Example: Streets of Berlin

In OpenStreetMap, a single street often comprises multiple ways. This segmentation occurs when there are variations in parking regulations, lane structures, or surface quality along its length. This can complicate a straightforward query such as “How long is this street?” — but the script below makes this a breeze.

In this example, we’ll fetch all the streets in Berlin, add up the lengths of their segment, and alphabetize the results.

import geodesk

germany = geodesk.Features("germany.gol")

berlin = germany("a[boundary=administrative][admin_level=4][name=Berlin]").one
streets = germany("w[highway][name]")

street_lengths = {}
for street in streets(berlin):
    name = street.name
    street_lengths[name] = street_lengths.get(name, 0) + street.length

for name, length in sorted(street_lengths.items()):
    print (f"{name}, {round(length)}")

Notes

• The highway tag encompasses “roads” in the broadest possible sense — from multi-lane freeways to dirt paths. You can narrow down the results by explicitly listing the highway types, such as highway=primary,secondary,tertiary,residential

• We can’t use street_lengths[name] += street.length, as it requires that the key is already present in the dictionary (otherwise, a KeyError is raised). The get() method doesn’t have this restriction and allows us to specify a default value (0)