CODEWITHSUNDEEP
elasticsearch
Rog
Rog

Reputation: 164

How can I add two analysers to a field

I use the nGram plugin for my queries but also I want to add the Phonetic Analysis to the same field on my mapping.

I'm trying the following settings, and it works well with the nGram plugin but I don't know how can I add the dbl_metaphone analyzer on my mapping.

curl -XPUT "127.0.0.1/index" -d'
{
   "settings": {
      "analysis": {
         "filter": {
            "nGram_filter": {
               "type": "nGram",
               "min_gram": 2,
               "max_gram": 20,
               "token_chars": [
                  "letter",
                  "digit",
                  "punctuation",
                  "symbol"
               ]
            },
            "dbl_metaphone": { 
               "type":    "phonetic",
               "encoder": "double_metaphone"
            }
         },
         "analyzer": {
            "nGram_analyzer": {
               "type": "custom",
               "tokenizer": "whitespace",
               "filter": [
                  "lowercase",
                  "asciifolding",
                  "nGram_filter"
               ]
            },
            "whitespace_analyzer": {
               "type": "custom",
               "tokenizer": "whitespace",
               "filter": [
                  "lowercase",
                  "asciifolding"
               ]
            },
            "dbl_metaphone": {
               "tokenizer": "standard",
               "filter":    "dbl_metaphone"
            }
         }
      }
   },
   "mappings": {
      "books": {
         "_all": {
            "analyzer": "nGram_analyzer",
            "search_analyzer": "whitespace_analyzer"
         },
         "properties": {
            "descripcion":{
               "type": "string",
               "analyzer": "nGram_analyzer",
               "search_analyzer": "whitespace_analyzer"
            },
            "categories":{
               "type": "nested",
               "properties": {
                  "nombre":{
                     "type": "string",
                     "analyzer": "nGram_analyzer",
                     "search_analyzer": "whitespace_analyzer"
                  }
               }
            }
         }
      }
   }
}'

Any help?

Upvotes: 0

Views: 55

Answers (2)

paweloque
paweloque

Reputation: 18864

My solution to this problem would be to have a multi field to analyze the data in different ways. Then to combine the results at query time you could use a bool query with either should or must clauses.

So the mapping would look something like this:

   "mappings": {
      "books": {
         "properties": {
            "descripcion": {
               "type": "string",
               "analyzer": "nGram_analyzer",
               "search_analyzer": "whitespace_analyzer"
               "fields": {
                  "metaphone": {
                     "type": "string",
                     "analyzer": "dbl_metaphone"
                  }
               }
            }
         }
      }
   }

And a query would have to use the descripcion and descripcion.metaphone fields. The field descripcion.metaphone is the one analyzed with the dbl_metaphone analyzer.

Does this solve your problem?

Upvotes: 1

Vineeth Mohan
Vineeth Mohan

Reputation: 19253

Combo analyzer is an option here. It allows you to specify multiple analyzers and it clubs the tokens from both the analyzers are the end result.

Upvotes: 0

Related Questions