Trigger when a specific SPO list Column has been modified

[lardo5150]

So I found an article that describes using the "When an item or a file is modified" with "Get Changes for an item or a file".

Unfortunately, I am having issues understanding how to set this up properly.

What I want:

I have an SPO list called Engineers.  I have two columns, AssignmentDays and Assignment Hours.

I have a scheduled flow that runs each day and converts these choice columns into text that is more user friendly for viewing.

What i would like to happen is whenever I make a change to AssignmentDays or Assignment Hours, it kicks the flow off.

What is the proper way to set this up, or can it even be done?

For example, if I change the first one from 0,1,2,3 to 1,2,3,4, the flow will kick off for just that specific SPO item.

[tom_riha · Accepted solution]

Hello @lardo5150 ,

using the 'Get changes..' action will trigger the flow with every change just to check if one of the columns was changed. I prefer to not the flow at all if there's no change in the monitored columns, but for that you'll need another (hidden from users) columns in the list.

It'll serve as a backup column - you'll keep the previous value of the field in that column, and check with every update if the values changed (one of the monitored columns was updated) or not. 

I wrote a blog post on this some time ago.

[lardo5150]

Ok, I may need you to walk me through this.

I created a backup column for AssignmentHours and AssignmentDays.

both of those columsn are Choice columns, so I made the backup columns the same.

At first I did not think I needed an array, so i just did an update to the backup column with, for example, AssignmentHours value.

What happened though is if AssignmentHours was 7,8,9, it only added the 9 to the backup column.

What is the correct way to do this with an array and a choice column?

[lardo5150]

Example of what it looks like (the flow did not get to the days yet as I canceled it)

[lardo5150]

Quick update, got my backup columns made, going to work on the trigger condition now.

[lardo5150]

I could not save the flow.

The trigger was throwing an error about the expression I put in.

@not(equals(triggerOutputs()?['body/AssignmentHours'],triggerOutputs()?[body/AssignmentHoursBackup']))

[lardo5150]

@tom_riha Ok, I got this to work on two of my three flows.

On the third one, it keeps looping.  I turn it off, and then back on, and it kicks off like 5 more runs.

The other two only kick off when the main field and backup field are different.

Looking at my third flow:

These are the same, so it should NOT be running, but it is for some reason (for the same test person)

Here is my trigger:

@not(equals(triggerOutputs()?['body/EngineerType'],triggerOutputs()?['body/TypeofEngineerBackup']))

These are both choice columns, just like my other two.

For example, here is one from my day change flow that is EXACTLY the same, except for the trigger.

@not(equals(triggerOutputs()?['body/AssignmentDays'],triggerOutputs()?['body/AssignmentDaysBackup']))

[tom_riha]

Hello @lardo5150 ,

if it's a choice field then I'd go one step deeper and compare only the values.

@not(equals(triggerOutputs()?['body/EngineerType/Value'],triggerOutputs()?['body/TypeofEngineerBackup/Value']))

The choice fields contain also ID together with the value, this ID represents order of the choice in the field = if the fields are not 1:1 then the ID might be different = the field will be evaluated as different. It's safer to use only the Value which will be the same.

[lardo5150]

@tom_riha  I ended up deleting and recreating the column, exactly like the main column.

The three choices are

"Id":0,"Value":"On-Prem"

"Id":1,"Value":"Online"

"Id":2,"Value":"Hybrid"

I pulled three engineers from the list, each who are listed as the above.

In the output, I can confirm that the ID and Values are exactly the same for the main column and the backup.

I used this as my trigger condition:

@not(equals(triggerOutputs()?['body/EngineerType/Value'],triggerOutputs()?['body/EngineerTypeBackup/Value']))

As soon as I turned it on, it started to run a bunch of times.

Did I enter it correctly?

I thought about building the array with both the ID and Value, but for whatever reason, ID is not an option to pick.

[tom_riha]

Hello @lardo5150 ,

another approach could be to use a Multiple lines of text columns as the backup - it'll store all the information from the original column as a string and you don't have to deal with the ID etc.

For the running flows, I'd check with what information it did run. I think that if you turn off a flow, and then turn it on later, it'll remember all the changes that were done during the turn off time and process them. Maybe it's just processing some changes that happened in the meantime where the columns were not synchronised yet.

[lardo5150]

@tom_riha I thought of that, but the original column is a choice column.

So if I setup the backup column as a multiple lines of text, I have two questions.

How do I format the backkup column, and how do we compare the two?

We would be comparing Choice to Text, how does that work in the trigger?

[tom_riha]

You won't format it, just keep it as a plain text column. The trigger shouldn't care, it takes and compares the content of the column, which will be the same string.

You can test all the situations directly in the flow, just check the output of the trigger action and what values are in the two columns.