Developer forum

Here's a screenshot of the query

Hi Gaëtan

You would need to add an analyzer on the field in question to support this. That is done by creating a custom field using an analyzer that normalizes the danish characters:

http://doc.dynamicweb.com/documentation-9/platform/indexing-search/indexing-search#4786

BR Nicolai

Thanks, sounds interesting. I have looked up how to create a custom analyzer in Lucene but I don't know much about it what I have found is pretty esoteric... Do you have any concrete example I could look at ?

Thank you.

Here is an article:
https://www.codeproject.com/Articles/32175/Lucene-Net-Text-Analysis

Thanks, I had already read that (along with other explanations and documentation, but it hasn't helped me much, it's really opaque.

I guess I simply need a better knowledge of Lucene, but I don't have time to dig into it that right now. I'll try to find a workaround.

I've found this tutorial which more or less does what I want, I just replaced "GetSynonyms" with my own function to create the alternative spellings : http://www.opten.ch/blog/2014/05/30/writing-a-custom-synonym-token-filter-in-lucenenet/

Then I use it in my analyser like this :

public class DanishAnalyzer : Analyzer
    {
        public override TokenStream TokenStream(string fieldName, TextReader reader)
        {
            TokenStream tokenStream = new KeywordTokenizer(reader);
            tokenStream = new DanishFilter(tokenStream);

            return tokenStream;
        }
    }

Then I created a new field type with this analyser, and a new field with this field type which takes its source from name, and my query searches this field (match any).

But it still doesn't work.

Anyone with Lucene experience to give me a hand ?

Thanks