A browser-based user interface for creating, editing, exporting of static GTFS (General Transit Feed Specification Reference) feeds for a public transit authority.

Development Status : V 3.2.0
Windows binary is available too now. Download from Releases page.

This project is the result of a collaboration between WRI (World Resources Institute) and KMRL (Kochi Metro Rail Limited).

Initially developed for use by KMRL, the source code has been open-sourced so it can grow and get better with community inputs and help for creating GTFS feeds for other transit agencies too.

The GTFS data pre-loaded in the program is of Kochi Metro, Kerala, India which on March 16, 2018 became India's first transit agency to release its schedules data in the global standard static GTFS format as free and open data.

See the KMRL open data portal and some news articles: 1, 2, 3, 4.

This program adheres to the static GTFS (General Transit Feed Specification Reference) open transit data specs as published by Google Transit here: https://developers.google.com/transit/gtfs/reference/
It also implements a GTFS extension for translations of stops and routes names to facilitate multilingual use of the data.

Lead programmer up till April 2018: Nikhil VJ from Pune, India.

See this and many more GTFS related resources listed on Awesome Transit, a one-stop community listing for all things GTFS.

• GTFS Manager Runthrough video
• Demonstration of XML Import feature
• Demonstration of HMRL's CSV Import feature

Psst! Password is program.

1. (Updated, Oct 2018) Download the repo, unzip, and double-click GTFSManager.exe in the main program folder. It'll have an icon.

2. That's it, that should start the program! A dos box should open up giving console messages, and in a few seconds a new tab should open in your system's default web browser with the program loaded.

3. There will probably be a Windows Firewall popup. Click OK or cancel.. this program doesn't do any to-and-fro communication over internet, that's just a default question that comes up when a server is activated. (In our case we're creating a portable web server for the front-end and back-end partd of the program to communicate). It does need internet for loading the background tiles on web maps, but other than that the program can run without internet connection.

4. More notes on this, including some troubleshooting and watchouts, on the wiki.

See on the project wiki: Running on Ubuntu OS

See on the project wiki: Running on Windows OS with Python 3

Check out the repo and run

docker-compose up

See on the project wiki: Changind the password

The end output of this program is a gtfs.zip file having your transit agency's static GTFS data. See the Commit and Export GTFS section on the main page for the same.

From the main page, you can import your GTFS feed in a .zip file. Structure of a feed zip must be as per standard GTFS specs.

gtfs.zip
 ˪ stops.txt
 ˪ routes.txt
 ˪ trips.txt
 ˪ stop_times.txt
 ˪ ...

Some requirements, s'il vous plait !

1. Keep all your files with lowercase .txt extension
2. Keep them up at root level in the zip archive, not inside any folders.
3. The table in stop_times.txt must be sorted by trip_id (you can open it in LibreOffice, or rename as .csv and open in Excel) before importing.
4. Similarly, the table in shapes.txt must be sorted by shape_id before importing.

Kindly validate your GTFS zip prior to importing so you know in advance if there are any issues with the feed. And no worries, this tool is made FOR fixing bad feeds, the system will import whatever you give it. If you need to delete a lot of junk data in the feed, the Tools > Delete ID section will be your favorite place to be.

In the import process, the program creates a backup ZIP of the current data and then imports your data into its database. You can see the backup listed later on under Past Commits section on the home page.

Please see the Issues section for seeing existing program improvement efforts, feedback, questions. Please make sure you search through all the issues (click here for full list) before filing a new one : it might already be covered in another.

Invitation: This project invites active participation from professionals in the coding and GTFS fields to join in to take it forward. Please feel free to fork, write your fixes/enhancements and create a pull request.

• Schedules with frequencies currently not supported, this is desired as for some systems their schedules may be frequency-based rather than fixed times, plus using frequency greatly reduces the database/feed size by removing repetitive entries from stop_times table.
• Shapes are only accepted in .geojson format. The first entry in the file will be picked up and others will be discarded.
• Doesn't support calendar_dates.txt yet. For that and all such extra files: If you've imported it, it'll be there in the DB, and will be there in the feed when you export again. Just haven't gotten around to making their UI yet. Code contributions welcome in that regard.

Moved to Wiki: Technical Overview

This project stands on the shoulders of several solutions that have been shared open source. Sharing mentions below.

• Leaflet.js and its various plugins for plotting information on maps and keeping things interactive
• Tabulator.js for handling data in tabular form at the front-end
• Bootstrap for general page design
• Jquery and Jquery UI for UI components like accordions, event handlers and file upload.
• Papa.parse for CSV parsing
• Chosen.js for search-as-you-type dropdowns

• tornado for simple web server with asynchronous callback
• pandas for handling practically all the data to-and-fro, and for very efficient functions like unpivoting.
• HDF5 format support (tables) for enabling pandas to work with high compression and high read/write speed .h5 format files.
• tinyDB for portable JSON database
• pycryptodomex for providing password encryption without the hiccups
• Many python modules for various operations: json, os, time, datetime, xmltodict, collections, zipfile, webbrowser, pycrptodomex, shutil, pathlib, math, json.decoder

In addition to this, there are several code snippets used throughout the program that were found from online forums like stackoverflow and on various tech blogs. The links to the sources of the snippets are mentioned in comments in the program files. Here is a shoutout to all the contributors on these forums and blogs : Thank You!

Big thanks to :

• Srinivas from Hyderabad, India for connecting folks together and sharing guidance
• Devdatta from Pune, India for a sharing a very simple working example to learn about Tornado web server serving asynchronous requests
• Tony Laidig from the global GTFS community for helping build a docker image
• Sean Barbeau and other co-ordinators of the GTFS community for including this program on Awesome Transit and giving encouragement and feedback on the GTFS forum.