Diana is a portal for digital humanities projects, developed by the Gothenburg Research Infrastructure in Digital Humanities (GRIDH) at the University of Gothenburg. It is a database coordination solution, interface for data input, and a service for making data accessible through generated REST APIs.

Diana is a solution for coordinating multiple potentially interacting databases. It has the following structure:

├── apps
│   ├── __init__ .py
│   ├── litteraturlabbet
│   ├── norfam
│   ├── rwanda
│   ├── saga
│   ├── shfa
│   └── ...
├── diana
│   ├── __init__.py
│   ├── abstract
│   ├── asgi.py
│   ├── routers.py
│   ├── settings.py
│   ├── storages.py
│   ├── urls.py
│   ├── utils.py
│   └── wsgi.py
├── locale
│   ├── en
│   └── sv
├── static
│   ├── admin-interface
│   ├── rwanda
│   └── shfa
├── templates
│   ├── redoc.html
│   └── swagger-ui.html
├── README.md
├── requirements.txt
├── environment.yml
└── manage.py

In apps, Diana stores a number of so-called applications, or projects, as git submodules. These all have a corresponding GitHub repository where development is done independently.

Another folder, diana technically also qualifies as an app. It coordinates the rest of the applications, and contains utility functions and common, abstract classes, in abstract. This is also where the global settings are located.

The locale folder contains files for internationalization of text in the admin interface of Diana. In the static and templates folders you will find static files, such as images and html templates (for customized frontends handled by Django itself) respectively.

To install the Digital Humanities Portal, we advise using a Conda distribution, such as Anaconda or Miniconda. This allows simultaneous installation of necessary binaries.

Diana and its dependencies have been tested on Linux, e.g. Ubuntu 16.04+ and RedHat OS. Problems have been documented due to geographical dependencies on Windows and MacOS.

Clone the repository and change directory.

git clone https://github.com/CDH-DevTeam/diana-backend.git
cd diana-backend

Create a new conda environment from the environment.yml file using

conda env create -n diana -f environment.yml

This will also install the required GDAL dependency for geographical databases.

Activate the conda environment:

conda activate diana

Before installing Diana, it is advised to first look through the Django tutorial and documentation.

For a local installation, a configs folder with database credentials for your local database is needed. Your local database needs the postgis extension which can be added as postgres user with:

\connect <databasename>
CREATE EXTENSION postgis;

Also, a settings_local.py with local settings is needed. The description of it can be found in deployment instructions.

Launch Django by migrating all the initial settings,

python manage.py migrate 

and create a suitable superuser.

python manage.py createsuperuser 

Run it locally via:

python manage.py runserver

To also fetch the submodules of a specific branch and for more detailed instructions, please follow the instructions in the deployment instructions.

Localization is done by using the gettext_lazy translation utility, e.g.

from django.utils.translation import gettext_lazy as _

admin.site.site_header = _('admin.site.site_header')

where admin.site.site_header is an identifier to be translated into the local languages. By running

python manage.py makemessages -l <language-code>

a .po translation file is generated, with the format

#: directory/file.py:line
msgid "admin.site.site_header"
msgstr ""

where msgstr is the corresponding translation to be added, e.g. "Admin site" in English and "Adminsajt" in Swedish. When the .po file is complete, the translations need to be compiled with

django-admin compilemessages

which generates a binary .mo file.

• http://localhost:8000/admin/ - Admin interface for users and groups
• http://localhost:8000/api/ - REST API endpoints
• http://localhost:8000/api// - Endpoint for a certain project
• http://localhost:8000/api//documentation/ - Dynamic API documentation with ReDoc/OpenAPI