News for hackers. Stuff that matters.
Re:Ask HN: What interesting thought did you read on HN but c... by arjunbajaj on Sunday May 27, 2018 @01:54PM (#17168137)
A service that converts PDF research papers into a clean readable webpage.

Their example paper was a research paper with two columns and diagrams to show their service actually works well.

I searched a couple of times but couldn't find them again...

Re:Ask HN: What interesting thought did you read on HN but c... by orbuch on Sunday May 27, 2018 @04:59PM (#17169073)
Re:Ask HN: What interesting thought did you read on HN but c... by arjunbajaj on Tuesday May 29, 2018 @06:30PM (#17183364)
Yes that's the one! Thank you so much!
Re:Ask HN: What interesting thought did you read on HN but c... by allenz on Sunday May 27, 2018 @02:13PM (#17168220)
Re:Ask HN: What interesting thought did you read on HN but c... by arjunbajaj on Sunday May 27, 2018 @03:04PM (#17168522)
I have stumbled upon this, but unfortunately, this is not it.

That was a hosted web service and it extracted the text and figures from the research paper but didn't maintain the structure. You could read it on a phone too.

I remember it didn't work on all research papers but was slowly expanding its database of supported papers. There was a place to paste the URL of a research paper to be processed.

Re:Ask HN: What interesting thought did you read on HN but c... by severine on Sunday May 27, 2018 @02:15PM (#17168234)
Sounds like pdf2htmlEX. Discussion:

https://news.ycombinator.com/item?id=5657988

Pdf2htmlEX – Convert PDF to HTML without losing text or format (github.com) 161 points by coolwanglu on May 5, 2013 | 48 comments

http://coolwanglu.github.io/pdf2htmlEX/demo/demo.html (Demo of research paper)

http://coolwanglu.github.io/pdf2htmlEX/

Re:Ask HN: What interesting thought did you read on HN but c... by slow_donkey on Sunday May 27, 2018 @03:45PM (#17168762)
Fwiw I've looked into these sorts of tools in the past for a work project and none of them work very well. Converting HTML to PDF is a solved problem but the other way around never comes out clean.

If your document structure is p simple it might be fine but I wouldn't bother for anything else.