Answered

PDF has multiple tables, need to extract correct ones

Like (0) Share

Report

Posted on 25 Sep 2023 22:30:05 by StanM

Hi, I have a large PDF file which contains different types of tables. Each type of table is in the PDF multiple times. I would like to extract only one type of table, all occurences and write to an Excel sheet.

For example, I have a table of outstanding invoices for customers. Each table contains only the invoices for one customer. We have 30 customers, so there are 30 of these tables (table 1) in the PDF.

Within the same PDF there are tables for invoices we owe to our vendors. We have 5 vendors, so there are 5 of these tables (table 2) in the PDF.

I would like to extract only table 2 information. Above each of these tables is text stating "Invoice info for Vendor xxx" where xxx is a vendor number. I can Parse Text to find "Invoice info for Vendor", then retrieve the vendor number with Get Subtext. How do I then extract the table underneath the text? I have tried Extract Tables from PDF, but when I parse the text, I do not know what the index of the associated table is.

All responses (3)

Answers (1)

StanM 23 on 27 Sep 2023 at 13:43:23

Like (1)

Report

Re: PDF has multiple tables, need to extract correct ones

Thank you for the replies.
I took your advice, Agnius, and split the PDF into multiple 1-page PDFs and processed that way. If the heading of a particular page was what I was looking for, I then extracted that page's table.
Verified answer

Agnius Bartninkas 10,045 Most Valuable Professional on 26 Sep 2023 at 11:03:26

Like (1)

Report
Re: PDF has multiple tables, need to extract correct ones

If you do not have a license for AI builder and don't want to purchase it, I can think of a couple of alternatives for doing this:

You could split the document into a separate file per page and then process each resulting page as a separate document. This works best when:

Your tables do not span over several pages,

Each table starts in a new page - no pages with several tables

You could just use Parse text with regular expressions to retrieve the relevant values from your PDF using certain keywords and patterns.

Regex is not limited by the two limitations option 1 is limited by, so it is essentially a better option, but arguably more complex to build. Especially if you are not familiar with regex at all. But since this is custom per document, in order for us to actually provide some guidance, we would need to get sample data.

-------------------------------------------------------------------------
If I have answered your question, please mark it as the preferred solution. If you like my response, please give it a Thumbs Up.

I also provide paid consultancy and development services using Power Automate. If you're interested, DM me and we can discuss it.
OkanMTL 703 Super User 2024 Season 1 on 26 Sep 2023 at 07:42:37

Like (0)

Report

Re: PDF has multiple tables, need to extract correct ones

Have you tried Document Processing offered by microsoft? This suits your problem better as this uses AI to recognize tables from a File. After your Document Processing unit it build, you can run it from cloud or PAD as you like

Good luck