/chronam

This software project is no longer being actively developed at the Library of Congress. Consider using the Open-ONI (https://github.com/open-oni) fork of the chronam software. Project mailing list: http://listserv.loc.gov/archives/chronam-users.html.

Primary LanguagePython

chronam

chronam is the Django application that the Library of Congress uses to make its Chronicling America website. The Chronicling America website makes millions of pages of historic American newspapers that have been digitized by the National Digital Newspaper Program (NDNP) browsable and searchable on the Web. A little bit of background is needed to understand why this software is being made available.

NDNP is actually a partnership between the Library of Congress, the National Endowment for the Humanities (NEH), and cultural heritage organizations (awardees) across the United States who have applied for grants to help digitize newspapers in their state. Awardees digitize newspaper microfilm according to a set of specifications and then ship the data back to the Library of Congress where it is loaded into Chronicling America.

Awardee institutions are able to use this data however they want, including creating their own websites that highlight their newspaper content in the local context of their own collections. The idea of making chronam available here on Github is to provide a technical option to these awardees, or other interested parties who want to make their own websites of NDNP newspaper content available. chronam provides a core set of functionality for loading, modeling and indexing NDNP data, while allowing you to customize the look and feel of the website to suit the needs of your organization.

The NDNP data is in the Public Domain and is itself available on the Web for anyone to use. The hope is that the chronam software can be useful for others who want to work with and/or publish the content.

Install

System level dependencies can be installed by following these operating system specific instructions:

After you have installed the system level dependencies you will need to install some application specific dependencies, and configure the application.

Install dependent services

MySQL

You will need a MySQL database. If this is a new server, you will need to start MySQL and assign it a root password:

sudo service mysqld start
/usr/bin/mysqladmin -u root password '' # pick a real password

You will probably want to change the password 'pick_one' in the example below to something else:

echo "DROP DATABASE IF EXISTS chronam; CREATE DATABASE chronam CHARACTER SET utf8mb4; CREATE USER 'chronam'@'localhost' IDENTIFIED BY 'pick_one'; GRANT ALL ON chronam.* to 'chronam'@'localhost'; GRANT ALL ON test_chronam.* TO 'chronam'@'localhost';" | mysql -u root -p

Solr

The Ubuntu and Red Hat guides have instructions for installing and starting Solr manually. For developmeny you may prefer to use Docker:

cd solr
docker build -t chronam-solr:latest .
docker run -p8983:8983 chronam-solr:latest

Install the application

First you will need to set up the local Python environment and install some Python dependencies:

cd /opt/chronam/
virtualenv -p /usr/bin/python2.7 ENV
source /opt/chronam/ENV/bin/activate
cp conf/chronam.pth ENV/lib/python2.7/site-packages/chronam.pth
pip install -r requirements.pip

Next you need to create some directories for data:

mkdir /srv/chronam/batches
mkdir /srv/chronam/cache
mkdir /srv/chronam/bib

You will need to create a Django settings file which uses the default settings and sets custom values specific to your site:

  1. Create a settings.py file in the chronam directory which imports the default values from the provided template for possible customization:

     echo 'from chronam.settings_template import *' > /opt/chronam/settings.py
    
  2. Ensure that the DJANGO_SETTINGS_MODULE environment variable is set to chronam.settings before you start a Django management command. This can be set as a user-wide default in your ~/.profile or but the recommended way is simply to make it part of the virtualenv activation process::

     echo 'export DJANGO_SETTINGS_MODULE=chronam.settings' >> /opt/chronam/ENV/bin/activate
    
  3. Add your database password to the settings.py file following the standard Django settings documentation:

     DATABASES = {
         'default': {
             'ENGINE': 'django.db.backends.mysql',
             'NAME': 'chronam_db',
             'USER': 'chronam_user',
             'HOST': 'mysql.example.org',
             'PASSWORD': 'NotTheRealPassword',
         }
     }
    

You should never edit the settings_template.py file since that may change in the next release but you may wish to periodically review the list of changes to that file in case you need to update your local settings.

Next you will need to initialize database schema and load some initial data:

django-admin.py migrate
django-admin.py loaddata initial_data languages
django-admin.py chronam_sync --skip-essays

And finally you will need to collect static files (stylesheets, images) for serving up by Apache in production settings:

django-admin.py collectstatic --noinput

Load Data

As mentioned above, the NDNP data that awardees create and ship to the Library of Congress is in the public domain and is made available on the Web as batches. Each batch contains newsaper issues for one or more newspaper titles. To use chronam you will need to have some of this batch data to load. If you are an awardee you probably have this data on hand already, but if not you can use a tool like wget to bulk download the batches. For example:

cd /srv/chronam/batches/
wget --recursive --no-parent --no-host-directories --cut-dirs 2 --reject index.html* https://chroniclingamerica.loc.gov/data/batches/uuml_thys_ver01/

In order to load data you will need to run the load_batch management command by passing it the full path to the batch directory. So assuming you have downloaded batch_uuml_thys_ver01 you will want to:

django-admin.py load_batch /srv/chronam/batches/uuml_thys_ver01

If this is a new server, you may need to start the web server:

sudo service httpd start

After this completes you should be able to view the batch in the batches report via the Web:

http://www.example.org/batches/

Caching

After loading data, you will need to clear the cache. If you are using a reverse proxie (like Varnish) you will need to also clear that, as well as any CDN you have. Below is a list of URLS that should be cleared based on what content you are loading.

All pages that contain a LCCN are tagged with that LCCN in the cache headers. This allows for purging by specific LCCN tag if there is a update to a batch.

List of URLs to purge when loading new batch

  • All URLs tagged with lccn=<LCCN>
  • All URLs matching these patterns:
    chroniclingamerica.loc.gov/tabs
    chroniclingamerica.loc.gov/sitemap*
    chroniclingamerica.loc.gov/frontpages*
    chroniclingamerica.loc.gov/titles*
    chroniclingamerica.loc.gov/states*
    chroniclingamerica.loc.gov/counties*
    chroniclingamerica.loc.gov/states_counties*
    chroniclingamerica.loc.gov/cities*
    chroniclingamerica.loc.gov/batches/summary*
    chroniclingamerica.loc.gov/reels*
    chroniclingamerica.loc.gov/reel*
    chroniclingamerica.loc.gov/essays*
    

List of URLs to purge when loading new Awardees

  • All URLs matching chroniclingamerica.loc.gov/awardees*

List of URLs to purge when loading new basic data

  • All URLs matching chroniclingamerica.loc.gov/institutions*

List of URLs to purge when loading code

  • All URLs matching these patterns:
    chroniclingamerica.loc.gov/ocr
    chroniclingamerica.loc.gov/about
    chroniclingamerica.loc.gov/about/api
    chroniclingamerica.loc.gov/help
    

Run Unit Tests

For the unit tests to work you will need:

  • to have the batch_uuml_thys_ver01 available. You can use the wget command in the previous section to get get it.
  • A local SOLR instance running
  • A local MySQL database
  • Access to the Essay Editor Feed

After that you should be able to:

cd /opt/chronam/
django-admin.py test chronam.core.tests --settings=chronam.settings_test

License

This software is in the Public Domain.