Spanish address (C/ vs Calle)

[dominikkukacka]

The spanish Spotify HQ address is:

C/ Eloy Gonzalo 27, 6ª planta
28010 Madrid
Spain

which fails in Pelias. If I replace the C/ with Calle I get the right result.

Calle Eloy Gonzalo 27, 6ª planta
28010 Madrid
Spain

[dominikkukacka]

I found this list of abbreviations: http://wiki.openstreetmap.org/wiki/Name_finder:Abbreviations#Espa.C3.B1ol_-_Spanish shouldn't they all be used by in the schema?

[dianashk]

Hey, @dominikkukacka! Thanks for reporting this. We definitely don't do a great job with abbreviations in other languages at the moment. Elasticsearch has some support for different languages and we'll need to figure out the right approach to handling these in the future. The abbreviations list you linked to will come in handy with that.

[zuzust]

Same thing for Catalan: http://wiki.openstreetmap.org/wiki/Name_finder:Abbreviations#Catal.C3.A0_-_Catalan

Hope to see it working in the near future.

[orangejulius]

Hi all, Long time, no discussion on this issue. We are adding better configurability around abbreviations/synonyms in https://github.com/pelias/schema/pull/310. It is now possible to add more synonyms such as c/calle, c/carrer, etc in a relatively organized way. Other people have mentioned Spanish and Catalan abbreviations, so this would be useful. It would still require some testing of course :)

[dpprdan]

Common German abbreviation are also not handled well, cf. "-str." (only city centroid found) vs. "-straße" (exact match). This is only a problem with compound streetnames, not separate ones, i.e. "Heisenberg Str. 2, Münster" returns an exact match.

Maybe add this to the Alt-Names milestone?

On second thought: Is this the same issue or a separate one, @orangejulius?

[missinglink]

Common German abbreviation are also not handled well

🟢 Heisenbergstraße 2, Münster 🟢 Heisenberg-straße 2, Münster 🟢 Heisenberg straße 2, Münster 🟢 Heisenbergstrasse 2, Münster 🟢 Heisenberg-strasse 2, Münster 🟢 Heisenberg strasse 2, Münster 🟢 Heisenberg str 2, Münster 🟢 Heisenberg str. 2, Münster 🔴 Heisenbergstr 2, Münster

You're correct that this one variant isn't currently handled well.

There are some tests here which cover this feature: https://github.com/pelias/model/blob/master/test/post/seperable_street_names.js

Compound German words such as 'stadtbibliothek' and 'straßenbahnhaltestelle' will always be difficult to handle, particularly when also considering their abbreviated forms.

[dpprdan]

🟢 Heisenbergstraße 2, Münster 🟢 Heisenberg-straße 2, Münster 🟢 Heisenberg straße 2, Münster 🟢 Heisenbergstrasse 2, Münster 🟢 Heisenberg-strasse 2, Münster 🟢 Heisenberg strasse 2, Münster

This is great!

These aren't abbreviations, though, so this is a separate issue, it seems?

🟢 Heisenberg str 2, Münster 🟢 Heisenberg str. 2, Münster

Great again, I didn't claim otherwise

This is only a problem with compound streetnames, not separate ones, i.e. "Heisenberg Str. 2, Münster" returns an exact match.

Which indeed leaves us with

🔴 Heisenbergstr 2, Münster

(And, to match the separated names above, 🔴 Heisenbergstr. 2, Münster)

As a native speaker, I'd say this is a pretty common abbreviation, though, if not the most common abbreviation used in German addresses (cf "st" for street in English). It can be found in the wild and both are already in the tests. (Only in expand, not in contract).

Other (compound) abbreviations also don't work, a few examples:

1. 🔴 Sievekingpl. 3, Hamburg / 🟢 Sievekingplatz 3, Hamburg
2. 🔴 Fischerbr. 2, Braunschweig / 🟢 Fischerbrücke 2, Braunschweig
3. 🔴 Jungfernstg. 2, Leipzig / 🟢 Jungfernstiege 2, Leipzig
4. 🔴 Oderstg. 2, Flensburg / 🟢 Oderstieg 2, Flensburg
5. 🔴 Elbch. 2, Hamburg / 🟢 Elbchaussee 2, Hamburg
6. 🔴 Querg. 1, Soltau / 🟢 Quergasse 1, Soltau

1-3 are from https://github.com/pelias/model/blob/master/test/post/seperable_street_names.js, 4-6 from the OSM list mentioned above.