Forum Discussion
PowerQuery removes nullable binary field when saving data flow, making full folder load impossible
I have created a dataflow on PBI service that connects to a SharePoint folder, with the intention of having it grab all the CSV files in the folder. Everything looks good initially, I can see the CSV files, and the "Content" column is there on the far left:
After I hit Save & Close however, Power Query adds a remove step that get's rid of this column, so I cannot see it in Power BI Desktop:
I was hoping to load this dataflow in desktop multiple times, and then grab individual CSVs from it based on date, but with the content column removed, I cannot expand a CSV into a table in desktop. Is there any way around this? I was hoping to not have to create multiple dataflows for each CSV I want to grab.
Thanks!
MarlonK
I have experienced the same issue and expanded the Binary column in the Dataflow itself then connected from Desktop.________________________
If my answer was helpful, please click Accept it as the solution to help other members find it useful
Click on the Thumbs-Up icon if you like this reply 🙂
7 Replies
MarlonK
I have experienced the same issue and expanded the Binary column in the Dataflow itself then connected from Desktop.________________________
If my answer was helpful, please click Accept it as the solution to help other members find it useful
Click on the Thumbs-Up icon if you like this reply 🙂
Yeah, that seems to be the only choice.
Can you elaborate on this solution? I am not familiar with what you mean by "Expand the binary colum in the dataflow".
My thought was to use a view on the source system and conver the binary to varchar.Might be kind-of late, but I had the same problem and derived the answer from Fowmy response. What he means is to click on the [Binary] column of the row (csv file) that you want so the [Binary] field is not returned by the dataflow. If you have a query to gets the list of files and then other queries that reference that list, you can instruct the dataflow to not load the query that gets the list. You do this by right clicking on the query that contains the [Binary] content and unselect the Enable Load option. Bottom line is the dataflow will not allow you to return a binary data type out of your dataflow.
Computed entities are saved as CSV files in data lake storage, so binary values aren't going to be supported
I'm having the same issues currently, what would be a workaround for this?
