lentil is a Ruby on Rails Engine that supports the harvesting of images from Instagram and provides several browsing views, mechanisms for sharing, tools for users to select their favorite images, an administrative interface for moderating images, and a system for harvesting images and submitting donor agreements in preparation of ingest into ex…
Ruby HTML CSS JavaScript
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
app Fix image ID number casting for GA Dec 7, 2016
config add route for embedding images into an iframe Feb 26, 2014
db
doc add v1 upgrade documentation Oct 28, 2016
lib
script
test
vendor/assets Treat animatedimages css as scss Nov 4, 2016
.gitignore
.travis.yml
Gemfile Update dependencies Dec 7, 2016
Gemfile.lock Lentil 1.0.5 Feb 2, 2017
Guardfile
MIT-LICENSE Initial code commit May 10, 2013
README.md
Rakefile add Ruby 2.0 support Nov 21, 2013
lentil.gemspec

README.md

Lentil

Instagram has removed access to a large portion of their public API in favor of a focus on their advertising API. Lentil will work with pre-existing developer credentials until June 2016, but we are currently attempting to contact Instagram and Facebook to discuss blanket application approvals for academic uses. We have developed a packet of generalized Lentil-related text and video for Instagram API requests that you are welcome to reuse. Instagram has approved the necessary access for Lentil-based applications that used these materials in the request.

lentil is a Ruby on Rails Engine that supports the harvesting of images from Instagram and provides several browsing views, mechanisms for sharing, tools for users to select their favorite images, an administrative interface for moderating images, and a system for harvesting images and submitting donor agreements in preparation of ingest into external repositories. Built according to the principles of responsive design, lentil is designed for use on mobile devices, tablets, desktops, and larger screens.

Build Status Gem Version

History

lentil is extracted from the My #HuntLibrary project at NCSU Libraries. My #HuntLibrary was created as a platform to foster student and community engagement with the new James B. Hunt Jr. Library via social media imagery and to preserve and archive these images as part of the record of the Hunt Library launch. Images from this crowdsourced documentation effort will be selected to become part of our permanent digital collections, allowing the NCSU community to contribute to the historical record of the Hunt Library through image submissions as well as the use of voting tools.

Project Maturity

Although we are using this gem in production, this gem should be considered an early release.

Installation

lentil is also available through the pre-packaged Social Media Combine for easier setup alongside Twitter harvesting software.

lentil requires Ruby 2.3.1 or greater.

Create a new Rails app with Rails 4.2.x

gem install rails -v '~> 4.2.7'
rails _4.2.7.1_ new --skip-turbolinks your_app_name
cd your_app_name

In the example above, 4.2.7.1 should be the version of Rails 4.x that is installed by the gem command.

Add lentil and therubyracer (or another ExecJS runtime) to your Gemfile and run bundle update

gem 'lentil'
gem 'therubyracer'

If you would like to work directly from the master branch, use gem 'lentil', :git => 'git://github.com/NCSU-Libraries/lentil.git' instead.

Run the generator

bundle exec rails generate lentil:install

Start the server

bundle exec rails server

Visit http://localhost:3000

Upgrading

You may need to run bundle exec rake lentil:install:migrations to incorporate any new database migration files.

Harvest Images

  • You will need to define your instagram_client_id, instagram_client_secret, and instagram_access_token in config/lentil_config.yml. You can generate these strings by creating an Instagram API client for your application and following the Client-side (Implicit) Authentication steps.

  • If you haven't already, add an administrative user in development.

bundle exec rake lentil:dummy_admin_user

You will then have an administrative user with the following credentials:

bundle exec rake lentil:image_services:instagram:fetch_by_tag

You ought to see a message like: "15 new images added"

Note: The harvesting task currently does not utilize Instagram's result paging. It is important at this point to harvest images frequently in order to avoid missed content.

Harvest Videos

Video harvesting is builtin to the image harvesting feature. Only once a single task has to be run to migrate the existing data to accomodate for schema. After updating lentil to the latest release run

bundle exec rails generate lentil:install
bundle exec rake lentil:image_services:restore_videos

You ought to see a message like: "x record(s) updated"

Videos can be moderated like images. In the home page and animate views videos will autoplay, muted and looped

Scheduling tasks

There are several rake tasks that you should schedule to run on a recurring basis:

  • rake lentil:image_services:instagram:fetch_by_tag harvests image metadata for all tags within "harvestable" tagsets.
  • rake lentil:image_services:test_image_files[number_of_images,image_service] checks that the image content for approved images is still available. After 3 failures, Lentil will stop displaying the images. After 10 failures, Lentil will stop checking. Defaults to checking 10 Instagram images. Images will not be checked more than once per day.
  • rake lentil:popularity:update_score recalculates the popularity score based on "likes" and "battles."

