Awesome

Validate GTFS files

The General Transit Feed Specification (GTFS) defines a common format for public transportation schedules and associated geographic information.

This project is a validating tool for such files and can perform checks ranging from simple ones (the archive is not valid, a file is missing) to more complex ones (a vehicle is moving too fast).

Online tool

transport-validator is the tool used by the French National Access Point to validate GTFS files. If you want to use it online, you can validate your own files at this address.

Validation output

Validation output is twofold:

• it gives useful information about the validated file, under the metadata entry
• it lists a serie of validation items, with a corresponding severity, under the validations entry. When relevant, geographical data (GeoJSON) related to the issue is given to ease file debugging.

The output is by default formatted in json, but yaml is also available. See Options for more information.

{
    "metadata": {
        ...
    },
    "validations": {
       ...
    }
}

Metadata

Give useful information about the validated file content:

There is also a stats object inside with various statistics about the data:

Note: For the stops_with_wheelchair_info_count, the information can be specified at the stop level (wheelchair_boarding equal to 1 or 2), or inherited from its parent station. Can be null if the GTFS contains errors preventing to compute this field.

Example

    "metadata": {
        "start_date": "2020-11-02",
        "end_date": "2022-01-31",
        "networks": [
            "carSud"
        ],
        "modes": [
            "bus", "tramway"
        ],
        "issues_count": {
            "ExcessiveSpeed": 5,
            "CloseStops": 10,
            "NullDuration": 10,
            "MissingName": 1,
            "MissingCoordinates": 1215,
            "InvalidCoordinates": 1215,
            "DuplicateStops": 49,
            "IdNotAscii": 171
        },
        "stats": {
            "stops_count": 17,
            "stop_areas_count": 2,
            "stop_points_count": 9,
            "stops_with_wheelchair_info_count": null,
            "lines_count": 5,
            "routes_with_custom_color_count": 0,
            "routes_with_short_name_count": 0,
            "routes_with_long_name_count": 4,
            "trips_count": 11,
            "trips_with_bike_info_count": 3,
            "trips_with_wheelchair_info_count": 3,
            "trips_with_shape_count": 0,
            "trips_with_trip_headsign_count": 9,
            "transfers_count": 0,
            "fares_attribute_count": 2,
            "fares_rules_count": 4
        },
        "has_fares": true,
        "has_shapes": true,
        "has_pathways": false,
        "some_stops_need_phone_agency": false,
        "some_stops_need_phone_driver": false
    }

Validations

The "validations" key contains the actual validation results.

Severity

Each check is associated with a severity level.

List of checks

The validator performs a number of checks. The list of checks can be seen in the file issues.rs.

Here is a human friendly list of them :

Geojson information

When relevant for the check, geojson information is added for each check output, making the GTFS debug process easier.

Example

Here is a validation output containing one warning, triggered by a non Ascii Stop id:

"validations": {
    "IdNotAscii" : [
        {
            "severity": "Warning",
            "issue_type": "IdNotAscii",
            "object_id": "AllBél",
            "object_type": "Stop",
            "object_name": "",
            "related_objects": [],
            "geojson": {
                "features": [
                    {
                        "geometry": null,
                        "properties": {
                            "id": "AllBél",
                            "name": ""
                        },
                        "type": "Feature"
                    }
                ],
                "type": "FeatureCollection"
            }
        }
    ]
}

Another example showing the GeoJSON information for an information about two stops too close:

    "validations": {
        "CloseStops": [
            {
                "severity": "Information",
                "issue_type": "CloseStops",
                "object_id": "PH00320P",
                "object_type": "Stop",
                "object_name": "Baril Les Hauts",
                "related_objects": [
                    {
                        "id": "PH00320C",
                        "object_type": "Stop",
                        "name": "Baril Les Hauts"
                    },
                    {
                        "id": "MAGM",
                        "object_type": "Route",
                        "name": "MAGM-MagmaBus Navette Centre Ville St Philippe"
                    }
                ],
                "details": "distance between the stops is 0 meter(s)",
                "geojson": {
                    "features": [
                        {
                            "geometry": {
                                "coordinates": [
                                    55.71866572404381,
                                    -21.356751531407003
                                ],
                                "type": "Point"
                            },
                            "properties": {
                                "id": "PH00320P",
                                "name": "Baril Les Hauts"
                            },
                            "type": "Feature"
                        },
                        {
                            "geometry": {
                                "coordinates": [
                                    55.71866572404381,
                                    -21.356751531407003
                                ],
                                "type": "Point"
                            },
                            "properties": {
                                "id": "PH00320C",
                                "name": "Baril Les Hauts"
                            },
                            "type": "Feature"
                        },
                        {
                            "geometry": null,
                            "properties": {
                                "details": "distance between the stops is 0 meter(s)"
                            },
                            "type": "Feature"
                        }
                    ],
                    "type": "FeatureCollection"
                }
            }
        ]
    }

Installation

1. This project is written in Rust. You need to first install Rust on your machine.

2. Clone the project:

git clone https://github.com/etalab/transport-validator/
cd transport-validator

Run the validator

Run from a local directory

The release version can be run as:

cargo run --release -- --input test_data/unused_stop

It is the fastest and recommanded version to use when validating files.

If you are developping on the validator, you can also run a debug build:

cargo run -- --input test_data/unused_stop

The validator can read a zip file, or an url:

cargo run --release -- -i some_gtfs.zip
cargo run --release -- -i https://example.com/network.gfts

If you do not intend to run the validator as a dæmon, it can be compiled without dæmon support, saving on compile time and binary size:

cargo run --release --no-default-features -- -i some_gtfs.zip

Run as a dæmon

The validator can run as a HTTP dæmon to validate any file from a url.

For now the call is synchronous. Be aware that if the file is large, the time required to download the GTFS zip, the request might time out.

The command to launch the dæmon is:

cargo run --release

You can then ask for a validation:

curl http://localhost:7878/validate?url=https://example.com/gtfs.zip

Options

• --input or -i: Path (can be a directory or a zip file) or HTTP URL (file will be downloaded) of the GTFS file.
• --max-issues or -m: The maxium number of issues per type. Defaults to 1000.
• --output-format or -f: Output format (when using the validator in command line). Value by default is json, but yaml is also available.
• --custom-rules or -c: Path to a YAML file containing custom values to use during the validation.

Custom rules

Some values used during the validations can be customized by using the --custom-rules option and providing a path to a YAML file.

Available customizations are:

Example

max_bus_speed: 120
max_gondola_speed: 50

If you need to customize other values, please let us know.

Lint

To lint our code we use rustfmt

Install it running:

rustup component add rustfmt-preview

Lint your code running:

cargo fmt --all -- --write-mode=diff

Contact

Questions? Comments? Get in touch with the transport.data.gouv.fr's technical team: tech@transport.beta.gouv.fr

Alternatives

• Google Transitfeed
• Conveyal/Catalogue’s Datatool
• Chouette