Arrange the pieces of the world! You need to drag the shape of the territory to the right place.
Just like in childhood we collected pictures piece by piece, so here you can collect a country
from regions or whole continents from countries!
Furthermore second variation is the Quiz. In the Quiz you need find the country by flag, emblem or the capital.
This is just web application, so you need:
- nginx
- uwsgi
- supervisor
- python >= 3.8
- redis
- postgres >= 12
- nodejs >= 12.13
Install nginx, uwsgi, supervisor and other packages:
$ sudo apt install nginx uwsgi uwsgi-plugin-python3 supervisor redis-server
$ sudo apt install gdal-bin gettext build-essential python3-dev libpq-dev
Template config files for nginx, uwsgi and supervisor are placed in deploy
directory.
I hope python 3 with pipenv already installed :)
$ pipenv install
-
Install Postgres with PostGIS:
$ sudo apt install postgresql postgresql-contrib postgis
-
Install postgis extension:
$ sudo su postgres -c psql postgres=# create user geopuzzle with password 'geopuzzle'; postgres=# create database geopuzzle owner geopuzzle;
-
Create user and database:
$ sudo su postgres -c psql postgres=# \c geopuzzle geopuzzle=# create extension postgis; geopuzzle=# create extension postgis_topology; geopuzzle=# create extension postgis_sfcgal; geopuzzle=# create extension fuzzystrmatch; geopuzzle=# create extension address_standardizer; geopuzzle=# create extension postgis_tiger_geocoder;
-
Create
.env
file (based on.env_template
) in project root with environment settings. Required parameters were already initialed default values in template file; -
Create tables:
(venv)$ ./manage.py migrate
All environment variables are collected in .env file (.env_template is template for that). Some variables such as REDIS_HOST or DJANGO_SETTINGS_MODULE do not require explanation. But some are specific to the project:
- OSM_KEY need only for load polygons into database
- GOOGLE_KEY is production key for Google Maps API (preferred, but not required)
- RAVEN_DSN for collect errors to Sentry (optional)
- AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY need only for AWS settings usage. Can left them blank.
Install nodejs:
$ sudo apt install nodejs npm
$ npm init
$ npm run build
(geopuzzle) $ ./manage.py collectstatic
I find excellent service for load geojson files with polygons - https://osm-boundaries.com/. Just select the country (or many or with regions) and click 'Export'. URL must be like:
https://osm-boundaries.com/Download/Submit?apiKey=74ba654a378b8daf80f&db=osm20210531&osmIds=-5682946,-5682950&format=GeoJSON&srid=4326
And download content by that link. You've got zip archive with GeoJson files, unpack them into geojson
folder.
After that you can run management command (venv)$ ./manage.py update_regions
for load polygons into database.
Take into account - data from OSM is not perfect and some regions can have bad or recursive links to each other.
In most cases, restarting can help :)
Management command usually run via manage.py script, for example: ./manage.py import_region
.
cache
- import/export and recalculate cache (from postgres)import_region
- load one .geojson into databaseupdate_infobox
- update infobox from WikiDataupdate_regions
- load all .geojson into database fromgeojson
foldervalidate_infoboxes
- check that all required data in all infoboxes is filled and correct
Project has 2 Dockerfiles:
- /Dockerfile.backend - for server side (django tests and deploy)
- /Dockerfile.frontend - for client side (webpack build bundles and jest tests for future)
All images should be up to date with all installed dependencies. This allows you to significantly reduce the time to perform tasks CI. Command for update image (frontend, by example):
$ docker build -t tyvik/geopuzzle:frontend -f Dockerfile.frontend .
$ docker push tyvik/geopuzzle:frontend
Build and run:
$ docker-compose build
$ docker-compose up -d
Stop:
$ docker-compose stop
You can download sample database.
Unpack archive into pgdata
directory.
Or download dump as sql file.
Go to https://localhost:8000