Command line tool to extract pathways as GeoJSON from a GTFS dataset.
npm install -g extract-gtfs-pathwaysUsage:
extract-gtfs-pathways <path-to-pathways-file> <path-to-stops-file> <output-directory>
Options:
--quiet -q Don't log the written files.
--pathway-props -f A JS function to determine additional pathway properties.
Example: pw => ({isWalking: pw.pathway_mode === '1'})
Note: The argument will be eval-ed!
--node-props -F A JS function to determine additional node properties.
Example: n => ({isFoo: n.stop_id === 'foo'})
Note: The argument will be eval-ed!
Examples:
unzip -j gtfs.zip -d data/gtfs
mkdir extracted-pathways
extract-gtfs-pathways data/gtfs/pathways.txt data/gtfs/stops.txt extracted-pathways
Notes:
This tool will read a reduced form of stops.txt into memory.
stops.txt needs to be sorted by
1. parent_station: lexically ascending, empty first
2. location_type: numerically descending, empty first
You can use Miller (https://miller.readthedocs.io/) and the
Unix tool sponge to do this:
mlr --csv sort -f parent_station -nr location_type \
stops.txt | sponge stops.txt
You can use the docker.io/derhuerst/extract-gtfs-pathways Docker image.
Remember to mount (using -v) the directory you're passing as <output-directory> (see above), otherwise the file will be created within the Docker container's virtual file system (and removed after the extraction has finished).
unzip -j gtfs.zip -d gtfs # directory with GTFS input
mkdir extracted-pathways # output directory
docker run -it --rm \
-v $PWD/gtfs:/gtfs \
-v $PWD/extracted-pathways:/extracted-pathways
docker.io/derhuerst/extract-gtfs-pathways /gtfs/pathways.txt /gtfs/stops.txt /extracted-pathwaysBy default, extract-gtfs-pathways adds some style properties to the generated features, following the Mapbox GL JS naming scheme, e.g. line-color, line-width & circle-radius. If you open them with a tool that supports these style properties – e.g. view-geojson, it will be easier to tell the nodes & pathways apart.
You can use --pathway-props/-f and --node-props/-F to customize the pathways' and nodes' properties, respectively. As an example, let's define two functions that override some default properties:
const WALKWAY = '1'
const ESCALATOR = '4'
const ELEVATOR = '5'
const pwOpacities = {[WALKWAY]: .3, [ESCALATOR]: .8, [ELEVATOR]: 1}
const pathwayProps = (pw) => ({
'line-opacity': pwOpacities[pw.pathway_mode] || .5,
'line-width': 2,
})
const STOP = '0' // or empty
const ENTRANCE_EXIT = '2'
const BOARDING_AREA = '4'
const nodeColors = {
[STOP]: '#ff0000', '': '#ff0000',
[ENTRANCE_EXIT]: '#00ff00',
[BOARDING_AREA]: '#0000ff',
}
const nodeProps = (n) => ({
'circle-color': nodeColors[n.location_type] || '#444444',
})We minify the functions and declare them as Bash variables:
todo: make these the default?
pw_props='pw => ({"line-opacity": {"1": .3, "4": .8, "5": 1}[pw.pathway_mode] || .5, "line-width": 2})'
node_props='n => ({"circle-color": {"0": "#ff0000", "": "#ff0000", "2": "#00ff00", "4": "#0000ff"}[n.location_type] || "#444444"})'Then, we can use them:
extract-gtfs-pathways --pathway-props "$pw_props" --node-props "$node_props" gtfs/pathways.txt gtfs/stops.txt outAlternatively, we can also use geojson.io's naming scheme:
pw_props='pw => ({"stroke-opacity": {"1": .3, "4": .8, "5": 1}[pw.pathway_mode] || .5, "stroke-width": 2})'
node_props='n => ({"marker-color": {"0": "#ff0000", "": "#ff0000", "2": "#00ff00", "4": "#0000ff"}[n.location_type] || "#444444"})'- view-geojson – View an GeoJSON in the browser that has been piped into it.
- extract-gtfs-shapes – Command-line tool to extract shapes from a GTFS dataset.
- Awesome GTFS: Frameworks and Libraries – A collection of libraries for working with GTFS.
If you have a question or need support using extract-gtfs-pathways, please double-check your code and setup first. If you think you have found a bug or want to propose a feature, use the issues page.
