I think it should work in PDFs using Apple’s own markup solution. Just preview the PDF and copy the text.
With images, I don’t think we added it, but it may be worth doing.
We are working on search at the moment. We can take it along for consideration.
Thank you. What I meant is to use that not just for text copying, but for search in general. It would be amazing for stuff like scanned business cards, receipts etc (apple notes is good for it, but consolidating this all in 1 app would be great imo.)
I see. Yes, I was thinking along those lines too.
I’ve searched for similar posts so apologies if this has been addressed, but is search within handwritten notes and/or full OCR support on the roadmap?
It’s not supported yet. We do hope to support that at some point, certainly for search within PDFs, for example.
Handwriting is more challenging. There is no built in support for handwriting provided by Apple yet. They use it in some of their own apps, but do not expose it for use by 3rd party apps. We hope that will change in time, and we will certainly support it if it does.
Search within text-based PDFs, even before an OCR option, would already be a huge improvement.
The ability to search text in PDFs is a huge improvement and advantage for Agenda users, I believe.
Looking into this a bit as I was curious if Apple Intelligence may have added some options and came across another potential solution.
Assuming you are using PencilKit framework, are you able to convert the PKDrawing to a standard UIImage and then run it through the Vision framework’s OCR?
You can accomplish this outside of Agenda by taking a screenshot and using native OCR on that image, but this is extra steps and unecessary duplication of information being copied out of the image and pasted into the Agenda Note.
Yes, it is doable, but quite a bit of work. We actually already do the OCR you mention for scans, when we generate a PDF from the scan.
Handwriting won’t work well though. OCR with handwriting is quite poor. Apple and others make it work with handwriting by including the timing of strokes in the data. This gives much more information than just an image, and that is how it works. But again, only for a few languages I think.
It would certainly be good if we could add the contents of PDFs etc to the spotlight index, so you can find things with search. We will try to do this in future.