Contains all metro🚇, tram🚊 and non-regional bus lines🚍. It doesn't take into account specific day, only time of the day.
To use this search, clone this repo and build and run the only cmake target
available JourneyPlanner.
I recommend using the release mode since loading all the data takes a little while.
The usage is pretty straightforward - enter the name of the start stop🚏, end stop🚏 and the departure time🕑 and you'll get the connection if it exists. If you enter just a part of the stop name, you might get a hint of the name you want.
The search uses Raptor algorithm presented at https://www.microsoft.com/en-us/research/wp-content/uploads/2012/01/raptor_alenex.pdf
or in longer version at https://renatowerneck.files.wordpress.com/2016/06/dpw14-raptor.pdf.
The most important thing about the algorithm is that it is not based on Dijkstra algorithm or a graph,
but it takes advantage of the GTFS format of the data (https://developers.google.com/transit/gtfs/reference/).
The specific data used in this project come from Prague public transport (available with some documentation at https://pid.cz/en/opendata/), but the data are preprocessed and transformed for my purposes using the 🐍 preprocess.py script.
In order to understand how this program really works, it is kind of necessary to read about the Raptor algorithm and about the GTFS format, but I tried to provide a lot of comments in the code, so you can at least have a general idea of how it works.
Also, following is some information about all source files:
- These files contain
Stop,RouteandTripclasses, which are essential terms in GTFS
- Timetable class - provides an interface for getting all the data (the stops, routes, trips and other stuff)
- Raptor class - this class does the whole search, recreates the connection (gets all the lines used) and shows the result
- InputReader class - reads and takes care of all user input
- the entry point, just merges everything together