invisible text layer

[lbr991]

I'm trying to take a non-searchable pdf and convert it to searchable pdf by superimposing an invisible text layer.
I need the invisible text bboxes to align exactly with the original bboxes (extracted by Textract).
I also don't want to specify a font and font_size because then the bboxes wouldn't align perfectly.

Is this possible with pdfrw?

[sl2c]

Everything is possible with pdfrw, but there's no out-of-the-box solution. However, it can be done in a straightforward way using pdfrwx, see e.g.: https://github.com/sl2c/pdfrwx/blob/master/hocreditor.py — this is an example class that inserts OCR layer specified using hOCR format as invisible text in PDF.