Creating Task Resources Based on a Definition

This short guide should help you understand how you can create Task resources for use in Starting A Process Via Task Resources. We will employ the use of the free version of Forge to help with visualization. You are invited to create a free account and follow along, but we will include screenshots of relevant views either way. Remember that the free version of Forge must not be used commercially.
As an example, we will create a Task resource from the task-start-dic-process.xml profile.

1st Step: Removing Placeholders

task-start-dic-process.xml includes placeholders for the version and date elements. For the duration of this guide, you can either remove or comment these elements, so Forge does not try to perform type checking on them, which would result in an error and Forge not loading the file.

2nd Step: Differential Chain

If the resource profile is only available as a differential, like in our case, we will want to aggregate the changes made to the base resource (in this case Task) by all profiles to make it more readable. To do this, we first need all the profiles involved. We already have task-start-dic-process.xml in our StructureDefinition folder. It lists a resource called task-base in its baseDefinition element. This resource is part of the DSF and can be found here. Put it into the StructureDefinition folder. Since task-base has the original FHIR Task as its baseDefinition element, we are done with this chain. In forge, you should now be able to open the StructureDefinition folder and select the task-start-dic-process.xml profile. It should look something like this:

3rd Step: Building the Task Resource

We will now go through each element one by one and include it into our Task resource, provided it is mandatory (cardinality at least 1..1) according to the profile. It is important that you not use any placeholders like #{version} for resources not read by the DSF BPE server. This is the case if we want a Task resource for use with cURL. But, placeholders should be used in Draft Task Resources instead of actual values wherever possible, since those are read by the DSF BPE server. This guide will create a Task resource without placeholders.
We will start out with the base element for all Task resources:

<Task xmlns="http://hl7.org/fhir">

</Task>

Before we start adding any elements listed in Forge's element tree, we have to include the Task.meta.profile element. Its requirement cannot be seen here which is why we mention it specifically. This is the only instance you will not see it in the element tree. It should look like this:

<Task xmlns="http://hl7.org/fhir">
    <meta>
        <profile value="http://dsf.dev/fhir/StructureDefinition/task-start-dic-process|1.0"/>
    </meta>
</Task>

The first element which can be found in the element tree is the instantiatesCanonical element. To add it, we will create an XML element with the same name and the value according to URLs:

<Task xmlns="http://hl7.org/fhir">
    <meta>
        <profile value="http://dsf.dev/fhir/StructureDefinition/task-start-dic-process|1.0"/>
    </meta>
    <instantiatesCanonical value="http://dsf.dev/bpe/Process/dicProcess|1.0" />
</Task>

We can continue this process for all primitive elements like these. Just make sure you pay attention to use the correct data type (e.g. proper coding value for elements with coding type).

By now your Task resources should look something like this:

Suggested solution

<Task xmlns="http://hl7.org/fhir">
    <meta>
        <profile value="http://dsf.dev/fhir/StructureDefinition/task-start-dic-process|1.0"/>
    </meta>
    <instantiatesCanonical value="http://dsf.dev/bpe/Process/dicProcess|1.0" />
    <status value="requested"/>
    <intent value="order"/>
    <authoredOn value="2024-02-08T10:00:00+00:00" />
</Task>

Let us look at a more complex element like the requester element:

We will start the same way we started with primitive elements, by adding the requester element:

<Task xmlns="http://hl7.org/fhir">
    <meta>
        <profile value="http://dsf.dev/fhir/StructureDefinition/task-start-dic-process|1.0"/>
    </meta>
    <instantiatesCanonical value="http://dsf.dev/bpe/Process/dicProcess|1.0" />
    <status value="requested"/>
    <intent value="order"/>
    <authoredOn value="2024-02-08T10:00:00+00:00" />
    <requester>
     
    </requester>
</Task>

Then, we will add primitive elements to requester like we did before for Task:

<Task xmlns="http://hl7.org/fhir">
    <meta>
        <profile value="http://dsf.dev/fhir/StructureDefinition/task-start-dic-process|1.0"/>
    </meta>
    <instantiatesCanonical value="http://dsf.dev/bpe/Process/dicProcess|1.0" />
    <status value="requested"/>
    <intent value="order"/>
    <authoredOn value="2024-02-08T10:00:00+00:00" />
    <requester>
        <type value="Organization"/>
    </requester>
</Task>

Important to note here that the value for the status will always be requested for Tasks being posted using cURL and the type element for requester and recipient will always have the value Organization in the DSF context.

