PDFs can be tough to work with. I try to use the same basic strategy every time I use them. I always keep the PageNumber column, sorted ascending, and then add an index column.
Then, in order to get all of your columns, and so the column values remain in the correct positions, you have to sort the tables in the table column by greatest to least count of columns:
= Table.AddColumn(NameOfPriorStep, "ColumnCount", each Table.ColumnCount([TableColumn]))
Then, sort the ColumnCount column in descending order. In your case, you could also filter out any tables with 0 columns at this point.
Next, I duplicate this query, and filter to just the Table values ( as opposed to the Page values). Then I filter to just pages in the original query.
Only now do I expand the table column. The very next step is to sort again, this time on your index column. Now, you have all of your columns, all in the correct positions (sometimes, some special characters will cause you trouble, like bullet points), and your pages are in the right order.
I promise you, if you take these initial steps anytime you work with PDFs in Power Query, your remaining work will be much, much easier!
--Nate
I’m usually answering from my phone, which means the results are visualized only in my mind. You’ll need to use my answer to know that it works—but it will work!!
