board-game-scraper

Scraping data about board games from the web. View the data live at Recommend.Games! Install via

pip install board-game-scraper

Sources

• Board Game Atlas (bga)
• BoardGameGeek (bgg)
• DBpedia (dbpedia)
• Luding.org (luding)
• Spielen.de (spielen)
• Wikidata (wikidata)

Run scrapers

Requires Python 3. Make sure Pipenv is installed and create the virtual environment:

python3 -m pip install --upgrade pipenv
pipenv install --dev
pipenv shell

Run a spider like so:

JOBDIR="jobs/${SPIDER}/$(date --utc +'%Y-%m-%dT%H-%M-%S')"
scrapy crawl "${SPIDER}" \
    --output 'feeds/%(name)s/%(time)s/%(class)s.csv' \
    --set "JOBDIR=${JOBDIR}"

where $SPIDER is one of the IDs above.

Run all the spiders with the run_scrapers.sh script. Get a list of the running scrapers' PIDs with the processes.sh script. You can close all the running scrapers via

./processes.sh stop

and resume them later.

Tests

You can run scrapy check to perform contract tests for all spiders, or scrapy check $SPIDER to test one particular spider. If tests fails, there most likely has been some change on the website and the spider needs updating.

Board game datasets

If you are interested in using any of the datasets produced by this scraper, take a look at the BoardGameGeek guild. A subset of the data can also be found on Kaggle.

Links

• board-game-scraper: This repository
• Recommend.Games: board game recommender using the scraped data
• recommend-games-server: Server code for Recommend.Games
• board-game-recommender: Recommender code for Recommend.Games