Runoregi

Runoregi is the Web front-end for verse and poem similarity calculations on Finnic oral folk poetry.

Setup

Python dependencies can be installed with the usual command:

pip3 install -r requirements.txt

As Runoregi is a front-end, it requires access to a MySQL database containing all the data (poem texts, metadata, similarities etc.) in order to work. The access is configured by the following environment variables:

• DB_HOST -- hostname,
• DB_PORT -- port (must be given even if it's the default 3306),
• DB_USER -- username,
• DB_PASS -- password,
• DB_NAME -- name of the database.

A dump of the database, including SQL scripts to load the tables, is available in hsci-r/filter-db (currently not public).

In addition to the database access, the following further environment variables can be configured:

• DB_LOGGING -- if set (to any value), Runoregi will write information about the requested pages and their loading times to the table runoregi_log,
• VISUALIZATIONS_URL -- the base URL of the visualizations application -- needed for the links to the visualizations (maps, type treemaps)!

Running

For all the commands listed below, remember to set the environment variables (see above).

Local testing

The easiest way to run Runoregi for quick testing (e.g. if you're working with the code) is to run:

python3 wsgi.py

The interface will be available for local connections under localhost:5000.

Running with Gunicorn

To run Runoregi on a server and access it through the default HTTP port, it should be enough to run:

gunicorn wsgi

See Gunicorn documentation for more options.

Deployment on Rahti

In our project Runoregi is hosted on the CSC's Rahti service, with a Docker image built using Source-to-image. For deployment, you need to link to this repository, provide a pull secret (as long as the repository is private) and set the environment variables to access the database.

Docker images for deployment on an own platform might be provided in the future if needed.

License

MIT (see LICENSE)

Publications

There are no publications so far descibing specifically the front-end. You can cite Runoregi by referring to one of the publications that describe the back-end calculations:

• Janicki M., Kallio K. and Sarv M. 2022. Exploring Finnic written oral folk poetry through string similarity.. Digital Scholarship in the Humanities.
• Janicki M. 2022. Optimizing the weighted sequence alignment algorithm for large-scale text similarity computation. In: Proceedings of the 2nd International Workshop on Natural Language Processing for Digital Humanities – NLP4DH 2022.