Automated edits/Stereo

From OpenStreetMap Wiki
Jump to: navigation, search

Luxembourg address normalisation

To improve our addresses, match the official CACLR source, and prepare our address import and future QA work, I would like to perform a mechanical normalisation on the addr:housenumber tag in Luxembourg:

  • everything uppercase, e.g. "79a" becomes “79A"
  • remove whitespace, e.g. "79 - 79A" becomes "79-79A”
  • replace bad connectors for multiple addresses, e.g. “75/77” becomes “75,77”

Approximately 489 ways and 62 nodes will be affected when we run this. I plan on running this only once, using my normal account.

The script I’m using is at https://openstreetmap.lu/normalise-housenumbers.py

A filtered output, containing all the addresses (610 KB) is at https://openstreetmap.lu/housenumber-normalised-modified.osm

The full output, containing every object with an address in Luxembourg (38.3 MB) is at https://openstreetmap.lu/housenumber-normalised.osm

I would be very happy to get feedback on it, even if it’s “looks good to me"

add missing Serbian transliterations to place names in Kosovo

To improve the completeness of the data and match city limit signs, I would like to automatically add missing Serbian transliterations to placenames:

- Download places with a name:sr but no name:sr-Latn in Kosovo from overpass - Transliterate missing values using the cyrtranslit library from opendatakosovo - Changes are uploaded using JOSM

This will only affect place=* objects.

Approximately 1731 objects will be affected, 25 relations and 1706 nodes.

I plan on running this only once, using my normal account.

The script I'll use and a sample output can be found at https://gist.github.com/grischard/b90f57a898fdae3c788764b88803d850

I would be very happy to get feedback on it, even if it’s “looks good to me"