News for hackers. Stuff that matters.

A comment on Ask HN: Are there any AI-enhanced solutions for converting PDFs to ePub?. Parent comment

Re:Ask HN: Are there any AI-enhanced solutions for convertin... by 2big2fail_47 on Monday September 11, 2023 @03:52PM (#37472707)
When the PDF is very clearly structured it's working just fine. But let's say the layout consists of multiple columns and complex formatting the output gets very imprecise. If the material is scanned it won't function at all.
[ Parent | on HN ]
Re:Ask HN: Are there any AI-enhanced solutions for convertin... by svennek on Tuesday September 12, 2023 @01:40AM (#37477234)
That is really hard, as there are no such things as columns in PDFs, only text starting at different (x,y) positions.

Hence most (if not all) programs export the text in the order they appear in the file.

And if it is scanned, there is no text at all (but you could OCR it).

[ Parent | on HN ]