This is a project conceived in 2012 at the Complex Systems Computation Group at the Helsinki Institute of Information technology. This fork is a continuation and updating of the project carried out by the Digital Humanities Group at the University of Vienna.
Stemweb is a Python 3 / Django web app, using Celery for task queueing and with external dependencies on an SQL server (either ) and Redis.
To run the service , first create a .env file according to the pattern found in .env.example in this directory, setting appropriate values for at least STEMWEB_DBENGINE, STEMWEB_DBNAME, STEMWEB_DBUSER, STEMWEB_DBPASS, and STEMWEB_SECRET_KEY. You can then start the service with the provided docker-compose file, or manually.
If you are using Docker, you can now run the command docker compose up in this directory. Three containers should start:
- A MySQL container, running on
mysql:3306and not open to the outside - A Redis container, running on
redis:6379and not open to the outside - A Stemweb server, running on
stemweb:8000and mapped to localhost:8000
If you run the command docker compose --profile testing up instead, an additional container will start:
- A Stemweb testing client, running on
client:8001and mapped to localhost:8001.
If you wish to run the service independently and not containerised, what passes for setup guidance can be found in the Dockerfile in this directory.
It should now be possible to make requests to Stemweb according to the guidelines in the white paper.
If you are running the testing client (see above), you can make the following requests to see Stemweb in operation.
curl -X POST http://localhost:8001/request/01_rhm
curl -X POST http://localhost:8001/request/02_nj
curl -X POST http://localhost:8001/request/03_nnet
will make requests to the Stemweb service using the respective test data in client/requests. These will return responses that look like:
{"jobid": 1,"status": 1}
Eventually, a file result-{jobid}-{date}.json should appear in the client/received directory here, which is Stemweb's answer. This can also be called up with the command
curl http://localhost:8001/query/{jobid}