[OSM-talk] TIGER data abbreviations

[Matthias Julius]

"Karl Newman" <siliconfiend at gmail.com> writes:

> On Dec 4, 2007 1:39 AM, Martijn van Oosterhout <kleptog at gmail.com> wrote:
>> On Dec 4, 2007 1:32 AM, Karl Newman <siliconfiend at gmail.com> wrote:
>> > I've been leaving them alone so far. See my mail from earlier today
>> > about componentizing the names, which may be a potential solution.
>>
>> I'm wondering how componentised names are going to work for languages
>> where the street type is not written as a seperate word. Examples
>> would be:
>>
>> Van Mierenveldtlaan
>> Maria van Jesselaan
>> Houthaak
>> Raam
>>
>> In this case "laan" is the street type, the rest is the name,
>> including spaces. The last two have no street type at all. And I'm not
>> sure how you'd consider "Oostplantsoen". A "plantsoen" is a kind of
>> garden, would that count as a street type?
>>
>
> Obviously there are going to be cases where this breaks down. So where
> there is no street type, maybe that tag would be blank? And some other
> way to indicate that it should be abutted with the base_name. In
> German, I've seen abbreviations of the street type even when they're
> stuck together with the name (to take a couple examples from Vienna:
> Stephansplatz and Landstrasse become Stephanspl. and Landstr.) But I
> think there's value here, for several purposes (including the name
> finder). I know it's difficult, but we're a smart group, right?

At least for TIGER data this should be less of a problem. Here is one
example:

tiger|name_direction_prefix = W
tiger|name_base = Liberty
tiger|name_type = St

>From those tags it should be relatively easy to reassemble the name
using a look-up table for prefix and type. This should be pretty
unambiguous.

Matthias