Delete duplicates in a SharePoint List

[der_cosmicyes]

Hello all,

I have got a SharePoint List with about 1400 items. It only should be arround 800 items and I found out that many items have duplicates. I don´t know why and that´s also not the point here.
My idea is to create a Flow which will get each row of the list and goes through all other rows and checks if *every* column is equal. If yes then delete this row.

I found this here in the forum:
Solved: Delete Duplicate in SharePoint List - Power Platform Community (microsoft.com)
But this is not exactly what I need I think.
I created this flow (unfortunately it is german) - only 3 conditions as an example. In reallity there are 8 columns.

I did not run it yet as I am pretty sure that this Flow will delete everything - or am I wrong?
How can I achieve that the flow deletes only duplicates but leaves one entry of each?

Thank you in advance!

[fchopo]

Hello @der_cosmicyes 

You could do the following:

1) Get all items.

2) For each item.

3) Look for an identical one, using get items and an Odata Filter query expression.

4) If more than one element is found (check the lenght of the ouput of the previous step) -> there is a duplicate.

5) Delete the item.

Have a look at this example:

Therefore, in your "Get Items 2" action, you should write an OData Filter query with the 8 fields to compare.

Have a look at this sources on how to work with this kind of expressions: 

Every Power Automate (MS Flow) Filter Query You Ever Wanted To Know As A Functional Consultant – DIY...

OData Filter query in SharePoint Get items in Flow (sharepains.com)

Hope it helps!

Ferran

[der_cosmicyes]

Thank you Ferran,
Thank you! This sounds a little complicated to a PowerAutomate Newbie. I haven´t heard of OData Filters but I am willing to dive deeper into it.
But one question:
Is there no way to combine "my" solution with the idea to check if more than 1 item is found?
I think my solution would delete everything but if I can check if more than 1 one was found this could be the solution. What do you think?

[fchopo · Accepted solution]

Hello @der_cosmicyes 

In your solution you should another apply to each. I mean, when do the comparison, you are comparing the values of the same item, so it will be always true, and therefore, it will delete all items.

You need an algorithm like that:

1) Get ALL items.

2) For each item.

3) Get ALL items AGAIN

4) For each item in step 3, compare it against the item in step 2.

5) If the values are the same, remove it.

If you find the use of OData filters too complicated, you could do something like this:

Hope it helps!

Ferran

[der_cosmicyes]

Thank you Ferran,
I have not tried yet but this seems very good to me!
Espacially the condition ID(1) is NOT equal to ID(2) is a very good idea.
I will try as soon as possible and then Accept as Solution (I hope 😉 )

Best regards

[der_cosmicyes]

Hi Ferran,
ok, I know why you wanted me to use OData Query: Performance! 🙂
Unfortunately I could not solve it using OData.
I am using now your second approach (which is really very slow...) but it works.
The only thing I changed ist to use ID of 'Get Items' and not of 'Get Items 2':

I set up a very small list with Items
* Paul McCartney
* John Lennon
* George Harrison
* Ringo Starr
* Paul McCartney

and used your second approach.
When deleting ID of 'Get Items 2' both Pauls are gone and the Beatles are a Trio 😉
Using ID of 'Get Items' worked.
You helped me very much - thank you!

[Jltitus]

@fchopo how would you remove all but one? I want to remove duplicates but keep one.

[malditoponi]

Same here! I need to remove all items but one. How can this be achieved?

Thank you!

[Anonymous]

I solved this recently. I have an automation that finds duplicates from a Sharepoint Online List and removes duplicates, leaving the original one.

The start of the flow is the same as the example before, so you use Odata to filter the items and check that the length of the results is more than 1. There are some changes I made after that and one change at the beginning of the flow. I created a variable. The variable contains the Sharepoint List data and the variable is updated after each filtering. At the end of the flow, I extract the List item ID's from the Array and use those ID's in the "Delete item" part. My flow is in finnish, but i'll try to explain with a picture:

Hope this helps!

[benksyong]

Thank you on your sharing.

Would it be okay for your to share more detail how the ODATA FILTERED ITEMS configured? 

I've tried, but it does not work. I'm new in Power automate, I wish to learn more how it works to improve the performance. 

[Anonymous]

Sure!

Select the "Get items" step, and open the advanced settings. Then type the code in the Odata filtering field. My code is below, but you can use whatever suits your needs. The field names are case sensitive, so keep that in mind.

Does this answer your question?

[benksyong]

Thank you @Anonymous . 

I would like to learn the logic. Would it be okay to get your input from few items below: 

1. From the

   if Title has to match with Sharepoint List column's header, then, '' is refer in Dynamic content with same column header? 

