FTR from FTR — Direct Dependency


  1. Download zen-ftr CLI: Download link

In this guide, we will create an FTR that represents a SNOMED CT subset that includes codes descending to "Angina (disorder)" (194828000). The FTR on which we depend is stored remotely in a GitHub source-code repository. You can modify the source-url field in the FTR manifest to provide a desired URL, which can be a file path or network URL. Main use case for this extraction engine — design new ValueSets.

Create a directory project with following structure:

  zen-package.edn #Package deps declaration
    system.edn #System entrypoint importing a ValueSet definion
    angina.edn # ValueSet definition
{:deps {}}
{:ns system
 :import #{aidbox angina}
 {:zen/tags #{aidbox/system}}}

This ValueSet definition conforms to the zen.fhir ValueSet schema and includes a :ftr property. The :ftr property contains an FTR manifest that defines an FTR-dependency source through the :source-url property, which allows the creation of an expanded version of the ValueSet to be stored in the resulting FTR. In addition, the specifies the tag to be selected within the provided FTR dependency. For more information on the FTR manifest, please refer to this page.

{ns angina
 import #{zen.fhir}
 {:zen/tags #{zen.fhir/value-set}
  :zen.fhir/version "0.6.14"
  :uri "angina-vs"
  :ftr {:module            "nanosubset"
        :source-url        ""
        :ftr-path          "/tmp"
        :tag               "prod"
        :source-type       :ftr
        :extractor-options {:target-tag "prod"
                            {:compose      {:include [{:valueSet [""]
                                                       :filter [{:op       "is-a"
                                                                 :property "concept"
                                                                 :value    "194828000"}]}]}
                             :id           "nanosubset"
                             :name         "nanosubset"
                             :resourceType "ValueSet"
                             :status       "active"
                             :url          "nanosubset"}}}}}

Initialize this directory as a git repository and commit your initial set-up:

git init && git add . && git commit -m "init"

Generate FTR

Replace <PATH_TO_JAR> placeholder with absolute path to zen.jar.

Execute command listed below in project directory:

java -jar <PATH_TO_JAR> build-ftr

Commit FTR directory:

git add . && git commit -m "Build ftr"

Now you can run Aidbox with the following configuration project and use FHIR Terminology API methods like $validate-code/$lookup on generated angina-vs ValueSet. Resource validation performed when someone invocates a FHIR REST operations will also validate ValueSet binding via FTR.

For detailed instructions about using Aidbox with Aidbox configuration project, please refer to this page.

Instruct Aidbox to load terminologies into the DB

Set the following environment variable:


By default, Aidbox does not load terminologies into the database as that can take a lot of disk space. This means that full terminology functionality won’t be available until you enable it manually. BOX_FEATURES_FTR_PULL_ENABLE environment variable is just for that. When you set it to true, Aidbox will load terminologies into the database on the next startup and start functioning as a fully-featured terminology server.

Further steps

For guidance on development and production usage, visit the links below:

For customizing Aidbox startup behavior when using FTR, read about FTR environment variables.

Last updated