How to PDF pages using Power Automate Desktop where some invoices are over two pages rather than the usual one page.

[Nerfino80]

Hi,

I am having an issue where I built a flow where it can split PDF document invoice into single pages based on heading. The issue I am facing is how do I do it where some invoices are over two pages. 

So the heading will be on page 1 and the next heading is on page three. I need page 1 and 2 to be built into a single PDF document. 

Invoice will be coming from different people so the heading will change but the variable will be passed from the cloud flow alongside the page number. Then PAD will loop and extract the pages by text reference. 

I have attached my PAD flow with page number from Cloud Flow alongside two invoices. One invoice covering one page the other is over two pages then on the third page and forth page it covering the invoice on a single page for both three and four. 

[Deenuji]

@Nerfino80 

could you please explain with some screenshot or sample pdf document of your use case so that we can try and propose appropriate solution for you. Thanks.

Thanks,
Deenuji Loganathan 👩‍💻
Automation Evangelist 🤖
Follow me on LinkedIn 👥

-------------------------------------------------------------------------------------------------------------
If I've helped solve your query, kindly mark my response as the solution ✔ and give it a thumbs up!👍 Your feedback supports future seekers 🚀. If you'd like to appreciate me, please write a LinkedIn recommendation 🙏

[VishnuReddy1997]

Hi @Nerfino80 ,

It that headings are same or any common points in PDF.

As mentioned @Deenuji ,Please share the screenshot or PDF File.

(Note:- if you got your solution you can mark as solution and gives kudos)

Thanks & Regards

Vishnu Reddy

[Nerfino80]

I have updated with screenshot thank you for looking into this

[Nerfino80]

Screenshot have been added thank you. 

[Deenuji]

@Nerfino80 
I understand that the first invoice consists of one page containing the invoice information, while the second invoice spans two pages. Could you please clarify what exactly you're trying to achieve here? I'm not entirely clear on your objective. Could you elaborate a bit more?


Thanks,
Deenuji Loganathan 👩‍💻
Automation Evangelist 🤖
Follow me on LinkedIn 👥

-------------------------------------------------------------------------------------------------------------
If I've helped solve your query, kindly mark my response as the solution ✔ and give it a thumbs up!👍 Your feedback supports future seekers 🚀. If you'd like to appreciate me, please write a LinkedIn recommendation 🙏

[Nerfino80]

What I am trying to do here is when the invoice comes into sharepoint cloud flow will pick that up and it a 25 page PDF documents with multiples invoice on one documents. I am trying to split this one documents into multiples documents based on the heading. I can get it to work when the heading and invoice on one page but across two pages is where I am struggling.  

[IWantToKnow]

Hi @Nerfino80 ,

When you say you were able to split the PDF into single files you meant that you did it with indicating the text to split to and from or simply split the file into single pages?

[Nerfino80]

Yeah each invoice got a header starting with "Short" in a one documents over three pages and I managed to split this into 3 different PDF documents based on the header "Short" so I ended up with 3 invoice document from 1 document. 

Just Headers spaced out over two pages is where I am struggling

[Agnius]

Split them into individual files per page, and build a Loop condition while 1=1 (infinite loop). Then use Get files in folder to pick up your files inside this loop. If %Files.Count% = 0, use Exit loop to stop looping.

Then if it's not equal to 0, continue the flow. Add a For each loop to iterate through the files. Use Extract text from PDF to get the text from the current item and store it in a variable. Continue looping until the header value changes. When it does, use Exit loop to exit the For each, and merge the files you've iterated through using Merge PDF files. Once done, delete the original files, and move the merged file elsewhere, so it is not picked up during the next iteration of the while loop. 

It's an ugly workaround, but it can be made to work. We've done that for a client in the past.

[Nerfino80]

Thank you for sharing this Agnius, I have attempted this are you able to share an image or build of how you done it so I can try replicate it please. 

[Agnius]

Nope, not really, as the specific solution was under an NDA and GDPR protection due to it being related to processing personal information of employees at the client.

