Sometimes LLM’s can be quite helpful, especially if you have to solve a problem quite quick. That was the case some weeks ago when my son needed a paper version of the flashcards he generated in Goodnotes.
My wife uses Goodnotes for some years now, learning vocabulary with the children. The program works good and is also available on the iPads we get from the munincipal for the school use of the kids.
Obviously there is a catch, you can only export the flash cards in Goodnotes own format. No export as pdf, csv, txt etc. And that was quite bad in this situation.
So i downloaded a .goodnotes file and had a look. These seem to be standard zip files which contain a bunch of .pb files of unknown propose. Some things in this file is plaintext, some things seem to be binary. So it should be possible to extract that stuff and throw away all the junk. But honestly a LLM could do that way faster than my, and also i suck when it comes to regex, which is needed for this task for sure.
So i head over to Claude, gave it an example export and told it to write an html file to convert the flashcards to pdf. Honestly that worked quite well after two iterations so that i was finally able to create the document my son needed.
You can upload the goodnotes file, than it shows you he pairs so you can correct it if needed and after it you can save it as PDF or csv.
I would upload that piece of html to Codeberg, but since they voted againt „AI Slop“ on their pltform (and i want to respect this) i do not want to risk my other repos by beeing banned for something like this.
If you want to try the script yoou can either download the zipped html from here or you use the hosted version at: https://flashcardtopdf.sengotta.net/
I hope that this may help someone. If your want to improve the code feel free to do so.
