I’ve done It before. It’s just highly dependent on the PDF. What software made the pdf? What format was it in prior to being a PDF? Is it a scanned image of text? Or does it have actual text.
Another option, that might work better, might not, though it would be more work, is use something like Microsoft Word first. Word can open a PDF and “convert” it to Word format. After that, you spend a lot of time fixing all the formatting which will probably be fucked up.
And then, once you have a nice clean Word doc, convert that to epub.
I’ve had some trouble with PDFs that were just images of pages of text (easy way to tell, assuming you’re on linux, is run pdftotext on it and see if you get anything). There’s a utility called pdfsandwich that will use Tesseract to OCR the images and add text to the PDF.
I guess better check latest version. But pdf is a pain, it places letters geometrically and loses information about stuff like text flow. You have to use heuristics to put the words and sentences back together.
Calibre should be able to convert from one to the other. Pandoc as well if you prefer CLI things.
Calibre does not do the job
I just checked and tested it, works for me. Is there anything unusual about the PDF?
I’ve done It before. It’s just highly dependent on the PDF. What software made the pdf? What format was it in prior to being a PDF? Is it a scanned image of text? Or does it have actual text.
Another option, that might work better, might not, though it would be more work, is use something like Microsoft Word first. Word can open a PDF and “convert” it to Word format. After that, you spend a lot of time fixing all the formatting which will probably be fucked up.
And then, once you have a nice clean Word doc, convert that to epub.
I’ve had some trouble with PDFs that were just images of pages of text (easy way to tell, assuming you’re on linux, is run
pdftotexton it and see if you get anything). There’s a utility calledpdfsandwichthat will use Tesseract to OCR the images and add text to the PDF.That might help too.
Thx crude scans are the only way you can get a lot of the more fringe books on https://annas-archive.li/ etc.
Pandoc can do epub to PDF but not the other way.
Hm. I stand corrected. I could swear it did both ways.
I guess better check latest version. But pdf is a pain, it places letters geometrically and loses information about stuff like text flow. You have to use heuristics to put the words and sentences back together.