Next, we will add the identifier element and its primitive sub-elements just like we started out doing it for the requester element. The identifier.value in this case will be dic.dsf.test. To understand why, take a look at the topic on organization identifiers:

<Task xmlns="http://hl7.org/fhir">
    <meta>
        <profile value="http://dsf.dev/fhir/StructureDefinition/task-start-dic-process|1.0"/>
    </meta>
    <instantiatesCanonical value="http://dsf.dev/bpe/Process/dicProcess|1.0" />
    <status value="requested"/>
    <intent value="order"/>
    <authoredOn value="2024-02-08T10:00:00+00:00" />
    <requester>
        <type value="Organization"/>
        <identifier>
            <system value="http://dsf.dev/sid/organization-identifier"/>
            <value value="dic.dsf.test" />
        </identifier>
    </requester>
</Task>

Notice that requester.identifier.system has a Fixed value annotation. You can see what the value is supposed to be by clicking on the system element in Forge or looking at the XML for the right Task profile. The right side will have all information about that element, including the actual value for Fixed value.

You should now be able to fill out all elements in your Task resource until you reach the slicing for Task.input. Your Task resource should look something like this:

Suggested solution

<Task xmlns="http://hl7.org/fhir">
    <meta>
        <profile value="http://dsf.dev/fhir/StructureDefinition/task-start-dic-process|1.0"/>
    </meta>
    <instantiatesCanonical value="http://dsf.dev/bpe/Process/dicProcess|1.0" />
    <status value="requested"/>
    <intent value="order"/>
    <authoredOn value="2024-02-08T10:00:00+00:00" />
    <requester>
        <type value="Organization"/>
        <identifier>
            <system value="http://dsf.dev/sid/organization-identifier"/>
            <value value="dic.dsf.test" />
        </identifier>
    </requester>
    <restriction>
        <recipient>
            <type value="Organization"/>
            <identifier>
                <system value="http://dsf.dev/sid/organization-identifier" />
                <value value="dic.dsf.test" />
            </identifier>
        </recipient>
    </restriction>
</Task>

Slicings are a bit different from regular elements. Let us look at the slice message-name:

If we were to continue including slices to the Task resource like we did so far, we would add a message-name element to our XML like this:

<Task xmlns="http://hl7.org/fhir">
    ...
    <input>
        <message-name>
            ...
        </message-name>
    </input>
</Task>

This approach however, would not work. FHIR processors do not use the name of the slice to map entries in your Task resource to the correct slice. They use discriminators. Discriminators define the elements a processor needs to distinguish slices by. You can see how the discriminator is configured by selecting the input element in Forge. In our case, a processor would look at the values for input.type.coding.system and input.type.coding.code to determine which slice this element belongs to. This only works because input.type.coding.system and input.type.coding.code are present in all slices and have a Fixed value. You can learn more about discriminators here.
All this means is that we effectively ignore the name of the slice as an element and start adding elements like we did before:

<Task xmlns="http://hl7.org/fhir">
    ...
    <input>
        <type>
            <coding>
                <system value="http://dsf.dev/fhir/CodeSystem/bpmn-message" />
                <code value="message-name" />
            </coding>
        </type>
        <valueString value="dicProcess" />
    </input>
</Task>

Now you should be able to add all remaining mandatory elements to your Task resource on your own. In the end, it should look something like this:

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

creating-task-resources-based-on-a-definition.md

creating-task-resources-based-on-a-definition.md

Creating Task Resources Based on a Definition

1st Step: Removing Placeholders

2nd Step: Differential Chain

3rd Step: Building the Task Resource

Files

creating-task-resources-based-on-a-definition.md

Latest commit

History

creating-task-resources-based-on-a-definition.md

File metadata and controls

Creating Task Resources Based on a Definition

1st Step: Removing Placeholders

2nd Step: Differential Chain

3rd Step: Building the Task Resource