mauriceqch / edt-utc

Web application to parse and convert UTC timetables

Notes

This repository is unmaintained and probably outdated by now. It is kept online as it can be an interesting starting point to understand the text format.

Ruby version

This app uses Ruby 2.2.0.

Deployment on Heroku

$ heroku login
$ heroku create
$ git push heroku master
$ heroku run rake db:reset

Local setup

$ bundle install
$ rake db:reset
$ rails server

How does it work ?

The main objective of this application is to parse the timetable text sent by the University of Technology of Compiègne. Here is a sample text :

EI03       C 1    LUNDI... 18:45-19:45,F1,S=RN104
EI03       D 1    SAMEDI..  8:15-12:15,F1,S=RN104

LG30       D 2    MERCREDI 10:15-12:15,F1,S=FA413
LG30       T 3 A  LUNDI... 13:00-14:00,F1,S=FA410

SR01       C 1    LUNDI... 14:15-16:15,F1,S=FA104
SR01       D 1    LUNDI... 16:30-18:30,F1,S=FA506

MT12       C 1    LUNDI...  8:00-10:00,F1,S=FA202
MT12       D 1    MARDI... 16:30-18:30,F1,S=FA518

The format is the following :

COURSECODE       TYPE GROUPNUMBER WEEK   DAYOFTHEWEEK STARTHOUR-ENDHOUR,FREQUENCY,S=CLASSROOM

Currently, the parsing is done using this regex :

(\w+)\s*([a-zA-Z]+)\s*(\d*)\s*([A|B]{0,1})\s*(\w+)[\.]*\s*(\d+:\d+)-(\d\d:\d\d),F(\d),S=\s*(\w*).*

The application tries to match the regex with each of the lines after some pre-processing. If the regex fails to recognize the text in the line, then a fallback method tries to parse the line using ruby.

Each line is converted into a non-persisted (not stored in database) Course object.

Authentication

The application uses Devise and the University's CAS to handle authentication.