Pdf error

At a glance

The post is a request for advice from community member Logan M about an issue. The comments suggest that the community members are having trouble loading PDF files on a remote server, but the files load fine locally. They discuss potential solutions, such as using a different PDF parser package or loading the PDF directly and converting it to a document object. One community member provides example code for creating a document object from a PDF text string. However, there is no explicitly marked answer to the original question.

SSenna

Hi @Logan M can you drop some wisdom abt this 😩 please

14 comments

LLogan M

Lol yea I saw this earlier today but tbh I have no idea

You are running on a remote sever, and the pdf won't load on the server but it loads locally?

SSenna

Yea

SSenna

or is there a version of llamaindex that uses different pdf parser package? I believe theres a version where it uses PYPDF2?

SSenna

and it loads txt file just fine. but pdf file, its always empty file

LLogan M

You could always load the pdf yourself with a pdf library of your choice, and the convert to a document object

WOW

yeah that helps

Lol it does?

how to convert to document object?

SSenna

bcs i can load it just fine with python reader

LLogan M

Plain Text

from llama_index import Document

document = Document("my pdf text string", doc_id="optional doc id", extra_info={"optional": "info dict"})

LLogan M

You can shove the entire pdf into one document, or split it any way you like and create many document objects

LLogan M

The doc id and extra info are optional (I think I made that clear lol but just making sure)

SSenna

Thanks

Add a reply

Find answers from the community

Pdf error