entities(
 
     Example: /bertron/geo/nearby?latitude=47.6062&longitude=-122.3321&radius_meters=10000
     """
-    db = mongo_client["bertron"]
+    db = mongo_client[cfg.mongo_database]
 
     # Check if the collection exists
     if "entities" not in db.list_collection_names():
@@ -215,7 +219,7 @@ def find_entities_in_bounding_box(
 
     Example: /bertron/geo/bbox?southwest_lat=47.5&southwest_lng=-122.4&northeast_lat=47.7&northeast_lng=-122.2
     """
-    db = mongo_client["bertron"]
+    db = mongo_client[cfg.mongo_database]
 
     # Check if the collection exists
     if "entities" not in db.list_collection_names():
@@ -274,7 +278,7 @@ def get_entity_by_id(id: str):
 
     Example: /bertron/emsl:12345
     """
-    db = mongo_client["bertron"]
+    db = mongo_client[cfg.mongo_database]
 
     # Check if the collection exists
     if "entities" not in db.list_collection_names():

src/tests/conftest.py

@@ -0,0 +1,35 @@
+r"""
+This module contains `pytest` fixture definitions that `pytest` will automatically make available
+to all tests within this directory and its descendant directories.
+
+From the `pytest` documentation:
+> The `conftest.py` file serves as a means of providing fixtures for an entire directory.
+> Fixtures defined in a `conftest.py` can be used by any test in that package without
+> needing to import them (`pytest` will automatically discover them).
+Source: https://docs.pytest.org/en/stable/reference/fixtures.html#conftest-py-sharing-fixtures-across-multiple-files
+"""
+
+import pytest
+
+from config import settings as cfg
+
+
+# Note: We use `autouse=True` so that this fixture is automatically applied to each test
+#       within its scope (since we are in a `conftest.py` file, its scope consists of
+#       the current directory and all descendant directories).
+@pytest.fixture(autouse=True)
+def patched_cfg():
+    r"""
+    A `pytest` fixture that temporarily patches the application configuration
+    so it references a test database.
+    """
+
+    test_database_name = "bertron_test"
+    main_database_name = cfg.mongo_database
+    assert main_database_name != test_database_name, (
+        "The main database name matches the test database name. "
+        "Reconfigure your environment to ensure they differ."
+    )
+    cfg.mongo_database = test_database_name
+    yield cfg
+    cfg.mongo_database = main_database_name