Skip to content

Latest commit

 

History

History
131 lines (86 loc) · 5.14 KB

README.md

File metadata and controls

131 lines (86 loc) · 5.14 KB
Raw

kleineAnfragen.

Collecting kleine Anfragen from Parlamentsdokumentationssystemen for easy search- and linkability.

Development

For a simple and quick development environment, docker-compose is used. Install docker and docker-compose, then run:

docker-compose up

docker-compose downloads the required services (postgres, elasticsearch, redis, ...) as docker containers and links them with the app. If you want to look into postgres or elasticsearch, uncomment the ports section in docker-compose.yml.

You may be required to execute database migrations. Try this:

docker-compose run web rails db:migrate
docker-compose run web rails db:seed

To get a rails console, run:

docker-compose run web rails c

Importing papers from the public database dump

If you want to develop with already scraped data, you can use the public available data dumps from the kleineAnfragen.de data page. Download the latest kleineanfragen-....sql.bz2 from there and put it into tmp/dump/.

To begin importing the data, you have to first enter a docker container:

docker run -v $(pwd)/tmp/dump:/tmp/dump --rm --network kleineanfragen_default -it kleineanfragen_database bash

Inside this one-off throwaway container, import the data with following commands

bzcat /tmp/dump/kleineanfragen-*.sql.bz2 | psql -h database -U kleineanfragen import
pg_dump -h database -U kleineanfragen -d import --data-only | psql -h database -U kleineanfragen -d kleineanfragen
psql -h database -U kleineanfragen import -c "DROP SCHEMA public CASCADE; CREATE SCHEMA public; GRANT ALL ON SCHEMA public TO postgres; GRANT ALL ON SCHEMA public TO public;"
exit

Normalizing Names with Nomenklatura

For normalizing names of people, parties and ministries, we use Nomenklatura.

If you want to use nomenklatura while developing, you need to edit docker-compose.yml:

  • Uncomment the nomenklatura link
  • the NOMENKLATURA_ environment variables
  • the whole nomenklatura image
  • set GITHUB_CLIENT_ID and GITHUB_CLIENT_SECRET to those of a new Github OAuth application.

After your next docker-compose up login to your nomenklatura instance (reachable at http://localhost:8080) and get the API key from the profile link. Insert it into docker-compose.yml.

kleineAnfragen needs multiple Datasets with the following identifiers that must be created in Nomenklatura:

  • ka-parties
  • ka-people-XX (replace XX with a two letter state)
  • ka-ministries-XX (replace XX with a two letter state)

Troubleshooting

You just git pulled and now kleineanfragen doesn't start anymore? Try docker-compose rm web and docker-compose build web — this rebuilds the container that the application is running in.

Dependencies

  • ruby 2.5.5
  • postgres
  • elasticsearch (for search)
  • redis (for sidekiq)
  • nodejs (for asset compiling)
  • tika (for extracting text from pdfs)
  • Nomenklatura (for normalization of people names, ministries and parties)
  • Poppler / pdftoppm (for thumbnailing)
  • image_optim binaries (for compressing thumbnails)
  • s3 compatible storage like s3ninja (see contrib/s3ninja for the modified dockered version)

Configuration

  • config/application.rb

    Please change the config.x.user_agent to your own email address.

  • .env

    In development, the environment variables are set in docker-compose.yml. For development without docker-compose (or production), create .env and fill it with these:

    export DATABASE_URL="postgres://user:pass@localhost/kleineanfragen"
export ELASTICSEARCH_URL="http://127.0.0.1:9200/"
export SECRET_KEY_BASE="FIXME"
export SECRET_SUBSCRIPTION_SALT="FIXME"
export S3_ACCESS_KEY="FIXME"
export S3_SECRET_KEY="FIXME"
export REDIS_URL="redis://localhost:6379"
export TIKA_SERVER_URL="http://localhost:9998"
export NOMENKLATURA_HOST="http://localhost:9000"
export NOMENKLATURA_APIKEY="FIXME"

  • config/fog.yml

    This file contains the connection details to your s3 server/bucket. Test uses the tmp folder, so you don't need a connection to a running s3 compatible storage.

Jobs

Jobs are run by ActiveJob / Sidekiq.

You may need to prefix them with bundle exec, so the correct gems are used.

The typical arguments are [State, LegislativeTerm, Reference]

  • Import new papers

    rails 'papers:import_new[BE, 17]'

  • Import single paper

    rails 'papers:import[BE, 17, 1234]'

  • Other

    The two import tasks should be enough for daily usage, if you need to (re-)upload the papers to s3 again or extract the text / names, you can use these:

    rails 'papers:store[BE, 17, 1234]'
rails 'papers:extract_text[BE, 17, 1234]'
rails 'papers:extract_originators[BE, 17, 1234]'
rails 'papers:extract_answerers[BE, 17, 1234]'