Bias on name field match

[xeruf]

First of all, thanks for the great tool, this is saving me today!

In an exemplary query I encountered, there is some gibberish before "Hamburg". But regardless of the language parameter, it matches "Novo Hamburgo" higher, ignoring the exact match of the word Hamburg. Maybe this can be nudged slightly, or do you think this result is somehow appropriate? RElated to #193

❯ curl 'https://photon.komoot.io/api/?q=Luftfrachthof-Geb%20Hamburg&limit=1' | jq '.features[0].properties'
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   435    0   435    0     0   1329      0 --:--:-- --:--:-- --:--:--  1334
{
  "osm_type": "R",
  "osm_id": 242616,
  "extent": [
    -51.1787892,
    -29.6378807,
    -50.9341246,
    -29.8302389
  ],
  "country": "Brasil",
  "osm_key": "place",
  "countrycode": "BR",
  "osm_value": "city",
  "name": "Novo Hamburgo",
  "county": "Região Geográfica Intermediária de Porto Alegre",
  "state": "Rio Grande do Sul",
  "type": "city"
}

❯ curl 'https://photon.komoot.io/api/?q=Luftfrachthof-Geb%20Hamburg&lang=de&limit=2' | jq '.features | map(.properties)'
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   782    0   782    0     0   2821      0 --:--:-- --:--:-- --:--:--  2823
[
  {
    "osm_type": "R",
    "osm_id": 242616,
    "extent": [
      -51.1787892,
      -29.6378807,
      -50.9341246,
      -29.8302389
    ],
    "country": "Brasilien",
    "osm_key": "place",
    "countrycode": "BR",
    "osm_value": "city",
    "name": "Novo Hamburgo",
    "county": "Região Geográfica Intermediária de Porto Alegre",
    "state": "Rio Grande do Sul",
    "type": "city"
  },
  {
    "osm_type": "N",
    "osm_id": 1680910488,
    "country": "Deutschland",
    "osm_key": "railway",
    "city": "Hamburg",
    "street": "Nordsteg",
    "countrycode": "DE",
    "district": "St. Georg",
    "osm_value": "station",
    "postcode": "20099",
    "name": "Hamburg Hauptbahnhof",
    "type": "house"
  }
]

❯ curl 'https://photon.komoot.io/api/?q=Hamburg&lang=de&limit=1' | jq '.features[0].properties'
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   332    0   332    0     0    277      0 --:--:--  0:00:01 --:--:--   277
{
  "osm_type": "R",
  "osm_id": 62782,
  "extent": [
    8.1044993,
    54.02765,
    10.3252805,
    53.3951118
  ],
  "country": "Deutschland",
  "osm_key": "place",
  "countrycode": "DE",
  "osm_value": "city",
  "name": "Hamburg",
  "type": "city"
}

[xeruf]

Maybe a flag whether this query is for geocoding (treat query as complete) or for autocompletion would be useful.

[lonvia]

I can't reproduce that result. It might have been just a transient data error. Results always contain the city of Hamburg for me as first result.

[xeruf]

Just entered the exact command above again, still receiving Novo Hamburgo

[lonvia]

This is once more a problem with lang=default: https://photon.komoot.io/api/?q=Luftfrachthof-Geb%20Hamburg&limit=1&lang=default

[lonvia]

I've pushed an experimental version of Photon to https://photon.komoot.io which reduces the boost for default slightly. Lets see if it causes other complaints of results becoming worse.

As a general rule, sending a lang= parameter with your queries when using them from scripts, is probably always a good idea.