Strange characters when extracting text

[dasapa]

Hello,

I have extracted data from several invoices, everything has gone perfectly, but with one when I extract text it shows characters like the following:

**_++IIKKB=B?=J?KJHK=HJ=@ +IKJH=!="&&  +IKJH= !=$#'!    +@A=ECK=+=@A=CK= -LK?E?IK -LK?E?IK 6EFKI6EFCAAH= !=$#" '& 6EFKI6EFAIFA?BE? !=$"&$ %%&

_**

Any suggestions if it can be fixed?

Thank you

[samkit-jain]

Hi @dasapa Thanks for your interest in the library. Could you please share the PDF so that we can investigate further? What happens when you use pdftotext? Is the text captured correctly?

[dasapa]

Hi Samkit,

thanks for the answer, there are two pdf files in wetransfer link. What is pdftotext? other librarie?

https://wetransfer.com/downloads/932b33efbce9be5fb51972d908bf5d7c20200927190237/e3355946c178b7b3430d4d07d60807be20200927190304/ee1009

[samkit-jain]

Thanks for sharing the PDFs @dasapa pdftotext is a utility to extract text from PDFs.

In the PDF, if you try and copy-paste, gibberish comes out. It is sometimes done intentionally by the PDF creators to prevent people from copy-pasting text from the PDF in which the unicode mapping (what character a glyph represents) is not included in the PDF or is incomplete. One workaround could be to run OCR on the PDF.

[SAIVENKATARAJU]

I am having the same issue. In my case most of the lines are replacing with question mark. most of the lines are getting lost

[afierro23]

@SAIVENKATARAJU were you able to figure out why you were getting the question marks? I am having a similar issue on some pdfs.

Thanks!

[SAIVENKATARAJU]

@afierro23 In my case, I am not able to do because the pdf creators itself use custom encoding techniques to avoid copying, We should go with OCR if required.