# Tokens outside the ngram size
**URL:** https://discuss.elastic.co/t/tokens-outside-the-ngram-size/6499
**Category:** Elasticsearch
**Created:** [January 25, 2012, 7:19pm UTC](https://discuss.elastic.co/t/tokens-outside-the-ngram-size/6499 "2012-01-25T19:19:26Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@John\_Downey](https://discuss.elastic.co/u/John_Downey)
#### Post date: [January 25, 2012, 7:19pm UTC](https://discuss.elastic.co/t/tokens-outside-the-ngram-size/6499/1 "2012-01-25T19:19:26Z")
I am in the process of setting up elasticsearch in our infrastructure. The initial setup includes a custom analyzer with an ngram filter (min 3/max 20) set as the index\_analyzer on two mappings. We have one free-form text field that searches will be entered from (Google style). Currently this text is being used to search the \_all field with a query\_string query.
While testing I noticed that tokens smaller than the ngram size are not being matched during searches. My expectation is that perfect matches would still result in the document matching.
Is there a way to configure the analyzer so that the \_all field can be searched with ngrams and still find perfect matches outside the ngram size?
--
John Downey
---
### Author:  [@kimchy](https://discuss.elastic.co/u/kimchy)
#### Post date: [January 25, 2012, 9:48pm UTC](https://discuss.elastic.co/t/tokens-outside-the-ngram-size/6499/2 "2012-01-25T21:48:09Z")
You mean the \_all field to have two different analyzers? No, it can only have one.
On Wednesday, January 25, 2012 at 9:19 PM, John Downey wrote:
> I am in the process of setting up elasticsearch in our infrastructure. The initial setup includes a custom analyzer with an ngram filter (min 3/max 20) set as the index\_analyzer on two mappings. We have one free-form text field that searches will be entered from (Google style). Currently this text is being used to search the \_all field with a query\_string query.
>
> While testing I noticed that tokens smaller than the ngram size are not being matched during searches. My expectation is that perfect matches would still result in the document matching.
>
> Is there a way to configure the analyzer so that the \_all field can be searched with ngrams and still find perfect matches outside the ngram size?
>
> --
> John Downey
---
### Author:  [@system](https://discuss.elastic.co/u/system)
#### Post date: [July 6, 2017, 3:41am UTC](https://discuss.elastic.co/t/tokens-outside-the-ngram-size/6499/3 "2017-07-06T03:41:28Z")