Kazimirski

The Kazimirski web application aims to facilitate crowdsourced transcription of scanned pages into digitized text, when OCRs do not produce reliable output (ex: handwritten text, old typography, or a combination of both such as in the Kazimirski dictionary).

Features

• Side-by-side transcription UI embedding Internet Archive's viewer
• Custom text input logic handling specificities of bi-directional text in a dictionary
• Integration of Trix WYSIWYG editor with minimal formatting & server-side markup sanitization
• Simple checkout-submit-review workflow for contributed pages
• Straightforward role system:
  • Transcribers can submit pages
  • Reviewers can review, correct and accept pages
  • Admins can access the management backend
• Dashboard with overview of overall progress
• Email notifications
• Custom CAPTCHA with bi-directional text

Setup

Clone code

git clone git@github.com:francoisbruneau/kazimirski.git

Install VirtualBox, Vagrant and Ansible

• VirtualBox
• Vagrant
• Ansible

Config

Create a .env file in the project folder and add:

RACK_ENV=development
PORT=5000
CAPTCHA_9ad39f737b804013809c6945dbd23355=answer1
CAPTCHA_ffa971d9f127408b88c47c734b7ddfbd=answer2
CAPTCHA_71074c5ee3ee4b0485ef6a860f55828a=answer3

Start Vagrant box

vagrant up

Start server

ssh Vagrant box

vagrant ssh

Create tables and seed them with sample data

rake db:setup

Run mailcatcher if you want to test emails locally

mailcatcher --http-ip=0.0.0.0

Run server

foreman start

Go to http://localhost:5000

Acknowledgements

Thanks to @noefroidevaux for the base Vagrant/Ansible/Ruby/Rails setup: https://github.com/noefroidevaux/rails-workshop