No description
Find a file
2019-04-23 23:02:57 +02:00
examples add comments to example travelynx.conf 2019-04-16 21:13:14 -04:00
lib Fix checkin/checkout of cancelled journeys 2019-04-23 23:02:57 +02:00
public/static Primitive auto-refresh of current journey status 2019-04-23 22:27:13 +02:00
t Finish transition from DBI to Mojo::Pg 2019-04-22 13:42:41 +02:00
templates Primitive auto-refresh of current journey status 2019-04-23 22:27:13 +02:00
.gitignore Use travelynx.conf for configuration and secrets 2019-04-13 12:18:26 +02:00
.travis.travelynx.conf Travis CI: Add database tests 2019-04-22 07:01:59 +02:00
.travis.yml Note that we need at least postgresql v9.4 2019-04-22 07:22:53 +02:00
cpanfile Finish transition from DBI to Mojo::Pg 2019-04-22 13:42:41 +02:00
cpanfile.snapshot Add cpanfile for dependency management with Carton 2019-04-21 18:15:43 +02:00
index.pl Initiate transition to a Mojolicious MVC web application 2019-03-21 22:12:45 +01:00
README.md Finish transition from DBI to Mojo::Pg 2019-04-22 13:42:41 +02:00

travelynx - Railway Travel Logger

travelynx allows checking into and out of individual trains, thus providing a log of your railway journeys annotated with real-time delays and service messages. At the moment, it only supports german railways and trains which are exposed by the Deutsche Bahn IRIS Interface.

Dependencies

  • perl >= 5.10
  • Cache::File (part of the Cache module)
  • Crypt::Eksblowfish
  • DateTime
  • DateTime::Format::Strptime
  • Email::Sender
  • Geo::Distance
  • Mojolicious
  • Mojolicious::Plugin::Authentication
  • Mojo::Pg
  • Travel::Status::DE::IRIS
  • UUID::Tiny
  • JSON
  • Geo::Distance::XS (speeds up statistics)
  • JSON::XS (speeds up API and statistics)

Dependencies On Docker

  • cpanminus
  • build-essential
  • libpq-dev
  • git
  • ssmtp

Setup

First, you need to set up a PostgreSQL database so that travelynx can store user accounts and journeys. It must be at least version 9.4 and should use a UTF-8 locale. The following steps describe setup on a Debian 9 system, though setup on other distribution should be similar.

  • Write down a strong random password
  • Create a postgres user for travelynx: sudo -u postgres createuser -P travelynx (enter password when prompted)
  • Create the database: sudo -u postgres createdb -O travelynx travelynx
  • Copy examples/travelynx.conf to the application root directory (the one in which index.pl resides) and configure it
  • Initialize the database: perl index.pl database migrate

Your server also needs to be able to send mail. Set up your MTA of choice and make sure that the sendmail binary can be used for outgoing mails. Mail reception on the server is not required.

Finally, configure the web service:

  • Set up a travelynx service using the service supervisor of your choice (see examples/travelynx.service for a systemd unit file)
  • Configure your web server to reverse-provy requests to the travelynx instance. See examples/nginx-site for an nginx config.

You can now start the travelynx service, navigate to the website and register your first account.

Please open an issue on https://github.com/derf/travelynx/issues or send a mail to derf+travelynx@finalrewind.org if there is anything missing or ambiguous in this setup manual.

Updating

It is recommended to run travelynx directly from the git repository. When updating, the workflow depends on whether schema updates need to applied or not.

git pull
chmod -R a+rX . # only needed if travelynx is running under a different user
if perl index.pl database has-current-schema; then
    systemctl reload travelynx
else
    systemctl stop travelynx
    perl index.pl database migrate
    systemctl start travelynx
fi

Note that this is subject to change -- the application may perform schema updates automatically in the future.

Usage

For the sake of this manual, we will assume your travelynx instance is running on travelynx.de

travelynx journey logging is based on checkin and checkout actions: You check into a train when boarding it, and check out again when leaving it. Real-time data is saved on both occasions, providing an accurate overview of both scheduled and actual journey times.

Checking in

You can check into a train up to 10 minutes before its scheduled departure and up to 3 hours after its actual departure (including delays). I recommend doing so when it arrives at the station or shortly after boarding.

First, you need to select the station you want to check in from. Navigate to travelynx.de or click/tap on the travelynx text in the navigation bar. You will see a list of the five stations closest to your current location (as reported by your browser). Select the station you're at or enter its name or DS100 code manually.

Now, as soon as you select a train, you will be checked in and travelynx will switch to the journey / checkout view.

Checking out

You can check out of a train up to 10 minutes before its scheduled arrival and up to 3 hours after its actual arrival. This ensures that accurate real-time data for your arrival is available. I recommend checking out when arriving at your destination or shortly after having left the train.

Once checked in, travelynx.de will show a list of all upcoming stops. Select one to check out there. You can also check out at a specific station by navigating to "travelynx.de/s/station name" and selecting "Hier auschecken".

If you forgot to check out in time, or are departing the train at a station which is not part of its documented route (and also not part of its documented route deviations), or are encountering issues with travelynx' real-time data fetcher, the checkout action will fail with an error message along the lines of "no real-time data available" or "train not found".

If you use the checkout link again, travelynx will perform a force checkout: it will log that you have left the train at the specified station, but omit arrival time, delay, and other real-time data. At the moment, this data cannot be specified manually.

Testing

The test scripts assume that travelynx.conf contains a valid database connection. They will create a test-specific schema, perform all operations in it, and then drop the schema. As such, the database specified in the config is not affected.

Nevertheless, bugs may happen. Do NOT run tests on your production database. Please use a separate development database instead.

Run the tests by executing prove.