Canonical features for OpenStreetMap

The goal of this project is to maintain a canonical list of commonly used features for suggesting consistent spelling and tagging in OpenStreetMap.

Watch the video from our talk at State of the Map US 2019 to learn more about this project!

You can browse the index at https://nsi.guide/ to see which features are missing Wikidata links, or have incomplete Wikipedia pages.

When mappers create features in OpenStreetMap, they are not always consistent about how they name and tag things. For example, we may prefer McDonald's tagged as amenity=fast_food but we see many examples of other spellings (Mc Donald's, McDonalds, McDonald’s) and taggings (amenity=restaurant).

Building a canonical feature index allows two very useful things:

• We can suggest the most "correct" way to tag things as users create them while editing.
• We can scan the OSM data for "incorrect" features and produce lists for review and cleanup.

The name-suggestion-index is in use in iD when adding a new item

Currently used in:

• iD (see above)
• Vespucci
• JOSM presets available
• Osmose
• Go Map!!, as they use iD presets

• Read the project Code of Conduct and remember to be nice to one another.
• See CONTRIBUTING.md for info about how to contribute to this index.

We're always looking for help! If you have any questions or want to reach out to a maintainer, ping bhousel on:

• OpenStreetMap US Slack (#poi or #general channels)

• Node.js version 10 or newer
• git for your platform

• Clone this project, for example: git clone git@github.com:osmlab/name-suggestion-index.git
• cd into the project folder,
• Run npm install to install libraries

The files under dist/* are generated:

• dist/collected/* - Frequently occuring tags collected from OpenStreetMap
• dist/filtered/* - Subset of tags that we are keeping or discarding
• dist/presets/* - Preset files for use in OSM editors, custom JOSM presets, etc.
• dist/taginfo.json - List of all tags this project uses (see: https://taginfo.openstreetmap.org/)
• dist/wikidata.json - Cached data retrieved from Wikidata

The files under config/*, data/*, and features/* can be edited:

• config/*:
  • config/filter_*.json - Regular expressions used to filter the OpenStreetMap tags into keep/discard lists
  • config/match_groups.json - Groups of OpenStreetMap tags that are considered equivalent for purposes of matching
• data/* - Data files for each kind of feature, organized by topic and OpenStreetMap tag
  • data/brands/amenity/*.json
  • data/brands/shop/*.json
  • data/transit/route/*.json
  • and so on…
• features/* - GeoJSON files that define custom regions where the features are allowed
  • hawaii.geojson
  • quebec.geojson
  • scotland.geojson
  • and so on…

👉 See CONTRIBUTING.md for info about how to contribute to this index.

• npm run build
  • Processes any custom locations under features/**/*.geojson
  • Regenerates dist/filtered/* keep and discard lists
  • Any new items from the keep list not already present in the index will be merged into it
  • Outputs many warnings to suggest updates to data/**/*.json

https://nsi.guide/ is a web application written in ReactJS that lets anyone browse the index.

• The source code for this app can be found under app/*
• npm run appbuild will rebuild it.

• npm run wikidata - Fetch useful data from Wikidata - labels, descriptions, logos, etc.
• npm run dist - Rebuild and minify the generated files in the dist/ folder.
• npm run - Lists other available commands

This takes a long time and a lot of disk space. It can be done occasionally by project maintainers. You do not need to do these steps in order to contribute to the index.

• Install osmium commandline tool and node package (may only be available on some environments)
  • apt-get install osmium-tool or brew install osmium-tool or similar
  • npm install --no-save osmium
• Download the planet
  • curl -L -o planet-latest.osm.pbf https://planet.openstreetmap.org/pbf/planet-latest.osm.pbf
• Prefilter the planet file to only include named items with keys we are looking for:
  • osmium tags-filter planet-latest.osm.pbf -R name,brand,operator,network -o filtered.osm.pbf
• Run node collect_all.js /path/to/filtered.osm.pbf
  • results will go in dist/collected/*.json
• A new challenge:
  • Attempt an npm run build. Now that unique id properties are generated, it is possible that this command will fail.
  • This can happen if there are multiple new items that end up with the same id (e.g. "MetroBus" vs "Metrobus")
  • You'll need to just pick one to keep, then keep trying to run npm run build until the duplicate id issues are gone.
  • git add . && git commit -m 'Collected common names from latest planet'

name-suggestion-index is available under the 3-Clause BSD License. See the LICENSE.md file for more details.