Hello, I'm currently working in developing a Cloud Flow that involves fetching details from Planner Tasks. One of the essential criteria is to ob

Getting Microsoft Planner Task Comments and Last Change Date

Task comments are posts sent to plan's O365 group owner conversation. Each task has a dedicated thread.

To get them, you'll have to go through Graph API. I'll first show you how to use Graph API in Power Automate then I'll show you how to query for a task comments.

Graph API call using Power automate:

Use "Invoke an HTTP request" action (preauthorized version)

Once action added, it invites you to create a connection, enter "https://graph.microsoft.com" in both "Base resource URL" & "Microsoft Entra ID" fields:

The flow to illustrate how to get task comments:

First get a task

Get task's plan (I renamed the "Invoke an HTTP request"):

https://graph.microsoft.com/v1.0/planner/plans/@{outputs('Get_a_task')?['body/planId']}

Get comments:

https://graph.microsoft.com/v1.0/groups/@{body('Get_plan')?['owner']}/threads/@{body('Get_a_task')?['conversationThreadId']}/posts

Select desired data (simplify json properties path for next actions):

From: body('Get_comments')?['value']

HTML: item()?['body']?['content']

Created: item()?['createdDateTime']

SenderName: item()?['sender']?['emailAddress']?['name']

SenderEmail: item()?['sender']?['emailAddress']?['address']

At this point we've got these outputs:

As you can see, comments have HTML embedded, so we need to clean it.

For each comment from select outputs:

items('Apply_to_each')?['HTML']

Outputs for one comment:

I hilighted my comment, but you can see extra comment added by planner in thread, let's clean that:

join(reverse(skip(reverse(split(outputs('Html_to_text')?['body'], decodeUriComponent('%0A'))), 7)), decodeUriComponent('%0A'))

This expression splits outputs by '\n' (decodeUriComponent('%0A')), then reverse the resulting array, skips the 7 lines we don't want, reverse back to get comment lines in right order and finally join the array using '\n' to get a string.

Now, let's build a JSON object with all needed data:

Comment: outputs('Extract_comment')

Created: items('Apply_to_each')?['Created']

SenderName: items('Apply_to_each')?['SenderName']

SenderEmail: items('Apply_to_each')?['SenderEmail']

Finally, get an array of all comment info using a handy trick to join all together. Instead of initializing a variable, and updating the variable inside the apply to each loop, you can place a compose action just outside the loop that gets outputs of the one belonging to the apply to each. Credits to Peter Veenstra: https://sharepains.com/2019/07/09/compose-actions-inside-apply-to-each-steps-in-microsoft-flow/