Add leading/trailing space to elasticsearch tokenizer ngram

Question

I am trying to produce ngram features with elasticsearch analyzer, in particular, I would like to add leading/trailing space to the word. For example, if the word is "2 Quick Foxes", the ngram features with leading/trailing space will be:

" 2 ", "2 Q", .....," "Fox", "oxe", "xes", "es "

PUT my-index-000001
{
  "settings": {
    "analysis": {
      "analyzer": {
        "my_analyzer": {
          "tokenizer": "my_tokenizer"
        }
      },
      "tokenizer": {
        "my_tokenizer": {
          "type": "ngram",
          "min_gram": 3,
          "max_gram": 3,
          "token_chars": [
            "letter",
            "digit"
          ]
        }
      }
    }
  }
}

POST my-index-000001/_analyze
{
  "analyzer": "my_analyzer",
  "text": "2 Quick Foxes"
}

Answer 1

You could add two pattern replace character filters -- one for the leading whitespace, the other for the trailing:

PUT my-index-000001
{
  "settings": {
    "index": {
      "analysis": {
        "analyzer": {
          "my_analyzer": {
            "tokenizer": "my_tokenizer",
            "char_filter": [
              "leading_space",
              "trailing_space"
            ]
          }
        },
        "tokenizer": {
          "my_tokenizer": {
            "type": "ngram",
            "min_gram": 3,
            "max_gram": 3,
            "token_chars": [
              "letter",
              "digit",
              "whitespace"       
            ]
          }
        },
        "char_filter": {
          "leading_space": {
            "type": "pattern_replace",
            "pattern": "(^.)",
            "replacement": " $1"
          },
          "trailing_space": {
            "type": "pattern_replace",
            "pattern": "(.$)",
            "replacement": "$1 "
          }
        }
      }
    }
  }
}

Notice the added whitespace to the token_chars of my_tokenizer -- the above won't work without it.