2. During Odata filtering Dynamic content selection, should I refer to 'Get items' or 'Get items 2'? why? what logic behind?
3. If I've 8 columns in my SharePoint List that I need to match with logic if all 8 column contents are identical, then consider the line item is duplicated. In this case, if that means i need to repeat all 8 column entry in OData filtering with example,
   • Title eq 'Title' AND column2 SP header eq 'Dynamic content: column2' AND column3 SP header eq 'Dynamic content: column3' AND....... column 8 SP header eq 'Dyanamic contect: column 8' 
4. if the 'Create the variable' is necessary step to define which specific item should be deleted?

[Anonymous]

I'll go trough this step by step.

The first step in the flow is to get the items without any filtering. Let's say that's called "Get items 1". You put the result of that into a variable (because you need to define the variable at the start of the flow) and let's say the variable is called YourVariable:

Then you add the "apply to all" step, and under that you create the Odata filtered get items -step. Let's call that "Get items 2". The logic in the Odata filtering is that you are comparing two lists of items. So basically "If the title of get items 2 is equal to a title in get items 1 and Endtime in get items 2 is equal to Endtime in get items 1". This of course returns all the values, but it groups the items based on that Odata filtering. So you can have an item with one result, or an item with 4 results, or however many duplicates that filtering can find.

After the filtering is done, you update the variable (YourVariable) you created at the start of the flow with the results of the filtering. Then you create an if-statement step in which you check the "length" of the variable items. Length in this case means, that how many results does one item contain. When you add the if-statement, add the code:
Length(variables('YourVariable')) and select "is greater than 1" from the dropdown menu. This skips all the results that are not duplicates.

The next step is "Compose". What we are doing here is a step, that skips the first result of YourVariable after the length has been checked. At this point all the non-duplicate entries are not present and we are left with ONLY the duplicates, we need to skip the first result, so that there will be one result left in the sharepoint online list. The code for the Compose -step is: skip(variables('YourVariable'),1)

Now it's time to update YourVariable again with the results of the Compose -step.

The final step is to get all the ID-numbers from YourVariable and delete the corresponding items from the Sharepoint Online list. You add the "delete items" step and in the item ID field you add the following code:

variables('YourVariable')[0]['ID']

This finds all the ID's and removes the list items based on those values. I'm not completely sure about the [0], but I think it removes null values, if there are any. I found this in an example and it was not explained. Anyway, the Flow works very well in my case.

[benksyong]

hi,  I ran into problem as below step: Column 'status does not exist'

but the Status column indeed remained in my SharePoint List, and my Odata filter entry: 

With example data in SharePoint list, the duplicated item is 

Title = a AND Status = aaa. 

Expected result post-automate is to remove one of the item named Title= a AND Status = aaa. ...still can't get it right. seek advise. 

Regards,

Yong

[kymramos]

Thanks for sharing this.

I thought it was working beautifully until I realized it's creating duplicates for each array and deleting them.

My list, which was originally at 2k (no duplicates because the other flow only run once), now only has 1 row...

Any suggestions? I followed your step exactly.

[kymramos]

Thanks for this solution! Have you figured out a quicker version of this? My version takes more than 2 minutes to delete a duplicate entry.

Here's the flow I built:
(Overview)

(First pair of "get items from SharePoint" + "apply action if condition was met")

(Second pair)

(Third and last pair)

*A replicate of the first pair, but instead, referencing the version of the SharePoint list (product of the series of Japanese-English translations).

New data gets loaded into the query (Power BI) whenever it gets refreshed, and so whenever the button is clicked, I think Power Automate always starts from the very first row of data and hence the duplicate entries generated. Is there a way to tell Power Automate to only create items that are not currently in the SharePoint list so that it won't have to generate duplicate entries?

I hope my explanation was clear!

Please let me know if you'd need me to expand.

Thanks a bunch!

[Slowpoke]

Hi,

I used this method, but I have a problem when a title has an apostrophe in it. It breaks the ODATA query.

Do you have an idea of an expression that would not break the "Get the items"?

Thanks by advance

[takolota]

@kymramos 

Yes, there is a faster & more efficient way to do this here: https://powerusers.microsoft.com/t5/Power-Automate-Cookbook/Find-and-Remove-Duplicates/m-p/2191403#M...

[Dsa2]

it doesn't work

@fchopo 

[Krono]

Hi,

you need to check what name/code "Get the items" creates for you columns.

The link below shows you, where you can find the correct spelt name of your columns (Post from: ScottShearer 08-16-2019 09:56 AM)

https://powerusers.microsoft.com/t5/General-Power-Automate/OData-filter-column-with-space/td-p/34241...

Does it help?

Regards