In this post I am going to show you how you can use an Azure Function to return an item from the SharePoint Online term store and use that value inside a Microsoft Flow app.
As an example of where this might be useful think of something like an approval workflow using Microsoft Flow that can lookup an approver’s email address from the SharePoint term store.
The basic architecture looks like below:
The first thing to do is to download the source code from this GitHub repository.
The code includes an Authentication context that supports both App Only authentication and SharePoint authentication through AppRegNew.aspx.
This can be configured by changing the Azure Function app setting called AUTH_TYPE from AppOnly to SPAuth.
Creating Azure AD App
We need our Azure Function to be able to authenticate with SharePoint so we will create a new Azure AD app in the portal and configure it’s application permissions to allow SharePoint access.
Setting up App Only auth is quite lengthy and there is pretty good documentation over on docs.microsoft.com you can follow the steps in this article to create an Azure AD app https://docs.microsoft.com/en-us/azure/azure-resource-manager/resource-group-create-service-principal-portal and then follow this article to configure it for app only authentication https://docs.microsoft.com/en-us/sharepoint/dev/solution-guidance/security-apponly-azuread.
Once you have the Azure AD app created and your app only certificate, make sure you put the .pfx file inside the .\Cert directory inside the solution and set its ‘Copy to Output Directory’ property to ‘Copy Always’. You will also need to update the name of this certificate in the Azure Function app settings in the CERT_NAME setting.
Open the permissions for the newly created AAD app in the portal and make sure the Office 365 SharePoint Online permissions are configure as below:
You will need to make a note of the Application ID for this AAD app and update the AUTH_ID app setting in the Azure Function once created.
Deploying Azure Function
In the Azure Portal create a new Function App as shown below and make a note of the App Name.
Next we want to deploy the function to Azure. We can either download the publish profile from the portal or use the Azure Functions Core Tools library https://www.npmjs.com/package/azure-functions-core-tools.
To use the core tools to deploy make sure you first build the solution. Open a command prompt and CD to .\bin\debug\net461 folder.
After installing the tools with
npm i -g azure-functions-core-tools
func azure functionapp publish leetestsptaxlookup
Replace leetestsptaxlookup with the function name you created earlier.
Update the functions app settings with the values in the local.settings.json file replacing the values with the ones that relate to your tenant.
Configure the Azure Function taxonomy lookup
The Azure Function will lookup from the term store based on the value in the TERM_GROUP app setting.
For instance if our term store structure look like below:
We would set our TERM_GROUP setting to be the value “Lee;Settings;SiteRequestApprover”. Notice the structure is TermGroup -> TermSet -> Term all separated by a semi colon.
Creating Microsoft Flow App
Open Microsoft Flow and create a new Flow. You can pick whichever trigger you want to use. Create a new HTTP action and configure as shown below:
The HTTP action just makes a GET request to the Azure Function passing in two parameters; siteUrl and code.
siteUrl is the URL of the SharePoint site you want to use as a context for obtaining the taxonomy values. You need to pass your host key to the code parameter. You can view your host keys when you visit the settings page for your function in the Azure portal.
When your Flow is run the Azure function will return the value from the term store inside the body of the HTTP response as show below:
This Body value can then be used as a parameter to other actions. This means we can use a value from the term store to drive decisions within our Microsoft Flow app. This could be something like an approver for a workflow for example. In reality any value can be retrived from the term store and we can host multiple Azure functions for different values.