Parse text column by column?

[Rapid1898-code]

Hello - is it somehow possible to extract the attached pdf column by column?
(when i read this pdf-file using extract_text() the text is outputted line by line - but i would need it column by column so i get all the person-information in the correct order)

inp.pdf

[jsvine]

This is something where you'll want to write some custom logic. I'd look something like this:

1. Identify the x-positions separating the columns — either by hardcoding them, or by examining the characteristics of the objects in page.lines.
2. Use those positions to define four bounding boxes, one for each column on the page
3. Use cropped = page.crop(bbox) to zoom in on each bounding box, and then cropped.extract_text() to get the column's text