This library is able to divide (German) a concatenated street names consisting of the pure street name, the house number, and its affix into its single parts.

Some streets contain a number as suffix or infix themselves so that it is impossible to decide whether this number is a house number or belongs to the street name. An example is "Straße 101" in Berlin. At first glance this seems to be a street "Straße" with house number "101", but in reality the number "101" is part of the street name. To make this decision unambiguous the street "Straße 101" is added to a list of "special streets". This list can be found in the file "specialstreets.txt" and can/must be updated periodically.

More Examples:

input	street	house no	affix
Gartenstr. 25a	Gartenstr.	25	a
Brückenstr. 12a-13c	Brückenstr.	12	a-13c
Straße 101 Nr. 12	Straße 101	12
C 3 54	C 3	54
In den 30 Morgen 34b	In den 30 Morgen	34	b

In the examples above you can see the very strange street names of the city Manheim. The inner of the city in organized as a matrix with names like A4 or C3.

Usage

val streetDivider = StreetDivider()
println(streetDivider.parse("Gartenstr. 25a"))

-> Location(street=Gartenstr., houseNumber=25, houseNoAffix=a)

Try it out

You can download (downloads/streetdivider.jar) a small executable jar with UI and test the street decomposition with your own street names...