Custom analyzer don't match with ASCII folding filter values

Elastic Stack Elasticsearch
1

I'm using a custom rebuilt turkish analyzer like this:

"analysis": {
      "filter":{
        "turkish_lowercase": {
          "type":       "lowercase",
          "language":   "turkish"
        },
        "turkish_stop": {
          "type":       "stop",
          "stopwords":  "_turkish_" 
        },
        "my_ascii_folder":{
          "type" : "asciifolding",
          "preserve_original" : true
        }
      },
      "analyzer": {
        "rebuilt_turkish": {
          "tokenizer": "standard",
          "filter": [
            "turkish_lowercase",
            "my_ascii_folder"
          ]
        }
      }
    }

My data is in turkish language. I'm using data folding to match for all possible inputs(for example assuming icralik and ıcralık as same, ı is a different vowel than i)

My text contains this sentence:

...kartlarının borcundan dolayı icralık duruma düşmesi üzerine...

When I test my text with the _analyze api using my custom analyzer, I get this result as I expected ( icralık is the original version inside the text and icralik is the folded one):

image
image384×590 11.2 KB

When I search for "icralık duruma" I can get the result without a problem:

image
image1515×493 101 KB

But when I execute the same search query with folded text "icralik duruma" I get no results:

image
image914×456 26.3 KB

What am I doing wrong?

2

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.