But the idea was more or less as follows:

1. Extract PDF file pages to new PDF file - each page into a separate file, named sequentially
2. Set variable %PreviousValue% to "" (empty value)
3. Loop condition -> While 1=1
   1. Get files in folder and store it into %Files%
   2. If %Files.Count% = 0
      1. Exit loop
   3. End
   4. Create new list for the pages to be merged and name it %List%
   5. For each %File% in %Files%
      1. Add item to list to add %File% to %List%
      2. Extract text from PDF and store it into %ExtractedPDFText%
      3. Parse text (likely with regex) to get the specific value that needs to be compared between pages - in our case it was the person ID and store it into %CurrentValue%
      4. If %PreviousValue <> "" AND CurrentValue <> PreviousValue%
         
         1. Set variable %PreviousValue% to %CurrentValue%
         2. Exit loop
      5. End
      6. Set variable %PreviousValue% to %CurrentValue%
   6. End
   7. If %List.Count% > 1
      1. Merge PDF files to merge %List%. Put it in some other directory, so it does not get picked up in the next iteration of the loop.
      2. Delete file(s) to delete %List%
   8. Else
      1. Move file(s) to move %List%, as it does not need to be merged due to only having one file. 
      2. Rename file(s) optionally, if you want to set a different name to the file.
   9. End
4. End

[Deenuji]

@Nerfino80 

Could you kindly provide a preview of the data from both page 1 and 2, ensuring any sensitive information is hidden? Additionally, I'd like to verify if the Debit Note Number follows a standard naming convention. Could you please confirm this?

[Nerfino80]

Standard naming convention can be anything from pages to prices etc nothing complex at all. 

[Deenuji]

@Nerfino80 

If you provide some screenshot(Hide only sensitive data) like above, it will help me to write some action to assist on your use case(Want to see the structure of the pdf extracted text).

[Nerfino80]

This is the flow I built based on your instructions and I think I am nearly there it just the merging of the files is where it failing. I can see the files in the list. 

Are you happy to read through my flow images to ensure that I have set everything out correctly. The other part I am confuse about is the currentvalues and previousvalues. 

I appreciate you all taking your time out to help me. 

[Deenuji · Accepted solution]

@Nerfino80 

Try the below approach and code for your use case and sample pdf also attached for your testing.

Source code:

