# Lucene Document Field Tokens **URL:** \ **Category:** Jira Server\ **Tags:** lucene\ **Created:** [July 23, 2019, 5:46pm UTC](https://community.developer.atlassian.com/t/lucene-document-field-tokens/30597 "2019-07-23T17:46:02Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![linklefebvre](https://sea1.discourse-cdn.com/atlassiandeveloper/user_avatar/community.developer.atlassian.com/linklefebvre/32/49416_2.png) [@linklefebvre](https://community.developer.atlassian.com/u/linklefebvre)\ **Post date:** [July 23, 2019, 5:46pm UTC](https://community.developer.atlassian.com/t/lucene-document-field-tokens/30597/1 "2019-07-23T17:46:02Z")
I was wondering if any other developers had to use a custom Analyzer / Tokenizer or simply custom Tokens for their Lucene Document Fields. We have a field let’s say “Car List” which would benefit from indexing the car “Brand”, “Name” and “Color” all in one field. Our goal is to index three tokens in one field: “General Motors”, “Acadia”, “Dark Blue”. If we index as a text field, it will create a token for “General”, “Motors”, “Acadia”, “Dark”, “Blue”. This will return result if searching for a Blue General Acadia (which we don’t want to return results, only for Dark Blue General Motors Acadia). If you had a similar situation, how did you guys implement it? Thanks for your time. ---
**Author:** ![linklefebvre](https://sea1.discourse-cdn.com/atlassiandeveloper/user_avatar/community.developer.atlassian.com/linklefebvre/32/49416_2.png) [@linklefebvre](https://community.developer.atlassian.com/u/linklefebvre)\ **Post date:** [July 24, 2019, 5:06pm UTC](https://community.developer.atlassian.com/t/lucene-document-field-tokens/30597/2 "2019-07-24T17:06:57Z")
What we ended up doing is creating our own `TokenStream` that builds a Field with the three tokens `General Motors` , `Acadia` and `Dark Blue`. When searching using a PhraseQuery, we were able to search by exact token in the index.