PIPAS persister

This repository provides an implementation of the PIPAS Persister, its PostgreSQL database and RESTful interface.

A test deployment runs on the Heroku cloud platform. The application can be reached at http://pipas-persister.herokuapp.com/

Services implemented so far

Please see the complete list with documentation at /resources/.

GET  /scheduling/problem                  get the current scheduling problem to solve
GET  /scheduling/solution                 get the current solution to the scheduling problem
PUT  /scheduling/solution                 update the current scheduling solution
GET  /patients/                           get the list of registered patients
GET  /appointments/{uuid}                 get information about a specific appointment
PUT  /appointments/{uuid}                 update information about an appointment
GET  /deliveries/{uuid}                   get information about a specific delivery
POST /deliveries/                         add information about a new delivery
GET  /service/planning                    get the current service planning
GET  /service/availabilities              get information about service availabilities
GET  /service/load                        get information about service (nurses and beds) load
GET  /treatments/{uuid}                   get information about a specific treatment
POST /treatments/                         create a patient and enrol her in a new treatment
GET  /treatments/{uuid}/unavailabilities  get patient unavailabilities information about a specific treatment
PUT  /treatments/{uuid}/unavailabilities  add patient unavailability information about a treatment
GET  /treatment-plans/                    get the list of available treament plans

Database seeding service

The database state can be reinstalled from scratch using the dedicated service:

PUT /testing/database?dataset=... put a specific dataset on the DB

Currently, the following datasets are available:

initial-state       almost empty, with a treatment plan and a few patients
mid-state           same patients, a few appointments in the past, more in the future
contention          same patients but only two beds per week day (and nothing the weekend)
big                 200 patients to be scheduled. No appointment so far

In practice:

curl -X PUT http://pipas-persister.herokuapp.com/testing/database?dataset=...

Project structure

/config      private configuration files on specific deployments
/lib         ruby code of the logical database facade and RESTful interface
/migrations  database migrations files (managed by Sequel)
/resources   formal description of RESTful resources
/seeds       definition of test datasets
/tasks       Rake tasks (e.g. test:unit, db:migrate)
/test        Unit and acceptance tests
/views       WLang views used for documentation generation

Overview of Rake tasks

test              Run all tests
test:unit         Run unit tests only
test:acceptance   Run acceptance tests only

db:drop           Remove the database and database user
db:create         Create the database (empty) and its user
db:seed[from]     Install particular seeds (defaults to 'initial-state')
db:migrate        Run migrations to use the latest schema
db:rebuild        Rebuild the database from scratch (create + migrate + seed)

How to install?

Dependencies

• Make sure that you have a recent PostgreSQL (>= 9.2)
• Make sure that you have a recent ruby (>= 1.9.3)
• Make sure that you have the 'bundler' gem installed (gem install bundler)

Cloning the project and installing dependencies

git clone https://github.com/blambeau/pipas-persister.git
cd pipas-persister
bundle install --bin-stubs

Installing the database

• Make sure that postgresql administration tools are available (dropuser, createuser, dropdb, createdb)
• Make sure the current user may use them without providing a password (otherwise, see tasks/db.rake and change the pg_cmd method, for instance to use sudo su postgres -c '...')
• Copy config/database.example.yml to config/database.yml and update to match your configuration

Then, and only then,

bin/rake db:rebuild