Extract text from specific regions in a PDF

[michalt1]

Looking for a solution to extract text from 3 zones in PDF files (2 in header and 1 in footer). I built a cloud flow using Encodian which works but I will need a paid subscription to use this on a large scale which I am trying to avoid if possible. I am struggling a bit with the PAD flow and keep getting the below error:

Microsoft.Flow.RPA.Desktop.Modules.SDK.ActionException: Can't rename file ---> System.ArgumentException: Illegal characters in path.
at System.IO.Path.CheckInvalidPathChars(String path, Boolean checkAdditional)
at System.IO.Path.GetExtension(String path)
at Microsoft.Flow.RPA.Desktop.Modules.File.Actions.RenameFilesAction.<>c__DisplayClass106_0.<GetNewFileNameResolver>b__0(FileObject file)
at Microsoft.Flow.RPA.Desktop.Modules.File.Actions.RenameFilesAction.Execute(ActionContext context)
--- End of inner exception stack trace ---
at Microsoft.Flow.RPA.Desktop.Modules.File.Actions.RenameFilesAction.Execute(ActionContext context)
at Microsoft.Flow.RPA.Desktop.Robin.Engine.Execution.ActionRunner.Run(IActionStatement statement, Dictionary`2 inputArguments, Dictionary`2 outputArguments)

The text I am trying to extract is English and nothing complex, what am I missing?

[seinecle]

Hi,

No relation to Encodian but I just happen to publish a free web app doing just that: text extraction from specific regions in pdfs. Feel free to try it and send me some feedback:

https://nocodefunctions.com/pdf_region_extractor/pdf_region_text_extractor_tool.html

[Agnius]

The action that is failing is not the OCR action, but the Rename file(s) action. It's failing because you're trying to include some characters that are not allowed in Windows file systems.

I suggest using a Replace text action on your result to replace the any illegal characters before using it in Rename file(s). Try using this (you can copy and paste this directly into PAD):

Text.Replace Text: OCRText TextToFind: $'''[<>:\"/\\\\|?\\*]''' IsRegEx: True IgnoreCase: False ReplaceWith: $'''%''%''' ActivateEscapeSequences: False Result=> OCRText

This will replace any illegal character with nothing.

You could also insert a whitespace or an underscore there, if needed, instead of %""%.

-------------------------------------------------------------------------
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.

[michalt1]

Thanks this is very nice!! I would need some way to do this automatically through a flow since it would be multiple files a day, so would love if there was a way to build this into a flow.

[Agnius]

The code I provided can be copied and pasted directly into your PAD designer to create the Replace text action in your flow. Try it.

-------------------------------------------------------------------------
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.

[takolota]

@Agnius @michalt1 

There’s also now this template to extract data from any part of any PDF in any structure using a cloud flow with text recognition & GPT processing: https://powerusers.microsoft.com/t5/Power-Automate-Cookbook/Extract-Data-From-PDFs-and-Images-With-G...