[HELP] Inconsistent Reading order

[emtee14]

Describe the bug
What's the problem?
When processing two PDFs that are visibly identical layout-wise except for different text value one pdf gives the text as a simple left-to-right block of text whilst the other break the text into two columns. I just wanted to know if there is a way to make it read all of it as a block of text and not try to interpret columns.

I've linked two of the PDFs that have the issue. On the 2nd page of both is the balance sheet I'm trying to read.

Expected behaviour (text read as a block) - bit.ly/3zG1LJO
Undesired behaviour (text split into two columns) - bit.ly/3UbvO5Q

• OS: MacOS
• Python version: 3.10
• OCRmyPDF version: 14.0.4
• Platform: ARM

[jbarlow83]

The reading order is inferred by the PDF viewer, unless there is markup in the PDF to indicate the appropriate reading order (such as making a tagged PDF). Adding this information is beyond the capabilities of OCR engines at the moment.

[emtee14]

Ah okay thank you. I'm using pypdf to then extract the text so that must be where the issue lies