Temple University Libraries Oral History Collections

The Temple University Libraries

Getting Started

Prerequisites

Here are the basic steps to get TUL_Hydra up and running.

git clone git@github.com:tulibraries/tul_ohist.git
git submodule init
git submodule update
bundle install
cp config/contentdm.yml.example config/contentdm.yml

Edit config/contentdm.yml. Replace samples with your ContentDM parameters

Complete the installation.

rake db:migrate

# The following step is missing from Hydra instructions.
# Do it or you'll be reindexing whenever you restart jetty.
rails generate hydra:jetty

rake jetty:config
rake jetty:start
rails s -d

Set up periodic housekeeping

On the production server, all queries are saved in the database. Set up a cron tob to remove saved searches older than 7 days. Log onto the TUL_Ohist server and edit /etc/crontab and add the line:

0 2 * * * cd /var/www/tul_ohist && RAILS_ENV=production bundle exec rake blacklight:delete_old_searches[7]

Oral History Object Architecture

Oral history objects are a composite of two or more repository asset types in the digitalMetadata datastream. These components consist of transcripts, and images, and possibly audio files.

Each oral history object has a master identifier (MID). Within the digitalMetadata datastream are filenames for the transcript, photograph, and audio. Those files are pulled from the filesystem and displayed or provided for playback in the object's show page.

The assets that are unique to the oral history item usually are composed of the MID with a three character suffix, (e.g. Q01).

The MID might be : AOHWMPJZ2014120022 The associated transcript_file would be AOHWMPJZ2014120022Q01.pdf

These files are located on a network server mounted at https://libdigital.temple.edu/pdfa1/Oral Histories/, accessed with the local_file_url helper method and rendered via the PDF_Object viewer.