Custom vector maps with accents

[Alder]

Hello everyone,

I was able to configure my custom vector map with Brazilian cities.
I have an index that one of the fields is the city and I was able to make a join between the index field and the custom vector map.
But it didn't work for cities that have accents.

example:
In my index has the city "Belem" and in the custom vector maps has the city "Belém"

How do I do to the join work with accented words?

[flash1293 (Joe Reuter)]

Hi @Alder

how exactly do you join the data? Could you provide your code and a few sample documents?

[Alder]

I am using the "custom vectror shapes" of the "maps" feature. I attached an image with join. Cities that do not have accents are ok.
The custom vector shape was built based on the file br_muns.geojson from the link below:

https://download2.exploratory.io/maps/br_muns.zip

[thomasneirynck]

Hi @Alder,

The term-join uses an exact string match, and the city-name with an accent and the city-name without are different. This is why this is failing.

I would consider following work-arounds:

• Do you have an ID-field in your index that you can use to join on? That br_muns.geojson file seems to have a few ID-fields. Are there matching ids in the documents of your index?
• Could you add a field to the features in br_muns.geojson where all special characters are replaced? Here is an example in Javascript to replace accented characters with the unaccented one https://stackoverflow.com/questions/990904/remove-accents-diacritics-in-a-string-in-javascript/37511463#37511463

[Alder]

Hi @thomasneirynck,

Thanks for the quick response and help.

I have normalized both my index and the file br_muns.geojson with the code below and I think it will work now.

public static String removerAcentos(String str) {
return Normalizer.normalize(str, Normalizer.Form.NFD).replaceAll("[^\p{ASCII}]", "");
}

Thanks everyone from elasticteam.