Skip to content

openstate/open-rechtspraak

Repository files navigation

Open Rechtspraak

Requirements

  • docker-compose
  • Make (to access dev shortcuts in Makefile)

Stack

Back end

  • Python 3.10
  • Flask
  • Postgres 13

Front end

  • Bootstrap 5
  • For JS see package.json
  • Webpacker for compilation

Getting started

  • You can initialize the full development environment with make init.
  • Pre-commit hooks are configured, and it is recommended you install them through pre-commit install.
  • For Python linting, we use various linters as specified in pre-commit-config.yaml. For javascript, we use ESLint with the rules in eslintrc.json. Make sure you configure these linters before contributing, because CI tests might fail if you don't.
  • Seed is available with make seed.
  • Visit localhost:5000.

Did you make any changes to the code base?

  • Run make test to run the test suite.
  • Ensure you add tests for new functionalities.

Periodically scraping data

Scraping data from people is done in two steps:

  1. Retrieving a list of people with make import_people from the https://namenlijst.rechtspraak.nl/Services/WebNamenlijstService/Zoek service. These people are saved.
  2. We enrich the details from people through https://namenlijst.rechtspraak.nl/Services/WebNamenlijstService/haalOp/?id=<rechtspraak_id> where rechtspraak_id is a unique ID assigned by namenlijst.rechtspraak.nl to the person.

Similarly, two commands exist for scraping verdicts and enriching them.

  1. Retrieving a list of verdicts (default = verdicts of the past three days) with make import_verdicts.
  2. Enriching verdicts with make enrich_verdicts in two steps: (1) collecting additional metadata of the verdict and (2) checking whether a person is mentioned in a verdict.

Warning: The number of verdicts imported by make import_verdicts can be high (ca. 50k a year). Consider this carefully when you use the --start_date and --end_date cli params.

Protecting a scraped person from public eyes

The People model has an attribute called protected. Toggling that attribute to True removes the person from all public listings (i.e. the API search endpoint) and blocks access to their detail page.

Config for production

These env variables are required for production:

  • ENV = production
  • FLASK_ENV = production
  • DATABASE_URL = url to the database, formatted similar to postgresql://user:password@localhost:port/database_name (alternatively you can use the individual parts of the database url in separate environment variables as found in .env.dist)
  • SECRET_KEY = a randomly generated string that is used for encryption
  • LOG_LEVEL = a valid level from Python's logging module; generally WARN or higher is recommended in production

How to deploy this app to production

  1. docker-compose up -d
  2. Compiling new assets: docker exec ors_webpack_1 npm run prod
  3. Running the latest migrations: docker exec ors_app_1 sh ./release/tasks.sh
  4. Ensure a cron is running with the jobs mentioned in release/cron.sh

Sitemap

  • A self-updating sitemap is included at /sitemap.xml. You can submit this sitemap to search engines to be indexed faster.

Contact details

Questions? Contact developers@openstate.eu.

Local dev

Without Flask in Docker, but a Docker based DB:

  • cp .env.dist .env
  • export $(cat .env | xargs)
  • export DATABASE_URL=postgresql://ors2user:ors2@localhost:54322/ors2