Claire Pritchard
April 2017 - Present
Letterpress is a Django web application for the management of transcriptions and images of letters and related documents. I originally created it to help manage my transcriptions of 19th-century correspondence which I was using as input for natural language processing experiments.
The name comes from an early office technology which people used to duplicate documents and store them in a letterpress copybook.
- Highly customized Django Admin
- Full-text searching with Elasticsearch
- SQLite database
- Place mapping with OpenStreetMap and OpenLayers 6
- Export to text files
- Word frequency statistics by month, with Bokeh charts
- Word clouds for writers, date ranges, etc.
- Make sure you have write access to the SQLite database file
db.sqlite3
- Set up a Django Admin user with the command
shell python manage.py createsuperuser
. - Elasticsearch index can be created or updated manually with the Django management command
push_to_index
, otherwise updates are automatic when the model is saved. - Text searches are fuzzy by default. For exact match, enclose search terms in quotes.
- Build the
django
andelasticsearch
services:
docker-compose -f docker-compose.yml build
- Run the
django
andelasticsearch
services:
docker-compose -f docker-compose.yml up
Photograph of copy press from Letterpress Commons, map marker from Maps Icons Collection, other images public domain. Thanks to Hipster Ipsum and Bacon Ipsum for some of the gibberish text in my unit tests.