I’m feeding a bunch of PDF reports into an LLM for a side project, but the tables come out as garbled text with all the columns merged together. Has anyone found a reliable way to extract them cleanly?
I agree, that is annoying. Honestly, I try selecting the table in the PDF reader, copying it, and pasting it into Excel or Sheets. If it lands in one column, I use Text to Columns with fixed width or tab delimiter, then save as CSV. To be fair, scanned pages or image tables usually will not copy cleanly, and then I am not sure there is a quick fix.