Job posting builder with bias analysis.
This is more or less a Jekyll site but with a lot of javascript and a Sinatra backend. Requires Ruby 2.2.2.
After cloning, run bundle install. Create a new file .env, copy in the contents of .env.example, and replace all the your_xxxx_here values with your own values.
SINATRA_SESSION_SECRET is an arbitrary string (for development)SKILLS_ENGINE_ID and SKILLS_ENGINE_SECRET would come from your SkillsEngine developer account. NOTE: PostingPro does not currently rely on this API and these values are optional if you're not using it. The API calls remain in the code for reference.jekyll build --watch compiles the site (& sees html/css/js changes) and foreman start runs it at http://0.0.0.0:5000.
Make sure Postgres is installed: which psql should give a reasonable answer. Also create a file called .env with the line DATABASE_URL=postgres://localhost:5432/postingpro. (You may need to create a new DB in postgres called postingpro.)
Run rake db:migrate and rake db:seed to set up the database structure and fill it with sample data. If the data doesn't show up, try running rake db:reset and rake db:seed.
Commits to master automagically push to staging. From there, code can be pushed to production via the heroku pipeline.
If you want to test your local changes in a similar-to-heroku environment, run heroku local web.
Deploying your own instance: This app has only been tested on Heroku and may require additional configuration on other providers. Remember to set up the config vars and change the admin password and session secret to something more secure than the dev values.
To add or remove templates, access the management portal at by adding /manage to the Posting Pro path. This part of the website requires a username and password. It's set in the .env file for local deployments and as an environment variable on Heroku. Please ask the administrator of the Heroku deployment for credentials.
The template parsing currently requires a very specific template. See the guide for an explanation of how parsings works.
We rely on a modified fork of TextStatistics.js for readability analysis (fork).
We also ported the gender-decoder algorithm into Javascript for the "gender balance" analysis.