Hivesearcher

Web front-end for Hivesearcher — the search UI, the API documentation (/api-docs) and API-key registration (/api-register). A small Flask app serves the built single-page app and proxies a few endpoints to the search API.

• Frontend: React 18 + Vite (in react_app/)
• Backend: Flask + gunicorn (app.py)
• State: Redis (caches search responses and holds the indexed document count)

Configuration

The backend reads its settings from config.ini (or from environment variables — see below). Copy the example and fill in your own values:

cp config.ini.example config.ini

Section	Key	Description
ESEARCH_API	URL	Base URL of the search API
ESEARCH_API	TOKEN	API key for the search API
ESEARCH_API	REQUEST_TIMEOUT	Upstream request timeout (seconds)
ESEARCH_API	CACHE_TIMEOUT	Redis cache TTL for search responses (seconds)
REDIS	HOST / PORT / PASS	Redis connection
FLASK	DEVELOPMENT / DEBUG	Flask flags

config.ini is gitignored — never commit real credentials.

Environment variable overrides

Any option can be overridden with an env var named __ENV__<SECTION>_<OPTION>, for example __ENV__ESEARCH_API_TOKEN=... or __ENV__REDIS_HOST=.... This is useful for containers/CI where you'd rather not keep a config file on disk.

To update config, change config.ini (or the env vars) and restart the app.

Local development

Backend (Flask dev server on port 3002) — needs Redis running locally:

pip3 install -r requirements.txt
python3 app.py

Frontend (Vite dev server with hot reload):

cd react_app
yarn install
yarn dev      # dev server
yarn build    # production build into react_app/build
yarn test     # unit tests

In production the Flask app serves the built react_app/build directory.

Production (Docker)

The web image is built with the multi-stage Dockerfile-web (Node builds the SPA, Python runs Flask/gunicorn on port 5001):

docker build -f Dockerfile-web -t hivesearcher-web .

Run it, mounting config.ini at /deploy/config.ini (keep secrets out of the image) and giving it access to your Redis and search API:

docker run -d --name hivesearcher-web --restart always \
  -p 5001:5001 \
  -v "$PWD/config.ini:/deploy/config.ini:ro" \
  hivesearcher-web

• Update config: edit the mounted config.ini (or the __ENV__* vars) and restart the container.
• Deploy new code: rebuild the image and recreate the container. Because the config is mounted (not baked into the image), it is preserved across rebuilds.

Document counter (optional)

Dockerfile-counter builds a small worker (counter/main.py) that periodically writes the indexed document count to Redis (served by the /api/count endpoint). It uses the same config.ini / env configuration:

docker build -f Dockerfile-counter -t hivesearcher-counter .
docker run -d --name hivesearcher-counter --restart always \
  -v "$PWD/config.ini:/deploy/config.ini:ro" \
  hivesearcher-counter