Additionally, there are three optional rake tasks that support image file harvesting for longer-term retention:

  • rake lentil:image_services:save_image_files[number_of_images,image_service,base_directory] Lentil normally only stores image metadata, but this task will save the image file to a specified directory. This file will still not be used by Lentil, but the file retrieval will be noted in the Image model. Defaults to 50 Instagram images, saving to the path specified by the base_image_file_dir option from config/lentil_config.yml.
  • rake lentil:image_services:dump_metadata[image_service,base_directory] Extracts all image metadata from the database and writes them as json files. Defaults to saving to the path specified by the base_image_file_dir option from config/lentil_config.yml.
  • rake lentil:image_services:submit_donor_agreements[number_of_images,image_service] will submit a donor agreement (using the donor_agreement_text option from config/lentil_config.yml) as a comment on approved Instagram images that have been in the system for at least a week. Currently defaults to one image.

In order to submit comments, you will need to generate an access token for the user that will submit the comments. See the Instagram documentation and a discussion of access token expiration.

An example of a background task schedule using the whenever gem:

case @environment
when 'production'
  every 5.minutes do
    rake "image_services:instagram:fetch_by_tag", :output => {:standard => nil}
  end

  every 30.minutes do
    rake "image_services:test_image_files", :output => {:standard => nil}
  end
end

# All environments
every :hour do
  rake "popularity:update_score", :output => {:standard => nil}
end

Customization

  • A simple first customization to do is to override the about page by putting your own about page at app/views/lentil/pages/about.html.erb.
  • If you haven't already you'll have update your config/lentil_config.yml file with your application name.
  • You can change the look of your site by including some variables in your application.css.scss before importing the lentil CSS:
$gray_text: #34282C;
$lentil_yellow: #DAB416;
$lentil_blue: #08C;
@import "lentil";
  • In order to set image backgrounds in My #HuntLibrary, the following CSS is included in a file that is imported after lentil:
body {
    background: image-url("background/fins3.jpg") fixed;
}

.navbar-inner {
    background: #121212 image-url("nav/top_background.png") repeat-x;
}

div.header {
    background: #0C1021 image-url("nav/example_08.jpg");
}

Special displays

Embeddable iframe view

This is a simple responsive non-interactive image grid that is intended to be embedded as an iframe. You can find this at http://YOUR_HOST/images/iframe. Image sizes in the grid are determined using a range of possible image widths, with the smallest width defaulting to 64px and the largest width calculated as 1.5 times the smallest width. The smallest width may be configured by including an integer parameter to the iframe URL, for example: http://YOUR_HOST/images/iframe?smallest_width=40. Currently, values greater than 320 may result in loss of resolution.

Large animated view

This is an example of an animated image grid that is designed for non-interactive displays. We have used customized versions of this on e-boards as well as very large video walls. This view will require some customization based on your project and target display. You can find this at http://YOUR_HOST/images/animate.

Testing

  • Install all of the dependencies:
bundle install
  • Prepare the test database
RAILS_ENV=test bundle exec rake app:db:migrate
  • Run the tests
bundle exec rake test

Running the Test/Dummy Instance

Sometimes you just might want to see what the dummy instance actually looks like. You can do the following:

bundle exec rake app:db:schema:load
cd test/dummy
bundle exec rake db:fixtures:load FIXTURES_PATH=../fixtures
bundle exec rails s

Once you load the fixtures you'll be able to harvest anything that has the tag "#hunttesting"

cd test/dummy

Edit config/lentil_config.yml and set instagram_client_id to the client ID associated with your Instagram API client.

bundle exec rake image_services:instagram:fetch_by_tag

Heroku

Documentation for deploying lentil to Heroku may be found in the doc directory.

License

See MIT-LICENSE

Contact

Submit a GitHub issue or contact lentil@lists.ncsu.edu.

Authors (in alphabetical order)

  • Jason Casden
  • Bret Davidson
  • Aniket Lawande
  • Cory Lown
  • Jason Ronallo
  • Heidi Tebbe

Additional project team members

  • Brian Dietz
  • Jennifer Garrett
  • Mike Nutt (Product Lead for My #HuntLibrary project)