Before proceeding with the installation process, it's important to ensure that your system meets the necessary prerequisites. Failure to install the prerequisites may result in errors or unexpected behavior during the installation process.

Please check the documentation or installation guide for the software or package you are trying to install to determine the specific prerequisites:

Unix-like system with bash, tar and java installed

To generate SDK by your zen-project config you have to install zen-cli. This command will install the latest version of the zen-cli on your system. The -g flag tells npm to install the package globally, making it available to all projects on your system.

npm install -g @aidbox/zen-cli

In the case you do not have zen-project configured - follow the documentation.

After configuration of zen-package you have to install dependencies (e.g hl7-fhir-r4-core) by typing the following into a terminal window

zen-cli zen pull-deps

This command will download FHIR Definitions All the value sets, profiles, etc. defined as part of the FHIR specification, and the included implementation guides depends on added packages.

In terminal move to your zen-project folder and generate SDK by run the following

zen-cli sdk

After running the zen-cli get-sdk command, you should be able to find the aidbox-javascript-sdk.tgz npm package in the root directory of your project. This archive is generated based on your zen-project and includes all types of resources.

Then, when you get your SDK move this archive into your project and add SDK as dependency in package.json

{
  "dependencies": {
    "aidbox-sdk": "file:<PATH_TO_GENERATED_ARCHIVE_SDK>"
  }
}

Install dependencies

npm install

Important: Before we start we have to make sure that aidbox client is configured and your access policies provide granular access to resources you're trying to reach.

import { Client } from 'aidbox-sdk';

export const aidbox = new Client('<HOST_URL>', {
  auth: {
    method: 'basic',
    credentials: {
      username: '<CLIENT_NAME>',
      password: '<CLIENT_SECRET>',
    },
  },
});

This code creates a new instance of the SDK from the aidbox-sdk package, and sets it up to connect to an Aidbox server running on <HOST_URL>. You would need to replace <CLIENT_NAME> and <CLIENT_SECRET> with the actual client ID and client secret that you configured earlier, also SDK support two types of API (Aidbox, FHIR see difference) just provide one of them as third parameter.

By using aidbox-sdk in your project, you can easily interact with an Aidbox server and perform actions like reading and writing resources, searching for resources, and more.

Then you can use aidboxClient in wherever you want

import { aidbox } from '../aidbox-client';

async function getPatients() {
  return aidbox.resource.list('Patient');
}

list method accepts the name of the resource and is the basis for the subsequent complication of the request

Method where add additional parameters for searching

For example, you want to find all patients with name John, so

resource.list('Patient').where('name', 'John');

Or, you want to find all patients with name John or Steve

resource.list('Patient').where('name', ['John', 'Steve']);

Also, method where support prefixes for numbers and date, just pass it as third parameter

resource.list('Patient').where('birthdate', '2014-06-30', 'gt');

Method sort add additional parameters for sorting

For example, you want to display the oldest patients

resource.list('Patient').sort('birthdate', 'asc');

And also, you want to sort this data by patients name

resource.list('Patient').sort('birthdate', 'asc').sort('name', 'asc');

Method count used for make limit the number of resource returned by request

resource.list('Patient').count(10);

To request only a portion of the resources from a server, you can use the summary function. This function allows you to specify the type of summary you want to receive, such as true, false, text, data, or count.

By default, when you make a request to a server, it will return all resource.

The summary function accepts several types of summaries, each of which provides a different level of detail about the data. For example:

• true - limited subset of elements from the resource

• false - all parts of the resource

• text - only the text, id, meta, and top-level mandatory elements

• data - resources without the text element

• count - count of the matching resources, without returning the actual matches

  summary("data")

If you need to retrieve specific elements of a resource from a server, you can use the elements function. This function allows you to specify which parts of the resource you are interested in, and can help to reduce the amount of data returned by the server.

By using the elements function, you can customize your requests to retrieve only the data that you need, rather than requesting the entire dataset. This can be particularly useful when dealing with large datasets or when you only need a small subset of the available data.

To use the elements function, simply pass in the elements you want to retrieve as arguments. For example, if you only need the name and address fields from a resource, you can make a request using the following syntax:

resource.list('Patient').elements(['name', 'address']);

This request will return only the name and address fields for each resource, rather than the entire dataset. By reducing the amount of data returned, you can help to streamline your requests and improve the performance of your application.

The get function is a tool for retrieving a single resource by its ID.

To use the getResource function, you must pass in the resource type and ID as arguments. For example, if you want to retrieve a patient resource with the ID some-id, you can make a request using the following syntax:

resource.get('Patient', 'id');

The createResource function is used to create a new resource.

The first argument is a resource name, the second one is body of resource

resource.create('Patient', { active: true });

The update function is used to update a specific resource identified by its id with a partial set of data provided in the third parameter. This function allows for partial updates of a resource without having to send the entire resource object.

