Is the pdf output deterministic? Can I use the parsed text to reliably determine if the contents of a pdf has changed

[jahirulislam-forgent]

Basically title

I have a workflow for detecting updated files. The primary mechanism is byte hash. But sometimes pdf files have different metadata even though the content itself is 100% same

So is the parsing output stable? Can I use it to reliably detect if the content of a pdf file has changed?

[logan-markewich]

Afaik its stable yes.

[logan-markewich]

If OCR is triggered, the OCR text (not the text in the pdf itself) can change slightly across machines though, just due to floating point variance