Cuneiform Digital Palæography Project (CDPP)

Ubuntu setup notes:

#! /bin/bash

# Instrutions for flask setup from: https://www.digitalocean.com/community/tutorials/how-to-serve-flask-applications-with-gunicorn-and-nginx-on-ubuntu-16-04

# Required packages not included with DO base Ubuntu 16.04
sudo apt-get update
sudo apt-get install mysql-server nginx python-pip python-dev elasticsearch libmysqlclient-dev libssl-dev yui-compressor

# mysql setup
CREATE DATABASE glyph CHARACTER SET utf8 COLLATE utf8_bin;
CREATE USER 'glyph'@'%' IDENTIFIED BY 'glyph';
GRANT ALL ON glyph.* TO 'glyph'@'%';
exit;

# Import data
mysql -u glyph -p glyph < db_dumps/schema.sql
mysql -u glyph -p glyph < db_dumps/glyph_latest.sql


# Global PIP items:
sudo pip install virtualenv

# virtualenv items:
pip install gunicorn
pip install -R requirements.txt

Installation (manual, assumes an Apache server)

1. Ensure you have git and mod_wsgi installed on your box, install using yum if necessary
2. Install the Python Virtualenv package using yum
3. Clone this repository: git clone https://github.com/urschrei/CDP.git
4. cd to the directory (it's probably called CDP)
5. Create a new virtualenv called venv, by executing virtualenv venv
6. Activate it by executing source venv/bin/activate
7. Install the required Python libraries by executing pip install -r requirements.txt
8. Ensure that a MySQL database named glyph exists. Username and password should both be glyph. See the section on Character Sets, below, for encoding and collation settings
9. Execute fab build_db to create the database structure and populate it. The db can also be populated by first importing db_dumps/schema.sql, then importing db_dumps/glyph_latest.sql
10. Open the run.wsgi file in an editor, and modify the first line with the full path to the activate_this script. For example, if you've cloned this repository into the /var/www directory, the full path would be /var/www/CDP/venv/bin/activate_this.py
11. See here for a sample Apache WSGI virtualhost
12. Remember to reload Apache when you've created the new vhost.

Development Box Installation (will run on Nginx/Gunicorn in a virtual machine)

1. Install Vagrant and Virtualbox
2. Clone this repository, and cd into it
3. Run vagrant up. This will create a new virtual machine with all necessary packages
4. When it's completed, run vagrant ssh to access the VM
5. Run fab build_db to populate the database
6. Run fab run_app to start the application in debug mode
7. Access the site on localhost:5000
8. The site will run in production configuration on 10.11.12.15

Other fab commands:

• fab shell: open an IPython shell with an imported app context and db instance. Queries can then be run like so:
  • db.session.query(Cdp).join(Sign).filter(Sign.sign_ref == 'AK').all()

Database admin

The database can be accessed via SSH (using e.g. Sequel Pro):

1. The SSH host is 10.11.12.15
2. The SSH user and password are vagrant/vagrant
3. The MySQL user is root, the password is blank

The application user/pass are glyph/glyph, and the db is glyph

Elasticsearch

The search functionality requires Elasticsearch, running on port 9200.

In order to generate the index before first use, run fab elasticsearch. Record changes and insertions should propagate to the index automatically, so long as the ORM is being used (i.e. in the normal course of things). The import script may also be run manually from a fab shell: %run utils/elastic.py

A note on MySQL character sets

As we're using MySQL, please pay particular attention to your DB's character encoding and collation settings, if you're setting it up manually.

For encoding, utf8mb4 should be used. For collation, utf8mb4_bin should be used. Binary collation is required in order to distinguish between e.g. S and Š.

Citation

This project has a DOI: 10.5281/zenodo.11647

Cite as:
Stephan Hügel (2014). Cuneiform Digital Palaeography Project (CDPP) v0.2. ZENODO.
10.5281/zenodo.11647