Excel in OneDrive gets automatically corrupted

[Gjakova]

Another week, another issue...

I am using an Excel source in PowerApps which is saved in OneDrive. I created a "master sheet" meaning that nobody can edit the sheet. I have formatted the needed columns as a table. PowerApps added a new column named __PowerAppsId__ and so far it works.

But when I looked into my Excel file again, I get the following message:

Why does it keep removing my AutoFilter and Table? What the hell? I have never experienced stuff like this with Mendix and other low-code tools.

[poweractivate]

@Gjakova I don't think AutoFilter is supported in Excel Data Source of Power Apps. If you want Sort or Filter functionality I recommend Dataverse. If you prefer not go so far as to use Dataverse you can also start with SharePoint instead as well and you could still achieve that functionality.  You can use Filter from Power Apps Canvas App to perform the Filtering - but it should not be prefiltered on Excel data source by AutoFilter I believe - I don't think that's supported for use in conjunction with Canvas App and an Excel with an AutoFilter embedded in it.

You may be able to keep using Excel data source but you may have to take the AutoFilter out and use Filter function from Canvas App instead on the data source in the Canvas App that is representing your Excel file. 

Check if the above helps.

Also in general when I am using Power Apps I almost never use the Excel Data Source except in the very small prototypes for the demonstration only, and only as a very small introduction to Power Apps. In general I am using SharePoint or Dataverse - sometimes I use other Data Sources as well - but in general I try to stay away from Excel for anything except maybe for some very specific and bare simplest use cases.

I can't tell from the post but if you are trying to do analytics I also would recommend Power BI - there you can leverage the full power and definitely import Excel into there. You can transform the data however you want and ultimately you can leverage Power Apps Canvas App as well on the transformed data.

In any case some more details of your scenario could help, as I think someone from this forum might be able to point you in a good direction with the best approaches for your scenario. 

[Gjakova]

Hi @poweractivate , thank you! But the reason that I am using OneDrive is that the End Users make some specific calculations within Excel and it saves them hours of work. I now created a template for them and that template is basically a copy (column names) of my SQL table.

So all they have to do is copy & paste their results into my template and then Power Apps will read that Excel file, do some calculations on some field inside Power Apps and then Patch it to my SQL database.

Only problem is that if Users change the Excel template then the connection breaks, so I need to make sure that users can insert safely 100+ rows into the template so that Power Apps can read it.

[poweractivate]

@Gjakova After the Connection to Excel is made as data source, that file should not be edited again I think. If it was another data source it is designed to work when changing the data source from outside Power Apps too, but I am not sure that Excel file should be changed from outside Power Apps afterwards.

I wonder if what you can do is that for each new Excel file it can be made the Connection again, and you can use Data Source Environment Variables (which I think was introduced within the past few quarters, specifically the Data Source ones) https://docs.microsoft.com/en-us/powerapps/maker/data-platform/environmentvariables - then what happens the Environment Variable can point to the new Excel File and the Connection can be made (which makes that extra column __PowerAppsId__ automatically) and the data can be brought in.  I am not so sure of this approach, it is in theory. If it were me I probably wouldn't even do it this way, but I mention it in case it helps you.

As for adding the rows in Excel. I don't think it works to modify Excel file from outside at random. Also this is a bit risky because if someone download the Excel File, start changing it randomly and replace it, that can be a problem of data integrity. What happens if someone else just does that 5 minutes later? The data would be in inconsistent state. I feel like this manner of working of randomly taking copy of the underlying database, changing it, reuploading back, and the risk some other random person might do this with a stale copy of it would cause some problems if it were me. The above approach with the environment variables of data source would remove this problem, but it is very theoretical and I didn't try it - and also there would no longer be only one copy of the file.

Or....can it? You can take the orphaned files and just use Power Apps (or a Flow) to write in bulk each row to a Master Excel sheet. Again pretty purely theoretical but in case it helps you can consider it.

I would encourage someone else if they have any idea for this scenario they can reply, really I am not sure of this way of working and maybe would prefer if someone else could chime in what they would do in this kind of scenario where Excel File have to be the data source and this manner of working of changing the Excel File manually all the time has to be done.