SET PdfFilePath TO $'''C:\\Boot\\3page_invoice.pdf'''
File.GetPathPart File: PdfFilePath RootPath=> RootPath Directory=> Directory FileName=> FileName FileNameWithoutExtension=> FileNameNoExtension Extension=> FileExtension
SET Count TO 0
LOOP WHILE ($'''True''') = ($'''True''')
    Pdf.ExtractPages PDFFile: PdfFilePath PageSelection: Count + 1 ExtractedPDFPath: $'''%Directory%\\Temp PDF from a desktop flow''' IfFileExists: Pdf.IfFileExists.AddSequentialSuffix ExtractedPDFFile=> ExtractedPDF
    ON ERROR PageOutOfBoundsError
        GOTO Finish
    END
    File.Delete Files: ExtractedPDF
    Variables.IncreaseVariable Value: Count IncrementValue: 1
END
LABEL Finish
Display.ShowMessageDialog.ShowMessage Title: $'''Flow finished running...''' Message: $'''The selected PDF file has %Count% pages.''' Icon: Display.Icon.None Buttons: Display.Buttons.OK DefaultButton: Display.DefaultButton.Button1 IsTopMost: True ButtonPressed=> ButtonPressed2
SET PageCounter TO 1
SET FromPage TO 1
SET ToPage TO 0
LOOP LoopIndex FROM PageCounter TO Count STEP 1
    Pdf.ExtractTextFromPDF.ExtractTextFromPage PDFFile: PdfFilePath PageNumber: PageCounter DetectLayout: False ExtractedText=> ExtractedPDFText
    Text.ParseText.RegexParseForFirstOccurrence Text: ExtractedPDFText TextToFind: $'''(?<=Supplier\\s*)\\S+''' StartingPosition: 0 IgnoreCase: False Match=> Match
    IF Match = 1 THEN
        IF (FromPage > 0 AND ToPage > 0) = $'''True''' THEN
            Text.FromNumber Number: FromPage DecimalPlaces: 0 UseThousandsSeparator: False FormattedNumber=> FromPageText
            Text.FromNumber Number: ToPage DecimalPlaces: 0 UseThousandsSeparator: False FormattedNumber=> ToPageText
            Pdf.ExtractPages PDFFile: PdfFilePath PageSelection: FromPageText + '-' + ToPageText ExtractedPDFPath: $'''C:\\Boot\\invoice.pdf''' IfFileExists: Pdf.IfFileExists.AddSequentialSuffix ExtractedPDFFile=> ExtractedPDF
            ON ERROR PageOutOfBoundsError
                GOTO Finish
            END
            SET ToPage TO 0
            SET FromPage TO PageCounter
        ELSE IF (FromPage > 0 AND FromPage <> 1 AND ToPage = 0) = $'''True''' THEN
            Text.FromNumber Number: FromPage DecimalPlaces: 0 UseThousandsSeparator: False FormattedNumber=> FromPageText
            Pdf.ExtractPages PDFFile: PdfFilePath PageSelection: FromPageText ExtractedPDFPath: $'''C:\\Boot\\invoice.pdf''' IfFileExists: Pdf.IfFileExists.AddSequentialSuffix ExtractedPDFFile=> ExtractedPDF
            ON ERROR PageOutOfBoundsError
                GOTO Finish
            END
            SET ToPage TO 0
            SET FromPage TO PageCounter
        END
    DISABLE ELSE IF PageCounter = Count THEN
        DISABLE Text.FromNumber Number: FromPage DecimalPlaces: 0 UseThousandsSeparator: False FormattedNumber=> FromPageText
        DISABLE Pdf.ExtractPages PDFFile: PdfFilePath PageSelection: FromPageText + '-' + Count ExtractedPDFPath: $'''C:\\Boot\\invoice.pdf''' IfFileExists: Pdf.IfFileExists.AddSequentialSuffix ExtractedPDFFile=> ExtractedPDF
        ON ERROR PageOutOfBoundsError
            GOTO Finish
        END
    ELSE
        SET ToPage TO PageCounter
    END
    SET PageCounter TO PageCounter + 1
END
Text.FromNumber Number: ToPage DecimalPlaces: 0 UseThousandsSeparator: False FormattedNumber=> ToPageText
Text.FromNumber Number: FromPage DecimalPlaces: 0 UseThousandsSeparator: False FormattedNumber=> FromPageText
IF (FromPage > 0 AND ToPage > 0) = $'''True''' THEN
    Pdf.ExtractPages PDFFile: PdfFilePath PageSelection: FromPageText + '-' + ToPageText ExtractedPDFPath: $'''C:\\Boot\\invoice.pdf''' IfFileExists: Pdf.IfFileExists.AddSequentialSuffix ExtractedPDFFile=> ExtractedPDF
    ON ERROR PageOutOfBoundsError
        GOTO Finish
    END
ELSE
    Pdf.ExtractPages PDFFile: PdfFilePath PageSelection: FromPageText ExtractedPDFPath: $'''C:\\Boot\\invoice.pdf''' IfFileExists: Pdf.IfFileExists.AddSequentialSuffix ExtractedPDFFile=> ExtractedPDF
    ON ERROR PageOutOfBoundsError
        GOTO Finish
    END
END

Please try and let me know in case of any queries. Please build your own logic to keep new name for newly creating pdf files.

How to copy/paste the above code into your PAD?

Thanks,
Deenuji Loganathan 👩‍💻
Automation Evangelist 🤖
Follow me on LinkedIn 👥

-------------------------------------------------------------------------------------------------------------
If I've helped solve your query, kindly mark my response as the solution ✔ and give it a thumbs up!👍 Your feedback supports future seekers 🚀

[Nerfino80]

Thank you both for sharing, this seriously will help so many people with multi pages PDF 🙂