The first parameter specifies the name of the resource to be updated. The second parameter identifies the specific resource to be updated.

resource.update('Patient', 'c58f29eb-f28d-67c1-0400-9af3aba3d58c', {
  active: false,
});

The deleteResource function is used to delete a specific resource identified by its id.

resource.delete('Patient', 'c58f29eb-f28d-67c1-0400-9af3aba3d58c');

Aidbox subscription is a way to subscribe and get notifications about updating resources on the server. See our subscription sample for more details.

await client.subsSubscription.create({
  id: 'patient-created',
  status: 'active',
  trigger: { Patient: { event: ['create'] } },
  channel: { endpoint: 'https://aidbox/patient-created' },
});

Aidbox has the ability to extend its logs. There is an endpoint that accepts logs with the defined structure from your application. These logs are ingested into the elastic log.

await client.sendLog({
    type: "ui",
    message: {event: "YOUR_EVENT", id: "id", ...},
    v: "APP_VERSION",
    fx: "fetchUsers"
})

SDK has helpers to prepare data for bundle request.

client.transformToBundle(
  [
    {
      resourceType: 'Patient',
      name: [
        {
          given: [''],
          family: '',
        },
      ],
    },
  ],
  'POST',
);

Bundle requests could be a transaction or batch type. SDK uses the "transaction" type by default but you can change it by providing it in the second parameter.

const data = ArrayOfPatients.map(client.bundleEntryPost);

await client.bundleRequest(data, 'batch');

Queues are a valuable tool for achieving reliable, asynchronous, scalable, and retry-safe code execution. By using queues, we can ensure that tasks or messages are processed reliably, even in the face of failures or system disruptions. The asynchronous nature of queues allows tasks to be processed independently, enabling parallelism and reducing wait times.

import { Engine } from 'aidbox-javascript-sdk';

const client = new Engine('http:https://localhost:8888', {
  username: 'test',
  password: 'secret',
});

First of all we have to create task's schema with settings that apply restrictions on input and output arguments for entity like this one:

 SendMessage
{:zen/tags #{awf.task/definition zen/schema}
 :type     zen/map
 :keys     {:params {:type    zen/map
                     :require #{:phone :message}
                     :keys    {:phone   {:type zen/string}
                               :message {:type zen/string}}}
            :result {:type    zen/map
                     :require #{:status}
                     :keys    {:status {:type zen/string}}}}}

More information about task definition
Notice: we have to regenerate SDK package each time we made changes into configuration project

Next step is creating business logic that will be considered as a worker, the worker will be handling each task, we can

client.task.implement(
  'SendMessage',
  async ({ params, status }) => {
    const { message, phone } = params;

    try {
      await fetch('https://message-sending-server.com/', {
        method: 'POST',
        body: JSON.stringify({ message, phone }),
      });

      return { status: 'success' };
    } catch (error) {
      return { status: 'failure' };
    }
  },
  { batchSize: 5 },
);

The way to execute a single task with unique context

await client.task.execute('SendMessage', {
  phone: '+1234567890',
  message: 'Hi!',
});

Aidbox provides Workflow Engine module, so you're able to define your own sequence of async tasks on top of Task API. In many business processes, certain tasks depend on the completion of other tasks. Defining task dependencies through workflow implementation allows the developer to control the order in which tasks are executed.

import { Engine } from 'aidbox-javascript-sdk';

const client = new Engine('http:https://localhost:8888', {
  username: 'test',
  password: 'secret',
});

 CheckOutWorkflow
{:zen/tags #{awf.workflow/definition zen/schema}
 :type     zen/map
 :pool     decision-pool
 :keys     {:params {:type    zen/map
                     :require #{:clientId}
                     :keys    {:clientId {:type zen/string}}}
            :result {:type zen/map
                     :keys {:messageId {:type zen/string}}}
            :error  {:type zen/map
                     :keys {:message {:type zen/string}}}}}

More information about workflow definition
Notice: we have to regenerate SDK package each time we made changes into configuration project

await client.workflow.implement(
  'CheckOutWorkflow',
  async ({ params }, { execute, complete, fail }) => {
    if (params.event === 'awf.workflow.event/workflow-init') {
      const response = await fetch('https://server.com/get-client');
      const { phone } = response.json();
      return [
        execute({
          definition: 'SendMessage',
          params: { phone, message: 'Hi!' },
        }),
      ];
    }

    try {
      const response = await fetch('https://workflow-state.com');
      const data = response.json();

      if (
        params.event === 'awf.workflow.event/task-completed' &&
        data.step === 1
      ) {
        return [execute({ definition: 'UpdateInformation', params: {} })];
      }

      if (
        params.event === 'awf.workflow.event/task-completed' &&
        data.step === 2
      ) {
        return [complete({})];
      }
    } catch (error) {
      return [fail({ error })];
    }

    return [];
  },
);

await client.workflow.execute('CheckOutWorkflow', { clientId: '' });